GIFT OF
MICHAEL REESE

EXPLOSIVES

AND THEIR POWER

TRANSLATED AND CONDENSED FROM THE FRENCH

OF
M. BEETHELOT

BY C. NAPIER HAKE,

FELLOW OP THE INSTITUTE OP CHEMISTRY, INSPECTOR OF EXPLOSIVES TO
THE GOVERNMENT OP VICTORIA;

AND WILLIAM MACNAB,

FELLOW OP THE INSTITUTE OF CHEMISTRY.

WITH A PREFACE BY
LIEUT.-COLONEL J. P. CUNDILL, K.A.,

H.M. INSPECTOR OP EXPLOSIVES.

,

WITH ILLUSTRATIONS.

LONDON:
JOHN MURRAY, ALBEMARLE STREET.

1892.

LONDON : PRINTED BY WILLIAM CLOWES AND SONS, LIMITED,
STAMFORD STREET AND CHARING CROSS.

PREFACE.

THE great work of M. Berthelot has for some years been a
mine from which copious stores of valuable matter have been
obtained and translated into various languages.

So far, however, no English translation or adaptation of the
book, as a whole, has appeared.

The idea of making such a translation, or, rather, condensation^
of M. Berthelot's somewhat bulky volumes occurred some time
ago to Mr. Hake and myself. Circumstances, however, notably
the appointment of Mr. Hake to the Inspectorship of Explosives
in the Colony of Victoria, and a considerable pressure on my
own time, prevented our carrying out this project in the way
originally intended. But Mr. Macnab, then associated with,
and subsequently successor to, Mr. Hake in his London business,
has undertaken and carried out the larger portion of the very
laborious work involved, and thus it is really to his energy
and kindness that the work as it now appears is due. M.
Berthelot's reputation as a scientist is world-wide ; his atten-
tion was first especially drawn to explosives in the year 1870,
and his labours have been continued with little, if any, inter-
mission to the present time.

The great key-note of the work now translated is the applica-
tion of thermo-chemistry to the study of explosives. Though
not the first in this field, yet M. Berthelot has, in the extent
and variety of his researches, eclipsed his colleagues, and it is
mainly due to him that thermo-chemistry occupies the position
which it now holds in this department of science.

The book does not pretend to be a practical guide to
manufacture, but is, on the other hand, most valuable to the

IV PREFACE.

manufacturer and practical experimentalist in the indications
which it gives of the properties and. powers likely to be
possessed by an explosive already made, or by one in con-
templation.

Scores of useless and dangerous mixtures would never have
seen the light had the inventors known and profited by what
M. Berthelot has told us.

Since the publication of M. Berthelot's work, new explosives
have come prominently on the scene both for military and civil
purposes.

Perhaps the most noteworthy of these are the various so-
called " flameless " and " smokeless " explosives. To the first
of these belongs a group, whose main constituents are nitrate
of ammonium mixed with dinitrobenzol, or other nitro-
derivative of the benzol series. Such are Eoburite, Bellite,
Securite, and Ammonite, all of which are in use in this country
for blasting purposes, especially in fiery mines. To the second
class belongs the very numerous but not very varied group of
" smokeless " or quasi-smokeless powders. Of these, one or
another has been adopted by most nations for military purposes.
They are divisible into two distinct classes, viz. those which
consist of nitrocellulose as their main constituent, and those
which have not only nitrocellulose, but nitroglycerin as their
principal constituents.

To these two classes they alf practically belong up to the
present time, though there are almost innumerable variations in
added ingredients or details of manufacture. By far the oldest
is the simple nitrocellulose powder. Some forms of it have been
widely used for many years in the sporting world. The older
powders, however, though excellent for shot-guns, failed in the
uniformity of result so essential in a military arm, and the diffi-
culties have been but comparatively recently overcome.

The close attention which has been paid of late years to the
subject of explosives has not been without its effect on the
oldest of them. Gunpowder, not so very long ago a somewhat
haphazard mixture, has been made to take its place as an
explosive deserving and obtaining at least as much care in
its manufacture and treatment as the so-called " chemical
explosives."

Picric acid, too, under various names and in various shapes,
has advanced from the rank of a u&eful article of ordinary
commerce to that of a powerful destructive agent.

PREFACE. V

But of all these recent advances the germs may be found in
M. Berthelot's work, not necessarily in all cases originated by
him, but more or less worked out, examined, and compared,
and having, so to speak, the soil prepared for their subsequent
growth. As previously stated, a certain amount of omission
and condensation has been exercised, for the original volumes
consist rather of a series of essays than one connected work,
and this condensation became, to avoid repetition, not only
advisable, but necessary. Several portions, consisting of matter
of merely historical interest, such as the history and origin of
explosives, and the history of methods of extraction of saltpetre
in France, have been omitted.

M. Berthelot adheres to the older chemical notation; this
has been replaced by that more recently introduced and now
most commonly in use.

It should be added that this book has been produced with
the full consent of M. Berthelot, who has also suggested what
Mr. Macnab has carried out, viz. the addition of abstracts of
some of M. Berthelot's essays published since the appearance
of the main work, and principally relating to the propagation
of detonation in explosive gaseous mixtures, with further
studies on the " explosive wave " in solid and liquid bodies.

J. P. C.

CONTENTS.

BOOK I— GENERAL PRINCIPLES.

CHAPTER I.

The force of explosive substances — Maximum work — High and low explosives
— Distribution of energy — Mixtures and definite chemical compounds

Page 1-4

CHAPTER II.

Products of explosive decomposition — Seven modes of decomposition of
ammonium nitrate — Progressive heating or sudden decomposition — Dis-
sociation modifies the heat disengaged, volume of gas, diminishes
pressure — Calculated and actual temperature and pressure .. 5-13

.CHAPTER III.

Heat produced generally positive— Calculation of heat disengaged — Definition
of small and large calorie — Potential energy of an explosive — Practical
result 14-17

CHAPTER IV.

Pressure of gases — Temperature — Specific heat — Direct measurement of
pressure — Crusher guage — Theory of crushing manometers — Theoretical
calculations — Density of charge and specific pressure — Maximum effort —
" Characteristic product " 18-34

CHAPTER V.

Duration of explosive reactions — Origin of reactions — Sensitiveness of explo-
sive substances — Molecular rapidity of reactions — Increases with tempera-
ture, and with density of body — Influence of process of inflammation — of
shock — Combustion and detonation — Combustion effected by nitric
oxide — Decomposition of endothermal combinations, acetylene, cyanogen,
etc. 35-74

Vlll CONTENTS.

CHAPTER VL

Explosion by influence — Dynamite detonates neighbouring cartridges in
indefinite numbers — Explosion transmitted by water — Theory founded
on existence of the " explosive " wave — Abel's theory of " synchronous
vibrations" — Explanation of these experiments according to the two
waves, one mechanical, the other chemical — Chemical stability of
matter in sonorous vibration .. .. .. .. Page 75-87

CHAPTER VII.

Explosive wave — Analogies and differences between this and the sound wave
— Comparative rapidity of the two kinds of wave — Experimental arrange-
ments— Tubes of lead, glass, caoutchouc — Tables showing theoretical
and found velocities of the wave in various gaseous mixtures — Influence
of initial inflammation — Propagation of explosive wave quite distinct from
ordinary combustion .. .. .. .. .. .. 88-113

BOOK II.— THERMO-CHEMISTRY OF EXPLOSIVE
COMPOUNDS.

CHAPTER I.

General principles — General theorems on reactions — Theorems on formation
of salts ; organic compounds — Relative to variation of heat of combination,
with the temperature and pressure — Thermo-chemical tables .. 114-144

CHAPTER II.

Calorimetric apparatus— Calorimetric bomb — Heat of combustion of gases

145-159

CHAPTER III.

Heat of formation of oxygenated compounds of nitrogen— Energy of nitrates
to be explained on thermo-chemical grounds — Heat of formation of nitric
oxide — Combustion of cyanogen by oxygen and nitric oxide — Heat of
formation of nitrogen monoxide ; of dissolved and anhydrous nitrogen
trioxide and the nitrates ; of nitric peroxide — Formation of nitrogen
trioxide — Heat disengaged by successive fixation of equivalents of oxygen
— Heat of formation of dilute, monohydrated, and anhydrous nitric acid
— Hyponitrous acid and Irpponitrites — Character of nitric oxide 160-201

CHAPTER IV.

Heat of formation of the nitrates — Combustion of explosive mixtures contain-
ing nitrates — Gunpowder imperfectly utilizes energy of its components —
Saltpetre not a very good agent of combustion — Reason for superiority
of organic compounds derived from nitric acid .. .. 202-206

CONTENTS. IX

CHAPTER V.

Origin of the nitrates — Natural nitrification — Chemical and thermal conditions
of nitrification — Necessity for alkaline media and oxygen — Transforma-
tion of free nitrogen into nitrogenous compounds — Action of high and
low tension electricity — 'Importance of atmospheric electricity in fixing
nitrogen on vegetable tissues .. .. .. .. Page 207-236

CHAPTBR VI.

Heat of formation of hydrogenated compounds of nitrogen — Ammonia and
ammoniacal salts — Volatility of ammonium nitrate — Formation of hydro-
oxylamine, ethylamine, trimethylamine, oxamide, formamide .. 237-260

CHAPTER VII.

Heat of formation of nitrogen sulphide — nitrogen selenide .. 261-263

CHAPTER VIII.

Heat of formation of compounds formed by the action of nitric acid on organic
substances— Heat produced by their combustion inversely proportional
to heat produced by union of the acid with the organic principle — Nitro-
benzene — Dinitrobenzene — Chloronitrobenzene — Nitrobenzoic acid —
Picric acid — Nitric ether — Nitroglycerin — Nitromannite — Nitric deriva-
tives from complex alcohols — Nitrostarch— Gun-cotton .. 264-289

CHAPTER IX.

Diazo-compounds — Excess of energy which they contain— Diazobenzene
nitrate — Explosion— Products of decomposition .. .. 290-296

CHAPTER X.

Heat of formation, decomposition, and combustion of mercury fulminate

297-298

CHAPTER XL

Heats of formation of the cyanogen series— Cyanogen— Hydrocyanic acid-
Three methods adopted for measuring its heat of formation — Potassium
and ammonium cyanides — Mercury and silver cyanides — Double cyanides
of mercury and potassium ; silver and potassium — Potassium ferrocyanide
— Cyanogen chloride and iodide — Potassium cyanate .. 299-343

CHAPTER XII.

Oxygenated compounds of chlorine, bromine, and iodine — Thermal formation
of chlorates — Combustion effected by potassium chlorate disengages more
heat than by free oxygen — Successive degrees of oxidation of chlorine —
Perchloric acid and salts — Explanation of the stability of the dilute acid
and instability of the pure acid— Bromic and hypobromous acid— lodic
acid— Comparison of chlorates, bromates, and iodates .. 344-363

X CONTENTS.

CHAPTER XIII.

Metallic oxalates— Conditions under which they are explosive Page 364-366

BOOK III— FORCE OF EXPLOSIVE SUBSTANCES IN
PARTICULAR.

CHAPTER I.

Classification of explosives — Definition of explosives — General list — First
group : explosive gases — Second group : gaseous detonating mixtures —
Third group: definite inorganic compounds — Fourth group: definite
organic compounds, solid or liquid— Fifth group : mixtures of definite
explosive compounds with inert bodies — Sixth group : mixtures formed
by an oxidisable explosive compound and a non-explosive oxidising body
— Seventh group: mixtures with an explosive oxidising base — Eighth
group : mixtures formed by oxidisable and oxidising bodies, neither of
which are explosive separately .. .. .. .* .. 367-370

CHAPTER II.

General data respecting the employment of a given explosive — Chemical
equation — Heat of formation of bodies involved and their products
— Specific heats — Temperatures — Densities — Pressures — Empirical
measures of force of explosives — Practical questions relative to employ-
ment, handling, manufacture, and storage — Tests of stability 371-382

CHAPTER III.

Explosive gases and detonating gaseous mixtures — Their maximum work —
Comparison with work of solid and liquid explosives — Influence of
pressure and initial temperature — Temperature of inflammation — Mixtures
of liquefied gases— Gases and combustible dusts .. .. 383-401

CHAPTER IV.

Definite non-carburetted explosive compounds — Nitrogen sulphide and
chloride — Potassium chlorate — Ammonium nitrate, perchlorate, and
bichromate 402-417

CHAPTER Y.

Nitric ethers— Nitro-ethylic ether — Nitro-methylic ether— Dinitro-glycolic
ether— Nitroglycerin— Nitromannite 418-430

CHAPTER VI.

Dynamites — Necessity of special detonator— Classification — Dynamite proper
— Properties— Precautions when using— Rapidity of detonation— Gases
produced— Dynamite with ammonium nitrate, and with nitrocellulose
base 431-443

CONTENTS. xi

CHAPTER VII.

Gun-cotton and nitrocelluloses — Their composition — Conditions and tests of
stability — Heat of formation and of total combustion and of decomposition
— Variation of products according to density of charge — Gun-cotton
mixed with nitrates and chlorates .. +.. .. Page 444-460

CHAPTER VIII.

Picric acid and picrates— Potassium picrate mixed with nitrate and with
chlorate 461-467

CHAPTER IX.

Diazo-compounds — Mercury fulminate mixed with nitrate and with chlorate
— Diazobenzene nitrate — Sprengel acid explosives — Perchloric ethers —
Silver oxalate 468-476

CHAPTER X.

Powders with a nitrate base— "Reactions between sulphur, carbon, their
oxides and salts — Decomposition of the alkaline sulphites and hypo-
sulphites by heat — The charcoals employed in the manufacture of
powder — Total combustion powders — Service powders — Products of
combustion — Theory of the combustion of powder — Comparison between
theory and observation — Sporting and blasting powder— Powders with
sodium nitrate and barium nitrate base .. .. .. 477-517

CHAPTER XI.

Powders with chlorate base — Dangers of chlorate powders — Explanation of
their easy inflammation and shattering effects— Comparison between
nitrate and chlorate powders 518-526

CHAPTER XII.

Conclusions — Summary of the work .. .. .. .. .. 527-542

APPENDIX.

Abstracts of papers by MM. Berthelot and Vieille on "Detonating Gaseous
Mixtures," " The Rapidity of Detonation in Solid and Liquid Explosives,"
"The Explosive Wave," "The Different Modes of Decomposition of
Picric Acid and Nitro Compounds " 543-553

INDEX •• •• 555

ERRATA,

Page 124, add, w The large calorie is the unit employed, and the equivalents

represent grams."
„ 160, line 7, for "these two substances" read "nitre and other compounds

containing oxygen."
„ 288, line 2 from bottom} for " carbon monnade " read " carbonic oxide."

EXPLOSIVES AND THEIR POWER.

BOOK I.

GENERAL PRINCIPLES.

CHAPTEK I.

FORCE OF EXPLOSIVES IN GENERAL.

THE force of explosive substances is expressed by the pressure
which they exert, and by the work which they accomplish. In
a confined space, pressure results in the simple rending of the
envelope, without any subsequent work being effected. This is
exemplified by the fracture of a shell, through the freezing of
water contained in it, or the splitting of a rock by hydraulic
wedges. The effect of an explosive would be to disperse the
fragments of the shell, or to pulverise or displace the rock.
This subsequent action represents the mechanical work of the
explosive substance.

The pressure is due to the gases evolved, and is dependent on
their volume and temperature. The work done depends princi-
pally on the amount of heat disengaged, which is a measure of
the energy developed.

In other words, the maximum work that an explosive
substance is capable of producing, is proportionate to the
amount of heat disengaged during its chemical transformation.

This may be expressed in kilogrammetres by the formula
425 Q, where Q is the number of units of heat evolved.

This theoretical limit is never reached in practice, but still a
knowledge of it is indispensable, as it is the only absolute
point of comparison.

The effective transformation of this energy into work,
depends on the volume of the gases evolved, the amount of heat
generated, and on the law of expansion.

A fraction only of the energy can be actually realized in

2 FORCE OF EXPLOSIVES IN GENERAL.

practice, in the form of useful work, a considerable amount
being absorbed in heating the surrounding medium, in creating
in it wave-motion, and in various other ways. For instance, in
blasting rock, the useful work consists partly in shattering the
rock, and partly in displacing the shattered masses. The
remaining energy is absorbed by work, owing to (1) incomplete
combustion, (2) compression and chemical changes induced in
the surrounding material operated on, (3) energy expended in
the cracking and heating of the material which is not displaced,
(4) the escape of gas through the holes and fissures caused by
the explosion.

The calculation of the distribution of the energy of an
explosive between the mechanical work accomplished, the heating
of the surrounding medium, and the vibratory movement com-
municated to the ground or air, etc., is very complicated, and
will be treated of in a later portion of this work.

A knowledge of the special properties of explosives enables
us to judge, more or less, which particular explosive is likely to
be suitable for a particular class of work. In popular language,
they are divided into " High " and " Low," and of these two
classes, Dynamite and Gunpowder may be taken as the par-
ticular types, but no hard and fast line can be drawn between
them.

Generally speaking, we mean by " high " explosives, those in
which the chemical transformation is very rapid, and which
exert a crushing or shattering effect; a comparatively slow
chemical transformation and propelling effect being, on the
other hand, characteristic of a " low " explosive.

In mercury fulminate we have an extreme instance of rapid
chemical transformation, accompanied by intense local action,
and other phenomena common to this class of explosives.

The more common " high " explosives are bodies containing
a large amount of oxygen, and possessing a definite chemical
composition.

They are produced by the action of nitric acid on organic
substances forming nitric ethers (nitroglycerin, nitromannite)
or nitro-substitution compounds (picric acid and its derivatives).

In consequence of the intimate contact of the combustible
and the oxygen in such compounds, a more energetic and rapid
action is developed on explosion than that which would result
from a simple mixture.

Perchloric ethers and mercury oxycyanide produce analogous
effects, as also ammonium nitrate, bichromate and perchlorate
(under certain conditions), the acid giving up its oxygen, and
the ammonia its hydrogen.

Formerly the force of an explosive was deduced from the
weight of available oxygen which it contained; but this idea
is inaccurate, for oxygen does not necessarily enter into the

RELATION OP OXYGEN TO COMBUSTIBLE. 3

composition of an explosive substance. Take, for instance,
diazobenzol and nitrogen sulphide or chloride, bodies which are
formed with absorption of heat from their elements, and which
decompose with a reverse thermal action. *

An explosive compound may be employed either in a pure
state or mixed with an inert substance, as in the case of
dynamite, a mixture of silicious earth and nitroglycerin.

The effect of such mixture is to diminish the violence of the
explosion, and to give to it a propelling or rending action rather
than a shattering one.

Or an explosive compound may be mixed with a substance
which increases the force of the explosion, as in the case of
nitroglycerin mixed with an active base. And here it is well
to distinguish three fundamental cases, based on the relation
between the oxygen and the combustible elements in the ex-
plosive body.

This relation is either that of a total combustion, as in the
case of silver oxalate, resolvable by the explosion into carbonic
acid and metallic silver.

C2Ag204 = 2C02 + Ag2.

Or the oxygen is deficient, which is the case in potassium
picrate and in gun-cotton.

Or, on the contrary, the oxygen is in excess; which is the
case in nitromannite and nitroglycerin.

In the last case there may be an advantage in utilising the
whole energy of the explosive body by adding a combustible
such as carbon, or better, nitrocotton, an explosive in itself, in
suitable proportions.

In the second case, where there is a deficiency of oxygen, an
oxidising agent such as potassium nitrate, may be added to the
explosive.

Mixtures, however, in which total combustion takes place are
not always those which produce the greatest effect in a given
weight and under given conditions.

Gunpowder, for example, mixed with a quantity of nitre
sufficient for complete combustion, develops, weight for weight,
less gas, and consequently less pressure, and produces less effect
than ordinary powder, in which there is a deficiency of oxygen.

The effects which result from the substitution of one salt for
an equivalent salt, in explosive mixtures, deserve particular
attention. Let us confine ourselves to the nitrates and to
a simple substitution which does not change the nature of the
powder; for instance, sodium nitrate, or barium nitrate, for
potassium nitrate.

The substituted salt, in equivalent proportions, would hardly
change the amount of heat liberated nor the volume of the
gases in the case of total combustion.

4 FORCE OF EXPLOSIVES IN GENERAL.

But, supposing even that in an incomplete combustion, such
as that of gunpowder, no change in the chemical reactions were
produced, nevertheless the substitution of barium nitrate for
potassium nitrate would result in an increase of the absolute
weight of the mixtures, and consequently diminished pressure
and less heat evolved per kilogramme, for the reason that the
equivalent of potassium nitrate, KN03 being 101, and that of
barium nitrate 130'5, the weight of the oxidising agent neces-
sary to burn a given weight of combustible is increased one-third.

The equivalent of sodium nitrate, NalSTOg, being 85, there will
be a less weight of it required than of potassium nitrate.

The heat set free by this weight, which supplies an equal
amount of oxygen to the combustible, is, moreover, about the same.

The substitution of sodium nitrate for potassium nitrate is
therefore advantageous in this respect. Unfortunately, the
hygroscopic properties of sodium nitrate are against its general
application.

Copper nitrate, Cu(N03)2, would doubtless be preferable to any
other, because its equivalent is a little less than that of potassium
nitrate, and more especially as this salt, in its equivalent pro-
portions, supplies to the combustible bodies a fifth more oxygen
than the alkaline nitrates, in consequence of the total reduction
of the copper. This deserves attention, for potassium, sodium,
and barium remain after the explosion in the state of car-
bonates. By reason of this twofold circumstance, viz. the
lesser equivalent and the larger proportion of oxygen available,
the heat developed by the same weight of copper nitrate in
burning the same combustible is considerably in excess of that
produced by the alkaline salts. Unfortunately, copper nitrate
has such a strong affinity for water that it has hitherto been
found impossible to obtain it in the anhydrous form.

Lead nitrate, Pb(N03)2, and silver nitrate, AgN03, are, on the
contrary, easy to obtain anhydrous, and offer as oxidising agents
advantages equal, if not superior, to those of copper nitrate
when employed in equivalent proportions.

But weight for weight this advantage no longer exists,
because their equivalents (165*5 and 171) are too high. The
price of silver nitrate would, moreover, militate against its
general adoption, and lead nitrate gives off very dangerous
fumes in confined places.

It has been considered advisable to enter into these details
in order to show what a variety of conditions have to be
considered in order to produce an explosive applicable to a
particular class of work, and in which the nature and proportions
of the constituents are such as to develop the maximum effect.
In order to work successfully to this end it is necessary that all
experiments should be directed by certain laws deduced from
chemical and dynamical considerations.

CHAPTEK II.

1. CHEMICAL COMPOSITION.

1. THE composition of the products of explosion can be foreseen
whenever the explosive substance contains enough oxygen to
transform the elements into stable compounds, and at the
highest degree of oxidation, as in the case of nitroglycerin and
nitromannite. This limit corresponds also to the maximum
thermal effect. It is not always attained in practice, especially
by the mixtures which contain potassium nitrate, on account of
the rapidity of the chemical and mechanical reactions and of
the cooling.

The explosive decomposition of certain binary compounds,
such as nitrogen sulphide, gives rise also to known products.

2. On the contrary, when the oxygen does not suffice for
total oxidation, or when ternary substances (not containing
oxygen), such as diazobenzol, are in question, the products
formed generally vary with the conditions of the explosion,
temperature, pressure, expansion, mechanical effects, etc. This is
also the case with black powder, gun-cotton and potassium picrate.

Under these circumstances, the composition of the products
cannot be determined beforehand, but must be ascertained by
special analyses, and for each condition of the reaction.

3. In this connection, experiments may be given relative to
the influence of the initial temperature and the rapidity of
heating on the mode of decomposition of bodies, and especially
the seven different modes of the decomposition (some endo-
thermal, others exothermal) of ammonium nitrate, a definite
compound, which leads to more decisive conclusions than simple
mixtures.

4. The following are the seven different modes of decomposi-
tion which ammonium nitrate undergoes.

(a) The dissociation or partial decomposition of fused or even
gaseous ammonium nitrate into gaseous nitric acid and ammonia,
which seems to be first produced and at a low temperature.
It corresponds necessarily with absorption of heat, namely,

6 CHEMICAL COMPOSITION.

— 41,300 cal. when the solid nitrate is used, and about — 37,000
when the salt is fused.

(6) The formation of nitrogen monoxide from ammonium
nitrate at a higher temperature, and when the heat is carefully
regulated. The reaction : NH4N03 (solid) = N20 + 2H20 (gas)
develops + 10,200 cal., the fused salt about + 14,000 cal.

If the salt be supposed to be previously decomposed into
gaseous nitric acid and ammonia, and the action to have really
taken place between these two compounds, the formation of
nitrogen monoxide, HN03 gas + NH3 = N20 4- 2H20, would
develop + 51,500 cal.

(c) When rapidly heated, the explosive decompositions,
properly so called, of ammonium nitrate take place ; one of
them produces nitrogen and oxygen.

NH4N03 = N2 + 0 + 2H20 (gas).

This reaction develops from the solid salt + 30,700 cal. ; from
the fused salt, about -f 35,000 cal.

(d) Nitrogen and nitrogen dioxide are also formed —

2NH4N03 = N202 + N2 + 4H20,

and + 9200 cal. are given off when the salt is solid and about
+ 13,000 cal. when it is fused.

(e) Heat is also liberated when ammonium nitrate gives rise
to nitrogen, water and nitrogen tetroxide —

4KE4N03 = 3N2 + N204 + 8H20

-f 29,500 caL being set free from the solid salt, and + 33,500
cal. from the fused salt.

(/) The ammonium nitrate may also be conceived as being
transformed into nitrogen, water and nitrogen trioxide.

3HN4N03 = 2N2 + N203 + 6H20.

This reaction liberates + 23,300 cal. from the solid salt, and
about 4- 27,000 cal. from the fused ; but never takes place
alone, as nitrogen trioxide exists only in the dissociated state in
presence of nitrogen dioxide and nitrogen tetroxide.

(g) Lastly, ammonium nitrate can be resolved into gaseous
nitric acid, nitrogen and aqueous vapour under certain influences
such as spongy platinum.

5NH4N03 = 2HN03 + 4£T2 + 9H20

yielding + 33,400 cal. from the solid salt, and about + 37,500
cal. from the fused.

These different modes of decomposition of ammonium nitrate,
which may be distinct or simultaneous, or more exactly the
predominance of any one of them, depend on their relative
rapidity and on the temperature at which decomposition is pro-
duced. This temperature is not fixed, but is itself subordinate
to the rapidity of heating. It has been established by a great

EFFECTS OF SLOW AND RAPID DECOMPOSITION. 7

number of observations, that each mode of decomposition of a
given substance commences at a certain temperature, and in
a given time a limited weight of substance is decomposed.

Special stress is laid upon the singular property which
ammonium nitrate possesses of undergoing several distinct
modes of decomposition, according to the rapidity of heating
and the temperature to which the substance is raised. Of these
decompositions, some take place with liberation of heat, others
with absorption of heat.

5. A similar property is possessed by most bodies which liberate
heat during decomposition, and especially by explosive bodies,
properly so called. It is particularly manifested in proportion
to the difference of the local conditions developed by progressive
heating in a mass which is not instantaneously decomposed.

On the other hand, the sudden explosion of detonating sub-
stances, when they consist of a definite compound such as gun-
cotton, nitroglycerin, mercury fulminate, etc., and when the
explosion is readily brought about, the reaction being uniformly
distributed throughout the entire mass, appears likely, generally
speaking, to give rise to simple and stable products. The
extreme conditions of temperature and molecular vibration
which accompany the phenomena hardly allow of its being
otherwise in a molecularly homogeneous mass.

This is, in fact, what has been verified during the explosion
of gun-cotton, as studied by Sarrau and Vieille.

If previous observers have noticed more complicated decom-
positions, it is because the conditions have been such that
the mass underwent partial coolings, and was decomposed at
certain points by distillation rather than by true explosion.

From researches made in conjunction with Vieille on the
explosion of mercury fulminate, it has been established that this
substance is also decomposed in the most simple manner into
carbonic oxide, nitrogen, and mercury. With gunpowder the
diversity of local conditions of combustion cannot, under any
circumstances, be avoided, because a mechanical mixture of
three pulverised bodies can never attain the same degree of
homogeneousness as a true chemical combination.

6. However, each of the products of the explosion is none the
less formed according to a regular law"; all result, in short from
a small number of definite transformations occurring at various
points of the mixture, and the diversity of which is the conse-
quence of the variety of the local conditions.

If the products remained in contact for a sufficient time, they
would undergo reciprocal actions, which would bring them to
the state corresponding to the maximum heat liberated (at the
temperature and under the same conditions of the experiment) ;
but the sudden cooling which they experience prevents this
state from being realised.

8 CHEMICAL COMPOSITION.

The mode of expansion, the nature of the work accomplished,
and the more or less complete transformation of the heat into
work at the moment of explosion must necessarily play an
important part in this connection.

This diversity in the products helps to explain the very varied
effects which the explosion of one and the same body may
produce, according to the method of inflammation.

2. DISSOCIATION.

1. In order to have a clearer idea of the effects produced by
explosive substances, it is necessary to examine not only the
products obtained after cooling, but also those which are pro-
duced during the explosion, and starting from the moment when
the system reaches the maximum temperature. Now, these
first products are sometimes simpler than those which are
observed after cooling ; they result partly from the formation of
a lower compound. For instance, from a polysulphide splitting
up into sulphur and monosulphide, and partly from incomplete
combination, as in the case of a mixture of water vapour with
its elements, hydrogen and oxygen.

In the above connection it is indispensable to take account of
the phenomena of dissociation.

The quantities of heat and the gaseous volumes under dis-
cussion are calculated at 0° C. and 760 mm. This calculation
is admissible for explosive compounds which can be resolved
into their elements, such as nitrogen sulphide, or for those
which give simple and stable products, such as mercury fulmi-
nate, which can be completely decomposed into mercury,
nitrogen and carbonic acid. But it is inadmissible when car-
bonic acid, water vapour, potassium polysulphide, sulphate,
or carbonate, etc., are formed. In these cases the compounds
probably do not exist as such. At the high temperature developed
during the reaction they are, no doubt, replaced either wholly
or in part by simple combinations, perhaps even by their
elements. Consequently the quantity of heat corresponding to
the real reactions is less than the quantity measured or calcu-
lated from the products observed after cooling, and lowers the
maximum temperature, as well as the corresponding pressure.
This last point is worthy of closer examination.

2. The pressure of a gaseous system is always diminished by
the fact of dissociation.

At first sight, this would seem to be a paradox, as dissocia-
tion has the effect of increasing the volume of gases reduced to
0° and 760 mm., when there has been condensation in the act
of combination, as in the formation of water vapour or carbonic
acid. But, on closer examination, it will be found that in all
known cases of combination accompanied by condensation, the

DISSOCIATION. 9

heat developed by the reaction is such that it increases the
gaseous volume if the reaction take place under constant
pressure, it consequently increases the pressure if kept at
constant volume. The effects are such that the heat liberated
increases the gaseous volume in a proportion greater than the
condensation, the latter being calculated upon the hypothesis of
a total combination effected at the initial temperature of the
system. In other words, the pressure of a gaseous system
cannot diminish, generally speaking, through the fact of an
exothermal reaction, when it takes place at constant volume, and
gives rise only to gaseous products.

But dissociation being an endothermal reaction the increase of
gaseous volume due to that action is more than counterbalanced
by the diminution of volume due to the absorption of heat, and
consequently the pressure can never be increased by dissociation.

3. Let us calculate these changes.

The pressure depends on the temperature developed, and on
the state of condensation of the products. Let t be the tempe-
rature developed by the real reaction, taking place at a constant
volume, and supposing the whole of the heat liberated to have
been employed in heating the products. Let V be the sum of
the volumes of the gaseous bodies which form part of the initial
system, supposing them reduced to 0° and 760 mm.

At the temperature t the final system contains a certain
number of gaseous bodies. Further, let Vx be the reduced
volume which these bodies would occupy if they could be
brought without change of state to 0° and 760 mm.

V 1

The ratio of the reduced volumes ^ =— expresses the con-

V rC

densation produced by the reaction. It is applicable to every
pressure and temperature according to the ordinary laws.

An arithmetical value can easily be found for this ratio for
every chemical reaction of which the formulae are related to the
molecular volumes. For example —

1 2
H2 4- 0 = H20 (gaseous) gives -^ = -

i 2
CO + 0 = C02 (gaseous) gives L = -

K o

Now let us calculate the pressure developed during the
reaction, occurring at constant volume and at the temperature
t ; the initial temperature being zero, and the initial pressure h.

Admitting the laws of Mariotte and Gay-Lussac, the pressure
will become

h X \(l + at)
a being equal to ^}s, as is known.

10 CHEMICAL COMPOSITION.

This pressure will be superior to, less than, or equal to, the
initial pressure according as 1 + at is greater, less than, or equal
to&.

Q

We should note that t = — ; Q being the quantity of heat

0

developed in the reaction and c the mean specific heat of the
products between zero and t°.

4. Further, the pressure increases if the condensation is nil,
that is if k = 1 (chlorine and hydrogen ; combustion of cyanogen
by oxygen). It increases especially if expansion occurs, that is
when k < 1 (combustion of acetylene by oxygen) assuming that
Q is positive in every direct and rapid reaction between gaseous
bodies.

Now let Tc > 1 this condensation is always comprised between
certain limits for definite gaseous compounds, limits such that
K = 4, 3, 2, 1 J. Hence the fundamental condition,

1 + a < £ or Q

c

a condition which is necessary for a diminution of pressure,
cannot be realised except in quite exceptional cases, in which
the heat disengaged by an internal reaction is very slight, and
beyond the scope of any observed reactions. We can assure
ourselves of this by making the calculation by means of the
specific heats at constant volume deduced from specific heat at
constant pressure which Eegnault has determined for many
bodies.

5. The calculation may also be made in a more general
manner by admitting with M. Clausius that the specific heats
at constant volume have an identical value for the atomic
weights of the various simple bodies ; that this value is equal to
2 '4: a number which is found for H = 1, in fact, that it does
not change by the fact of combination.

Now, W being the quantity of heat disengaged in a reaction
between gaseous bodies in relation to the atomic weights, and
M the number of atomic weights which are engaged in the
reaction, the pressure will only diminish if we have

W < 655M(& - 1).

It is easy to see that this condition is not fulfilled in the
best known gaseous combinations. In making the calculation,
whether by the aid of this formula, or the foregoing, no example
has been discovered of diminution of pressure among the
numerous reactions which have been examined.

It should be noted that it is sufficient to make the calculation
for the supposed total reaction, the result being the same for the
supposed partial reaction, that is to say, in the case of dis-
sociation. This can be easily proved, for the uncombined

ACTUAL AND CALCULATED TEMPERATURE. 11

portion does not contribute any heat and only operates by the
difference between the specific heat of the compounds and the
sum of that of the components, a difference which is nil
according to the hypothesis of Clausius.

6. Without carrying this discussion any further, the following
general proposition may be deduced from it relative to chemical
combination.

When the heat disengaged in a reaction taking place between
gaseous bodies, and with the exclusive formation of gaseous pro-
ducts, is entirely applied to heating the products, then there is
always increase of pressure, the volume being constant.

This proposition has very important applications in the study
of explosive substances ; but, of course, it is applicable only to
gases forming gaseous products, for it is evident that the forma-
tion of a solid compound from gaseous components would cause
a reduction of pressure.

The influence of dissociation being thus marked by a lowering
in the pressure of the gaseous systems, it must also be observed
that its existence and effect should not be unduly exaggerated ;
the latter must be less than one would at first suppose on
account of certain compensations.

We will dwell a little on this matter on account of its great
importance.

7. The actual temperature which is developed in an explosive
reaction is in general less than the temperature calculated in
accordance with the specific heats of the gas, estimated at about
the normal pressure and ordinary temperature, since the specific
heat of greatly compressed gases is not constant. In fact, the
specific heat of gases formed with condensation increases with
the temperature, according to the facts observed by Eegnault
and M. E. Wiedemann on gaseous carbonic acid and other
compound gases. It must also increase with the pressure, at
one and the same temperature, in proportion as the gas
approaches the liquid state, the specific heat of a liquid being
nearly always greater than that of the same body in its gaseous
form, at the same temperature. An equal quantity of heat
applied to compressed gases, such as those which are produced
in explosive phenomena, will therefore produce less rise in
temperature than if their specific heat were constant, and equal
to that of the same gases at the normal pressure, as is generally
assumed in these calculations.

Hence a smaller increase in dissociation, which depends
chiefly on the temperature. It is further limited by another
circumstance, relative to the pressure developed.

8. Now, the actual pressure is not so much diminished as
one might judge from a calculation founded on the ordinary
laws of gases, and on the lowering of the theoretical temperature.
The laws of Mariotte and of Gay-Lussac are hardly applicable

12 CHEMICAL COMPOSITION.

in the case of such enormous pressures as those observed in the
combustion of powder. With greatly compressed gases the
pressure varies with the temperature much more rapidly than
would follow from these laws ; it approaches the rate observed
by physicists in the study of vapours. For a given temperature
the pressure is therefore generally higher than that which would
be given by calculating according to the ordinary laws of gases.
This tends to compensate in the calculation of pressures the
contrary influences exercised by the variation in the specific
heats.

Now, the phenomena of dissociation depend on the pressure,
as well as on the temperature. The state of combination of
elements, all things else being equal, is higher as the pressure
is greater — a relation which is easily conceived a priori, and
which is confirmed by experiments relative to the decomposition
of acetylene into carbon and hydrogen at different pressures by
the electric spark.1 But the pressures increase with the tempe-
ratures, and even much more rapidly, as has just been stated ;
the decomposing influence of the temperature can therefore be
compensated, either wholly or in part, by the opposite influence
of pressure.

9. The inverse action of these two classes of phenomena
remains such that a substance undergoing transformation at
constant volume without loss of heat, will tend towards a
certain limiting state ; the transformation of the first portions
will at first raise the temperature and pressure to the point at
which dissociation will limit the phenomenon. This is also a
theoretical maximum, since the mass is continually cooled by
radiation and conduction. But the greater the mass operated
upon, the nearer will this result be approached.

10. The phenomena of dissociation do not only exert their
influence on the maximum effort which the explosive substance
can develop, but they also come into play during the first period
of expansion. In proportion as the gases of the explosive
expand in acting on the projectile, they cool, in consequence of
which the elements enter into combination in a more complete
manner and with the formation of more complicated compounds.
From this there results a new disengagement of heat which
increases without ceasing during the whole of a period of
expansion.

Therefore, in general, the transformation effected in the bore
of a cannon cannot be regarded as adiabatic. The temperature
of the gases will not be lowered by a quantity any way pro-
portionate to the exterior work done, even independently of the
losses of heat due to exterior causes of cooling; seeing that
restoration of heat takes place through the chemical reaction,
during a considerable period.

1 " Annales de Chimie et de Physique," 4e se'rie, torn, xviii. p. 196. 1869.

CURVE OF ACTUAL AND THEORETICAL PRESSURES. 13

11. The true pressures will therefore always be greater, except
at the commencement, than those calculated from the quantity
of heat actually disengaged at the moment of maximum
temperature.

On the other hand they will be at first less than the pressure
calculated from the quantity of heat observed in the calorimeter
at the ordinary temperature. But this latter difference
diminishes, and finally disappears altogether in proportion as
the volume increases, the reactions becoming more complete.
The curve of the true pressures, expressed as a function of the
volumes, is at first more drawn out than the curve of the
theoretical pressures with which it finally coincides, when
the state of combination of the elements has become the same
as at the ordinary temperature.

12. To sum up, the quantity of heat and consequently the
maximum work which explosive substances can develop while
burning in a constant volume, may be calculated independently
of the phenomena of dissociation, provided the final state of
combination of the elements be exactly defined.

Thus the knowledge of the initial composition, and that of
the products determine the potential energy, whilst pressure and
expansion are subordinates to dissociation.

CHAPTEK III.

HEAT DISENGAGED.

1. THE total quantity of heat disengaged during an explosive
reaction, can be experimentally measured in a calorimeter.
The apparatus employed for this purpose will be described
further on. The quantity of heat is generally positive. There are,
however, certain reactions, such as that of tartaric acid on sodium
bicarbonate, which develop gas and at the same time produce
cold. The explosion of the containing vessel might thus
coincide with the latter phenomena. It would be the same
with the explosion of a vessel containing a compressed gas. But
these are exceptional cases, and outside of the ordinary applica-
tions of explosives.

2. The heat developed can be calculated after deducting the
mechanical effects, when the products of the explosive reaction
are exactly known, and when the heat of formation of the original
substances, as well as of the products, from the elements is also
known. It is only necessary to deduct the former quantity of
heat from the latter to obtain the heat developed daring the
explosion.

3. The calculations are made from the thermo-chemical data
contained in the tables (pp. 125-144). These tables are taken
from the author's " Essai de Mecanique Chimique."

4. The quantity of heat necessary to raise 1 grm. of water
from 0° to 1° is generally called a calorie. This unit is every-
where employed to represent the heat disengaged by the trans-
formation of 1 grm. of matter.

But the magnitude of the quantities of heat disengaged when
chemical reactions are referred to the equivalent weights
(expressed in grms.) has rendered necessary the use of a unit
a thousand times greater ; this is the large Calorie, the quantity
of heat necessary to raise 1 kgm. of water from 0° to 1°.

5. To find, for example, the heat disengaged by the detonation
of nitroglycerin, under constant pressure, in the open air,

2(C3H5N309) = 6C02 + 5H20 liquid + 3N2 + 0.
According to the tables, the heat disengaged by the union of

CALCULATIONS. 15

the elements of nitroglycerin, C3 + H5 + N3 + 09 = C3H5 (N03)3
liquid, amounts to + 98'0 Cal.

On the other hand, the formation of the products —

3(0 + 02) = 3C02 disengages + 94 x 3 = 282

i[5(H2 + 0) = 5H20] „ +34-5x5 = 172-5

Total 454-5

The heat disengaged by the explosion will therefore be
+ 454-5 - 98-0 = + 356-5 Cal.

This is the heat set free by the decomposition of one equivalent
of nitroglycerin under atmospheric pressure about the tempera-
ture of 15°.

6. If the decomposition take place in a closed vessel, under
constant volume, rather more heat will be set free; because
the gases developed by the nitroglycerin in the open air, effect
a certain amount of work in driving back the atmosphere, and
this work consumes a corresponding amount of heat.

The excess of heat resulting from an explosion in a closed
vessel may be calculated by the aid of the following formula :

(1) Qtr = Qtp + (N' - N)0-54 + 0'002£.

Q,tp expresses the heat disengaged at constant pressure, Qtr the
heat disengaged at constant volume, t the surrounding tempera-
ture.

N and N' are defined as follows : — Let I be the number of
litres occupied by the original gas in the closed vessel in which
the explosion has taken place, the gases assumed to be reduced
to 0° and 760 mm., and I' the number of litres occupied by the
gases after explosion, reduced to 0° and 760 mm. Eeplace I by
the expression 22*32N, and I' by 22-32N', in order to compare
the volume of the gases with that occupied by 2 grms. of
hydrogen H2, taken as unity, namely 22*32 litres.

The formula (1) establishes a general relation between the
heat of the reactions taking place at constant pressure and
those in constant volume. The author has demonstrated this in
his " Essai de Mecanique Chimique," torn. i. p. 44.

Let this formula be applied to the decomposition of nitro-
glycerin, this substance being taken at 15°. In this case we
may put N = 0.
Hence

N, = 3X4 + 5X2 + 3X2 + 1 = 29 = ^

4 4

Qfir = typ + 7-25 X 0-54 + 7'25 X 0*002 X 15 = Qtp +
4-13 = 360-6 Cal.

This quantity applies to the weight represented by the formula
C3H5(N03)3, namely 227 grms. One grm. will therefore give
1590 small calories.

16 HEAT DISENGAGED.

This figure is deduced from the heat of formation of water,
carbonic acid and nitroglycerin, the latter being taken from the
three following data : — the heat of combustion of glycerin,
which leads to its heat of formation ; the heat of formation of
nitric acid ; and lastly, the heat disengaged by the action of this
acid on the glycerin.

Sarrau and Vieille have measured directly the heat disengaged
by the explosion of nitroglycerin in a closed vessel, and have
found 1600 cal. for one grm.

The figures 1590 and 1600 have thus been obtained by the
two inverse methods just indicated, and they are as concordant
as can be expected, taking into consideration the small errors
inseparable from all experiments.

7. It should be remarked here, that the quantity of heat dis-
engaged by an explosive is only a fixed quantity, which can be
calculated beforehand, when the material undergoes total com-
bustion, otherwise the heat cannot be calculated for lack of
knowledge of the products of combustion, for these can vary
with the pressure, the manner of ignition, and many other
circumstances (pp. 6, 7).

Further, when working with closed vessels, the oxygen of the
air contained in the space plays a part when the combustion is
incomplete — its effect is greater the smaller the density of charge.
Thus, in calorimetric experiments it is advisable to operate in
an atmosphere of nitrogen when there is not total combustion.
The walls of the vessel, especially when of iron or copper, bear
a part in the chemical reaction which has often been overlooked.
These metals are oxidised at the expense of the air or of the
nitrates, or attacked by the sulphur, etc. Hence there are
subsidiary disengagements of heat which affect the determina-
tions. To avoid these troubles the author conducts all his
determinations in vessels lined with platinum.

8. In what has preceded it has been supposed that the
chemical reaction was not accompanied by any special mechanical
effect. But in general the object of explosions is to do certain
work ; the measure and valuation of this work ought to be made
in each particular case. Hence result most important but
complicated calculations for the theory of firearms, in which the
expansion of the gases plays an important part. Details of
these will be found in the memoirs of Sarrau, De Saint Eobert,
Noble and Abel, Sebert and Hugoniot, and other authorities
who have devoted special attention to ballistics.

9. Without any theory the sum of this work might be arrived
at by an inverse process, namely, by effecting the explosive
reaction in a calorimeter, and measuring the heat disengaged at
the instant in which the work is accomplished. The difference
between the quantity of heat disengaged in a reaction effected
without mechanical effects, and the same reaction with

POTENTIAL ENERGY. 17

mechanical effects, measures the heat consumed by these
mechanical effects. But it is not easy to carry out exact calori-
metric experiments under these conditions.

10. However, the heat disengaged measures the maximum
work which the explosive can accomplish acting under atmo-
spheric pressure. It suffices to multiply this quantity of heat
by 425, the mechanical equivalent of it to express this work in
kilogrammetres. This is the value of its potential energy.

The potential energy of an explosive must not be confounded,
as has sometimes been done, with the heat of combustion of a
substance combustible by air or oxygen ; for example, in com-
paring what has been called the potential of coal with the
potential of .powder. For the energy of the powder is contained
entirely in itself, while the energy of coal in combustion resides
not in the inflammable body alone, but in a system composed
of this body and the air necessary to bum it. Even in the case
of an explosive the total heat disengaged at the ordinary tem-
perature is not in general that which regulates the pressure
developed at the moment of explosion. This latter quantity of
heat corresponds solely to the formation of compounds actually
existing at the temperature and in the conditions of the
explosion ; that is to say, it is subordinate to dissociation. For
example, if at the temperature of explosion the carbonic acid is
dissociated to the extent of one-third into carbonic oxide and
oxygen, it would be necessary to deduct from the heat trans-
formable into work, the heat corresponding to the metamor-
phosis of this third consisting of carbonic oxide.

11. From what has been said it will be seen that it is very
interesting to compare the potential of an explosive with the
work which the gases developed by its explosion could accom-
plish in the ease of an indefinite expansion. This point has
hitherto only been experimentally studied in the case of powder ;
the discussion of the results observed would lead to questions in
mechanics which are foreign to the chief subject of this book.
It will only be added that, according to the most recent experi-
ments— those of Sebert and Hugoniot l — the ratio between the

total and potential work for powder would be ; ', or

305 kgm.

44 per cent. This ratio coincides approximately with the ratio
in weight of gaseous products to the saline products of the
explosion. In practice the limit of work which 1 kgm. of
powder can effect falls to 90,000 kgms., that is to say, below
one-third of its potential energy.

1 " Memorial de 1'Artillerie de Marine," torn. x. p. 184. 1882.

CHAPTER IV.

PRESSURE OF GASES.

§ 1. VOLUME OF GASES.

THE volume of gases formed and their temperature determine
the pressure developed when the explosive substance is decom-
posed in a constant volume. M. Berthelot proceeds to show
how these various data are obtained, and gives the usual
formulae found in text-books.

§ 2. TEMPERATURE.

1. The temperature developed by an explosive substance
can be directly measured, in principle at least. But, as a
matter of fact, this measurement, which is exclusively that
of very high temperatures, presents extreme difficulties, and
there is hardly any known case in which they have been com-
pletely surmounted. All that is known is, that the explosion
of powder develops a temperature higher than that required for
the fusion of platinum, that is to say, than 1775°.

2. The theoretical calculation of the temperature can be per-
formed in the following manner.

The temperature, T, developed in any reaction, such as an
explosion, is calculated by dividing the quantity of heat dis-
engaged, Q, by the mean specific heat of the products, c, esti-
mated between T and the surrounding temperature.

This expression is exact, provided the true specific heat be
introduced into it, as well as the quantity of heat corresponding
to the formation of the products which really exist at the
temperature and under the conditions of the explosion.

3. Theory further shows that the heat disengaged, and, conse-
quently, the temperature produced, are independent of the size

SPECIFIC HEAT. 19

of the receptacle in which the operation has taken place, when-
ever the chemical reaction remains the same.

Hence it is also the same with the ratio between the initial
pressure and the developed pressure at constant volume. This
is, in fact, what follows from Joule's law, provided that such a
law apply to gases so highly compressed as those with which
we are concerned.

4. We will now examine to what degree these various theo-
retical data are really known.

On the one hand, the products which exist at the maximum
temperature, and under the conditions of the explosion, are not
necessarily identical with those which are found after cooling.
At this high temperature, the component elements can be only
partially combined, or transformed into simpler compounds.
Therefore the heat disengaged at the moment of the explosion
will be diminished. On the other hand, the state of combina-
tion is the more advanced, and the dissociation less, as the
pressure developed is more considerable. In general, the
maximum temperature appears to be very much below the
theoretical.

§ 3. SPECIFIC HEAT.

1. The specific heat of gases, which is the base of all these
calculations, requires to be defined. For the sake of greater
simplicity, the specific heat really observable, at the ordinary
temperature, in products obtained after cooling, is adopted, this
specific heat being taken at constant volume, if the reaction
take place in a closed receptacle ; or at constant pressure, if we
operate at atmospheric pressure. The table of these specific
heats will be given on pp. 141-143.

2. However, these suppositions are not accurate. The specific
heat taken at the ordinary temperature, T, does not remain con-
stant at higher temperatures for compound bodies, whatever be
their state ; it is not even so for simple bodies in the liquid or
solid state. In reality the greater number of these specific
heats increase rapidly with the temperature. More especially,
the specific heat of gases compressed to several thousand atmo-
spheres, such as results from the explosion of powder or nitro-
glycerin, is unknown, and it doubtless varies extremely with the
temperature and pressure. Its variations should be similar to
those of liquids, from which the state of gases so compressed is
not very remote. Now, the specific heat of certain liquids, such
as alcohol, can be doubled between limits of temperature so
little separated as 0° and 150°, according to the experiments of
Eegnault,1 and those of Him.2 This is therefore a very uncertain
datum.

1 " Relation des experiences," etc., torn. ii. p. 272. 1862.

2 " Annales de Chimie et de Physique," 4e s<Srie, torn. x. p. 86. 1867.

C 2

20 PRESSURE OF GASES.

3. Attempts have been made to supplement it by an hypo-
thesis more arbitrary, doubtless, but convenient for calculations.
This hypothesis consists in regarding all compound bodies as
possessing, at a high temperature, a constant specific heat, inde-
pendent of temperature and pressure, and equal to that of the
sum of their gaseous elements, of course at constant volume.
This specific heat will be the same, and equal to 4' 8 for every
gaseous element of a weight such that it occupies the molecular
volume taken as unity.

4. The sum of the specific heats can therefore be found by
multiplying the sum of the molecular volumes concerned in the
reaction by 4*8, and dividing by the unit volume.

§ 4. PRESSURE.

The pressure developed at the moment of the explosive
reaction can either be calculated a priori, or directly measured.
This subject will be divided into four sections, viz. : —

Direct measurements (1st section).

Calculations (2nd section).

Density of charge and specific pressure (3rd section).

Lastly, the " characteristic product," a term of comparison
wholly deduced from purely empirical data.

First Section. — Direct Measurements.

1. Direct measurements are made with the aid of various
apparatus, some based on the static, others on the dynamic
method, that is, on the study of the law of the movement
imparted to a heavy body.

2. The earliest and simplest of all the apparatus is that of
Eumford (1792), who experimentally ascertained the weight
capable of keeping in equilibrium the pressure of powder gases.1
The results obtained by this instrument for densities of charge
comprised between 01 and 0*3 do not greatly deviate from the
most recent figures observed by Noble and Abel. Above these
densities Kumibrd's figures are excessive.

3. The Eodman punch (1857) and its modifications, as well
as the Uchatius eprouvette (1869), are based on the size of an
indent made on a copper disc by a steel punch fitted to a piston
acted on by the gases of the explosive substance. In the
apparatus of Meudon, successively improved by Colonels Mont-
luisant and Keffye, the " flowing " of a cylindrical mass of lead,
thrust by the gases into a conical channel of smaller dimensions,
is observed.

The crusher gauge of the English Commission on explosive
substances, deduces the pressure from the crushing of a copper

1 " Trait^ sur la poudre par Uppman et Meyer traduit et augmente par
Desortiaux," p. 562. 1878.

PRESSURE GAUGES.

21

cylinder. All these instruments should be checked by com-
parison, by studying the effects of well-known pressures and
drawing up corresponding tables.

4. The spring dynamometer of Le Boulenge may be mentioned,
and the manometric balances of Marcel Deprez based on
opposite pressures. An account of them will be found in the
" Traite sur la poudre " already quoted (p. 572).

In the same work will also be found a description of the
apparatus founded on the dynamic method, such as the experi-
ments made with the ballistic pendulum by Cavalli (1845-1860),
and by Neumann (1851), the use of Schulze's chronographs

Fig. 1.

(1864), Noble (1872), Noble and Abel (1874), the use of the
ballistic pendulum fitted with a metallic plate to measure
the effect of impact, by Ph. Hess, and the other similar ap-
paratus invented by that learned Austrian officer (1873-1879),
the use of Captain Eicq's recorder (1873), that of the Boulenge
monograph, of the accelerometer and accelerograph of Marcel
Deprez and Sebert (1873-1878), that of the velocimeter due to
Sebert, a learned officer to whom we owe so many ingenious
inventions, etc. However noteworthy these instruments may
be, their description would carry us too far, and it does not
enter into the scope of the present work.

22 PKESSUKE OF GASES.

5. The "crusher" employed in the experiments which the
author made in common with M. Vieille, will only be described.
In this is measured the crushing of a small cylinder of copper
placed between a fixed anvil and the head of a piston with a
base of known section which receives the action of the gases.

The eprouvette is of mild steel, 0*022 metres in internal
diameter, of a thickness equal to the calibre and of a capacity
of 24*3 cms. It is fitted at one end with a plug containing
the crushing apparatus which serves for the measurement of
the pressures, the other end is closed by a plug carrying the
firing arrangement.

In order to avoid all local action at the contact with the
metal, the charge is suspended in the centre of the eprouvette
in the form of a cylindrical cartridge, of a shape similar to the
interior of the chamber. A fine iron wire capable of being
brought to a red heat by electricity traverses the cartridge.

Fig. 1 shows the drawing of this eprouvette, frequently
employed in the investigations of the Commission des matieres
explosives. It consists of a cylindrical tube of mild steel,
A, strengthened externally, according to Schultz's system, by a

Fig. 2.— End view of the eprouvette.

steel wire 0'8 mm. in diameter wound on the tube at a tension
of 35 kilog. The strengthening covering consists of fifteen coils
of wire.

The cylinder is closed at both ends by the steel plugs BB',
the joint between the plugs and the tube being formed by
annular copper gas checks, cc'. The plugs are screwed into
two discs of wrought iron DD', which latter are held together
by six bolts EE' (Fig. 2).

The ignition of the charge is effected by the incandescence of
a metallic wire, stretched between two supports, W ', one of

THE CRUSHER GAUGE. 23

which is fixed on the plug, and the other on a central metallic
rod, n, which traverses the plug, from which it is isolated through
the interposition of a thin coating of shellac. The crusher, as is
well known, was proposed and applied in 1871 in England by
Captain Noble, in his researches on the combustion of powder.
It is fitted to the plug, B', and consists of a piston, a, of tempered
steel, easily fitting in a channel following the axis of the plug,
and of a cylinder of copper, r, 0-008 metres in diameter and
0*013 metres in height, placed between the piston-head and
stopper screwed into the plug.

6. In the method of calibration adopted by the naval artillery,
the cylinders are crushed under weights acting without initial
velocity, and the reduced heights of the crushed cylinders are
measured. From this, by interpolation, is derived a table
establishing empirical relations between these heights h, called
remaining heights, and the corresponding weights, R. Taking IT
to represent the maximum pressure developed in an experiment,
and w the area of the base of the piston, TT is calculated by the
ratio TTW = E. In order to keep the pressure within the limits
of the testing table, it is sufficient to vary the base of the piston.

The results obtained are compared by introducing into the
same chamber increasing weights of the explosive substance.
The ratio of the weight of the explosive to the internal volume
of the eprouvette, is termed the density of charge (see p. 28).

7. The theory of crushing manometers, such as the crusher
above described, has been examined in a most thorough manner
by Sarrau and Vieille.1 They first calibrated the apparatus by
crushing the cylinder progressively and slowly by very small
amounts, until it supported without permanent deformation a
given charge. From this was obtained a ratio between the
final charge, called the force of calibration (force de tarage), 9 and
the diminution in the height of the cylinder, that is, the corre-
sponding crushing E. K0 and K being constants independent of
the explosive, 0 varying from 1000 kgms. to 3500 kgms., we
have

(1) 0 = K0 + K£ : K0 = 541 : K = 535,

the units being the millimetre and kilogramme.

This relation being established, how can the resulting indica-
tion be applied to experiments ? Two limiting cases present
themselves :

(a) The development of the pressure is slow enough, and the
mass of the crushing piston small enough, to permit of the
forces of inertia being neglected ; in this case there is practically
equilibrium between the pressure developed by the explosion
and the resistance of the cylinder. The maximum pressure is

1 u Comptes rendus des stances de 1'Academie des Sciences," pp. 26, 130,
et 180. 1882.

24 PRESSURE OF GASES.

then equal to the force of calibration corresponding to the
crushing observed. It is given by formula (1).

(5) The development of the pressure is so rapid that the
displacement of the piston taking place during the development
of the maximum pressure may be disregarded, the piston
having, besides, a sufficient mass; in this case the movement
of the piston may be regarded as effected under constant
pressure from the start, and throughout the whole of its
duration. The calculation shows that the value of this pressure
is equal to a force of calibration corresponding to half the
crushing.

8. In practice, and for a given explosive, it has to be ascer-
tained whether the instrument works at one or other of these
limits, then to estimate the maximum pressure applicable to the
intermediate cases. Hence, it is necessary to register the
duration of the crushing, as well as the law of the movement of
the piston, and to compare the latter with the results given by
calculation for the movement of the piston crushing the cylinder
under the action of a force which would be a function of the
time. Theory shows that the phenomenon is ruled by the
ratio existing between the effective duration, T, of the crushing
taking place under variable pressure and the duration, r0, of this
crushing caused by a constant force acting on the piston without
initial velocity.

However, it is preferable to substitute for a correction which
is always somewhat doubtful, data obtained under experimental
conditions near one or other limit. We will give some results.

The authors have found for gunpowder, that the crushing

remains the same when — varies from 4'8 to 251, variations

TO

which depend upon the degree of aggregation of the powder
(dust, grain, cake, compressed blocks). We are, therefore, always
in the neighbourhood of the first limit ; that is to say, formula
(1) is applicable in all cases.

The maximum pressure of powder gases at the density of
charge 0'70 has thus been found equal to 3574 kgms. per
sq. cm. Powdered potassium picrate, on the contrary, was
so rapidly decomposed that no appreciable value could be
observed for T (expressed in ten-thousandths of a second).
The maximum pressure was found equal to 1985 kgms. under a
density of charge 0*30. The same salt in compressed blocks
gave a more appreciable duration of combustion ; or 0*0005 sees.
to 0*0006 sees, and a less amount of crushing. It is, therefore,
the other limit which must be applied, and this has been ex-
perimentally verified.

PRESSURE DEVELOPED. 25

With powdered gun-cotton (density of charge 0'20), r is also
inappreciable, and the maximum pressure equal to 1985 kgms.,
the weight of the piston having varied from 727 grms. to
42*7 grms. Dynamite (density of charge 0'30) is decomposed
slower than gun-cotton, but quicker than black powder; the
detonation being of course produced by the aid of fulminate.
It therefore supplies an intermediate case, in which the dis-
cussion of the measurements is more delicate. By employing
pistons of medium weight, and even light pistons, it is very
difficult to attain the lower limit (1), at least with a certainty
comparable to that of the experiments relative to the preceding
substances. On the other hand, towards the opposite limit

(2), the ratio - may be neglected by giving the piston a mass

^0

of 4 kgms. ; the crushing was then nearly double that obtained
with pistons weighing 3 '8 grms. and 6*9 grms. Hence, it can
be seen that the two limiting cases have been realized with
dynamite, as also the intermediate cases, by modifying the mass
of the piston.

The maximum pressure for a piston of 4 kgms. has been
found equal to 2413 kgms. per sq. cm. for the density of
charge 0'30. With a piston of mean weight, that is weighing
only 5 9 '7 grms., the density of charge still being 0'30, dynamite
and picrate give the same crushing; however, the maximum
pressures are very different.

What characterises the experiments made with dynamite is,
that the calculation made for very light pistons from formula (1),
and for very heavy pistons from formula (2), should give, and in
fact does give, the same figure for the value of the pressure
exerted.

It is clear from the above with what precaution the
crushers must be employed to measure the maximum pressures
of explosives. The study of these pressures should be made by
the new method of Sarrau and Vieille.

9. It should here be remarked that the measurements thus
obtained correspond only to a certain mean of pressures, a mean
which is capable of being considerably exceeded at certain
points. In reality, the gases suddenly developed by the
chemical reaction represent real whirlwinds in which there
exist jets of matter under very different states of compression,
and an interior fluctuation. This is shown by the mechanical
effects produced by these gases on solid substances, and
especially on metals, which are hollowed and furrowed in places
as if they had received the impress of an extremely hard solid
body.

The measurement of initial pressures in cannons likewise
manifests local irregularities and differences, sometimes enor-
mous, between the pressures observed at the same instant at

26 PBESSUKE OF GASES.

various points of the chamber where the combustion of the
powder takes place.

The pressure, then, is not uniform, and it may vary in an
almost discontinuous manner as well as the movement at first
communicated to the projectile.

Second Section. — Calculations.

In order to give a better idea of the value of the results
calculated by the ordinary laws of gases, such as there are
known at about the atmospheric pressure and the ordinary
temperature, we will give the experimental measurements made
on certain explosive bodies, which do not give dissociable
products (at least to an appreciable extent), and which give by
their decomposition elementary bodies or gases formed without
condensation : for example, carbonic oxide, of which the specific
heat is comparable to that of the simple gases. Such are
nitrogen sulphide, decomposable into sulphur and nitrogen, and
mercury fulminate, decomposable into mercury and carbonic
oxide. They will supply us with types for calculations of this
kind.

1. Take, for instance, 10 grms. of mercury fulminate de-
tonating in a capacity of 50 cms. (density of charge 0'2). The
heat liberated amounts to 114,500 cal. for the reaction,

C2HgN202 = 2CO + N2 + Hg ;

but the mercury being gaseous, the heat of vaporisation must
be deducted in calculating the pressure, say 15,400 ; which
reduces the heat available for increasing the pressure to
99,100 cal.

Taking for the value of the specific heat at constant volume
of carbonic oxide as well as for that of nitrogen and mercury,
each at its molecular weight, the figure 4'8 (a figure, moreover,
which is in accordance with experiment for the two first bodies),
and neglecting the deviation which exists between this number
and the specific heat of liquid mercury, we find for the tempera-
ture produced —

The volume of the permanent gases (nitrogen and carbonic
oxide) given by the reaction and reduced to 0° and 0'760 metres
will be 22-32 litres X 3.

At a temperature t it becomes

22-32 X 3 X (l + 2^).

To this should be added, starting from 360°, and at the pressure
0760 metres, a volume 22*32 lit. (l + ^73 j of mercurv vapour.

MERCURY FULMINATE AND NITROGEN SULPHIDE. 27

We will thus definitely obtain at the temperature t, supposed
higher than 360°, and at the normal pressure, a volume of gas

equal to 89'28 lit. (l + ^~\

At 5161° this would make 1776 litres from a weight of ful-
minate equal to 284 grms. Now, the capacity in which the
explosion took place being 50 cms. for 10 grms., would be
1*42 litres for 284 grms. Hence the corresponding pressure

would be, by Mariotte's law, — — =s 1251 atm,, or 1293 kgms.

per sq. cm. The experiment made with a crusher gave a
crushing c of 2*4 mms. Mercury fulminate belonging to the
class of explosives for which the duration of the development
of the pressure may be neglected (p. 23) compared with the
duration of working of the crushing apparatus, formula (2)

(p. 24) should be applied, that is, 541 + 535^. This gives 1183

4

kgms. per sq. cm. Water between 1293 and 1183 is hardly
one-twelfth.

Similarly for the density of charge 0*3 theory deduced from
Mariotte's law gives 1939 kgms., and the crusher 1871 kgms.

It should further be noted that the value 1183 leads to
the specific pressure 5915 kgms., while the value 1871 gives
6233 kgms. (pressure of unit weight in unit volume) ; figures
which are sufficiently close to each other to allow of the mean
being taken — 6100 kgms. in round numbers.

2. Take, again, nitrogen sulphide. This body has been ex-
ploded in a closed vessel, and it has been found that

NS ic= N + S, develops + 32,300 cal.

To calculate the pressure at the moment of explosion the heat
absorbed by the vaporization of the sulphur must be deducted.
If this transformation took place towards 448°, it would absorb
about 2600 cal., and there would remain + 29,700 cal. But .this
figure is still too high, the temperature of the sulphur being raised
during the explosion to a point at which this body resumes its theo-
retical gaseous density, instead of a triple density which it has at
448°. This new transformation absorbs a considerable quantity
of heat, which we will estimate provisionally after the analogy
of the polymers at 15,000 or 20,000 cal. for S4, or 8000 to
10,000 cal. for S2. We thus arrive at about 24,000 cal., a
figure which will be employed in default of fuller knowledge.
Let us admit that the sum of the specific heats at constant
volume of nitrogen and sulphur be equal to 4'8 at every
temperature, and neglect the differences between the theoretical
specific heat of sulphur and its real specific heat, in the solid
and liquid states, in order to simplify the calculations.

28 PRESSURE OF GASES.

The temperature of the system developed by the explosion
will then be —

4-8

The volume of the permanent gases considered at a sufficiently
high temperature and at the normal pressure, being here

t\

22-32 lit. (1+-)

at 4375° will be 380 litres, this volume being yielded by 46 grins,
of nitrogen sulphide.

Such a weight exploding in a capacity of 230 cms. would
develop by Mariotte's law a pressure equal to 1652 atm., or
1707 kgms. per sq. cm.

Now, the trial made with a crusher at a density of charge
0*20, and calculated by the old process, gave 1703 kgms. Here
the calculated pressure is practically equal to the pressure given
by the crusher. But according to the new theory it would be
necessary to correct the latter figure, and to take into account
the uncertainty of the estimation of the heat of transformation
of sulphur.

Hence it will be seen that direct experiments are necessary.
However, deviations of quite a different kind might have been
expected.

Third Section. — Density of Charge and Specific Pressure.

1. The relation between the number of grammes expressing
the weight of the explosive substance and the number of cubic
centimetres expressing the capacity in which the explosion
takes place is termed density of charge. Now, if bodies sus-
ceptible of being completely transformed into gas at the tempera-
ture of the explosion be operated upon, Mariotte's law shows that
the pressure developed should be proportional to the density of
charge. The temperature, moreover, would remain the same
in all cases.

2. This relation may be regarded as accurate for very low
densities of charge ; the ordinary laws of gases being applicable
between these limits. But it ceases to be so for medium den-
sities, starting from 01 to 0'2, as might be expected, owing to
the inaccuracy of Mariotte's and Gay-Lussac's laws for the
corresponding pressures.

3. However, strange to say, the relation again tends to exist
for high densities of charge, which are the most interesting for us.
This approximate coincidence results, doubtless, from some com-
pensation between the variation of the pressures, which is more
rapid than Mariotte's law would show, and the variation in the
specific heats, which increase with the temperature and pressure

DENSITY OF CHARGE AND SPECIFIC PRESSURE. 29

(see p. 11), instead of remaining constant, as we have assumed
in our calculations. The dissociation, moreover, must be nil, or
reduced to the minimum for such considerable pressures. This
phenomenon does not enter into the question for nitrogen
sulphide, a compound resolvable into its elements by explosion.

4. However this may be, this relation has been found to be
approached by Sarrau and Vieille in their researches on nitro-
glycerin and gun-cotton, substances furnishing no solid residue,
and such, moreover, that the products of their explosion are sus-
ceptible of dissociation.

5. The experiments the author has made in common with
M. Vieille on nitrogen sulphide, and on mercury fulminate
under the conditions where the explosive substance is entirely
changed into gas, and, what is most essential into non-dis-
sociated gases, confirm it in a most positive manner. For
example, mercury fulminate having been taken with densities
of charge equal to 0'20 and 0*30, the results given by the
crusher, calculated according to the new estimate of the force of
calibration (forces de tarage), show for a density of charge equal
to unity (1 grm. in 1 cm.) 5915 kgms. according to the first ex-
periment ; 6233 according to the second (see pp. 26, 27), figures
which are sufficiently near each other for us to admit the verifi-
cation of the law. Similarly with nitrogen sulphide, for the
density of charge 0'30 we have found, from the indication of the
crusher calculated in the ordinary way, a pressure of 2441
kgms., which gives 8140 for the density 1. A second experi-
ment made with the density 0'2, which reduced to unity gives
8500 kgms., shows scarcely any deviation.

Lastly, for gun-cotton, Sarrau and Vieille have found at
different densities of charge, figures fluctuating near a constant
value of about 10,000 kgms., according to their new theory.

6. All these figures verify the approximate proportion between
the pressure developed and the density of charge. Some of
these have been calculated by means of the indications of the
crushers, according to the old method of estimating the forces of
calibration, and by simply deducing the pressure from the re-
maining height of the crushed cylinder. But it is easy to show
that the same practical verifications may be arrived at, at least
for high pressures, by the new theory of Sarrau and Vieille.
Let us first suppose the pressure equal to the force of calibration.

0 = K0 + KE

in the case of explosive substances of which the action is not
too rapid (p. 23) ; K0 and K being constants independent of
the explosive substance. Hence it results that for high pressure
the pressure tends to become proportional to s. The indications
based on the force of calibration calculated on the old system
retain, therefore, their signification in this case, and it is the

30 PKESSUBE OF GASES.

same with the empirical relations which may be deduced from
these indications.

Now take an explosive substance of which the action is
extremely rapid (p. 24). In this case the pressure is equal to
a force of calibration corresponding to the half of the crushing,

TT

K0 + — c, the constants retaining the same value as above.
2

Here again, for one and the same substance, the high pressures
tend to become proportional to the crushing £ ; but the indica-
tions deduced from the calibration must be reduced by half.

7. Thus the limiting value of the pressure reduced to the
unit of density of charge appears to be a constant; let it be
called/; then

s being the pressure observed for a density of charge, A. This
constant is characteristic for each explosive substance, and may
be called specific pressure. It corresponds to one of the defini-
tions which has been given of the force of explosive substances,
viz. the pressure developed by unit weight of the substance
detonating in unit volume,

8. Maximum Effort. It should be observed, however, that
the specific pressure does not represent the maximum effort
which an explosive substance can develop. In fact, this effort
is that of a substance detonating in a space entirely filled, that
is, in a space equal to its own volume. Now the latter only
corresponds to the specific pressure for a body of which the
absolute density equals unity. It will therefore be less for a
body of which the density is less than unity, as in the case of
gaseous mixtures and explosive gases, as well as of certain
liquids. On the contrary, it will be greater for all solid ex-
plosive substances known up to the present.

It may be calculated, and in fact, from the preceding law it
is easy to estimate, the effort of a substance detonating in a
completely filled space, it being sufficient to multiply the
characteristic number of the pressures by the real density of
the pure substance. For instance, the density of mercury ful-
minate being equal to 4!42, this body would develop a pressure
of about 27,000 kgms. per sq. cm. by exploding in its own
volume : an enormous figure, and higher than that of all known
explosives.

9. Up till now, in the calculations of the specific pressure and
of the maximum effort, we have supposed that the explosive
substance is entirely transformed into gaseous products. But it
may happen that a portion of the substance keeps its solid
state, which is the case, for instance, with dynamite, a mixture
of nitroglycerin and silicious earth. The volume of the latter

NOBLE AND ABEL'S EXPERIMENTS

31

solid matter must then be deducted from that of the capacity in
which the explosion occurs.
We can put, more simply,

$i being the pressure observed (in kgms.), A the density of
charge (ratio between the number of grms. representing the
weight of the substance and the number of cub. cms. repre-
senting the capacity), a the volume expressed in cub. cms.,
of the solid or liquid products resulting from the combustion
of 1 grm. of explosive substance, measured at the temperature
of the explosion. j

Further, putting — = n ; /= s^n - a), n here expressing the

ratio of the capacity, expressed in cub. cms., to the weight L
of the substance expressed in grms. f

10. The relation thus modified has been verified, at least
approximately for dynamite, by Sarrau and Yieille.

It also represents the experiments of Noble and Abel on
the explosion of black powder. In fact, by supposing a = 0'68
and / = 2193 kgms., the numbers found by these authorities
give for pebble and K.L.G. powders,

Pressure per sq. cm.

^

Calculated.

235 kgms.
508
828
1207
1666
2230
2963
3869
5127
6926

At first sight it appears that these latter results tend to
exclude the hypothesis of the total vaporisation of the products
yielded by the explosion of black powder. However, the easy
vaporisation of potassium sulphide at temperatures lower than
1000° tends to encourage the supposition with respect to this
body that it assumes the gaseous form at the temperature of the
explosion of powder, and the experiments l of Bousingault would
also permit of our conceiving the gaseous state of potassium
sulphate and carbonate. This point, therefore, remains reserved.
There is all the more reason for this, as the co-efficient, a, can be

1 " Annales de Chime et de Physique," 4' sdrie, torn. xii. p. 428.

Density of
charge.

0-1

Measura

231 kg

•

i.
ms.

0-2

513

0-3

839

0-4

1220

0-5

1684

0-6 ...

2266

0-7

3006

0-8

... 3912

0-9

5112

1-0

6569

32 PRESSURE OF GASES.

explained equally well by the new laws applicable to the calcu-
lation of pressures in very highly compressed gases.

Instead of the above formula, the following may be em-
ployed : —

4030
Sl "

which gives somewhat higher results, but which would appear
preferable in some respects.1

11. In the calculations in Book III., it has been thought
useful, notwithstanding the previous reservations, to give the
calculation of the theoretical pressure according to the laws of
Mariotte and Gay-Lussac. But care has been taken to define

the result with reference to the density of charge — , instead of

n

merely taking the density 1.

This offers the advantage that the figure thus defined has
a physical signification for low densities of charge. For high
densities its value becomes more and more dubious. However,
it can still be employed in a certain number of comparisons, as
follows from what has preceded.

12. The permanent pressure will also be given, that is, the
pressure exerted by the permanent gases, produced by the ex-
plosion in a completely closed and resisting vessel, and reduced
to 0°. This pressure will always be calculated for a density of

charge _. In fact, it cannot exceed the tension of liquefaction

n
of the gases experimented upon.

Fourth Section. — " Characteristic Product"
1. Another simpler term of comparison deduced solely from
experimental data can be presented in the study of the pressure
developed by explosive substances, viz. the product of the
reduced volume of the gases, V0, by the heat liberated, Q, this
product being divided by the specific heat, c. The latter is
calculated by referring it to the weight of matter capable of
producing this volume and quantity of heat.
By this means is obtained the expression

which may be termed the " characteristic product."

2. It is sufficient to divide it by the actual volume, n (ex-
pressed in cub. cms.), of the capacity in which the unit weight
of the explosive substance has been placed, in order to refer

it to the density of charge - :

n

1 " Memorial de 1'Artillerie de Marine," torn. x. p. 187.

CHAEACTEEISTIC PBODUCT." 33

V0Q

nc

3. In the case where there exist along with the gases fixed
substances, so that the unit weight of the explosive substance
yields a quantity of fixed substance occupying a fraction of a

V 0

cub. cm., a, it will be necessary to replace -2-X- by

nc
V.Q

(n — a)c

4. The expression which has just been defined is almost
exactly proportional to the theoretical pressure for any two
explosive substances capable of being entirely changed into gas
at the temperature of the explosion.

Now, for a given substance, the theoretical pressure is given
by the expression

If the temperatures were reckoned from the absolute zero this
expression would become

V0Q

that is to say, that it would be identical, with the exception of
a multiple, with the characteristic product.

For another substance, enclosed in the same capacity, under
the same density of charge there will be

an expression which would become from the absolute zero

V'oQ'

273^' '

In reality action takes place at an initial temperature higher
than the absolute zero; but it should be noted that if the

quotient ^— represents a number much greater than unity,

the ratio of the theoretical pressures for two given substances,
that is

34: PRESSURE OF GASES.

will be practically the same as the simpler ratio,

Vo Q <f

5. In the case where the specific heats are the same, or very
nearly so, which is that of powders having a nitrate as base, and
a certain number of other explosive substances, this ratio
reduces itself to

6. In other cases, if it be supposed with some mathematicians
that the specific heat of a compound is equal in theory to the
sum of those of its elements, the ratio of the specific heats

c' a'

- might be replaced by the ratio * of the number of the atoms,

c q

that is, of the elementary units of the compound (each of these
units being referred to its atomic weight), or

4

But this formula is very open to dispute, owing to the inac-
curacy of the hypothesis relative to the specific heats (see p. 19).
On this point, we will only state that the specific heat of a
molecule of potassium sulphate would be according to theory l
equal to 24 x 7 = 16*8, while experiments have given, even
near the ordinary temperature, 3 3 '2, that is, the double. It
would be easy to give very numerous examples of the same
kind, derived from the study of solid and liquid compounds.

7. Owing to these disagreements, it is preferable to take for c
and c' their experimental values, and to admit that their ratio
remains nearly constant, notwithstanding the doubts which are
connected with the application of these values to very high
temperatures.

The ratio of the characteristic products

therefore only retains a purely empirical meaning, but it offers
the advantage of being calculable for the unit weight, from
simply experimental data, and without introducing any hypo-
thesis relative to the laws of gases. It furnishes the elements
of a first comparison between explosive substances, in the room
of a more perfect theory.

1 2-4 is the specific heat at constant volume of the simple gases.

CHAPTER V.
DURATION OF EXPLOSIVE REACTIONS.

§ 1. GENERAL IDEAS.

1. THE chemical transformation in a mass which explodes,
arises and is propagated with a certain rapidity, the knowledge
of which is of primary importance for theory, as well as practice.
In fact, the rapidity with which the gases are liberated depends
upon it, and consequently the velocity communicated to pro-
jectiles, as also the effects produced in blasting at the expense of
the rocks which it is desired to break up, or the obstacles to be
removed in military engineering. Now the heat liberated by a
given reaction may be almost entirely employed to heat the
gases and increase their pressure, if the reaction be very rapid ;
while it is dissipated to no purpose by radiation and conduction
if the reaction be slow.

In the former case the effects may be very various.

When an instantaneous decomposition takes place, a given
quantity of explosive substance crushes on the spot the portions
of rock with which it is in contact. Its energy is therefore
consumed in a work almost useless from the industrial point of
view, but which is sometimes desired in military engineering,
with a view to hollowing a primary chamber, destined to contain
a larger charge of explosive. If the development of the gases
be less sudden, though still extremely rapid, the same quantity
of explosive may on the other hand dislocate the rock by
producing extended fissures in it, and by hurling abruptly aside
the nearest portions of rock, which is in general the result aimed
at by miners.

This action is transformed in certain cases into a general
shaking, which causes the ground to tremble and considerably
displaces the centres of gravity of stones and other objects, thus
destroying the stability of masonry and fortified works.

Lastly, the same quantity of explosive sometimes reduces its
effects to elastic displacements, and an undulating movement

D 2

36 DURATION OF EXPLOSIVE REACTIONS.

of the ground, which are propagated to a distance without great
local destruction, the pressures developed having been exerted
sufficiently slowly to give the rock or wall time to displace itself
very slightly " en masse." In this case the explosive substance
will have produced scarcely any useful effect.

This question of the duration of reactions playing an essential
part in all questions relative to explosive substances, has led the
author to bring together here the principal considerations and
experiments to which it has given rise, experiments on which he
has been engaged for many years.

2. The chemical transformation of the explosive substance
has therefore to be defined from the threefold point of view of
its origin, its duration, and its propagation.

§ 2. ORIGIN OF EEACTIONS.

1. A reaction, once started, continues by itself, being propa-
gated either by simple progressive inflammation, or by almost
instantaneous detonation.

2. In all cases relative to the usual explosive substances, to
develop the reaction requires preliminary work,1 a sort of pre-
paration which is represented by the necessity of raising the
substance to a certain initial temperature, such as 315° for black
powder, 190° for mercury fulminate, etc. Indeed, if it were
otherwise, no explosive substance could be prepared beforehand
and stored in a magazine.

But to what point are these notions applicable to the cases in
which the reaction results from a shock, a sudden pressure, or
any mechanical influence ?

3. The author is of opinion that every explosive reaction
should be attributed to a preliminary heating, which is gradually
transmitted, directly or indirectly, raising successively all the
parts of the matter to the temperature of decomposition.

Shock pressure, friction, or mechanical effects are only
efficacious by causing this preliminary heating, and sometimes
propagating it in virtue of the direct or alternative transforma-
tions of the energy into heat, and according to the various
mechanisms, to which reference will be made in § 6.

4. This being granted, let it be noted that the decomposition
of one and the same substance can take place at very unequal
temperatures and velocities; a substance slowly decomposable at
a certain temperature being able to exist at much higher tempe-
ratures, though during a gradually shortening interval.

It is in this way that certain explosive substances are some-
times spontaneously decomposed with great slowness, from the
ordinary temperature, and only produce detonations if the
temperature be raised intentionally or by accident.
1 " Essai de M^canique Chimique," torn. ii. p. 6.

SENSITIVENESS OF EXPLOSIVES. 37

The author has, moreover, developed the whole of this theory
in another place,1 and recalls it in order to thoroughly fix the
ideas.

5. It plays a very important part in the explanation of the
mode of formation of the secondary compounds, produced by the
explosion of powder, several of these compounds being formed
at the very outset, at temperatures which would gradually
destroy them if they lasted long enough. But the suddenness
of the cooling keeps the compounds, such as formic acid,
ammonia, nitric acid, from the destruction which they would
quickly undergo if they were maintained in a constant manner
at the initial temperature of their formation. In fact, this sudden
cooling brings them to the temperature at which they are defi-
nitely stable.

§ 3. SENSITIVENESS OF EXPLOSIVE SUBSTANCES.

1. This sensitiveness depends both on condition of heating,
and on the mode of propagation of the reactions. It varies
according to circumstances. One substance is sensitive to the
slightest rise in temperature, another to a sudden pressure,
another to shock, properly so called, another detonates with the
least friction. Thus for example, silver oxalate detonates at
about 130°, nitrogen sulphide at about 207°, mercury fulminate
at a temperature near this, about 190°, and nevertheless the
fulminate is much more sensitive to shock and friction than
nitrogen sulphide and silver oxalate. There exist therefore
special properties, depending on the individual structure of each
substance, particularly for the solids, which favour decomposi-
tion under given circumstances. But there also exist general
conditions, which it will now be useful to state.

2. The sensitiveness exhibited by the same substance
increases with the initial temperature at which the operation
is performed, that is, a temperature nearer to that at which the
body commences to be spontaneously decomposed, the explana-
tion of this being that the heat liberated by the reaction proper
undergoes less loss by radiation, and that it raises to the desired
degree a greater weight of the non -decomposed substance.

A portion of the sensitiveness will be rendered still greater
if this limit be exceeded, that is, if the conditions prevail under
which a slow decomposition may be transformed by the least
heating into a rapid decomposition.

A substance taken near and especially above this limit may
be said to be in the state of chemical tension.

For example, celluloid, a body which does not detonate
under the hammer at the ordinary temperature, acquires the
property of detonating when heated to about its softening point,
1 " Essai de M^canique Chimique," torn. ii. p. 58 and following.

38 DURATION OF EXPLOSIVE REACTIONS.

viz. towards 160° to 180°, a point which is near the temperature
of the rapid decomposition of the substance.

3. When two different explosive substances are compared,
which are decomposed at the same temperature and with similar
rapidity, their relative sensitiveness to shock and friction, at a lower
temperature, depends on the quantity of matter over which, from
the first instant, the work of the shock or of the friction is dis-
tributed; that is to say, it depends on the cohesion of the
substance, which regulates at the point struck the transforma-
tion of energy into heat, and consequently, the temperature
developed around this point.

4. Cohesion, also, generally intervenes in the case of direct
inflammation ; the same quantity of heat produced by the com-
bustion of the first portions, being able to raise to the degree of
decomposition the temperature of a small quantity of matter, to
which it is exclusively applied, while if it be distributed over a
greater mass the temperature of the latter will not be raised to
the degree requisite for the decomposition to be propagated.

5. The mass heated remaining the same, and the substances
being different, the sensitiveness depends on the initial temperature
at which decomposition commences. This temperature being con-
siderably lower, for example, for potassium chlorate than for
the nitrate, the chlorate powder will be more sensitive in this
respect.

6. The sensitiveness depends furthermore on the quantity of
heat liberated by decomposition ; that is to say, that the sensitive-
ness will be greater, all things else being equal, if the reaction
liberates more heat.

7. The same quantity of heat will produce different effects on the
same weight of matter according to the specific heat of the latter.
For instance, potassium chlorate, the specific heat of which is
0*209, substituted for an equal weight of potassium nitrate, of
which the specific heat is 0*239, in the composition of an
explosive mixture should give, and in fact does give, a more
sensitive powder than the nitrate.

This condition contributes, with the lower temperature of
decomposition and the absence of cohesion, to render the chlorate
powders eminently dangerous.

§ 4. MOLECULAR KAPIDITY OF THE KEACTIONS.

First Section. — General Phenomena.

1. The rapidity of a reaction must be regarded in a different
manner according as it is the question of a homogeneous system,
and especially of a gaseous system, submitted to identical con-
ditions of temperature and pressure in all its parts, or if the
system be submitted at one point to a rise in temperature or

SLOW AND BAPID REACTIONS. 39

to a shock capable of producing an explosion which is then
gradually propagated.

Let us first examine the former case. We shall distinguish
the molecular rapidity of the reactions, which is defined by the
quantity of matter transformed at a fixed temperature and con-
stant pressure, under invariable conditions, and the rapidity of
propagation of reactions.

For greater clearness the phenomena will be treated in a
general manner in the first section, then the molecular rapidity
of reactions in a homogeneous system, submitted to uniform
conditions and contained in an enclosure to which it cannot
yield, and from which it cannot borrow heat, will be specially
studied (second section). Lastly, a system also homogeneous,
but which can lose heat, will be examined (third section).

2. Suppose, first, a certain body, or a certain mixture, capable
of undergoing a chemical transformation. When the whole
mass is placed under the same conditions of temperature,
pressure, or of vibratory movement, etc., it seems as if the
reaction must be instantaneously developed in all the parts at
the same time ; the sudden explosions of nitrogen chloride and
of nitroglycerin would seem favourable to this idea at first
sight. However, a closer observation proves that molecular
reactions generally require a certain time for their accomplish-
ment, even when they liberate heat. Such is, for instance, the
decomposition of formic acid into hydrogen and carbonic acid,
which can be easily followed experimentally owing to the
slowness with which this decomposition is effected. Carried
out in a closed vessel and maintained at the fixed temperature
of 260°, it requires many hours. Nevertheless this reaction
liberates 5800 cal. per equivalent of formic acid, viz. 126 cal.
per gramme.1 *

3. The following are further examples of reactions liberating
a large quantity of heat, without, however, being instantaneous.

Thus, acetylene, changed into benzene towards a dull red heat
by a slow reaction, liberates at the same volume half as much
heat again as a detonating mixture formed of oxygen and
hydrogen in the proportions of water, viz. 85,500 cal. for
33-6 litres of acetylene (reduced to 0° and 0-760), instead of
59,000 cal., produced by the formation of gaseous water by
means of the same volume of electrolytic gas.

It is about four times as much as the heat liberated by a
chlorate powder, weight for weight, viz. 2192 cal. per 1 grm. of
transformed acetylene, instead of 590'6 cal. for 1 grm. of
potassium chlorate powder.

Cyanogen liberates three times as much (1435 cal. per 1 grm.)
as the same weight of chlorate powder, and this number is

1 " Essai de M&anique Chimique," torn. ii. p. 17, and especially p. 58 and
following.

40 DURATION OF EXPLOSIVE REACTIONS.

double that of the heat liberated by an explosive mixture formed
of electrolytic gas at the same volume, viz. 33*6 litres, or
112,000 cal. instead of 59,000, when the cyanogen is decom-
posed into free carbon and nitrogen by the electric spark.
Nevertheless, even though the carbon immediately commences
to be precipitated, the cyanogen does not detonate under the
influence of the spark, nor even of the voltaic arc, which is a
proof of the slowness of the reaction thus caused.

Under other conditions, however, the cyanogen and acetylene
can be decomposed with detonation into their elements, but it is
neither by simple heating nor by the action of the electric arc
or spark (see p. 67).

Instances might be multiplied of these facts l which comprise
the explosive bodies, properly so called, when maintained at a
temperature slightly less than that which determines the ex-
plosion. Silver oxalate, for instance, is slowly decomposed at
100°, while it detonates sharply at a slightly higher temperature.

4. In a word, every molecular reaction effected by simple
heating at a constant temperature in a homogeneous body and
submitted to conditions which appear identical for all its parts,
has a characteristic coefficient relative to the duration. This
coefficient depends on the temperature, the pressure, and the
relative proportions; it plays a very important part in the
study of the unequally destructive properties of explosive
compounds.

This will be exemplified by a few applications.

5. The longer or shorter duration of a reaction does not change
the quantity of heat liberated by the total transformation of a
given weight of explosive matter. But if the gases formed
gradually expand, for instance, as in a cannon, owing to the
change of capacity, increased by the flight of the projectile, or
owing to cooling due to contact with the walls of the vessel ;
under these circumstances the initial pressures will be less the
longer the transformation of a given weight of matter lasts.

On the contrary, when a very rapid transformation of the
whole mass in a closed vessel, allows the initial pressures to
attain the immensity of their theoretical limits, or to approach
it, it is difficult to construct vessels strong enough to contain
the gases of explosion.

6. This explains the influence of resisting envelopes and of
tamping, an influence which is especially apparent with slow
powders, but which is also observed with rapid powders, par-
ticularly in detonators.

At the moment of the explosion the pressure at first developed
around the ignited point tends to diminish, owing to the expan-
sion of the gases, and in proportion as the products are dis-
tributed throughout a more considerable space. If the gases
1 " Annales de Chimie et de Physique," 4e se'rie, torn, xviii. p. 142.

ACTION OF GASES AND WATER AS TAMPING. 41

retained the whole of their heat the pressure after some instants
would depend solely on the extent of the space. The pressure
would be greater the more limited the space; the maximum
pressure corresponding to the explosion of the substance in its
own volume, and the molecular rapidity of the reaction exerting
no influence. But this is an extreme limit, owing to the losses
of heat which the products of the explosion continually undergo
by contact, conduction, and radiation ; hence results a cooling,
which lowers the temperature, and therefore the pressure, as
well as the rapidity of the chemical reaction. The initial
pressure tends to approach this limit, the more rapid the powder,
the smaller the capacity enclosing the powder, and the more
resisting the walls of this capacity, which enables them to hold
the compressed gases during a longer period.

7. The same will happen, not only when an explosive body is
placed in a fixed and resisting capacity, but also if it be placed
in a thin envelope, or under a stratum of water, or even in the
open air. In fact, when the duration of reaction decreases
beyond measure, the gases liberated develop pressures which
increase with an extreme rapidity, so rapidly, indeed, that the
surrounding bodies, solid, liquid, or even gaseous, have not time
to put themselves in motion in order to yield gradually to these
pressures ; these bodies, therefore, offer to the expansion of the
gases resistances comparable to those of a fixed wall.

It is known that a film of water on the surface of nitrogen
chloride is sufficient to give rise to such effects. A drop of this
substance placed in a watch-glass may detonate without breaking
it, while if it be covered with a little water the glass is broken.
By operating on a slightly larger mass, even the plank upon
which the vessel is placed may be pierced under these conditions.

The same result may sometimes be attained by increasing the
mass of the explosive substance; the gases first ignited not
having time to dissipate, are then acting as tamping. This effect
becomes greater and greater, as the temperature of the substance,
and therefore the rapidity of the reaction, increases.

It is in this way that compressed dynamite and gun-cotton,
substances capable of being inflamed without danger by the aid
of an ignited body when operated upon in small quantities, have
sometimes caused terrible explosions, owing to the general
inflammation of a considerable mass.

To sum up, the nearer the duration of the reaction approaches
being instantaneous, the nearer the initial pressure, even in an
open vessel, approaches the theoretical pressure, the latter being
calculated for the case of a decomposition effected in a constant
capacity entirely filled by the explosive substance. It is thus
that the extraordinary destructive effects produced by mercury
fulminate, nitroglycerin, or compressed gun-cotton can be
appreciated.

42 DURATION OF EXPLOSIVE REACTIONS.

We shall now analyse the phenomena in a more precise
manner.

Second Section. — A homogeneous system submitted to uniform
conditions and contained in an enclosure to which it cannot yield,
and from which it cannot take any heat.

Let it be a homogeneous system capable of disengaging heat
by chemical transformation.

The case will first be examined where this system is sub-
mitted to uniform conditions in all its parts, and contained in an
enclosure to which it cannot yield, and from which it cannot take
any heat. Under these theoretical conditions, the mass of the
matter is of no importance.

1. The molecular rapidity of reactions in a homogeneous system,
everything else being equal, increases with the temperature.1
Indeed, it increases according to a very rapid law, as is proved
by the author's experiments on ethers,2 the rapidity being then
represented by an exponential function of the temperature, a
function of which the numerical value, in the formation of
acetic ether, is 22,000 times greater near 200° than near 7°.

2. The temperature of the system increases, at least up to a
certain limit, by the very effect of the reaction.

The temperature of the system increases, at first incessantly,
and does so up to a limit defined by the figure obtained by
dividing the heat liberated by unit weight by the specific heat
of the system.

Further, the rapidity with which the system tends towards
this limit goes on increasing according as the rise in tempera-
ture produced by the reaction itself is more considerable.

In a gaseous system contained in a fixed enclosure the
acceleration will become even greater, at least at the commence-
ment, and this owing to. the influence of pressure, which
necessarily increases by the fact of the rise in temperature.
In short, the author has established that, everything else being
equal, and while operating at a fixed temperature, the reactions
are affected more rapidly in liquid than in gaseous media.
In gaseous media, in particular, he has recognised that the
reactions are the more rapid the greater the pressure.3

3. The molecular rapidity of the reactions in a homogeneous
system, increases with the condensation of the substances, or more
simply, the rapidity of the reactions increases with the pressure in
gaseous systems*

Thus, in an enclosure supposed impermeable to heat the

1 " Essai de Me"canique Chimique," torn. ii. p. 64.

2 Ibid., p. 93. 3 Ibid., p. 94.

4 In solid or liquid systems, on the contrary, the pressure exerts little
influence according to the author's experiments, which is conceivable because
it hardly modifies the state of condensation of the substances.

EFFECTS OF INERT BODIES. 43

elementary rapidity of the reactions will go on incessantly
increasing, for the twofold reason that the temperature is con-
tinually rising and that the pressure of the gases is continually
increasing.

However, the influence of pressure will be more sensible at
the beginning than at the end of the experiment, seeing that the
non-combined part diminishes more and more, and that there
arrives a moment when the tension proper of this part, con-
sidered independently of the rest, ceases to go on increasing in
consequence of the heating.

4. The molecular rapidity of reactions in a homogeneous system
depends on the relative proportions of the components.

When operating at constant temperature, the combination is
greatly accelerated by the presence of an excess of either of the
components.

At constant temperature, the action is, on the contrary, checked
by the presence of an inert substance which acts by diminish-
ing the state of condensation of the substance.

At variable temperature the reactions are a fortiori retarded
by the presence of an inert body, such as the nitrogen of the
air, or the silica of ordinary dynamite, this inert body absorbing
the heat and lowering the temperature of the system without
exerting any influence tending to accelerate it by its presence.

At variable temperature the reaction is generally slower in
presence of an excess of one of the components than if equal-
equivalents be used, the necessity of heating this excess more
than compensating its accelerating influence.

It is clear that if the proportion of the inert matter be such
that the temperature of the system cannot be raised to the
degree necessary for the combination to be continued of itself
the reaction will cease to be explosive, and even to propagate
itself.

It is in this way that the character of explosive substances
may be changed by mixture with an inert body. The follow-
ing are some characteristic facts.

Dynamite of 75 per cent, is less shattering than pure nitrp-
glycerin. However, such dynamite cannot be employed in
charging shells, as the latter would explode in the bore of the
cannon, under the influence of the initial shock of the powder.
Dynamite of 50 or 60 per cent, can, on the contrary, be employed
in hollow projectiles, which may be fired from cannons without
danger.

This is not all. With dynamite of 60 per cent, the projectile
can explode at the point of arrival, even without a special fuse,
if it be arrested by a very resisting body, such as an armour
plate, the rise of temperature which results from the trans-
formation of the energy into heat, produced by the sudden
stoppage, being sufficient to cause the explosion. But if the

44 DURATION OF EXPLOSIVE REACTIONS.

charge of nitroglycerin be lowered to 30 or 40 per cent, the
shell charged with such dynamite will require the employment
of a percussion fuse in order to explode, as in the case of black
powder. It is true that such dynamite scarcely offers any
advantage over ordinary powder.

Another essential point is that not only the molecular
rapidity is diminished under these conditions, but also the
rapidity of inflammation and the rapidity of combustion of an
explosive substance are also extremely retarded when it is
mixed with an inert body in proportions approaching those
which correspond to the limits of inflammability. Consequently,
towards these limits inflammation becomes uncertain, com-
bustion is badly propagated, and the explosive character of the
phenomenon ceases to be manifest.

Third Section. — A homogeneous system submitted to uniform
conditions but capable of losing heat.

1. These general relations being established for a system
where all the heat which it liberates is employed to raise the
temperature, we come to the real case, that in which the system
still supposed homogeneous and submitted to uniform conditions at
the outset yields a portion of its heat to the surrounding bodies by
radiation or conduction. The mass of the substances employed,
which does not come into question in principle in the first case,
here plays an essential part.

In short, whenever the rapidity of the reactions is not great,
a part of the heat produced will be gradually dissipated, and the
elevation of temperature will soon attain a certain limit. This
limit will be that at which the loss of heat produced by the
external actions is equal to the gain due to the internal reactions
of the system, the reaction will then take place with a certain
rapidity constant or nearly so, without however becoming
explosive.

This is the case of a substance fusing under ordinary con-
ditions, and it is also the case, generally in a marked degree of
slowness, of a small quantity of an explosive substance which
is spontaneously decomposed.

But if the mass operated upon be increased, supposing it
contained in a fixed capacity, the quantity of heat lost by
radiation or conduction at a given temperature of the system
will be less ; the total quantity of heat retained in the interior
at the end of a given time will therefore be increased. Thus
the temperature of such a system must be higher whether it
tend towards a new limit superior to the preceding or whether
its increase becomes more and more rapid and finally explosive,
owing to the correlative increase in the pressures.

This same correlative accelerating of the pressures and of the
rapidity of the reactions plays an important part in the inter-

SPONTANEOUS DECOMPOSITION. 45

pretation of the effects of tamping, as has been said before
(pp. 40, 41).

It is, further, in this way that every fusing substance may be
transformed into a detonating substance, when the mass of it
contained in a given capacity is increased, of course without
any change being made either in the orifices or the form of the
capacity.

The difference between the various modes of decomposition
of an explosive substance, according as its mass is more or less
considerable, deserve particular attention, for it is continually
occurring in practice.

2. This is noticeable even when there is an escape for the
gases of the explosion, provided the explosive mass be large
enough. It is thus that the decomposition of a fusing sub-
stance, taken at constant weight, and contained within a given
capacity, may change into explosion, when the orifice of this
capacity is contracted in such a manner that the inward pressure
and temperature may increase beyond a certain limit.

3. The same remark applies to spontaneous decomposition of
great masses of matter. Slow at first at the ordinary tempera-
ture, they quicken under the influence of that very increase in
temperature to which they themselves give rise. It may also
happen that this rise in temperature changes the character of
the decomposition by causing a fresh reaction after the initial
one, throwing off more heat. The rise in the temperature of
the mass hence still further increases, even to the extent of
producing a tumultuous reaction, and a general explosion.

4. These facts, often observed in laboratories, have been
quoted to account for the spontaneous explosions of gun-cotton
and nitroglycerin, and their tendency is to cause us to regard
as especially dangerous any explosive substance in which the
process of decomposition has commenced.

5. These considerations demonstrate the cause of general
explosions, not only of explosive substances contained in very
solid vessels, but even in vessels whose resistance is very slight,
such as wooden cases or thin metallic envelopes, and again of
explosions of substances piled up in the open air when the
accumulation of these substances permit of a rise in the
temperature and of a gradual acceleration in the reaction
(see p. 41).

6. General explosions may also occur with substances divided
into very small quantities, if these small quantities are
sufficiently close to one another to constitute a large mass in
the aggregate, and if the mechanical effects admit of accumula-
tion and to produce a common result.

The precautions, therefore, both for storage and use, should be
taken just as though all the individual portions of the explosive
substance were collected into one single mass. Herein lie the

46 DURATION OF EXPLOSIVE EEACTIONS.

consequences which theory has pointed to as possible, and the
accidental realisation of which has been proved by terrible
catastrophes.

Thus, for instance, the experiments made by the Birmingham
Chamber of Commerce on the transport and storage of amorces
have shown that capsules each containing 0'015 grms. of
fulminate, do not explode en masse either under the influence
of a shock, or when crushed by the wheels of a locomotive, or
when placed in an incandescent muffle, or in the centre of a
burning furnace.

Yet this is not so if the charge of fulminate contained in the
capsules be considerably increased. The feeling of security
created by these first experiments no longer exists even in
England, in consequence of the explosion of a vessel loaded
with amorces in the Thames.

Experience has in fact shown that the explosion of one single
strong fulminating capsule is sufficient to cause the explosion
of all the capsules placed in the same case. If the case itself
explodes, all the adjoining cases will also explode.

It is on account of analogous phenomena that the small
fulminating caps (amorces), used as children's playthings, have
only too often given rise to serious accidents.

At Yanves, near Paris, a child was amusing itself by letting
off one of these amorces between the blades of a pair of scissors,
when two packets of six hundred amorces that were lying on
the table exploded at the same moment. The child was killed,
the chair shattered, and the flooring staved in.

The explosion which occurred in Paris, in the Rue Beranger,
on May 14, 1878, may also be mentioned, in a store con-
taining amorces intended for children's toys. These amorces
were composed as follows : — One kind, called single, of a
mixture of potassium chlorate (12 parts), amorphous phosphorus
(6 parts), lead oxide (12 parts), and resin (I part) ; the others,
called double, consisted of a mixture of potassium chlorate
(9 parts), amorphous phosphorus (1 part), antimony sulphide
(1 part), flowers of sulphur (0'25 part), and nitre (0*25 part).
The latter, more sensitive to friction, averaged O'Ol grm. in
weight. From six to eight millions of these amorces pasted
on paper slips, in lots of five each, were piled up in the ware-
house in boxes. A few of these having become ignited by an
accident, - the origin of which was never clearly ascertained,
caused the whole to explode. One building suddenly gave
way, the fapade being blown out, and the stonework hurled
some distance. One stone, measuring a cubic metre, was thrown
to a distance of fifty-two metres. A great part of the adjoining
building was also destroyed, fourteen persons were killed on the
spot, and sixteen received injuries.

These terrible effects are explained when we consider that

RAPIDITY OF PROPAGATION OF REACTIONS. 47

the weight of the entire explosive matter contained in the
amorces amounted to about 64 kgms., and that its force, owing
to the composition of this matter, was equal to a force of
226 kgms. of black powder.1

It is essential, that persons having explosive substances
under their charge should never lose sight of the conviction
that, from the facts and general truths which have just been
stated, preventive measures should always be prescribed on the
hypothesis of a total explosion.

§ 5. KAPIDITY WITH WHICH KEACTIONS ARE PROPAGATED.

1. Let us now examine the case of a homogeneous system, but
subject to different conditions in its various parts, such as those
resulting from ignition at one point, or from a local shock, con-
ditions to which some of the facts quoted in the preceding
paragraph may be ascribed.

For propagating the transformation in a mass which explodes,
and is not subject to the same conditions in all its parts, the
physical conditions of temperature, pressure, etc., which have
incited the phenomenon at one point must reproduce them-
selves successively from layer to layer throughout every part of
the whole mass.

On this head attention may be called to the numerous
experiments made by artillerymen, as to the rapidity of the
combustion of ordinary powder, and as to that of gun-cotton ;
the rapidity varying according to the physical construction and
chemical composition of the powders. We will firstly sum
up these results, as well as those observed with mixtures of
gaseous explosives, that is to say, the observations relating to
the rapidity of combustion of mixtures of oxygen and hydrogen,
or of carbonic oxide, or of hydrocarbon gases.

We shall then speak of the new and unexpected results
furnished by the study of gun-cotton and nitroglycerin ; describe
the new conception of the employment of caps, and the hitherto
unknown distinction between the simple ignition and the
genuine explosion of explosive matters will be discussed ; a
distinction which the author's recent experiments led him to
extend to gaseous mixtures themselves. Then it will be
attempted to apply these differences to theoretical conceptions.
Thus we shall be led on to the notion of the explosive wave,
which will form the subject of a special chapter.

2. According to Piobert,2 the rapidity of the combustion of
powder in the open air, observed on vertically placed prisms of
known length, the lateral surfaces of which were greased for the

1 These facts have been taken from the report presented by the Committee
of Inquiry.

2 Piobert, " Traite d'Artillerie," partie thSorique.

48 DURATION OF EXPLOSIVE REACTIONS.

purpose of ensuring the regularity of the phenomenon, has been
found to be between 10 and 13 mms. per second for service powder.
It varies, moreover, in an inverse ratio to the apparent density
of the powder, in which the fire propagates itself in successive
layers.

3. The velocity of combustion in this sense, that is to say, the
velocity with which the reaction is transmitted into the interior
of a single explosive mass, should not be confounded with the
velocity of inflammation, that is, with the time necessary for the
propagation of the same reaction in a whole formed out of a
series of small masses or grains placed side by side. In order
to characterise the difference between these two velocities, we
shall give the experiments made by Piobert with hollow iron
semi-cylinders of various lengths. The velocity of inflammation,
measured in the open air, varies in the old service powders
from 1*5 metre to 3'4 metres per second ; in tubes closed at one
end, from 0*30 metre to 1*5 metre, according as the grain is
0*2 mm. or 2*5 mms. in diameter. The increase is in propor-
tion to the size of the grain. These figures are very different
from those given above for the velocity of combustion in a
single mass.

4. The combustion velocity of powder depends on the pressure
of air or of the surrounding gases.

At the end of the seventeenth century, Boyle made some
experiments on the combustion of powder in a vacuum, and
observed that, under these circumstances, the grains of powder
projected on a red-hot iron fused with detonating. If we try
a certain number of grains at the same time, explosion results ;
doubtless because the conditions of local pressure are changed
round each grain which deflagrates.

Huyghens repeated the same experiment by igniting powder
by means of a lens which concentrated the solar rays.

If the heating is progressive, as would be obtained by the
radiation from a lighted coal, the sulphur may either be sub-
limed, thus destroying the homogeneity of the compound, or the
powder can be fused according to Hawksbee (1702).

These experiments have been often tried with various modi-
fications, such as the employment of a platinum wire brought to
red heat by electricity, for the purpose of igniting the powder
in a vacuum (Abel). Bianchi has thus observed that gun-
cotton slowly decomposes in a vacuum before exploding, and
this result has since been extended to nitroglycerin by Heeren
and AbeL

Mercury fulminate on the other hand explodes in a vacuum
when placed in contact with a brass wire heated to redness ;
but the explosion does not propagate itself to grains which are
not contiguous, as it would do under atmospheric pressure.

5. Not only does a vacuum prevent the explosion of the powder

VELOCITY OF COMBUSTION. 49

but every diminution in pressure retards it. In 1855, Mitchell
observed that fuses burn slower on high mountains ; some very
exact experiments were made in this respect by Frankland in
1861 in his own laboratory, and afterwards by De Saint-Robert
in the Alps. Under pressures varying between 0'722 metre
and O405 metre, that is to say, below atmospheric pressure, the
velocity of the combustion of powder would be practically
represented, according to De Saint-Robert, by such a formula as
the following : „

V = A/,

A being a constant, and p expressing the pressure.

These effects are to be attributed to the greater or less velocity
with which the heated gases escape before having had time to
heat the adjacent portions of the solid matter. This is equiva-
lent to saying that pressure lessens the number of gaseous
particles brought up to a high temperature which at every
moment come in contact with the solid particles not yet ignited
and share with them their energy, so as to place themselves in
equilibrium of temperature.

Whatever the pressure, if it operates under a constant
volume, the initial temperature of these particles remains prac-
tically the same ; at least, so long as the chemical reaction is
not modified. But if operating under a constant pressure, the
case will be otherwise, the temperature being lowered lay the
expansion of the gases.

6. On the other hand, the combustion velocity of powder
increases with great rapidity, once we obtain the considerable
pressures which are produced in cannons and guns. Thus, for
instance, Captain Castan estimated the combustion velocity of
powder, in the interior of large bore cannons, at 0*32 metre
per second, instead of about O'l metre in the open air.

7. The combustion velocity of other explosive substances has
not formed the object of such numerous experiments as that of
black powder ; moreover, it gives rise to fresh observations and
to a theory of an entirely different order, as will presently be
stated.

Piobert estimated the combustion velocity of non-compressed
gun-cotton at eight times more than that of service powder; which
estimate was applied to a progressive combustion effected
without detonation.

8. These researches were extended to explosive gaseous
mixtures. In 1867, Bunsen 1 estimated the velocity of combus-
tion for electrolytic gas (hydrogen and oxygen) at 34 metres per
second, and at only one metre per second for the mixture of
carbonic oxide and oxygen in equivalent proportions, these
mixtures being taken under atmospheric pressure. He deter-

1 " Annalee de Chimie et de Physique," 4e se*rie, torn. xiv. p. 449.

E

50 DURATION OF EXPLOSIVE REACTIONS.

mined its rate of flow through a narrow orifice, ignited the jet
and endeavoured to discover what was the minimum rate of
liow at which the flame would remain stationary at the orifice
without going back into the interior. Mallard1 made several
experiments on various mixtures of air and of marsh gas, or of
coal gas; he found that the velocity of combustion defined
as above rapidly diminishes in proportion as the ratio of the
gases which do not contribute to the combustion is increased, the
maximum speed being 0'560 metre per second in the case of
a mixture of eight parts by volume of air and one part of
marsh gas. If a mixture contain twelve parts of air and one
part of marsh gas it will descend to O04 metre. With coal
gas and air the maximum velocity obtained has been almost
double. Mallard and Le Chatelier have recently returned to
this question by other processes which have given them very
varied results according to the mode of combustion. This will
be referred to later on, and the existence of detonating velocities
for the same gaseous mixtures reaching almost up to 3000
metres per second, and the causes of these differences will be
shown.

9. The study of new explosive substances has in fact led to
a fuller knowledge of the mode of propagation of chemical
reaction in the interior of a mass in combustion ; and it has
radically modified the ideas which prevailed on this question.
At one time, when black powder was the only known explosive,
the only point studied was how to ignite it, the effects of con-
secutive explosion not appearing to depend on the process of
ignition. But nitroglycerin and gun-cotton have evinced a
singular diversity in this respect.

10. To form a correct conception of them, mention must first
be made of the phenomena of shock and of other analogous
causes capable of producing deflagration.

Shock could hardly of itself affect the decomposition of a
substance which absorbs heat, unless recourse be had to colossal
masses animated by enormous energy and concentrating all their
action on a very small quantity of matter, which is very difficult
to effect. For instance, the energy of a weight of 1630 kgms.
falling from a height of 1 metre, would be necessary in order to
decompose 1 grm. of water, assuming that by any artifice the
totality of this energy could be transmitted to a gramme of
water.

On the other hand, if the decomposition of the substance
disengages heat, we can conceive that a limited energy would
suffice to provoke it, provided it were applied in its entirety to
a very small quantity of matter whose temperature it raised to
the desired degree for determining reaction.

Thus a few heavy strokes of a hammer on potassium chlorate
1 " Annales de Mines," torn. viii. 3e livraison. 1871.

EFFECTS OF SHOCK. 51

wrapped in a sheet of platinum and placed on an anvil, will be
sufficient to give rise to the formation of very sensible traces of
potassium chloride, whereas under similar conditions potassium
sulphate gives no indication whatever of decomposition. But
the decomposition of potassium sulphate into potassium sulphide
and oxygen absorbs heat, whereas the decomposition of potas-
sium chlorate into potassium chloride disengages heat (11,000
cal. for KClOa).

11. This condition, however, is not sufficient for shock to
produce a detonation. It is further necessary that the energy
developed by the decomposition of the first portions should be
able to communicate itself to the neighbouring parts, so as to
determine step by step the decomposition of the whole mass.
The most favourable condition is evidently that one in which
the particles of the explosive substance are in movement and
animated by an energy of such a nature that their sudden
stoppage would transform this force into heat in the interior of
the substance itself. The substance is thus heated in a uniform
and sudden manner ; if the proper temperature be attained, the
explosion occurs immediately. Such conditions may be realized
on the sudden stoppage of a bomb-shell charged with dynamite
which meets with a resisting surface (see p. 43). In an opposite
sense it may be noted that the shock of a hammer which is
hardly sufficient to produce on some isolated points the desired
conditions with pure potassium chlorate is, on the contrary,
very efficacious with nitroglycerin. Even the fall of a weight
of 4*7 kgms. from a height of 0*25 metre on to a drop of nitro-
glycerin covering a surface of 2 sq. cms. is sufficient to cause
the explosion of this substance.1 On the other hand, nitro-
glycerin mixed with a silicious earth constitutes dynamite, a
substance which is very slightly susceptible to shock, because
the porous and cellular structure of the silica militates against
the immediate and local communication of energy to a very
small portion of nitroglycerin when regarded apart from
the rest.

Further, the explosion of black powder causes nitroglycerin
to explode, whereas it does not produce the explosion of
dynamite, at least in the open air and with weak charges.

But this inertness disappears under the influence of certain
particularly violent shocks, such as that of mercury fulminate.
Again, the explosion of nitroglycerin is very different, according
as it is pure or mixed with another body or is operated on by a
simple shock, by contact with a body in weak ignition, or in
active ignition, or, again, by the aid of an ordinary fuse; or,
finally, by the contact of a strong priming of mercury fulminate.

1 Ch. Girard, Millot et Vogt (" Comptes rendus des stances de 1'Acade'mie
des Sciences," torn. Ixxi. p. 691).

E2

52 DURATION OF EXPLOSIVE REACTIONS.

§ 6. MULTIPLICITY OF THE MODES OF COMBUSTION.

1. According to the method employed for ignition, dynamite
may be quietly decomposed without any flame ; or it may burn
briskly ; or, again, give rise to an explosion properly so called,
either moderate, or capable of dislocating rocks, or even of
locally crushing them and of producing the most violent
effects.

2. The substances which determine these latter effects have
received more especially the name of detonators. Noble was
the first to recognise the character of these when experimenting
on nitroglycerin in 1864, and from thence he deduced the con-
venient method of exploding this substance effectually by
means of a priming of mercury fulminate.

Gun-cotton does not present less variety. Abel has published,
with reference to this matter, since the year 1868, some very
curious experiments, which similarly tend to establish a great
diversity between the conditions of deflagration in this substance
according to the mode of detonating it.1 Eoux and Sarrau
have generalised these phenomena by distinguishing what they
have termed the explosions of the first and of the second order,
a real distinction, yet one which appears insufficient, by reason
of its too absolute character.

3. However strange this diversity may appear at first sight,
the author holds nevertheless that thermo-dynamic theories are
capable of accounting for it by a suitable analysis of the
phenomena of shock.

In fact, the variety of explosive phenomena depends on the
speed with which the reaction is propagated, and on the more
or less intense pressures which result therefrom.

Take the simplest case, that of an explosion determined by
the fall of a weight from a certain height. At first one would
be inclined to attribute the effects observed to the heat dis-
engaged by the compression due to the shock of the suddenly
arrested weight. But calculation shows that the stoppage
of a weight of several kgms. falling from a height of
0*25 metre or 0*50 metre cannot raise the temperature of the
explosive mass more than a fraction of a degree, if the resultant
heat be uniformly distributed throughout the entire mass.
Such mass could therefore never reach to a high temperature,
such as 190° and 200°, for instance, in the case of nitroglycerin,
a temperature to which it would seem necessary to raise the
entire mass suddenly in order to produce explosion.

It is by a different mechanism that the energy of the weight
transformed into heat becomes the origin of the effects observed.

1 "Comptes rendus des stances de I'Acad&nie des Sciences," torn. Ixix.
pp. 105 a 121. 1869.

53

It is sufficient to admit that pressures which result from the
shock exercised on the surface of nitroglycerin are too sudden
to distribute themselves uniformly throughout the whole mass,
and that consequently the transformation of the energy into
heat takes place more especially in the first layers reached by
the shock. If this be sufficiently violent, these layers may thus
be suddenly raised up to about 200°, and they will immediately
decompose, producing a great quantity of gases. The produc-
tion of gas is, in its turn, so sudden that the striking body has
not time to displace itself, and the sudden expansion of the
gases of the explosion produces a first shock, doubtless, more
violent than the first on the layers situated below. The energy
of this new shock changes itself into heat in the layers which it
first reaches. It determines their explosion, and this alternation
between a shock developing an energy which becomes changed
into heat, and a production of heat which raises the temperature
of the heated layers up to the degree of a new explosion capable
of reproducing a shock, transmits the reaction from layer to
layer throughout the entire mass. The propagation of the
deflagration thus takes place by virtue of phenomena comparable .
to those which give rise to a sound wave ; that is to say, by
producing a true explosive wave which travels with a rapidity
incomparably greater than that of a simple inflammation,
induced by the contact of a body in ignition and effected in
conditions under which the gases freely expand as they are
produced. We shall define this explosive wave and study its
characteristics in Chapter VII.

4. The reaction induced by a first shock in a given explosive
substance propagates itself with a rapidity which depends on
the intensity of the first shock, seeing that the energy of the
latter transformed into heat determines the intensity of the first
explosion, and consequently that of the entire series of consecu-
tive effects. Now the intensity of the first shock may vary
considerably according to the mode in which it is produced.
The effect of the blow of a hammer may vary in its dura-
tion, for instance, from yJ0 to -nyj-<ro °f a second, according to
whether the blow be given with a hammer with a flexible
handle, or with a block of steel, as shown by Marcel Deprez's
experiments.

It follows, then, that the explosion of a solid or liquid mass
may develop itself according to an infinite number of different
laws, each of which is determined c&teris paribus by the original
impulse. The more violent the initial shock, the more sudden
will be the decomposition induced, and the more considerable
will be the pressures exercised during the entire course of this
decomposition. One single explosive substance may therefore
give rise to the most diverse effects according to the process of
ignition.

54 DURATION OF EXPLOSIVE REACTIONS.

5. Effects vary similarly according to whether the substance
is pure or associated with foreign substances, and again accord-
ing to the structure of the latter. This is shown by dynamite,
a mixture of nitroglycerin with silica ; which has lost a great
part of its sensitiveness to ordinary shock, yet it remains explo-
sive under the shock of a ball, and particularly under the shock
of mercury fulminate.

6. Gun-gotton impregnated with water or paraffin becomes
equally non-sensitive to a shock. In order to explode it a small
supplementary cartridge of dry gun-cotton powder primed with
fulminate must be employed.

If a small quantity of camphor be incorporated with nitro-
cellulose its susceptibility to explosion by shock is almost
completely annihilated, at least at the ordinary temperature ; to
such an extent that this association constitutes a substance now
employed for various purposes under the name of celluloid.

7. Blasting gelatin, which is made of nitroglycerin and nitro-
cotton, sometimes with the addition of camphor, constitutes an
elastic mass very slightly sensitive to shock, and which in like
manner requires an auxiliary cartridge of dry gun-cotton also
primed with fulminate.

8. The change introduced by the camphor and resinous
matters in the explosive properties of such substances, results
from the modification supervening in the cohesion of the mass.
The mass has acquired a certain elasticity and solidity of parts,
by reason of which the initial shock of the detonator is at. once
propagated throughout a much larger mass (see p. 38). Besides,
a portion of the effects of the shock is expended in the work of
tearing up and separation, a small portion of it remains, which
is susceptible of heating the parts directly struck, and this
heating is, moreover, distributed throughout a large mass.
Hence a sudden and local rise in temperature, capable of deter-
mining chemical and consecutive mechanical actions, becomes
more difficult ; it will require the employment of a detonator
of much greater weight. This results from the preceding theory.

9. But camphor, on the other hand, should not, and does not,
in fact, exercise, as experience proves, any specific action on a
discontinuous powder, such as potassium chlorate powders. It
is this which also explains the fact that blasting gelatin, when
frozen, recovers a sensitiveness to shock which may be compared
with that of nitroglycerin ; the continuity of it has been
destroyed by the crystallisation of this substance.

10. Hence we see the importance of primers, until recently
regarded as simple agents intended to communicate flame to
powder. In fact these primers, provided only their bulk
be sufficient, regulate by their nature the character of the
initial shock, and consequently the character of the entire
explosion. In this case they take the name of detonators,

COMBUSTION AND DETONATION. 55

properly so called. Pure mercury fulminate is especially em-
ployed in this respect ; it is the most powerful of detonators, that
is to say, its shock is more violent and more sudden than that
of any other substance, which is explained by the suddenness of
its decomposition, together with the extraordinary magnitude
of the pressure which it would develop when detonating in its
own volume (almost 26000atm.). At page 52 a certain number
of characteristic facts nave been cited relative to this specific
influence of primers. We shall return to this subject.

§ 7. COMBUSTION AND DETONATION.

1. Progressive combustion has particularly preserved the
name of combustion; the name detonation being reserved to
almost instantaneous combustion with expansion of gas. From
this, again, we get the distinction proposed by Sarrau between
the explosions termed of the first order, such as those of black
powder, which are really ordinary combustions, and the ex-
plosions termed of the second order, or detonations properly so
called, such as that of nitroglycerin, produced by a strong
priming of mercury fulminate.

Nevertheless, the facts known do not, in the author's opinion,
compel us to admit a difference in the nature, and a line of
absolute demarcation between the two orders of phenomena;
they tend rather to place these latter under an aspect showing
an indefinite variety comprised between two extreme limits : —

(a) The detonation of the explosive substance in its own
volume reaching the maximum of temperature and pressure,
and consequently the maximum of speed which chemical
reaction realised under these conditions involves. This effect is
produced when the substance retains the totality of its energy,
that is to say, of the heat developed in the chemical transforma-
tion up to the moment when this latter propagates itself to the
adjacent portions. Detonation is especially produced by a very
sudden shock. Gases formed at the point where the shock is
produced at first have not, so to speak, time to become dis-
placed, and they immediately communicate their energy to the
parts in contact ; the action is thus propagated into the entire
mass with a sort of regularity, producing in it a veritable
explosive wave.

, The velocities of propagation which have been measured with
dynamite and compressed gun-cotton belong to this order of
detonations, and they are very different from those of combustion
of black powder. For instance, the Austrian artillerists have
observed a velocity exceeding 6000 metres per second when
detonating a dynamite cylinder 67 metres in length ; Colonel
Se'bert has observed velocities of 5000 to 7000 metres (a mean
velocity of 6138 metres) in pulverulent gun-cotton compressed

56 DURATION OF EXPLOSIVE REACTIONS.

into long leaden tubes. Further on it will be seen that the
author, together with M. Vieille, has measured velocities of
several thousands of metres per second in explosive gaseous
compounds.

(b) Progressive combustion transmitting itself, step by step,
under the conditions in which the cooling due to conductivity
by contact with inert substances, etc., lowers the temperature to
the lowest degree compatible with the continuation of the re-
action ; all heat is thus dissipated with the exception of a very
small fraction necessary to propagate the reaction in the
adjacent parts. The velocity of combustion in explosive gases
measured by Bunsen (p. 49) is attributable to this mode of
inflammation.

In the case of solid or liquid explosives the propagation of
simple inflammation is rendered more difficult by the movement
of the gases, which distribute themselves throughout a large
space around the point inflamed, instead of acting in a volume
equal to, or slightly different from that of the original bodies ;
they thus share their temperature with a large mass of the sub-
stance up to such a point that the latter cannot be raised to the
desired degree for it to commence decomposition. Thus we
often see the substance dispersed by the gases without ex-
periencing total combustion, and even without undergoing any
change. This happens particularly with explosive substances
not confined within an envelope which concentrates the action
of the gases and gives to it a common resultant (p. 40).

This is the case with nitroglycerin, which is found unaltered
in the vicinity of progressive deflagrations, and it also occurs
with dynamite placed on the ground in a thin layer. Damp
gun-cotton, which is not inflammable, has also furnished
numerous instances of this dispersion resulting from the use
of an insufficient detonator. It is by reason of this special
action of gases that a simple inflammation of a dynamite
cartridge, owing to the use of a badly placed fuse or of ful-
minate not in sufficiently close contact, should be avoided,
inflammation thus preceding the direct action, which ought to
be produced in immediate contact with the fulminate.

2. Between these two limits an entire series of intermediate
states are observed, and they are unlimited in number as the
various modes of inflaming dynamite demonstrate. This is
proved by the influence of a sufficiently strong tamping (p. 40),
which transforms an inflammation into a true detonation.

Finally, we might here cite the inequality of the effects
produced by successive explosions of charges of the same
agent detonating by influence at limiting distances beyond
which the explosion would no longer propagate itself (see
further on).

This variety in the phenomena is due to two orders of causes,

DIFFERENT MODES OF DECOMPOSITION. 57

the one being mechanical and the other more particularly
chemical.

3. From a mechanical point of view it is conceivable that
between the two limits of progressive combustion and detona-
tion the intermediate modes of propagation and reaction may,
according to circumstances, be produced (p. 53), combustion
tending to transform itself more or less quickly into detonation.
But only the two limits should be regarded as constituting
regular standards. This will be more fully defined in the
chapter relating to the explosive wave.

4. Chemical phenomena themselves may vary, at any rate
under certain conditions. In fact, the mode of decomposition is
not unique, unless the explosive substance contains sufficient
oxygen to cause a total combustion, as happens in the case of
nitroglycerin and nitromannite. It is further essential that
this total combustion should have really taken place; which
does not necessarily occur, especially in slow inflammations,
effected at as low a temperature as possible and in which in-
complete reactions may at first become developed.

5. But it often happens that oxygen is deficient, or that the
first reaction gives rise to a bad distribution of this oxygen, as
in the case where nitroglycerin burns slowly, producing a
nitrous vapour, and fixed or gaseous matters, incompletely
burned. Under these circumstances the possible decompositions
are manifold; their number depends on the temperature, on the
pressure, and on the quickness of the heating. We have already
pointed out this case for ammonium nitrate (p. 6) ; it is
generally observed in organic substances decomposable by
heating.1 Mixtures such as black powder are equally susceptible
of it.

6. Among the numerous decompositions of the same sub-
stance, those which develop the greatest heat are those which
also generate the most violent explosive effects, all things else
being equal. This is evident, since the volume of gases (re-
duced to 0° and 760 mm.) reaches its maximum value at the
same time. But it takes place in other cases, the decomposition
being always followed by a diminution of pressure, as has been
elsewhere shown (p. 8).

On the other hand, these are not generally the reactions
which manifest themselves at the lowest temperature possible.
If, therefore, the explosive body receive in a given time a
quantity of heat which is insufficient to carry its temperature
up to a degree which corresponds to the most violent reactions,
it will experience a decomposition capable of disengaging less
heat, or even of absorbing it ; and it will by this decomposition
become completely destroyed without developing its most
energetic effects.

1 " Essai de M&anique Chimique," torn. n. p. 45.

f>8 DURATION OF EXPLOSIVE REACTIONS.

The contrary will happen if the body be rapidly heated up to
the temperature corresponding to the most energetic reaction.

7. In fine, the multiplicity of possible reactions involves
a complete series of intermediate phenomena, especially as,
according to the mode of heating, it may happen that several
decompositions will succeed one another progressively. This
succession of decompositions gives place to effects even more
complicated, as Jungfleisch has pointed out, when the first
decomposition, instead of producing a total elimination of the
decomposed part (changed into gaseous or volatile substances),
results in a division of the primitive substance into two parts ;
the one gaseous, which becomes eliminated, and the other solid
or liquid, which remains exposed to the consecutive action of
heating. The composition of this residue being no longer the
same — which happens, for instance, with nitroglycerin which
has at first disengaged a portion of its oxygen in the form of
nitrous vapours — the effects of its consecutive destruction may
be completely changed.

8. Such are the causes, some chemical, some mechanical,
owing to which nitroglycerin and compressed gun-cotton each
produce such different effects, according as they are inflamed by
the aid of a body feebly ignited by a flame, by an ordinary
fuse, or again by the aid of a cap charged with mercury
fulminate.

For example, Eoux and Sarrau have found that the necessary
charges for breaking a bomb shell, vary cceteris paribus, in an
inverse ratio to the following numbers, the value of which is
calculated by taking gunpowder as unit.

Detonation. Inflammation.

Nitroglycerin lOO 4-8

Compressed gun-cotton 6-5 3-0

Picric acid 5-5 2-0

Potassium picrate 5-3 1*8

The weight of the bursting charge with black powder, itself
under the influence of nitroglycerin primed with fulminate, has
been reduced in the proportion of 4'34 to 1.

This inequality in the force of the same powder according to
the method of ignition, is also partially attributable to the
cooling produced by the walls in a slower reaction ; but
generally it results from a change occurring in the chemical
reaction.

9. The diversity of the effects is less marked with non-com-
pressed gun-cotton, because the influence of the initial shock is
exercised on a smaller quantity of matter, and particularly
because the propagation of the successive reactions in the mass
develops therein weaker initial pressures and a less direct
transformation of energy into heat transmitted to the explosive
body. The cause of this is the interposed air. Consequently

EFFECTS OF DIFFERENT PRIMINGS. 59

the explosive wave can only be produced with difficulty in such
a substance.

Compressed gun-cotton itself is less compact than nitro-
glycerin owing to its structure. This is the reason why
pressures which are due to shocks should become sensibly
attenuated by the existence of interstices. Gun-cotton is there-
fore more difficult to explode than nitroglycerin. Nitroglycerin
explodes by the fall of a weight from a lesser height, by the
use of a priming charged with gun-cotton, or a mixture of
fulminate and potassium chlorate, etc. ; whereas gun-cotton does
not explode under the influence of nitroglycerin, nor under the
influence of a mixture of fulminate and chlorate, it requires the
more sudden shock of pure mercury fulminate.

This latter agent is also less efficacious if it be employed
exposed than if it be placed in a thick copper or tin covering ;
it is less efficacious in an envelope made of paper or tinfoil,
than in a copper envelope; it is still less efficacious if the
priming be not in contact with the gun-cotton. Finally, if it
T5e~plaeed in a leaden tube, an elastic substance which at once
yields to pressure, its effect becomes nullified.

Nitroglycerin is less explosive under the influence of a
priming of fulminate if it be inflamed before the explosion of
the fulminate, the previous inflammation producing a certain
void between the two (p. 56). The absence of immediate contact
between the dynamite contained in the cartridges and the
priming of fulminate is prejudicial for the same reason, the
shock being partially deadened by the interposed air. The
sensitiveness to the action of the fulminate is greater in
dynamite, containing liquid nitroglycerin, than in that con-
taining frozen nitroglycerin, which is similarly explained by
the absence of homogeneity in congealed dynamite, in which
nitroglycerin is partially separated from the porous silica owing
to its solidification.

10. All these phenomena are explained by the more or less-
considerable value of initial pressures, by their more or less
sudden development, and by their more or less easy communica-
tion to the rest of the mass ; that is to say, by the conditions
which regulate the energy transformed into heat in a given time
in the interior of the first layers of the explosive substance
which are reached by the shock (see pp. 52, 53).

The quantity of energy thus transformed depends therefore
both on the suddenness of the shock, and on the greatness of
the work which it is capable of developing, Now here we have
two data, which vary with each explosive substance. For
instance, the most suitable primings are not always those in
which the explosion is the most instantaneous. Abel has
recognised that nitrogen chloride is not very efficacious in
inflaming gun-cotton; nitrogen iodide, so sensitive to the least

60 DURATION OF EXPLOSIVE REACTIONS.

friction, remains absolutely powerless with gun-cotton. Now
nitrogen chloride is precisely one of the explosive bodies of
which we are treating here, which develop the least heat con-
sequently the least work, for a given weight, owing to the high
figure of the equivalent of chlorine. We therefore see that it is
necessary to use more of it by way of priming. As to nitrogen
iodide, according to the analogies taken from iodo-substitution
compounds,1 and from the great weight of the equivalent of
iodine, its explosion should develop much less heat and much
less work for the same weight than even nitrogen chloride ; its
impotence is therefore easily understood.

§ 8. COMBUSTION EFFECTED BY NITRIC OXIDE.

It is advisable to consider here the conditions which deter-
mine the commencement of reactions, conditions which are of
fundamental importance in the study of explosive substances
and on the knowledge of which the study of the combustion
effected by nitric oxide throws a very special light.

1. Nitric oxide contains more than half its weight of oxygen,
and this oxygen, in connection with a combustible body, disen-
gages 21,600 cal. more than free oxygen (0 = 16 grm.). It there-
fore seems that nitric oxide should be a more active burning
agent than free oxygen. Nevertheless, this only happens under
peculiar circumstances, noticed by chemists from the commence-
ment of the nineteenth century, which have given rise to
experimentSj which are produced in every course of lectures,
but have not as yet been properly explained. The author has
resumed this study, which appears to throw a good deal of light
on the work which precedes reactions, and on the manifold
equilibriums of which one and the same system is susceptible.

2. Let us place in the presence of free oxygen two gases sus-
ceptible of combination with it, in the same proportions of
volume, such as nitric oxide and hydrogen previously mixed in
equal volumes, N02 + H2 4- 02, nitric peroxide is immediately
formed, hydrogen being unaffected (respecte). This preference
is manifested, evidently, owing to the inequality of the initial
temperature of the two reactions, nitric peroxide being formed
in the cold ; whereas water is produced only at about 500° to
600°.

3. Nevertheless, this explanation is less decisive than it
appears to be, since the combination of nitric oxide and hydro-
gen liberates a great quantity of heat (1900 cal.), say two-
thirds of the heat of formation of gaseous water (29,500 cal.).
Now this heat should raise the temperature of the system to the
degree at which oxygen and hydrogen combine.

In order fully to demonstrate the phenomenon, the experi-
1 " Annales de Chimie et de Physique," 4* s£rie, torn. xx. p. 449.

HYDROGEN AND NITRIC OXIDE. 61

ment was repeated, doubling the quantity of oxygen, so that the
proportion of this element sufficed for the combustion of the
hydrogen and also nitric oxide.

The reaction produced under these conditions does not give
rise to the combustion of hydrogen, the nitric peroxide being
formed alone, whether the nitric oxide be introduced into the
mixture made beforehand of oxygen and hydrogen, or the oxygen
be introduced into a mixture previously formed of hydrogen and
nitric oxide.

Now the temperature developed by this formation would be
927°, according to the calculation based on the known specific
heats of the elements, and supposing that of the nitric peroxide
to be equal to the sum of its components. It seems difficult to
explain these facts, except by supposing the real temperature to
be much lower ; that is, by attributing to the nitric peroxide
a specific heat greater than that of the elements, as is the case
with the chlorides of phosphorus, arsenic, silicon, tin, titanium,
etc., in the gaseous state,1 and probably increasing with the
temperature, as in the case of carbonic acid.

This has, in fact, been verified experimentally by the author
and M. Ogier.2 The temperature, calculated according to these
new data, falls to 700°, and even lower.

There is, moreover, no exceptional property of the nitric
oxide to prevent combustion. In fact, if the inflammable
temperature of a mixture of oxygen and combustible gas, such as
oxygen and phosphoretted hydrogen, be notably lower, it will
suffice to introduce a few bubbles of nitric oxide, in order to
ignite it at once.

4. When these experiments, with a mixture of hydrogen and
nitric oxide, are carried out over mercury, a complication takes
place which corresponds to a new distribution of the oxygen,
the mercury intervening as a third combustible body, forming
basic nitrates and nitrites.

The quantity of oxygen absorbed is almost double, but the
hydrogen remains unconsumed.

5. These facts being admitted, let us see what happens when
we try to ignite a mixture of hydrogen and nitric oxide.
Berthollet and H. Davy found that ignition does not take
place, either under the influence of the electric spark, or the
influence of a body in a state of combustion. Further, a lighted
match is extinguished in a similar gaseous mixture. If some-
times the hydrogen of this mixture becomes ignited, it is outside
the test tube and at the expense of the oxygen in the atmo-
sphere.

Nevertheless, the flame of the match, or the electric spark,
brings about, at the point heated, decomposition of nitric oxide

1 " Essai de Me'canique Chimique," torn. 1. pp. 336 and 400.

2 " Bulletin de la Socie'td Chimique," 2' s^rie, torn, xxxvii. p. 335.

02 DURATION OF EXPLOSIVE REACTIONS.

into its elements ; for this decomposition takes place at from
500° to 550°, according to the author's experiments.1 But the
oxygen is gradually taken up by the surplus of oxide, without
uniting in any notable proportion with the hydrogen, as has
been shown.

6. The reaction between hydrogen and nitric oxide takes
place, howeyer, when it is excited by a series of sparks, but
gradually and locally. In fact, at the end of ten minutes the
mixture of nitric oxide and hydrogen in equal volumes, NO +
H2, was reduced to one half in these conditions. After some
hours the nitric oxide had disappeared, but there remained
several hundredth parts of free hydrogen, and a basic salt had
been formed at the expense of the mercury. This latter forma-
tion proves that the oxygen set free by the sparks was seized, to
a fractional extent, by the nitric oxide to produce nitric per-
oxide, a gas, the presence of which was quite manifest. This
nitric peroxide gas, in its turn, is partially destroyed by the
hydrogen under the influence of the spark, whilst another
portion oxidises the mercury, and thus a portion of the oxygen
is withdrawn from the ulterior reaction of the hydrogen.

In a word, the formation of nitric peroxide is intermediate
between the decomposition of the nitric oxide and the oxidation
of at least a portion of the hydrogen. We have then, —

(1) NO = N + 0.

(2) NO + 0 = N02.

(3) N02 + 2H2 = 2H20 + N.

Therefore, in order that the hydrogen may be regularly
oxidised, it is not the nitric oxide which it is necessary to
decompose, but the nitric peroxide, a very stable compound, the
destruction of which requires an extremely high temperature.
This accounts for the fact that the combustion induced by flame
or electric sparks is not propagated.

7. The same experiments were repeated with a mixture of
nitric oxide and carbonic oxide :

NO + CO.

According to "W. Henry, this mixture is also not ignited by a
lighted match, which is extinguished in it, nor by electric
sparks.

Nevertheless, it was observed that a series of sparks continued
during some hours decomposed it completely. Only half of the
carbonic oxide is thus converted into carbonic acid, and the
combustion is so imperfect that a little carbon is precipitated on
the platinum wires, as if pure carbonic oxide had been employed.
The surplus oxygen of the oxide forms first of all nitric per-
oxide, and then basic salts of mercury.

Here again, the temperature produced by the spark was
1 " Annales de Chimie et de Physique," 5* G^rie, torn. vii. p. 197.

GASES WHICH DO NOT IGNITE IN NITKIC OXIDE. 63

sufficient to burn the carbonic oxide, but all round the path of
the spark the temperature fell rapidly to a point at which it
could still decompose the nitric oxide, without igniting the
carbonic oxide.

The oxygen formed at the expense of the former compound
thus produced nitric peroxide gas with the surplus.

8. The contrast will be remarked between this experiment
and the sudden combustion of carbonic oxide produced by
mercury fulminate detonating in nitric oxide (see further on).
The fact is that the latter agent sets free at once all the oxygen
of the oxide without passing through the state of nitric per-
oxide.

9. Let us examine more closely the list of gases and other
bodies capable of burning direct at the expense of the nitric
oxide by simple inflammation, or electric sparks, and seek the
causes of the difference which exists between the reaction of
these bodies and of those which do not burn immediately.

The following do not ignite : —

Nitric oxide and hydrogen in equal volumes,

Nitric oxide mixed in the same way with carbonic oxide,

NO + CO.

Nitric oxide mixed with marsh gas,
4NO + CH4.
Nitric oxide mixed with methyl chloride,

3NO + (CH3C1).

And even nitric oxide mixed with methylic ether,
6NO + (CH3)20.

The combination of these mixtures does not take place by
contact with a small flame, nor under the influence of electric
sparks.

Sulphur also, when simply ignited, is extinguished in nitric
oxide.

This absence of combustion is especially remarkable with
methylic ether, which takes the same quantity of oxygen and
disengages nearly the same quantity of heat as ethylene, a gas
which burns, on the contrary, at the expense of the nitric oxide.
The two mixtures occupy, moreover, the same volume.

10. On the other hand, the contact of a lighted match ignites
the following mixtures when formed according to equivalent
ratios of volume : —

Nitric oxide mixed with cyanogen,

2NO + CN.
Nitric oxide mixed with acetylene,

5NO + C2H2.
Nitric oxide mixed with ethylene,

6NO + C2H4.

64 DURATION OF EXPLOSIVE REACTIONS.

These combustions, started by a small flame in a test tube, are
gradual and progressive, and only produce very feeble explosions
like that of carbonic oxide and oxygen.

By means of a powerful electric spark combustion also takes
place, and with singular violence, which shows the difference in
the mode of propagation of the chemical action.

Here it may be mentioned that phosphorus burns briskly in
nitric oxide, that the same takes place with boiling sulphur,
with carbon previously made incandescent, and that carbon
bisulphide also burns briskly in this gas ; these are well-known
experiments.

11. The principal cause of these diversities is the difference
in the temperatures developed by the combustible bodies burning
at the expense of the nitric oxide.

The theoretical calculation of these temperatures may be made
by admitting, in the ordinary way, that the specific heat of a
compound gas is equal to the sum of its elements, and that each
of the latter taken at its molecular weight possesses the same
specific heat as hydrogen, that is, 6 '8 for H2 = 2 grms. at
constant pressure. Temperatures thus calculated are certainly
not the real temperatures, yet it may be admitted that the order
of relative amounts is the same, and that is sufficient for our
comparisons.

Mixtures which do not ignite.

Theoretical temperature
of combustion.

NO + H2 (water, gaseous) 5900°

NO + CO ,. 6600°

3NO + CH3C1 (water, gaseous) 5700°

4NO + CH4 (water, gaseous) 6300°

6NO + (CH8)20 (water, gaseous) 6000°

2NO + S taken at 15° 6600°

Mixtures which do ignite.

Theoretical temperature
of combustion.

2NO + CN ... ... 8500°

5NO + C2H2 (water, gaseous) 8700°

6NO + C2H4 (water, gaseous) 7400°

6NO + CS2 ... 7500°

2NO + C ... 8200°

5NO + P2 10200°

4NO + PH3 8400°

2NO -j- S previously heated to 450° ... 7050°

It will" be observed that the theoretical temperature of
combustion of sulphur, taken at about 15° by nitric oxide, is
very near the limit ; it therefore does not burn. On the contrary,
if the sulphur be contained in a heated receptacle and kept at
a temperature of about 450° by boiling, the nitric oxide being
rapidly raised by contact with the vessel to about the same
temperature, and thus the temperature of combustion of the

AMMONIA AND NITRIC OXIDE. 65

mixture be raised, then the sulphur should burn in nitric
oxide.1

This is what is noticed, as we know, in operating with
sulphur placed in a small crucible previously brought to a
red heat.

The temperatures of combustion estimated in this way are
generally very near those estimated by the employment of free
oxygen, the excess of heat produced by the decomposition of
nitric oxide being compensated by the necessity of heating the
nitrogen. All these figures, however, do not express absolute
values, yet they may be regarded as marking the relative order
of temperatures of combustion.

12. This table, understood in this manner, shows that the
property of burning at the expense of nitric oxide under the
influence of a flame or electric spark, depends more especially
on the temperatures developed. The comparison of ethylene
with methylic ether is particularly decisive in this respect,
since the relations of volume between the combustible and the
combustive gas are exactly the same, and the heats disengaged
(451,100 cal. and 443,800 cal.) do not sensibly differ, but
methylic ether also contains the elements of water, which lowers
the temperature of combustion.

In short, among the bodies comprised in the table, none of
those which develop a theoretical temperature below 7000° will
ignite, whereas all the bodies which develop a higher tempera-
ture either burn or detonate. It is possible that this circum-
stance is connected with the previous formation of nitric
peroxide at the expense of nitric oxide (see p. 62), and con-
sequently with the necessity for a very high temperature in
order to regenerate, at the expense of the nitric oxide, the
oxygen which is indispensable to combustion.

13. Instead of destroying nitric peroxide by heating it to an
excessively high temperature, it can be decomposed by a
chemical reaction at a lower temperature, which lowers the
theoretical limit of the temperature of combustion.

This is precisely what happens in the case of ammonia gas.
This gas, in fact, mixed with nitric oxide,

3NO + 2NH3,

ignites with a match, and, according to W. Henry, detonates
under the influence of the electric spark. The theoretical
temperature of combustion of the mixture (5200°) is, however,
less than all the foregoing temperatures. But, on the other
hand, nitric peroxide reacts even when cold on ammonia gas,
and the reaction develops itself still more simply by the intro-
duction of oxygen into a mixture of nitric oxide and of ammonia
gas. When cold it will produce both nitrogen and ammonium

1 "Essai de M^canique Chimique," torn. i. p. 331.

F

66 DURATION OF EXPLOSIVE REACTIONS.

nitrate,1 which at a high temperature resolves itself into nitrogen
and water. Therefore we definitely obtain

2NO + 0 + 2H3N = 4N + 3H20 disengages (water, gaseous)

+ 98,000 cal.

Every portion of nitric oxide destroyed by the spark with the
formation of free oxygen, determines, therefore, a new reaction
which disengages heat, and easily propagates the combustion of
the system, which does not take place in gases which do not
exercise a special reaction on nitric peroxide.

§ 9. DECOMPOSITION OF ENDOTHEKMAL COMBINATIONS,
ACETYLENE, CYANOGEN, ETC.

1. So far, we have treated more especially of the combustion
and detonation of mixtures and combinations containing such
combustible elements as carbon, hydrogen, sulphur, and the com-
bustive elements such as oxygen. But as the theories which we
are considering are based essentially on the disengagement of
heat and the development of the gases produced by transforma-
tion, they lead to consequences of a very special character, which
are very interesting as regards the decomposition of endothermal
combinations such as acetylene, cyanogen, and nitric oxide.

Acetylene, cyanogen, and nitric oxide are, in fact, formed
from their elements with the absorption of heat. This absorp-
tion amounts to

A -61,100 cal.2 for acetylene (C2H2 = 26 grms.)
A - 74,500 cal. for cyanogen (2CN = 52 grms.)
A -31 ,600 cal. for nitric oxide (NO = 30 grms.)

If we succeed in rapidly decomposing these gases into their
elements, such a quantity of heat reproduced inversely will
raise the temperature up to 3000° in acetylene and nitric oxide,
up to 4000° in cyanogen, according to a calculation founded on
known specific heats of the elements.

The proper figures for this calculation are as follows. We
will admit for the mean specific heat of carbon C2 = 24 grms.,
the value 12 ; for that of hydrogen H2 = 2 grms., 6'8 at constant
pressure, and 4'8 at constant volume, these latter values being
equally applicable to nitrogen N2 = 28 grms., and to oxygen
02 = 32 grms. at the same volume. We thus find,

For acetylene decomposed under constant pressure 3300°,
under constant volume 3640°.

For cyanogen decomposed under constant pressure 3960°,
under constant volume 4375°.

1 See author's remarks, "Annales de Chimie et de Physique," 59 sdrie,
torn. vi. p. 208.

2 This figure refers to carbon as diamond ; in amorphous carbon such as is
precipitated at the time of decomposition we should obtain 6000 cal. less.
The same remark applies to cyanogen as the mean specific heat of carbon.

DECOMPOSITION OF ENDOTHERMAL COMBINATIONS. 67

For nitric oxide decomposed under constant pressure 3200°,
under constant volume 4500°.

It is understood that the calculation of these temperatures is
subordinate to the presumed constancy of the specific heats.
Whatever opinion is held in this respect it is certain that it
gives an idea on temperature more probable in the present case,
where it is a question of an elementary decomposition, than in
reactions in which compound bodies are formed, such as in the
combustions of hydrogen or carbonic oxide, combustions which
are limited in their progress by the dissociation of compound
bodies.

2. However, it has not been possible up to the present to
effect the explosion of acetylene, or cyanogen, or of nitric
oxide.

Whereas hypochlorous gas detonates under the influence of
slight heat, when in contact with a flame, or a spark, in spite
of the smaller amount of heat liberated, + 15,200 cal. (for
C120 = 87 grms.), which can only raise the elements of this gas
to '1250°, on the other hand, acetylene, cyanogen, and nitric
oxide do not detonate either by simple heating or by contact
with flame, nor even under the influence of the spark or even
the electric arc.

These differences are important. The diversity which exists
between the mode of destruction of endothermal combinations
is due in each given reaction to the necessity of a kind of pre-
paration, and a certain amount of preliminary work. The
author has, besides, examined l the characters and the generality
of this preliminary work in the production of chemical re-
actions. Now the work necessary for resolving the compounds
named into the elements does not appear to consist in a simple
heating, slow and progressive in its nature, at least within the
limits of the temperature above pointed out. In fact, acetylene,
cyanogen, and nitric oxide never explode, as far as the author's
experience goes, no matter to what temperature they are raised.

It is not that these compound gases are absolutely very
stable — they in fact decompose frequently, and even according
to experience at a dull red heat, either with the formation of
polymers (benzene by acetylene), or with a fresh distribution of
their elements (nitrogen, monoxide, and nitric peroxide, by
nitric oxide) 2 — but they do not explode in spite of the very great
liberation of heat accompanying these changes, probably by
reason of the slowness of their action, nor do they explode,
which is stranger still, under the influence of electric sparks, in
spite of the excessive and sudden heat which these latter develop.
Carbon, however, on the passage of the sparks, is precipitated at
once from acetylene or cyanogen, while hydrogen and nitrogen

1 " Essai sur la M^canique Chimique," torn. ii. p. 6.

2 " Annales de Chimie et de Physique," 5" s£rie, tom. vi. p. 198.

F 2

68 DURATION OF EXPLOSIVE REACTIONS.

are liberated. The electric arc accelerates in a peculiar degree
the decomposition of the cyanogen in which it is produced, yet
without rendering it explosive.1 Nitrogen and the oxygen of
the nitric oxide also separate on the passage of the electric
spark. As a matter of fact the oxygen of this latter gas
becomes united with the excess of the surrounding oxide and
generates nitric peroxide. A portion of the hydrogen and of
the carbon liberated at the expense of the acetylene also
reunites under the influence of the electricity so as to recon-
stitute this hydrocarbon, the whole forming a system in
equilibrium.2 To these circumstances might be attributed
the absence of the propagation of the decomposition, but this
explanation is not sufficient for the cyanogen, which becomes
entirely decomposed,3 without possibility of reconversion.

Nor does this suffice for arseniuretted hydrogen, a gas
decomposable, according to Ogier, with liberation of 36,700 cal.
(AsH3 = 78 grms.).

This latter gas is so very unstable that it is continually
decomposing at normal temperature, if kept in sealed glass
tubes. It is well known with what facility even the last trace
is decomposed by heat in Marsh's apparatus. A series of
electric sparks will also completely destroy it. Nevertheless,
arseniuretted hydrogen does not explode, as the author has
shown, either under the influence of progressive heat or under
that of the electric sparks.

3. Thus, in the endothermal combinations already enumerated,
there exists a condition, associated with their molecular con-
stitution, which prevents the propagation of the chemical action
under the influence of mere progressive heating or of the
electric spark, at least so long as the temperature remains below
certain limits.

We are aware that the study of explosive substances presents
circumstances which are analogous. The simple ignition of
dynamite, for instance, would not suffice to cause its explosion ;
on the contrary, Nobel has shown that explosion is produced by
the influence of special detonators, such as mercury fulminate,
and which are susceptible of developing a very violent shock.
The thermodynamic theory has already been given (p. 53) of
these effects, which appear to be due to the formation of a
veritable explosive wave, which wave is totally distinct from
the sonorous waves, properly so called, since it results from a
certain cycle of mechanical, calorific, and chemical actions, which
reproduce themselves step by step, transforming themselves one
into the other; this is shown in the experiments which the

1 " Comptes rendus des stances de I'Acad&nie des Sciences," xcv. p. 955.

2 '" Annales de Chimie et de Physique," 4" se*rie, torn, xviii. pp. 160, 199.

3 That is, it does not contain any trace of a hydrogenated body susceptible
of forming hydrocyanic acid, which, on the contrary, gives rise to equilibrium.

DETONATION OF ACETYLENE.

author made, together with Vieille, on mixtures of hydrogen
and other combustible gases with oxygen. It has in like
manner been shown that the great effectiveness of mercury
fulminate as a detonator is explained, not only by the rapidity
of decomposition in this body, but more particularly by the
enormous pressure which it develops when exploding in its
own volume, pressures far above those of all known bodies,
and which may, according to our tests, be estimated at over
27,000 kgms. per sq. cm.

This led to the detonation being attempted of acetylene,
cyanogen, and arseniuretted hydrogen under the influence of
mercury fulminate, and the
trials were completely suc-
cessful. The following are
the details.

4. Acetylene. — Introduce
a certain volume of acety-
lene, from 20 c.c. to 25
c.c., for instance, into a
small test tube, E, the walls
of which should be very
thick. In the centre of the
gaseous mass place a small
cartridge, K, containing a
small quantity of fulmi-
nate (about O'l grm.), and
traversed by a very thin
metallic wire in contact at
the other end with the iron
fitting of the test glass, an
electric current will bring Pi
this wire to a red heat.
All this is supported by a
tube containing a second
wire fused into the tube,
and extending outwards as
far as F. The capillary
glass tube CO, in the form
of an inverted syphon, is

Fig. 3.

fixed into a plug D, which closes the test tube.

Fig. 3 shows the system in readiness ; Fig. 4 shows the glass
tube provided with its inner wire.

Fig. 5 shows the steel plug in its natural size, with the hole
T, into which the above-named cap is screwed.

Fig. 6 gives in its natural size the steel cap P, through which
there is a passage for the tube, which is cemented into this cap
along with the second wire.

This arrangement will permit of the test tube being filled

70

DURATION OF EXPLOSIVE REACTIONS.

with gas over mercury, and of there introducing the wires
fitted with their fuses and adjusted to the plug. This is then
closed by a bayonet fastening, and the detonation is effected
under a constant volume.

For this purpose, the current is passed ; the fulminate goes

off, a violent explosion is
caused, and a large flame ap-
pears in the test tube. After
cooling, the glass will be found
filled with black finely divided
carbon, the acetylene has dis-
appeared, and free hydrogen
remains.1 Unscrew the cap P
under the mercury, remove it
with the capillary tube, and
then collect and examine the
gases contained in the test
tube.

The acetylene is thus purely
and simply decomposed into its
elements —

C2H2 = C2 -f- H2.

Scarcely a trace of the original gas
will be found, and, if any, it will not be
more than a hundredth of a cub. cm., and
this is doubtless some portion not reached
by the explosion.

The reaction is so rapid that the small car-
tridge of thin paper which enveloped the fulmi-
nate will be found torn, but not burned, even
FiS- 4- in its thinnest fibres ; and this is explained, if we
note that the time during which the paper re-
mained in the explosion centre was about ^^^wcf °f a
second, according to the thickness of the paper and the known
data relative to the rapidity of this order of decomposition.

The carbon set free exhibits the same general conditions as
that obtained in a tube at a red heat ; it is mainly amorphous
carbon, and not graphite ; it dissolves almost totally when treated
several times with a mixture of fuming nitric acid and potassium
chlorate. ^ Nevertheless, in this way, it gives a trace of graphite
oxide, which proves that it contains a trace of graphite, pro-
duced doubtless by the transformation of the amorphous carbon
under the influence of the excessive temperature to which it has
been subjected.

The author has, in fact, shown that amorphous carbon heated
up to about 2500° by electrolytic gas commences to change into

1 Mixed with nitrogen and carbonic oxide proceeding from the fulminate.
and which- have been formed independently.

DETONATION OF CYANOGEN.

71

graphite, and that the lamp-black precipitated by the incom-
plete combustion of the hydrocarbon also contains a trace
of it.1

5. Cyanogen. — The same test carried out with cyanogen is

Fig, 5.

Fig. 6.

equally successful ; the cyanogen detonates under the influence
of the fulminate, and resolves itself into its elements.
2ONT = C2 4- Na.

Thus we can produce free nitrogen, and amorphous carbon in
a highly divided state similar to what is obtained by the
electric spark. This carbon marks paper as plumbago will do.
Yet it is by no means real graphite, because it will almost
totally dissolve, if repeatedly treated with a mixture of fuming
nitric acid and potassium chlorate. Still one trace of graphitic
oxide, left as a residue, bears witness to the existence of a trace
of graphite, as in the case of acetylene.

This test is not always successful. Sometimes the explosion
of the fulminate takes place without precipitating the carbon of
the cyanogen.

Nitro-diazobenzene, which was also tested by using it as
a detonator instead of fulminate, decomposed without causing
the cyanogen to explode. Even the mode of decomposition of

1 " Annales de Chimie et de Physique," 5* seVie, torn. xix. p. 418. The
voltaic arc produces a more complete transformation; but then the effects
of the heat become complicated by those of electricity (p. 419).

72 DURATION OF EXPLOSIVE REACTIONS.

nitro-diazobenzene was different under circumstances in which
the detonator is destroyed at a slight pressure, from its decompo-
sition in the calorimetric bomb under a high pressure, as
observed by Vieille and the author. Instead of obtaining all the
oxygen from the compound in the state of carbonic oxide at the
same time as free nitrogen and a nitrogenous carbon of a very
porous and dense nature, on this occasion, along with the
nitrogen, only one-fourth of the volume of the theoretical car-
bonic oxide was observed, along with some phenol and a tarry
substance.

6. Nitric oxide. — This body explodes under the influence of
mercury fulminate, but the phenomenon is more complicated
than with the former gases, the carbonic oxide produced by the
fulminate burning, at the expense of the oxygen of the nitric
oxide, to form carbonic acid. This combustion appears to have
taken place at the expense of free oxygen, and not of nitric per-
oxide formed transitorily. In fact, the mercury is not attacked,
contrarily to what always happens when this gas appears for a
moment.

We therefore have,

NO = N + 0
CO + 0 = C02.

The combustion even of carbonic oxide is characteristic, for
nitric oxide, mixed with carbonic oxide, does not explode either
by simple inflammation, or by the electric spark.

7. Arseniuretted hydrogen. — Arseniuretted hydrogen has ex-
ploded under the influence of the fulminate, and has become
absolutely resolved into its elements, arsenic and hydrogen.

AsH3 = As + H3.

8. Here will be given experiments on the sudden decomposi-
tion of nitrogen monoxide into nitrogen and oxygen. This
decomposition, which liberates -f 20,600 cal. (N20 = 44 grins.),
may be caused by the sudden compression of 30 c.c. of this
gas reduced to ^^ of their volume by the sudden fall of a ram
weighing 500 kgms.1

On the other hand, nitrogen monoxide only decomposes
gradually under the influence of progressive heat or of electric
sparks.

9. All these tests are in reference to gases. But solid or
liquid eridothermal combinations offer the same variety. While
nitrogen chloride and iodide explode under the influence of a
slight heat, or of slight friction, nitrogen sulphide requires to be
heated up to 207°, or requires violent concussion, in order to
explode and to become resolved into its elements. It then
liberates 4- 32,300 cal. (NS2 = 46 grms.), according to tests
which the author has made along with Vieille.

1 " Annals de Chimie et de Physique," 5e s&ie, torn. iv. p. 145.

ATTEMPT TO DECOMPOSE CHLORINE. 73

10. Potassium chlorate, itself a body which liberates -f 11,000
cal. (KC103 = 122-6 grms.) when decomposing into oxygen and
potassium chloride, may undergo this decomposition at an
ordinary temperature, if struck violently with a hammer on an
anvil, after being enveloped in a thin sheet of platinum. It has
been found, in fact, that in this way an appreciable quantity of
chloride is found. Pure chlorate, in a state of fusion, explodes
much more easily, and sometimes of itself, if the heating be too
sudden. This detonation has been the cause of more than one
accident in laboratories.

11. As a further instance may be mentioned celluloid (a
variety of nitro-cotton, mixed with various substances). At
ordinary temperatures it is a very stable substance. The author,
however, observed that this body explodes when brought up to
the temperature at which it softens, and in this state struck
with a hammer on an anvil.

Generally speaking, compounds and explosive mixtures
become more and more sensitive to shocks in proportion as they
approach the temperature of their initial decomposition (see
p. 37).

12. Two other experiments were made, which it may be
useful to point out, in spite of their negative character. One of
them consisted in exploding the fulminate in an atmosphere of
gaseous chlorine. Assuming the compound nature of chlorine
regarded as an endothermal radical containing oxygen, one
would have been able to observe the products of the decomposi-
tion caused by the explosion of the fulminate, yet the results
were negative, as, of course, was to be expected, in accordance
with received ideas. The chlorine had scarcely been introduced
into the atmosphere when the fulminate exploded of itself, yet
the chlorine was not destroyed.

This gas having been subsequently absorbed by agitating
it with mercury, carbonic oxide and nitrogen remained in the
proportion of gaseous volumes answering to the fulminate ; that
is to say, without any excess of carbonic acid, or of any other
product formed at the expense of the chlorine.

13. An attempt was also made to destroy glucose, on the
assumption that fermentations are exothermal operations.1 A
strong capsule of fulminate, containing 1*5 grms. of this body,
was exploded in a metallic cartridge completely filled with an
aqueous 20% solution of glucose. But the result was negative.

14. In fine, acetylene, cyanogen, and arseniuretted hydrogen,
that is to say, gases formed by the absorption of heat but
which do not explode by simple heating, may be caused to
explode under the influence of a sudden and very violent
shock, such as that which results from the explosion of the
mercury fulminate. This shock, in reality, only reaches a

1 " Essai de Mdcanique Chimique," torn. xi. p. 55.

74 DURATION OF EXPLOSIVE REACTIONS.

certain stratum of the gaseous molecules, to which it communi-
cates an enormous energy. Under this shock the molecular
edifice loses its relative stability, for which it was indebted to
a special structure. Its interior connections having become
broken, it crumbles, and the initial force becomes immediately
strengthened by everything which answers to the heat of the
decomposition of the gas. Hence a fresh shock, caused by the
adjacent stratum, which also causes its decomposition, the
actions co-ordinate themselves, reproduce and propagate one
another, step by step, with similar characteristics, and in an
extremely short interval of time, after the manner of the
explosive wave, until the total destruction of the system is
complete.

These are the phenomena which bring to light the direct
thermo-dynamic relations existing between chemical and
mechanical actions.

CHAPTEE VI.

EXPLOSIONS BY INFLUENCE.

§ 1. EXPERIMENTAL OBSERVATIONS.

1. So far we have studied the development of explosive
reactions either from the point of view of their duration in
a homogeneous system, all the parts of which are maintained at
an identical temperature, or from their propagation in an equally
homogeneous system which is fired directly by means of a body
in ignition or by a violent shock. But the study of explosive
substances has revealed the existence of another mode of pro-
pagating reactions in explosives ; this propagation taking place
at a distance and through the medium of the air or of solid bodies
which of themselves do not participate in the chemical change.

We now refer to explosions by influence, which hitherto have
been suspected from certain known facts in connection with the
simultaneous explosion of several buildings, widely separated, in
catastrophes at powder works.

Attention has been especially called to this class of phenomena
by the study of nitroglycerin and gun-cotton.

2. We will first cite some characteristic facts.

A dynamite cartridge exploded by means of a priming of
fulminate causes the explosion of cartridges in its vicinity, not
only by contact and by direct shock, but even at a distance.
An indefinite number of cartridges in a straight line or regular
curve can also be exploded in this way.

The distances at which explosion will propagate itself are,
comparatively speaking, considerable. Thus, for instance, with
cartridges contained in stiff metallic cases, and placed on firm
ground, the explosion caused by 100 grms. of Vonges dynamite
(75% of nitroglycerin, 25% of randanite, that is to say of silica
in a very finely divided state), communicates itself to a distance
of 0-3 metre, according to Captain Coville's tests. D being the
distance in metres and C the weight of the charge in kgms.,
the tests of this officer have given D = 3'0 C.

With cartridges resting on a rail he obtained D = 7*0 C.

76 EXPLOSIONS BY INFLUENCE.

On a loose or free soil the distances were less. When the
cartridge was suspended in the air, detonation did not take place
by influence, probably because the cartridge not being fixed
could easily recoil, thus diminishing the violence of the shock.

However, there are trials on record which show that air is
sufficient to transmit detonation by influence, although less
easily, and when dealing with large masses.

With dynamite containing less nitroglycerin (55% of nitro-
glycerin, and 45% of Boghead ashes) placed in cartridges of
a similar nature and laid on the ground, the trials made by
Captain Pamard gave shorter distances: D = 0'9 C.

If metallic casings having less resistance be used, the distance
to which the explosion propagates itself is similarly reduced.

Dynamite when merely spread about on the ground even
ceases to propagate the explosion.

Experiments made in Austria have given similar results.
They have shown that the explosion communicates itself both
in the open air with intervals of 0*04 metre and through deal
planks 0*018 metre thick. In a leaden tube, with a diameter
equal to 0*15 metre and 1 metre long, a cartridge placed at one
extremity will cause the explosion of another cartridge placed
at the opposite end.

The transmission of the explosion is more easily effected in
tubes of cast iron. Joints lessen the susceptibility of trans-
mission.

3. The explosion thus propagated may grow weaker from one
cartridge to another and even change its character. Thus
according to experiments made by Captain Muntz at Versailles
in 1872, a first charge of dynamite when exploding direct had
made a crater in the ground the radius of which was 0*30 metre.
The second charge, which exploded by influence, produced a
hollow merely of 0*22 metre ; the effect of the detonation had
therefore become lessened. This diminution should become
manifest particularly towards the limit of the distances at which
the influence ceases.

In the same way four tinplate screens were placed at intervals
of 0'040 metre, and a small cylinder of gun-cotton was placed
against each of them, the whole fixed on a board. At a distance
of 0*015 metre in front of the first screen, a similar cylinder
was exploded. All the cylinders exploded, but a progressive
diminution was observed in the cavities produced in the board
placed below each cylinder.

According to these facts, propagation by influence depends
both on the pressure acquired by the gases and on the nature of
the support. It is not even necessary that this support should
be firm.

It has been ascertained that these effects are not generally
due to simple projections of fragments of casing or of the neigh-

EXPLOSIONS TRANSMITTED BY WATER. 77

bouring substances, although such projections often play a
certain part. In this respect, the real character of the effects
produced is shown more particularly from tests made under water.

4. In fact, when experimenting in water, below a depth of
1-30 metres a charge of dynamite weighing 5 kgms. will cause
the explosion of a charge of 4 kgms. situated at a distance of
3 metres.

The water therefore transmits the explosive shock, at any
rate to a certain distance, in the same way as a solid body.
This transmission is so violent that fishes are killed in ponds
within a certain radius by the explosion of a dynamite cartridge ;
this process is sometimes employed by fishermen, but has the
disadvantage of destroying all the fish.

5. Similar trials have been made by Abel with compressed
gun-cotton. According to his observations the explosion of a
first block determines the explosion of a series of similar
blocks. This propagation has also been studied under water ;
the explosion of a torpedo charged with gun-cotton causing the
explosion of neighbouring torpedoes placed within a certain
radius.

Sudden pressures transmitted by water have even been
measured by the aid of the lead crusher at different distances,
such as 2'50 metres, 3'50 metres, 4*50 metres and 5*50 metres.
They decrease with the distance, as might be expected. Besides,
experience proves that the relative position of the charge and the
crusher is immaterial, and this is in accordance with the principle
of equal transmission of hydraulic pressures in all directions.

6. Explosions of fulminating substances, propagating them-
selves suddenly to a great number of amorces, belong to the
same order of explosions by influence.

The explosion in the Eue Beranger has been previously
mentioned (p. 46). The experiments made on that occasion by
Sarrau showed that amorces, similar to those which caused this
catastrophe, will burn successively by simple inflammation
during a fire without giving rise to a general explosion, whereas
the explosions of some of these amorces each containing O'OIO
grm. of explosive matter, if produced by a sudden pressure,
determines, by influence, the explosions of neighbouring packets
even when not contiguous, and when situated at a distance of
0*15 metre. A general explosion, therefore, can be easily pro-
duced by influence under these conditions.

§ 2. THEORY FOUNDED ON THE EXISTENCE OF THE EXPLOSIVE

WAVE.

1. It follows from these facts, and particularly from experi-
ments made under water, that explosions by influence are not
due to inflammation, properly so called, but to the transmission

78 EXPLOSIONS BY INFLUENCE.

of a shock resulting from enormous and sudden pressures
produced by nitroglycerin or gun-cotton, the energy of which
shock is transformed into heat in the explosive substance (see
pp. 36, 57).

2. In an extremely rapid reaction, the pressure may approach
the limit corresponding to the matter exploding in its own
volume; and the disturbance due to the sudden development
of pressures, nearly theoretical, may propagate itself either by
the mediation of the ground and of the supports, or through
the air itself, when projected en masse, as has been shown by
the explosions of certain powder mills, gun-cotton magazines,
and also by some of the experiments made with dynamite and
compressed gun-cotton. The intensity of the shock propagated
either by a column of air or by a liquid or solid mass, varies
according to the nature of the explosive body and its mode of
inflammation ; it is more violent the shorter the duration of the
chemical reaction and the more gas there is developed ; that is
to say, a stronger initial pressure and a greater heat, or, in
other words, greater work for an equal weight of explosive
substance (see pp. 40, 41).

3. This transmission of the shock is more easily effected by
solids than by liquids, and more easily by liquids than by gases ;
in the case of gases it takes place all the more easily if they are
compressed. It is propagated all the more easily through solids
when these are hard ; iron transmits better than earth, and hard
earth better than soft soil.

Any kind of junction has a tendency to weaken, especially if
any softer substance intervene. Hence the employment as a
receptacle of a tube formed of a goose quill, will stop the effect
of mercury fulminate, whereas a copper tube or capsule trans-
mits this effect in all its intensity.

Explosions by influence propagate themselves all the more
easily in a series of cartridges, if the casing of the first deto-
nating cartridge is very strong ; this allows the gases to attain
a very high pressure before the bursting of the casing (p. 40).

The existence of an air-space between the fulminate and the
dynamite, will, on the other hand, diminish the violence of the
shock transmitted, and consequently that of the explosion. As
a general rule, the effect of shattering powders is lessened when
there is no contact.

4. In order to form a complete idea of the transmission by
supports of sudden pressures which give rise to shock, it is well
to bear in mind the general principle whereby pressures in a
homogeneous mass transmit themselves equally in all directions,
and are the same over a small surface, whatever may be the
direction. The explosions produced under water with gun-cotton
show, as has been said above, that this principle is equally ap-
plicable to sudden pressures produced by explosive phenomena.

TWO ORDERS OF WAVES. 79

But this ceases to be true when passing from one medium to
another.

5. If the chemically inactive substance which transmits the
explosive movement be fixed in a given position on the ground
or on a rail on which the first cartridge has been placed, or
again, held by the pressure of a mass of deep water, in which
the first detonation has been produced, the propagation of the
movement in this matter could scarcely have taken place except
under the form of a wave of a purely physical order, a wave, the
character of which is essentially different to the first wave
which was present at the explosion, the latter being both of a
chemical and physical order, and having been developed in the
explosive body itself. While the first or chemical wave pro-
pagates itself with a constant intensity, the second, or physical
wave, transmits the vibration starting from the explosive centre,
and all around it, with an intensity which diminishes in inverse
ratio to the square of the distance. In the immediate neigh-
bourhood of the centre, the displacement of molecules may
break the cohesion of the mass, and disperse it, or crush it by
enlarging the chamber of explosion, if the experiment be carried
out in a cavity. But at a very short distance, and the greatness
of this depends on the elasticity of the surrounding medium,
these movements, confused at first, regulate themselves, so as to
give rise to the wave properly so called, characterised by sudden
compressions and deformations of the substance. The amplitude
of these undulatory oscillations depends on the greatness of the
initial impulse.

They progress with an excessive rapidity, at the same time
constantly decreasing in intensity, and they maintain their
regularity up to points at which the medium is interrupted.
There these sudden compressions and deformations change their
nature, and transform themselves into an impelling movement,
that is to say, they reproduce the shock. If then they act on a
fresh cartridge they will cause it to explode. This shock will
further be attenuated by distance, owing to the decrease thus
introduced into its intensity. Consequently the character of
the explosion may be modified. The effects will thus diminish
up to a certain distance from the point of origin, beyond which
distance the explosion will cease to produce itself.

When the explosion has taken place in a second cartridge the
same series of effects is reproduced from the second to the third
cartridge, but they depend upon the character of the explosion
in the second cartridge and so on.

6. Such is the theory which appears to the author to account
for explosions by influence, and for the phenomena which
accompanies them. It rests on the production of two orders of
waves, the one being the explosive wave, properly so called,
developed in the substance which explodes, and consisting of a

80 EXPLOSIONS BY INFLUENCE.

transformation incessantly reproduced from chemical actions
into calorific and mechanical actions, which transmit the shock
to the supports and to contiguous bodies ; and the other purely
physical and mechanical, which also transmits sudden pressures
around the centre of vibration to neighbouring bodies and by a
peculiar circumstance to a fresh mass of explosive matter.

The explosive wave, once produced, propagates itself without
diminishing in force, because the chemical reactions which
develop it regenerate its energy proportionately along the whole
course ; whereas the mechanical wave is constantly losing its
intensity in proportion as its energy, which is determined ^only
by the original impulsion, is distributed into a more considerable
mass of matter.

7. A different theory than this was at first proposed by
Abel, namely, the theory of synchronous vibrations, of which it
will be well to speak now. According to this authority the
determining cause of the detonation of an explosive body resides
in the synchronism between the vibrations produced by the
body which provokes the detonation and those which would be
produced by the first body when detonating, precisely as a
violin-string resounds at a distance in unison with another
chord, set in vibration. In support of this, Abel cited the
following facts. In the first place, detonators appear to be
special for each kind of explosive substance. For instance,
nitrogen iodide, which is very susceptible to shock and friction,
does not appear to be able to cause the detonation of compressed
gun-cotton. Nitrogen chloride, so easily explosive of itself, only
produces detonation when a weight ten times that of the neces-
sary fulminate is employed. In the same way nitroglycerin does
not cause the detonation of gun-cotton in sheets on which the
envelope containing it is placed. In this way 23'3 grms. of
nitroglycerin have been made to detonate without success. On
the other hand, the inverse influence is proved, 775 grms. of
compressed gun-cotton having detonated nitroglycerin enclosed
in an envelope of thin foil at a distance of 0'02 metre. A
priming formed of a mixture of potassium ferrocyanide and
potassium chlorate will not cause gun-cotton to detonate
(according to Brown).

Finally, according to Trauzl, a much greater weight of a
priming made of a mixture of mercury fulminate and potassium
chlorate. should be taken than if it were formed of fulminate
alone. Nevertheless the heat liberated by unit weight is one-
fifth greater with the former mixture.

8. Champion and Pellet have adduced the following experi-
ments in support of this ingenious hypothesis : they fixed on
the string of a contra-bass particles of nitrogen iodide, a
substance which detonates by the slightest friction. They
then caused the strings of a similar instrument situated at a

DETONATION AND SYNCHRONOUS VIBRATIONS. 81

distance to vibrate ; detonation took place, but only for sounds
higher than a given note, which note represented sixty vibrations
per second. They then took two conjugate parabolic mirrors
fixed 2'5 metres apart, and they placed along the line of foci at
different points a few drops of nitroglycerin or grains of nitrogen
iodide, then they caused the detonation of a large drop of nitro-
glycerin on one of the foci ; they observed that the explosive
substances placed on the conjugate focus exploded in unison, to
the exclusion of similar substances placed at other points. A
coating of lamp-black placed on the surface of mirrors served
to prevent any reflection and the concentration of the calorific
rays.

9. None of these tests, however, are conclusive, and several
of them appear absolutely contrary to the theory of synchronous
vibrations. In the first place it may be remarked that the fact
of a certain musical note being capable of determining each
kind of explosion has never been established properly; it is
only below a certain note that the effects cease to be produced,
whereas they take place by preference, and whatever be the
explosive body, in the sharpest notes. Besides, the effects cease
to be produced at distances incomparably less than the
resonance of the chords in unison, which proves that detonations
are functions of the intensity of mechanical action rather than
of the character of the vibration which determines them. De-
tonation also ceases to be produced when the weight of the
detonator is too slight, and consequently when the energy of the
shock is attenuated. The specific vibrating note, however,
which determines explosion should always remain the same.
For instance, cartridges of 75 per cent, dynamite cease to
explode when the capsule contains less than 0'2 grm. of
fulminate ; the explosion only being insured in any case at the
regulation weight of 1 grm. This confirms the existence of a
direct relation between the character of the detonation and the
intensity of the shock produced by one and the same detonator.

If it were true that gun-cotton could explode nitroglycerin
by reason of the synchronism of the vibration transmitted, it is
difficult to understand why reciprocal action does not take
place ; whereas the absence of reciprocity is easily explained by
the difference in the structure of the two substances, which
plays an important part in the transformation of energy into
work (p. 38).

10. This same diversity of structure and the modifications
which it introduces into the transmission of the phenomena of
shock, and the transformation of mechanical energy into calorific
energy, may be quoted in order to account for the facts observed
by Abel.

The difference between the energy of pure fulminate and that
of fulminate when mixed with potassium chlorate is not any

82 EXPLOSION BY INFLUENCE.

less easy to explain ; the shock produced by the former body
being more sudden by reason of the absence of all dissociation of
the product, which is no other than carbonic oxide ; this absence
should be opposed to the dissociation of carbonic acid which is
produced in the second case. Probably also the formation of
potassium chloride disseminated in the gases produced with the
aid of potassium chlorate serves to attenuate the shock, like the
silica in dynamite.

11. All the effects observed with nitrogen iodide are explained
by the vibration of the supports, and by the effects of the
resulting friction, this substance being eminently susceptible to
friction.

12. The experiment with the conjugate mirrors is accounted
for quite as fairly by the concentration of movements of the
air in the focus, and consequently by the mechanical effects
resulting therefrom.

13. Lambert has further shown in experiments carried out on
behalf of the Commission des substances explosives that in the case
of the explosion of dynamite cartridges when produced in cast-
iron tubes of large diameter, there did not appear to be any
difference as far as regards detonations caused by influence
between the nodal and internodal parts of the tube.

14. Being anxious to clear up the question altogether, by
eliminating the influence of the supports and the diversity of
cohesion and of the physical structure of solid explosive sub-
stances, the author has undertaken special tests on the chemical
stability of matter in sonorous vibration. A summary of the
result will now be given.

§ 3. CHEMICAL STABILITY OF MATTER IN SONOROUS VIBRATION.

1. A large number of chemical transformations are now
attributed to the energy of ethereal matter, animated by these
vibratory and other movements which produce calorific, luminous,
and electric phenomena. This energy, when communicated to
ponderable matter, produces therein decompositions and combi-
nations. Is it the same with the ordinary vibrations of ponder-
able matter — that is to say, with sonorous vibrations which are
transmitted according to the laws of acoustics ? The question
is a very interesting one, and touches especially on the study of
explosive substances.

The ingenious experiments above recorded have been published
by Noble and Abel, as well as by Champion and Pellett, and
many authorities admit that explosive bodies may detonate
under the influence of certain musical notes, which would
cause them to vibrate in unison. However seductive the theory
may be, the results obtained so far do not, however, establish it
beyond dispute. Explosions of dynamite and gun-cotton by

SONOKOUS VIBRATIONS.

83

influence are explained more simply, as has been said above, by
the direct effect of the shock propagated by gases at short
distances, beyond which they do not propagate themselves in
any way. As to nitrogen iodide, which is the subject of the

principal observations relative to explosions by resonance, it is
a powder so sensitive to friction that it may be asked whether
its detonation does not take place by shock and by the friction
of the supports, the real seat of resonance in unison.

G2

84

EXPLOSION BY INFLUENCE.

2. The author deemed it expedient to make fresh researches
with gases and with liquids, which substances are more suitable
for propagating the vibratory movement, properly so called,
than a powder. Substances were selected decomposable with

liberation of heat, so
as to lessen the im-
portance of the part
played by the vibratory
movement, in propa-
gating reaction with-
out compelling it to
do all its work in virtue
of its own energy.
Finally, experiments
were made on un-
stable bodies, and even
during a state of con-
tinuous decomposition
which it was merely a
question of accelerat-
ing; these apparently
are the most favour-
able conditions. The
whole question was to
make the substance
resound into chemical
transformation. The
trials were carried out
by two processes which
correspond to vibra-
tions of very unequal
rapidity, namely : —

1st. By means of a
large horizontal tuning
fork moved by an elec-
tric interruptor, and
one of the arms of
which was loaded with
a bottle of 250 cms.
capacity, containing
the gas or liquid, the
other arm bearing an
equivalent weight. The
effective vibration of
the bottle has been verified, as also that of the liquid, otherwise
manifested by ordinary optical appearances. This arrangement
has supplied about 100 simple vibrations per second (Fig. 7).
2nd. By means of a large horizontal glass tube sealed at both

OZONE, ARSENIURETTED HYDROGEN. 85

ends, holding about 400 c.c., 60 cms. long and 3 cms. wide,
placed in longitudinal vibration by the friction of a horizontal
wheel provided with a moist piece of felt. This very simple
appliance, which Koenig has arranged, produced, during experi-
ments on ozone, 7200 simple vibrations per second, according
to observations taken by this expert (Fig. 8).

The sharpness of this note is almost intolerable.

The following are the results observed with ozone, arseniu-
retted hydrogen, and sulphuric acid in the presence of ethylene,
oxygenated water, and persulphuric acid.

3. Ozone. — The oxygen used contained such proportions of
ozone as 58 mgrms. per litre, a degree easily obtainable with the
author's appliances. With the tuning fork (100 vibrations), a
state of vibration having been maintained for an hour and a
half, the amount of ozone in the gas remained constant, both
with dry ozone and with ozone mixed with 10 c.c. of water.
This latter did not either lower the degree of the ozone or
supply oxygenated water.1

With the tube and wheel (7200 vibrations), the state of vibra-
tion being maintained for half an hour, the degree of dry gas
did not vary. The absorption of the ozone was determined
subsequently by standard solution of arsenious acid : the dimi-
nution in the strength of the latter was found equivalent to
171 div. of permanganate ; while this diminution was precisely
171 in an equal volume of the same gas analysed previous to
the test.

Now, ozone is a gas which is transformed into ordinary
oxygen with liberation of heat ( — 14,800 cal. for Oz. =
24 grms.), and it became transformed spontaneously in a slow
and continuous manner, passing from 53 mgrms. to 29 mgrms.
in 24 hours, when it was left to itself in the conditions above
given. Nevertheless, it may be seen that its transformation
was not accelerated by a movement which caused it to vibrate
7200 times per second for half an hour. Its spontaneous
decomposition could not therefore be attributed to these
sonorous vibrations which constantly traverse all bodies in
nature.

Such an absence of reaction is not, on the other hand,
explicable by an inverse influence, for a similar tube filled with
pure oxygen did not modify the strength of the arsenious
solution after similar vibration and for a similar space of time.

4. Arseniuretted Hydrogen. — A similar vibratory movement
communicated to a tube filled with this gas, and afterwards
sealed, did not modify it; nevertheless, in the space of 24
hours, the tube began to be covered with a coating of metallic

1 In these experiments it will be well to guard against the alkalinity of
glass, which will rapidly destroy the ozone. When using pulverised glass
one is specially exposed to this accident.

86 EXPLOSION BY INFLUENCE.

arsenic, as a tube does which is filled with the same gas and
which has not undergone any vibration. This gas reduces itself
into its elements, liberating, according to Ogier, + 36,700 cal.,
which explains its instability. We see, therefore, that it is not
increased by the sonorous vibrations.

5. Ethylene and Sulphuric Acid. — The author endeavoured to
accelerate the slow combination of these two bodies, which is so
easily effected under the influence of continuous agitation and by
the concurrence of shocks produced by a mass of mercury, by
having recourse to the vibratory movement. This slow com-
bination is exothermal.

A bottle of 240 c.c. containing pure ethylene, and also 5
c.c. to 6 c.c. of sulphuric acid and mercury, has been set in
vibration by a tuning fork (100 vibrations per second) ; the acid
vibrated and was pulverised on the surface ; yet at the end of
half an hour the absorption of gas was slight, and very nearly
the same as in a similar bottle kept immovable in a distant room.

6. Oxygenated Water. — 10 cc. of a solution containing
9 '3 mgrms. of active oxygen, placed in a bottle of 250 c.c.
capacity, are not altered in degree by the effect of the movement
of the tuning fork (100 vibrations per second) kept up for half
an hour. Yet the liquid actually vibrated and lost 0'9 mgrms.
of oxygen every 24 hours ; 10 c.c. of a solution containing
6 '3 mgrms. of active oxygen set in vibration (7200 vibrations)
in a tube of 4 c.c. full of air for half an hour gave afterwards
6*25 mgrms.

7. Persulphuric Acid. — Same results with the tuning fork
(100 vibrations); initial degree 13 mgrms., final degree 12'6
mgrms. With the tube (7200 vibrations), initial degree 3*6
mgrms., final degree 2*8 mgrms. The difference here appears
slightly to exceed the rapidity of spontaneous decomposition,
this rapidity being greater than with oxygenated water, but it
scarcely ever exceeds the limit of error.

The results observed with these liquids merit all the more
attention since it has been possible to assimilate these systems
a priori to the liquids containing oxygen in a state of super-
saturated solution, a solution which agitation, and particularly a
vibratory movement, will reduce to its normal state. In fact,
the foregoing liquids will certainly hold a certain quantity of
oxygen in this state, as may be easily proved ; but this amount
of oxygen does not act either on the permanganate or on the
potassium iodide employed, and it should be studied apart. As
a matter of fact, it does not intervene here in any equilibrium
of dissociation capable of being influenced by the separation of
the oxygen and the oxygenated water. It would doubtless be
otherwise in a system in a state of dissociation, and the
equilibrium of which would be maintained by the presence of a
gas actually dissolved; but then it would no longer be a question

SONOKOUS AND EXPLOSIVE WAVES. 87

of a direct influence of the vibratory movement on chemical
transformation. The tests made with gases such as ozone and
arseniuretted hydrogen are not subject to this complication;
they tend to do away with the hypothesis of a direct influence
of sonorous vibrations, even when very rapid, of the gaseous
particles on their chemical transformation.

8. It has been said that there is among the incessant and reci-
procal shocks of gaseous particles, when in motion in an enclosed
space, a certain number which are susceptible of raising the
particles which undergo them to very high temperatures. If it
were really so, a mixture of oxygen and hydrogen elements,
which combine towards 6500°, would become gradually trans-
formed into water, ammonia gas, decomposable at about 800°,
would slowly change into nitrogen and hydrogen, etc. The author
never observed anything like this in these gaseous systems
preserved for a period of ten years. If this effect does not
take place, it is probably due to the loss of energy in each
gaseous particle regarded individually, and even its total
energy remains comprised within certain limits.

9. In fine, matter is stable under the influence of sonorous
vibrations, whereas it transforms itself under the influence of
ethereal vibrations. This diversity in the mode of action of
two kinds of vibrations is not surprising if we consider to what
extent the sharpest sonorous vibrations are incomparably slower
than luminous or calorific vibrations.

10. Yet there appears little doubt that the propagation of
explosion by influence is caused by virtue of an undulatory
movement; a complex movement of a chemical and physical
order in the explosive substance which is transformed, whereas
it is purely physical in intermediate substances whose nature is
not changed. What also distinguishes this kind of movement
from sonorous vibrations, properly so called, is the extreme
intensity, that is to say, the greatness of the energy which it
transmits. It is thus that the explosive wave propagates itself
in the substance which explodes, not by reason of a single shock,
the energy of which would become weaker as it propagates
itself, but by reason of a series of similar shocks incessantly
reproduced, and which, as they continue, regenerate the energy
throughout the wave. On the other hand the propagation by
air or by supports is effected solely by reason of the energy of
the last shock communicated by the explosive substance, an
energy which is no longer regenerated and which rapidly
weakens by distance.

The explosive substance does not detonate because it transmits
the movement, but, on the contrary, because it stops it, and
because it transforms its mechanical energy on the spot into
calorific energy capable of suddenly raising the temperature of
the substance up to a degree which causes its decomposition.

CHAPTER VII.

THE EXPLOSIVE WAVE.

§ 1, GENERAL CHARACTERISTICS.1

1. THE study of the various modes of decomposition of
explosive substances, and especially that of detonation as com-
pared with combustion, and of explosions by influence, leads to
the admission of the existence of a wave motion peculiar to and
characteristic of explosive phenomena; this is the explosive
wave. It will be more accurately and completely denned by
showing how it is propagated in gaseous systems. The results
of the experiments which the author undertook in conjunction
with M. Vieille led to the examination of the rate of propagation
of the explosion in gases, the physical constitution of which
gives to these researches a peculiar theoretical interest. In the
experiments the conditions of the phenomena, the pressure of
the gases, their nature and relative proportion, and the form,
dimensions and nature of the vessels in which they are con-
tained, were varied. They confirmed the existence of a new
kind of wave motion of a compound nature, i.e. produced by
a certain concordance between the physical and chemical
impulses in the matter undergoing transformation. The wave
motion once produced is then propagated from layer to layer
throughout the whole mass, in accordance with the successive im-
pulses of the gaseous molecules brought to a more intense state
of vibration by the heat given off in their combination and
transformed with but very slight displacement of their original
position. Similar phenomena may be developed in explosive
solids and liquids.

Such effects are comparable to those of a sound wave, but
with this important difference, that the sound wave is transmitted
onwards by degrees with little active energy, a very small excess
of pressure, and with a velocity which depends solely on the

1 " Comptes Rendus des stances de l'Acade*mie des Sciences," torn, xciii.
p. 21 ; torn. xciv. pp. 101, 149, 822 ; torn. xcv. pp. 151 and 199.

CHAKACTEKISTICS OF THE WAVE. 89

physical constitution of the vibrating medium, this velocity
being the same for all kinds of vibrations. But, in the case of
the explosive wave, it is the change of chemical constitution
which is propagated communicating to the moving system
enormous energy and considerable excess of pressure. The
velocity of the explosive wave is also much greater than that
of sound waves transmitted through the same medium.

The explosive phenomenon is not reproduced periodically, it
gives rise to one single characteristic wave, whereas the
phenomenon of sound is generated by a periodical succession of
waves resembling one another.

The characteristics of this new wave are —

(1) It is propagated uniformly, as shown in the experiments
made with oxyhydric, oxycarbonic, and oxycyanic mixtures,
which were made successively in tubes of lead, gutta-percha
and glass, with lengths varying from 40 to 30 and 20 metres.

It is certain that disturbances are produced near the
extremities of the tubes. However, they do not extend far
under the conditions of the experiments ; in fact, the experiments
made with the tube closed, open at either or both ends, gave
the same velocity, which remained the same for a given length.

(2) The velocity of the explosive wave depends essentially on
the nature of the explosive compound, and not on the composi-
tion of the tube containing it (lead, gutta-percha).

(3) The influence of the diameter of the tube on the velocity
of the wave is not appreciable between diameters of 5 mms.
and 15 mms. It is, however, manifest in a capillary tube,
but the diminution, even in this- extreme case (2390 metres
instead of 2840 metres), is not excessive. In short, the velocity
depends less and less on the diameter in proportion as the
increase of the latter leaves more liberty to the individual
movements of the gaseous particles and diminishes the friction
against the sides of the tube.

These conclusions are in accordance with those of M.
Regnault on the velocity of the sound wave in tubes.1

(4) The velocity of the explosive wave is independent of
pressure, between the limits 1 and 3, as referred to the pressure
of the atmosphere. This is a fundamental property, for it
establishes the fact that the rate of propagation of the explosive
wave is governed by the same general laws as the velocity of
sound.

(5) The theoretical relation which exists between the velocity
of the explosive wave and the chemical nature of the gas which
transmits it is more difficult to establish, this velocity depend-
ing on the temperatures, and these not being the same in the
combustion of two different systems.

The inequality of the temperatures results from the unequal

1 " Me*moires de 1'Acade'mie des Sciences," torn, xxxvii. p. 456.

90 THE EXPLOSIVE WAVE.

magnitude of the quantities of heat ; for instance, 68,200 cal.
for CO + 0 ; 59,000 cal. for H2 + 0, supposing the water to be
in the gaseous form ; it also results, for the same quantity of
heat, from the inequality of the specific heats. The calculation
of these temperatures remains doubtful, on account of dissocia-
tion and uncertainties surrounding the value of specific heats at
high temperatures.

An idea of the theoretical relation that regulates the velocity
of the explosive wave may be formed, however, if it be noted
that the total energy of the gas, at the moment of explosion,
depends on its initial temperature, and on the heat given off
during the combination itself. These two data determine the
absolute temperature of the system, which is in proportion to
the energy of translation (Jrav2) of the gaseous molecules.
That is to say, the excess of energy communicated to the mole-
cules by the act of the chemical combination is simply the heat
given off in the reaction ; the pressure exercised by the molecules
on the sides of the vessels is the immediate translation of it,
according to the most recent theories.

Thus a point is reached where mechanical notions and
thermal notions tend to intermingle.

To formulate this, the rate of translation of the molecules at
the moment of combination is proportional, according to the
relation of the energy, to the square root of the ratio of the
absolute temperature T, to the density of the gas as compared
with air, or, as M. Clausius expresses it,

0 = 29-354 metres \/-
P

In reality, the physical notion of the temperature T does not
enter into this estimation of the velocity, and the formula
simply expresses the fact that the translating energy of the
molecules of the gaseous system produced by the reaction, and
containing all the heat developed by the latter, is proportional
to the energy of translation of the same gaseous system, con-
taining only the heat which it retains at zero.

This formula has been verified, approximately at least, for
a score of gaseous compounds, differing greatly in their com-
position (as described hereafter).

2. Thus it seems, that in the act of explosion, a certain
number of gaseous molecules amongst those forming the portion
that is first ignited, are hurled forward with the velocity corre-
sponding to the maximum temperature developed by the
chemical combination, the shock which they impart determines
the propagation of this combination into the next section, and
the movement is reproduced from section to section with a
velocity if not identical with, at least comparable to, that of the
molecules themselves.

EXPERIMENTAL ARRANGEMENTS.

91

The transmission of the energy, under these conditions of
extreme rapidity of action, is perhaps effected with greater
facility between gaseous molecules of the same nature, in virtue
of a kind of unison causing similar movements, than between
the molecules of gas and the enclosing vessel.

The action is not the same, as will be shown, in cases where
the system in ignition has time to lose a portion of its heat,
which is communicated to foreign gases or to bodies in the
vicinity not capable of undergoing the same chemical trans-
formation.

§ 2. EXPEEIMENTAL AERANGEMENTS.

1. The mode of procedure adopted in this study is very
simple. It consists,

(1) In filling with a detonating mixture under a given pressure,
a tube of great length (about 40 metres, Figs. 9 and 10).

(2) In effecting the ignition at one of the extremities, by means
of an electric spark (Fig. 11).

(3) In interrupting, by means of the flame itself, two electric

Fig. 9.— Tube with its interrupters.

currents, placed at certain points in the tube, the interval
between which is exactly defined by two couplings which
connect the consecutive portions of the tube (Figs. 13 and 14).
The currents are transmitted along very narrow strips of tin
(Fig. 12), gummed upon paper, and held by the couplings
between the two insulating discs of leather, which have a hole
in the centre, so as to establish the complete continuity of the
bore. These strips are arranged normally to the direction of
the flame.

A grain (about '010 of a gramme) of mercury fulminate
exploding on contact with the flame, destroys the strip and
interrupts the current.

Potassium picrate has also been used to produce the same
effect.

The gaseous compound is ignited by means of an electric spark,
either at the beginning of the tube or at some given point.

92

THE EXPLOSIVE WAVE.

2. These arrangements will now be described in detail.

The tube has sometimes been laid in a single horizontal

straight line, and sometimes in a succession of parallel rows, as

shown in Fig. 9. The tube is represented as fixed upon a

vertical wooden frame. It is provided with two terminal taps,

A and B, and an intermediate
interrupter, C.

Fig. 10 represents one of
the terminal taps without
any additional mechanism.

Fig. 11 represents a tap
with a lateral tube enclosing
an insulated metallic wire.

Fig. 10.— Terminal Tap.

The spark is made to flash between the wire and the metallic
casing of this pipe.

Fig. 12 shows the arrangement of one of the strips to be
broken by the explosion ; s s is the strip of tin, p p is the slip

of paper on which
the tin is glued.

The tin is ex-
posed at the point
V of the tube T,
which is shown in
section.

The grain of ful-
minate is placed
at i.

Fig. 13 repre-
sents the section of

Fig. 11. — Tap and Apparatus for igniting by
Electricity.

the coupling at right angles to the axis of the tube.

The coupling is marked C C C C. It is formed of four semi-
circular pieces facing each other in pairs, two only being shown
in the figure ; they are clamped together and round the tube T,
by means of the screws E E.

Fig. 14 represents a section following the axis of the tube.

The tube T T T T shown in this figure is not the tube of
gutta-percha itself, but a brass tube of the same section, on
which the gutta-percha tube is fitted, either on one side
only, or on both sides at once, as in Fig. 9. This arrange-
ment is necessary for clamping the coupling and fixing the
interrupters.

C C C C are the four parts of the coupling, the screws not
being shown in order not to complicate the figure. The channel,
V V, serves for the passage of the gas.

The strip of tin, s s, is held in position by small metallic
supports, r r, on which the wires conveying the electric current
are fixed.

Between the portions of the coupling, C C, are the two discs

EXPEKIMENTAL ARRANGEMENTS.

93

of insulating leather, shown here only in section, their projection
being given in Fig. 12 (see letter T).

The grain of fulminate is always at i.

The time that elapsed between the two interruptions was
estimated by means of the Le Boulenge chronograph, this
instrument being capable of measuring to the ^o~o~<T(J °f a second.

Fig. 12.— Strip of tin.

Fig. 13.— Coupling section at right angles to the
axis of the tube.

The chronograph (Figs. 15 and 16) consists of two funda-
mental parts.

a. The chronometer T (Fig. 15), a long cylindrical rod sus-
pended vertically, provided with zinc casing tubes, E, and held
by magnetic attraction to the extremity of an electro-magnet,
M, through which passes the first current destined to be
broken.

1. The registering apparatus, T (Fig. 16), a similar cylinder

94

THE EXPLOSIVE WAVE.

held by the electro-magnet, M, through which passes the second

current also destined to be interrupted.

A catch, C, consists of a knife edge (a circular milled head of

hardened cast steel), mounted upon a spring which may be held

firm, or tightened by the handle of a lever.

The chronometer, on its circuit being broken, becomes

detached and falls vertically freely; the second circuit being

next broken, the registering apparatus falls in its turn, comes in

contact with the free extremity of the lever and disengages the

catch, the knife is projected forward, strikes the chronometer

in its course, and imprints upon its
casing a mark, the position of which
enables the rapidity of the pheno-
menon to be calculated. The details
of this calculation, with corrections,
will be found in the "Traite sur
la poudre," etc., traduit et augmente
par Desortiaux, pp. 538 and 542.
1878 (Dunod).

This method was found preferable
to the registration by mechanical
processes; the latter are subject to
irregularities which are of great im-
portance in such rapid phenomena.

The use of too short tubes for con-
taining the gases was avoided, as this
would exaggerate errors and expose
the experiments to those well-known
disturbances which arise in the
vicinity of the source of the waves.
Eeference will be made to this point
as it is one of great interest, and it will be shown at
ne time that the variation in pressure of the gases is

propagated with exactly the same rapidity as the ignition of

the detonators.

Fig. 14. — Section following the
axis of the tube.

again,
the same

§ 3. GENERAL CONDITIONS OF THE EXPERIMENTS.

Our experiments had reference,

1. To the arrangement of the tube ;

2. To its .composition ;

3. To its characteristics, whether open or closed ;

4. To its length ;

5., To the initial pressure of the gaseous compound ;

6. To the composition of this compound which was varied
sometimes by introducing an inert gas, and sometimes by modi-
fying the nature of the combustible gas.

Arrangement of the tube. — The first experiments were made

HYDROGEN AND OXYGEN.

95

with rectilinear and horizontal leaden tube 42*45 metres long 1
and '005 of a metre in diameter.

It was filled with an electrolytic mixture of hydrogen and
oxygen, under atmospheric pressure. After each experiment,
the tube was dried by causing
a current of dry air to circulate
through it for several hours.

The following table gives all
the experiments. The extreme
results have not been eliminated
from it as is sometimes done.

Time observed Velocity per

in seconds.

second.

(1)

0-014633

... 2901-0 metres.

(2)

0-014597

... 2908-1 „

(3)

0-013914

... 3050-9 „

(4)

0-015047

... 2821-2 „

(5)

0-015816

... 2675-5 „

(6)

0-014752

... 2877-6 „

(7)

0-014782

... 2871-8 „

(8)

0-015253

... 2783-1 „

Mean 0-014860 2861-1 „

The mean variation in one ex-
periment amounts to 79 metres,
the maximum variation to -f- 190
metres and - 186 metres, corre-
sponding to intervals of time
equal to + 0-00095, or at the
maximum nearly joW °f a
second, the mean error being
half this quantity. With the
oxyhydrogen mixture, the mean
length measured on the rod of
the chronographs is equal to
•0448 metre, which figures give
a clearer idea of the degree of
exactness attained in measure-
ments of this order. With the
mixture of carbonic oxide and
oxygen, this length amounted
to 107 metre.

The mean error in the experi-
ments is ten times as great as
that recorded by the chrono-
graph. This arises, not from
the instrument itself, but from
the unequal delays occurring in
the process of interruption employed. It is known that such
1 All the lengths are taken between the two interruptions.

Fig. 15. — Le Boulenge Chronograph-
chronometer.

96

THE EXPLOSIVE WAVE.

errors exist in all processes of this nature, and their magnitude
must be estimated in each instance. In this case it amounted
to 2 '8 per cent, of the quantity measured, on an average, and
to 6 '6 in extreme cases.

Some trials made with a vertical tube, shorter it is true, gave
the same velocities as with the hori-
zontal tube.

The rectilinear arrangement of the
tube such as was employed at first,
required too extensive a space, and
could only be effected in the open air
and under conditions that were difficult
to maintain and vary in prolonged ex-
periments. For this reason the idea
was conceived of laying the tube in the
laboratory itself, in parallel horizontal
rows separated by bends with a con-
siderable radius of curvature ; the whole
was fixed upon a vertical frame (see
Fig. 9, p. 91).

In this operation the length of the

*ube was -'^dby •? of a metre.

bringing it to 43135 metres.
The detonation was repeated under these new conditions,
which gave for the velocity per second —

Metres.

2860-4
2712-9
2791-5

Average ... 2788-3

These figures are rather lower than in the preceding experi-
ment, but they do not fall below the mean limits of error. It
may therefore be assumed that the velocity is the same in the
bent tube as in the straight one, and the general mean, 2841
metres, will be adopted.

Composition of the tube. — The unexpected magnitude of
this velocity, which is intermediate between the velocity of
sound in the detonating gaseous compound and in the metal
constituting the tube, gave rise to some doubts. Was it really
the rate of propagation of the detonation that was being
measured, or was it not rather some particular vibratory
movement propagated by the metal, arising from the explosion,
produced at its extremity ?

It seems, however, hardly probable that a propagation of this
nature could cause the detonation of the fulminate, when we
consider the weakness of the movement thus transmitted, and
again the intervention of the leather discs T, between the metal

VELOCITIES. 97

and the strips of tin (Fig. 12). We may also mention the
absence of detonation in certain grains of fulminate which had
been slightly greased by accident, a circumstance which retarded
the heating without otherwise modifying the explosive property.
Again, when the flame is extinguished in its course, as we
observed with the capillary glass tube, the furthest registering
apparatus remains intact. We could not however feel assured
on this point until we succeeded in reproducing our experiments
and obtaining the same velocities in a tube of caoutchouc, a
substance which could not be suspected of propagating the
vibratory movement like metals. This was rendered possible
by the fact that the internal combustion of the gaseous mixture
is so rapid that it does not affect the material of which the tube
is made. This caoutchouc tube was 40109 metres long, and
several mms. thick, its external diameter being 5 mms., and it
was capable of supporting either a vacuum or an interior pressure
of several atmospheres without any appreciable deformation.
It was fixed upon the frame already described, in parallel lines
(Kg- 9).

These were the results : —

Velocity per second.

2685 metres.
2911 „
2994 ,„
2672 „

2788 „

Mean ... 2810 ,,

This mean agrees with the value, 2841, obtained with the
leaden tube, within the limits of error.

The propagation of the explosive phenomenon is thus inde-
pendent of the composition of the tube, provided that the
internal diameter remains the same.

We now come to some experiments made with a system of
glass tubes, the total length of which was 43*24 metres, but the
mean internal diameter only '0015 metre. These were capillary
tubes, each 2 metres in length, connected end to end by means of
caoutchouc tubes, the whole being fixed upon the frame before
mentioned (p. 91). The bends were made with the same glass
tubing. We found — velocity per second, 2403 and 2279 metres ;
mean, 2341 metres.

These figures are rather lower than the foregoing ones, no
doubt owing to the difference in the diameter ; the propagation
of the explosion being impeded in a capillary tube, as is the
case also with the propagation of sound. The experiments
made in glass enable us to see the propagation of the flame.
Working in darkness, we see the entire length of the tube
lighted up at the same moment, the eye not being able to
perceive the progress of the flame.

H

98 THE EXPLOSIVE WAVE.

It sometimes happens that the flame does not reach the end
of the tube, probably in consequence of the insufficient heating
of the sections in advance of the mixture in ignition. One of
the trials led to some important observations on this point. The
flame being stopped in its course, without however going out,
the aqueous vapour, condensed behind it, produced a backward
draught upon the gas, and a return of the flame was observed
very clearly towards its starting-point, this movement lasting
during a very appreciable interval of time, a second, perhaps,
for a distance of 2 metres. This shows well the difference
between the progressive combustion of the gaseous compound
and its detonation properly so called.

Diameter of the Tubes.— In order to investigate more fully
the influence of the diameter of the tubes, it was thought
advisable to make fresh measurements with a leaden tube, the
internal diameter of which was equal to 15 mms., i.e. three
times as great as the preceding one, and 30 '43 metres long.
Three experiments gave 2754, 2975, 3019 metres; mean,
2916 metres.

The experiments made with a leaden tube, the diameter of
which was equal to 5 mms., having given 2841 metres, we see
that the velocity is to all intents independent of the diameter
of the tubes, reckoning from 5 mms.

It must be noted, however, that in a capillary glass tube
(diameter 1*5 mms.), the velocity was found to be equal to
2341 metres, i.e. it was somewhat lower.

Closing of the Tube. — It may be asked is the rate of pro-
pagation of the detonation the same whether the tube be
open or closed ? It is only the latter case that strictly fulfils
the conditions of combustion within a constant volume. To
meet this question, experiments were made (with the caoutchouc
tube), leaving open first the orifice farthest from the point of
ignition ; then the one nearest to this point ; and, finally, both
at once.

Three experiments of this nature gave —

Velocity per second.

The farthest orifice only being open ... 2645 metres.
The nearest orifice only being open ... 3052 ,,
The two open together ... ... 2766 „

Mean ... 2821 „

The mean, with the same tube entirely closed, was 2810.

Thus the velocities have been found to be to all intents the
same in all four cases. We see by this that the propagation of
the detonation is so rapid that while it is taking place the gases
are not projected forward, and have not time to escape from the
tube to any appreciable extent — at least, in narrow tubes. This
is explained by the fact that the detonation proceeds more

INFLUENCE OF DETONATORS. 99

rapidly than sound in the same gases, taken at the ordinary
temperature. The condensation of the aqueous vapour, which
is effected behind the flame, is also of little importance, since
the time is too short to allow of its taking place to any
appreciable extent.

Influence of the Detonators. — Do the minute detonators, em-
ployed for interrupting the electric current of the registering
apparatus, help to regulate the propagation of the inflammation ?
In order to answer this question, it was sufficient to measure, not
the time that elapsed between the destruction of two fulminate
interrupters, placed at opposite extremities of the tube, but the
interval between the breaking of the induction current of the
coil that produced the spark at the beginning of the tube and
the ignition of the fulminate interrupter placed at the farthest
end of it.

The intervals of time observed, for a length of 40*054 metres,
were —

•012556, -012288, '012904 sees.
Mean = -012583 sees.

But there are errors in these figures, arising from delays in
the registration, delays unequal in principle, as two different
kinds of signals are involved. The difference between these
two delays was calculated by measuring the time that elapsed
between the signal of the spark and that of an interrupter
•05 metre away. This time is negative, i.e. the delay in the
signal of the spark is greater than that in the signal of the
interrupter. Three experiments gave —

•001559, -001968, -002129 sec.
Mean = -001885 sec.

This correction, added to the above experiments, gives
"014468 sees., bringing the velocity to 2770 metres per second.
The experiment made with two similar interrupters gave
2810 metres, the agreement of which result shows that the
velocity observed is independent of the detonators.

This is still more clearly shown in some experiments here-
after described (p. 108), in which the propagation of the pressures
had been registered by effecting the initial ignition by means of
an electric spark. In fact, the propagation of the pressures
starting a few centimetres from the beginning of the tube
proceeds with a velocity of about 2700 metres, a rate in accord-
ance with the results above mentioned.

Length of the Tube. — It now remained to ascertain whether
the propagation of the explosion takes place uniformly in
the tubes. This is clearly proved in the following experi-
ments.

With the caoutchouc tube 5 mms. in diameter these results
were obtained —

H2

100

THE EXPLOSIVE WAVE.

Distance of interrupters.

40-109 metres
29-982 *

Mixture (H2 + 0).

Velocity.

2810
/2692\
\2716/

Mean.

2704

Distance of interrupters.
Metres.

40-059 ...
299821 ...
20-0922 .

Mixture (CO + 0).

Velocity.

Metres.

1096
1140
1187

Metres.

1068
1121
1183

Metres.

1104

Mean.

Metres.

1089
1130
1185

Again with the glass tube 1*5 mms. in diameter —
Mixture (H2 + 0).

Distance of interrupters.

43-340 metres 3
20-944

Velocity.

2341 metres.
2433

In all these measurements, the differences between the
velocities measured with unequal lengths do not exceed the
limits of error.

Pressure. — The pressure was varied in the ratio of about
1 to 3. The caoutchouc tube (40*054 metres long) was employed,
and with three different gaseous mixtures.

Mixture (H2 + 0).

Pressure (expressed by the height
of a column of mercury).

0-560 metre
0-760 „
1-260 „
1-580

Velocity.

2763 metres.
2800 „
2776 „
2744

Mixture (CO + 0).

Pressure.

0-570 metre
0-760 „
0-854 „

1-560

Velocity.

1120 metres.
1089 „
1072 „

/1140\

\1124J »

1152

1 A section of tubing 13 metres long filled with the same compound was
afterwards added on ; i.e. the interrupter was placed in the path of the flame,
and not at the extremity of the tube.

2 The interrupter was placed half-way along the tube, in the path of the
flame.

3 This time the interrupter was placed at the end of the tube.

SPECIFIC VELOCITY. 101

Mixture of cyanogen and oxygen (CN + 02)

Pressure. Velocity.

0-388 metre ......... 2171-4 metres.

0-758 „ ...

0-878 „

Same conclusions.

Thus, as far as the experiments went, the rate of propagation
of the detonation, either with the mixture of hydrogen and
oxygen, or with the mixture of carbonic oxide and oxygen, is
practically independent of the pressure, like the velocity of
sound and the rate of translation of the gaseous molecules,
which are analogous phenomena.

§ 4. SPECIFIC VELOCITY OF THE EXPLOSIVE WAVE.

1. We have now established the fact that the explosive wave
is propagated uniformly and that its velocity is independent
of the pressure, and also of the composition and diameter of the
tubes, beyond a certain limit. Thus, this velocity constitutes
for each inflammable compound, a true specific constant, the
knowledge of which is of great interest, from the point of view
of the theory of the movements of gases, and also from that of
the employment of explosive substances. For this reason it
was thought expedient to go more deeply into the study of this
question, extending our operations to a large number of mixtures
differing greatly in their composition.

2. Each experiment was repeated two or three times ; it was
generally performed in the caoutchouc tube, 40 metres long,
with an internal diameter of 5 mms., and of great thickness (as
already described, p. 97). The results obtained are shown in
five tables, containing the most remarkable cases. In these tables,
the first column gives the composition of the initial mixture ;
the second, the density of the products of combusition p, as
compared with that of air, taken as unit ; the third, the number,
N, of molecular volumes of the elements (supposing them to be
gaseous) entering into reaction, the volume being —

TT

N [ 22-32 litres X - — X (1 X at)] ;

the fourth column gives the heat, Q, given off by the reaction,
the water being supposed to be in a gaseous form * ; the fifth
column gives the square root of this quantity, VQ 5 the sixth

contains the quotients — — —' 6*8 being the constant of the

1 This quantity was measured near zero ; but it would be little different at
zero in the cases considered here, especially if the specific heat of the com-
pound were estimated as the sum of the specific heats of its elements.

102 THE EXPLOSIVE WAVE.

specific heats of the elements under a constant pressure ; this is
the theoretical temperature, T, of the reaction ; the seventh
gives the theoretical values, 0, of the mean rate of translation
per second of the gaseous molecules constituting the products of
the combustion, a rate calculated for the temperature T, accord-
ing to the formula of Clausius (see p. 90).

P

This is a velocity which we propose to compare with the ex-
perimental velocity of the explosive wave, U, which is shown in
the eighth column.

3. The temperature, T, is here calculated from the specific
heats of the elements under a constant pressure. The results
thus obtained, agree in general with the observations (columns
7 and 8) ; they agree far better than if the calculation were
made from the specific heats at a constant volume, although
the latter method would, on first thought, seem the more
plausible.

We may account for the intervention of the specific heats
under a constant pressure, if we consider that the combustion,
in passing from layer to layer, is preceded by the previous
compression of the layer of gas that it is about to transform.
From that time the combustion takes place under a constant
pressure throughout the tube. It might be thought that the
temperature, T, must be increased by the elevation of tempera-
ture produced by this previous compression. But we must
take into account the fact that the combustion of each layer
produces, at the same time as heat, the work necessary for
compressing the following layer ; i.e. it loses in this way exactly
the same quantity of heat as it has gained by its own com-
pression. In short, as regards elevation of temperature, the
effect is the same as if we were working under a constant
pressure. The agreement ' between the figures calculated and
the numbers observed confirms this view of the phenomena.1

The fact is, the physical conception of the temperature, T,
does not enter into this estimate of the velocity, and the calcu-
lation simply shows that the energy of translation of the mole-
cules of the gaseous system produced by the reaction, and

1 It is assumed here that the specific heat of a compound gas under a
constant p'ressure is the sum of the specific heats of its elements, which
assumption is, in reality, only true when the volume is constant, or, for the
two specific heats, in the case of gases formed without condensation. But
this underestimate in the case of the specific heat under a constant pressure,
of gases formed with condensation, is to a certain extent compensated by the
fact that the specific heat of these gases rises with the temperature : this is
proved by the study of the specific heats of carbonic acid, nitrogen monoxide,
etc. (" Essai de Mdcanique Chimique," torn. i. p. 440). This hypothesis may,
therefore, be admitted for a first approximation.

THEOEETICAL AND FOUND VELOCITIES.

retaining all the heat thereby developed, is proportional to the
energy of translation of the same gaseous system, containing
only the heat that it retains at zero.
4. Here follow the series of tables : —

TABLE I. — ONE COMBUSTIBLE GAS ASSOCIATED WITH OXYGEN.

Number

.of 1

Nature of the
mixture.

Density
of the
pro-
ducts.

mole-
cular
volumes
of the
ele-

Heat of
1 combus-
tion (water
gaseous).

Theo-
retical
velocity.

Velocity found by
experiment.

ments.

Q T

P

N

Q

VQ

N X6-8"1

e

U (per sec.).

cal.

degrees.

metres.

metres.

Hydrogen ....

0-622

1-5

59,000

243

5780

2831

2810 (p. 97)

H2+0

Carbonic Oxide . .

1*529

1-5

68,200

261

6700

1941

1089 (p. 100)

2CO + O2
Acetylene or Ethine .

1-227

4-5

308,100

555

10,070

2660

(I&82-5

C2H2 + 05

f918fi-ft^

Ethylene ....
C2H4 + O6

1-075

6-0

321,400

567

7880

2517

[2232-4/2209*5

Ethane ....

0-985

7-5

359,300

598

7050

2483

i9<uJ-4J2363

C2H6 + 07

'9Q1Q <>\

Methane . . .

0-924

4-5

193,500

440

6320

2427

22fiO-Ol

2 CH4 + 08

£t£i\j\J \J 1

Cyanogen . . .

1-343

4-0

262,500

512

9650

2490

2195 (p. 101)

According to the figures in this table, the theoretical velocity
is very near the velocity found by experiment for hydrogen.

For the hydrocarbons and for cyanogen, this theoretical
velocity is rather too high, the discrepancies being comprised
between five and twelve hundredths, i.e. the formula keeps
within an approximate value.

For carbonic oxide the discrepancy is much greater, exceed-
ing forty hundredths ; thus the formula is not applicable to this
. gas (see p. 107).

It will be seen that it remains approximate, even for gases
that are formed with absorption of heat, and that give rise, upon
their formation, to the highest temperatures of combustion, such
as cyanogen and acetylene.

It is also approximate for very different ratios of volume
between the combustible gases and the oxygen, such as 2 : 5, 6,
7, 8, in the series of the hydrocarbons, and 2 : 1 for the
hydrogen.

Lastly, it is approximate for very unequal ratios of condensa-
tion in the combination, such as a condensation of a third
(hydrogen), of a seventh (acetylene), or the absence of all con-
densation (ethylene, methane, cyanogen) ; or even an expansion
(ethane). In the calculation of these volumes the water is

104

THE EXPLOSIVE WAVE.

assumed to be in the gaseous state in the hydrocarbons, a con-
dition that does not enter into the case of carbonic oxide or
cyanogen.

Thus it seems to be an established fact that the proposed
formula represents approximately the velocity of the explosive
wave for hydro-carbon gases.

5. This conclusion may be extended to the mixtures formed
with these gases and hydrogen, or even carbonic oxide, as will
be shown, the hydrogen imparting to these mixtures a law of
detonation similar to its own.

TABLE II. — Two COMBUSTIBLE GASES ASSOCIATED WITH OXYGEN.

Number

of

Nature of the
mixture.

Density
of the
pro-
ducts.

mole-
cular
volumes
of the
ele-

Heat of
combus-
tion (water
gaseous).

Theo-
retical
velocity.

Velocity found by
.experiment.

ments.

Q T

P

N

Q

V Q

e

U (per sec.).

cal.

degrees.

metres.

metres.

Carbonic oxide and

hydrogen

1-075

3

127,200

357

6230

2236

2008

CO + H2 + 02

2CO + 3H2 + 05

0-985

7-5

313,400

560

6150

2321

\2245j

Ethylene and hydro-
gen ....
C2H4 + H2 + 07

0-985

7-5

380,400

617

7460

2551

J2411-4 U417
\2422 Jz*

C2H4 + 2H2 + 08

0-924

9

439,400

663

7180

2588

(2671 U,7q
\2487-5 /zo/y

(2184)

Ethane and hydrogen

0-924

9

418,300

647

6830

2522

{2227 [2250 J

C2H6 + H2 + 08

2339

1 Different preparations.
TABLE III. — ONE COMBUSTIBLE GAS ASSOCIATED WITH A COMPOUND COMBUSTIVE GAS.

Number

of

Nature of the
mixture.

Density
of the
pro-
ducts.

mole-
cular
volumes
of the
ele-

Heat of
combus-
tion (water
gaseous).

Theo-
retical
velocity.

Velocity found by
experiment.

ments.

<j m

P

N

Q

V Q

NX6 8=T

e

U (per sec.).

cal.

degrees.

metres.

metres.

Nitrogen monoxide

and hydrogen .

0-796

2-5

79,600

281

4680

2250

|2^J2284

N2O + H2
— and carbonic oxide
N20 + CO

1-250

2-5

88,800

298

5220

1897

)H02-5K106,
\1110-6/110bt

— and cyanogen

1-131

8

345,000

587

6340

2198

2035-5

2N2O + CN

Nitrogen dioxide and

1-194

8

349,000

591

8550

2485

The detonation

cyanogen

was not pro-

N202 + CN

pagated in

the tube.

ISOMERIC MIXTURES.

105

With the nitrogen monoxide, the velocity found by experi-
ment is near the theoretical value for compounds containing
hydrogen or cyanogen. With carbonic oxide we find the same
anomaly as with oxygen.

6. One of the most interesting examples, and one most strongly
confirming the theory, is the case of the isomeric compounds,
i.e. those in which the composition of the final system is the
same. In fact, in these cases the influence of the individual
nature of the combustible gases, and even that of the combustive
ones, is eliminated.

TABLE IV. — ISOMERIC MIXTURES.

Number

of

Nature of the
mixture.

Density
of the
pro-
ducts.

mole-
cular
volumes
of the
ele-

Heat of
combus-
tion (water
gaseous).

Theo-
retical
velocity.

Velocity found by
experiment.

ments.

Q T

P

N

Q

V Q

Nx6-8~J

6

U (per sec.).

First Group. — Hydro-Carbon Gases and pure Oxygen.

(1) Methane and Isomeric Mixtures.

cal.

degrees.

metres.

metres.

2(CH4 + 04). . .

0-924

9

387,000

622

6320

2427

2287

C2H6 + H2 + 08 .
C2H4 + 2H2 + 08 .

0-924
0-924

9
9

418,300
439,400

647
663

6830
7180

2522
2588

2250
2579

C2H6 + 0T

(2) Ethane and Isomeric Mixtures.

0-985 I 7-5 I 359,300 I 598 I 7050 I 2483 1 2363
0-985 I 7-5 | 380,400 | 617 | 7460 [ 2551 | 2417

Second Group. — Hydro-carbon Gases compared with
Oxy-carbon Mixtures.

(3) Ethylene and Isomeric Mixtures.

C2H4 + 06 . . .
2(00 -f- H2 + 02) .

1-075
1-075

6
6

321,400
254,400

567
504

7880
6230

2517
2236

2219-5
2208

C2H6 + 07 . . .
2CO + 3H2 + 05 .

(

0-985
0-985

1) Eth
7-5
7-5

ane and It
359,300
313,400

iomeri
598
560

c Mixture
7050
6150

8.

2483
2321

2363
2170

2CN + N2 + 04. .
2(00 + N2 + O) .

0

1-250
1-250

)) Cyai
5
5

aogen mix
262,500
136,400

edwit
512
370

h Nitroge
7720
4010

n and Isomeric Mixtures.

9qq4 |/2116-0\204Q.fi
2334 \1971-2/m
1661 | 1000 ?l

Third Group. — Compound Oxygen yielding Gases, compared
with Mixtures formed with pure Oxygen.

H2 -f N20 .
H2 + N2 + 0 . .

<<

0-796
0-796

J)Hyd
25
2-5

rogen.
79,600 1 281
59,000 | 243

4680
3470

2250
1935

2284
2121

CO + N2O

CO + N2 + 0 . .

c

1-250
1250

r) Cart
2-5
2-5

>onie Oxid
88,800
68,200

e.
298
261

5220
4010

1897
1661

1106-5
1000 ? l

1 The detonation is not usually propagated. However, this figure was found among
the author's notes without other detail.

106

THE EXPLOSIVE WAVE.

These compounds satisfy the law fairly closely, with the ex-
ception of the carbonic oxide. The isomeric compounds have
generally approximate velocities. They enable us to appreciate
more exactly the influence of the heat given off, Q, eliminating
that of the density, the specific heat of the products, and even
of individual composition, which are the same.

Thus, in order to make a comparison, it is merely necessary
to divide the velocities found by V Q!

Thus—

3-69

1st system
2nd „
3rd „
4th „
5th „
6th „
7th

•

3-68
3-95
3-91
3-93
3-99
8-13
3-67

3-48
3-92
3-98
3-88
2-70
8-73
3-83

It will be seen in general the coincidence is still more marked,
with the exception of the fifth system, in which carbonic oxide,
which does not satisfy the general theory, is compared with
cyanogen.

We will now examine the influence of inert gases, which do
not participate in the chemical reaction.

TABLE V. — COMBUSTIBLE GASES, OXYGEN AND INERT GASES.

dumber

of

Nature of the
mixture.

Density
of the
pro-
ducts.

mole-
cular
volumes
of the
ele-

Heat of
combus-
tion (water
gaseous).

Theo-
retical
velocity.

Velocity found by
experiment.

ments.

Q

P

K

Q

VQ

NX6-8

e

U (per sec.)

cal.

degrees.

metres.

metres.

Hydrogen and nitro-

gen

0-622

1-5

59,000

243

5780

2831

2810

H2 + N2 + 0

0796

2-5

59,000

243

3470

1935

2121

0-846

3-33

59,000

243

2610

1820

1439

0-267H -I- 0-733 air

0-868

3-80

59,000

243

2287

1505

1201

0-233H + 0-768 air 0-885

4-27

59,000

243

2042

1409

1205

0-217H + 0-783_air

0-895

4-56

59,000

243

1903

1389

The detonation

was not pro-

pagated.

Carbonic oxide and

nitrogen

CO+ O

1-529

1-5

68,200

261

6700

1941

1089

CO + N2 + O

1-250

2-5

68,200

261

4010

1661

1000?

Propagation

doubtful.

0-3CO + 0-7 air

1-165

4-33

68,200

261

2260

1236

The detonation

was not pro-
pagated.

LIMIT OF PROPAGATION OF DETONATION.

107

TABLE V. — COMBUSTIBLE GASES, OXYGEN AND INERT GASES— (Continued).

Number

of

Density

mole-

Heat of

Theo-

Nature of the
mixture.

of the
pro-
ducts.

cular
volumes
of the
ele-

combus-
tion (water
gaseous).

retical
velocity.

Velocity found by
experiment.

P

ments.

N

Q

A/T

NX6-8~T

0

U (per sec.).

cal.

degrees.

metres.

metres.

Methane and nitrogen
CH4 + 04
CH4 + 2N2 + 04

0-923
0942

4-6
6-5

193,500
193,500

440
440

6320
4378

2427
2002

2287
1858

CH4 + 4N2 + 04

0-951

8-5

193,500

440

3347

1744

1151

1 The detonation

CH4 + 7-52N2 + O4^
methane and air /

0-958

120

193,500

440

2371

1450

was not pro-
pagated.

Cyanogen and nitro-

gen
2CN + O4

1-343

4-0

262,500

512

9650

2490

2195

2CN + N2 -f O4

1-250

5-0

262,500

512

7720

2334

2044

2CN + 2 N2 + 04

1-194

6-0

262,500

512

6340

2152

{Im-T}1203'3

2CN+4N2 + 04

M27

8-0

262,500

512

4825

1920

The detonation

was not pro-

pagated.

Detonation was not effected in a
CO + N2 + O is doubtful.

mixture richer in nitrogen. The mixture

The general relations were the same, except for the com-
pounds that border upon the limit at which the detonation
ceases to be propagated, such as the mixture of cyanogen with
twice its volume of nitrogen, that of methane with four times
its volume of nitrogen, carbonic oxide, etc. With hydrogen
and an excess of nitrogen, there was also a decided fall in the
results.

7. To sum up, the velocity of translation of the gaseous mole-
cules, preserving the whole of the energy corresponding to the
heat given off by the reaction, may be regarded as a limit
representing the maximum rate of propagation of the explosive
wave.

But this velocity is diminished by the contact of gases and
other foreign bodies ; and also when the mass ignited at the
beginning is too small and too rapidly cooled by radiation ; and
again when the elementary velocity of the chemical reaction 1
is too feeble, as seems to be the case with carbonic oxide. Under
these conditions the wave slackens, and may even stop alto-
gether, the combustion being then propagated from layer to
layer at a much slower rate. Reference will be made to this
point again.

1 " Essai de Mecanique Chimique," torn. ii. p. 14.

108

THE EXPLOSIVE WAVE.

§ 5. ON THE PERIOD OF VARIABLE CONDITION PRECEDING
DETONATION AND THE CONDITIONS OF THE ESTABLISHMENT
OF THE EXPLOSIVE WAVE.

1. It is now proposed to study the conditions of the establish-
ment of the explosive wave, and the period of variable condition
preceding this establishment, a period analogous to that which
precedes the establishment of the sound wave.

2. The following process has enabled precise measurements
to be made of the variation of the velocities during very short

intervals of time, such
as *0003 of a second.

A revolving cylinder
gives the following re-
cord:

(1) The spark that
determines the initial in-
flammation at the mouth
of the tube; the trace of
this spark is shown at e
(Fig. 18).

(2) The movement of a
very light piston, placed
at the other extremity of
the tube, in which it moves
freely. This piston is shown
in Fig. 17 in projection
upon the revolving cylinder.
The details of its construc-
tion are here shown : i.e.
the tube, the piston fur-
nished with its pencil in-
tended to trace its course

upon the cylinder, and lastly the terminal cap of the piston
tube.

In this way is recorded the time that elapses between the
two phenomena and the law of the movement of the piston
(Fig. 18).

The delays are thus avoided which might result either
from the employment of a metallic manometer or from the
propagation of the phenomena to an auxiliary vessel. Each
number gives the average of from two to five experiments,
made with electrolytic gas (Ha + 0) in a caoutchouc tube
5 mms. in diameter. We will first study the velocities,
then the corresponding pressures, and lastly the limits of
detonation.

17. — Registration of variable
velocities.

INFLUENCE OF INITIAL INFLAMMATION.

109

3. Velocities (per second).

Mean velocities

instance irom
of inflamm.'
the pis

0-020 mel
0-050
0-500
5-250
20-190
40-430

me poii
ition to
ton.

res

ii

Durations
observed.

. 0-000275 sees.
. 0-000342 „
. 0-000541 „
. 0-002108 „
. 0-007620 „
. 0-015100 „

from the
beginning.

72-72 metres
.. 146-20 „
.. 924-40 „
.. 2491-00 „
.. 2649-00 „
.. 2679-00 „

in each
interval.

72-7 metres
.. 448-0 „
.. 2261-0 „
.. 3031-0 „
.. 2710-0 „
.. 2706-0 „

Hence it is seen that the velocity increases rapidly from the
starting point to the fifth cm., from which
point the numbers obtained may be re-
garded as almost constant, at least within
the limit of the errors of the experiments,
which have a very considerable relative
value at the commencement, for such short
intervals.

The establishment of a regular system
can only be effected successfully when the
sparks that inflame the compound are
strong enough. With feeble sparks, the
period of variable condition can be greatly
prolonged : over a space of 10 metres,
mean velocities of 2126 metres and even
661 metres were thus obtained. Analogous
phenomena are observed with the other ex-
plosive compounds. Electrolytic gas mixed
with nitrogen, for example H2 + 0 + 2N,
gave a velocity of 41 -9 metres per second
in the two first cms., 1068 metres in the
consecutive sections of 5*25 metres, and
1163 metres in the consecutive sections
of 10 metres.

The influence of the initial inflammation
is in this case still more marked, the velo-
city having fallen by accident to 445 and
435 metres, without any apparent change
in the power of the initial spark ; more-
over, the nature of its product, in this case,
indicated a different mode of combustion.

These discrepancies are not, in general,
observed1 with the process of registration
based upon the employment of the ful-
minate interrupters, which tends to prove
that the fulminate, by the sudden pressures

1 Mention may here be made of an experiment in which the compound
H2 + 0 + N gave an exceptional velocity of 1564-5 metres, instead of the
normal result 2121 metres ; probably on account of the exceptional weakness
of the priming.

110 THE EXPLOSIVE WAVE.

which it develops, helps the gaseous column to take up the
detonation at once, which result it would attain later with less
regularity by the ordinary inflammation.

4. Pressures. — These are deduced from the path traced by the
piston.

For explosive gas (H2 + 0), the piston, placed at 2 cms. from
the point of ignition, is projected forward at first by a pressure
of 500 to 600 grms. per sq. cm., but this pressure falls very
quickly, until it becomes nil and even negative (on account of
the condensation of the aqueous vapour) at the end of '0005 of
a second.

At -5 of a metre from the beginning, a pressure of 1-2 kgms.
was found.

At 5*25 metres from the point of inflammation the first dis-
placement of the piston took place under a pressure of about
5 kgms. per sq. cm. ; and this pressure at the end of '00125 of
a second, was still more than 3 kgms.

Now, at this moment, the inflammation progressed 2 '7 metres
in a similar tube, according to the velocities mentioned above.

It will be seen, then, that in this part of the tube a con-
siderable gaseous column, formed of aqueous vapour, is main-
tained at a high pressure, whereas at the beginning the pressure
produced in one section by the combustion of the mixture is
almost instantaneously annulled by the condensation of the
sections in front of it; otherwise, the increase in pressures
corresponds to the increase in velocity. It was found that the
maximum of pressure developed by the mixture H2 + 0, burn-
ing in a closed vessel, is about 7 kgms. In this case the cooling
influence of the sides of the vessel may be disregarded. In
abnormal cases, when the rate of propagation falls below 2000
metres the pressure falls at the same time, which shows plainly
the correlation of the two kinds of phenomena.

5. Limits of Detonation. — It is possibly due to similar causes
that certain explosions of firedamp attain an exceptional rate
of propagation and unusual violence. When the explosive wave
is not propagated, combustion may still take place to a certain
extent.

The limit of detonation in oxyhydrogen compounds is at
about 22 per cent, of hydrogen, whereas the ordinary limit of
combustion in mixtures of hydrogen and oxygen is at about 6
per cent, of hydrogen.

As the lower limit of detonation is approached the velocity
of the wave falls considerably below the theoretical velocity
(see above). The mixtures of cyanogen and nitric oxide such
as CN -f 2 NO show some points of interest. This compound,
contained in an eudiometer, is exploded violently by a powerful
spark. When ignited with a match it burns progressively.
But, on the other hand, we did not succeed in propagating the

PROPAGATION OF EXPLOSIVE WAVE. Ill

explosive wave through the tubes. Here is found the same
resistance to combustion that is characteristic of the compounds
formed with nitric oxide (p. 63), a resistance that only dis-
appears in compounds that are capable of developing an
excessive temperature. In short, in the experiments described
above, we did not observe any rate of propagation of the wave
below 1000 metres per second.

Moreover, the propagation of the wave ceased whenever the
theoretical temperature, T, of the compounds formed with free
oxygen fell below 2000° (for hydrogen or cyanogen associated
with nitrogen) or 1700° (for carbonic oxide or methane asso-
ciated with nitrogen) ; figures corresponding to a lower limit of
the energy of the molecules.

Finally, the propagation of the wave ceased every time the
volume of the products of combustion amounted to less than
the quarter (for hydrogen and nitrogen) or even the third (for
methane or cyanogen associated with nitrogen) of the total
volume of the final compound.

6. Taking all these observations into consideration, the pro-
pagation of the explosive wave is quite a distinct phenomenon
from ordinary combustion. It only occurs when the layer
ignited exercises the greatest possible pressure upon the next
layer, i.e. when the ignited gaseous molecules possess the
maximum velocity and consequently the maximum translating
energy ; which is simply the mechanical expression of the fact
that they preserve almost the whole of the heat developed by
the chemical reaction. This is shown by the approximate
agreement of the calculations based upon the theoretical
estimate of the translating energy with the values obtained by
experiment for the velocity of the explosive wave. It is also
shown by the correlative increase of the pressure and velocities
towards the point of ignition.

7. The first coincidence shows, moreover, that dissociation
has little influence in these phenomena ; perhaps because it is
restrained by the high pressure developed along the path of the
wave and by its short duration. If this were not the case, the
energy, and consequently the velocity, would fall far below
the value calculated.

The influence of dissociation seems also annulled by the fact
that the velocity of the wave is independent of the initial
pressure (without admitting that dissociation is independent of
the pressure).

8. It may, however, be remarked in conclusion, that it is the
undulatory movement which is propagated, and not the gaseous
mass which is transported with such great velocities. In fact,
the velocity of the wave is the same, as has been shown, in a
tube open at both ends, closed at one end and open at the other,
or even closed at both ends.

112 THE EXPLOSIVE WAVE.

This result is also obtained in the experiments with the
oxyhydrogen mixture, in which the same velocity was found
either for the propagation of the flame (as attested by the
destruction of the solid fulminate interrupters) or for the propa-
gation of the pressure (as shown by the piston). The tracings
also show that the pressure attains its maximum instantly
upon the contact of the ignited layer with the layer immediately
in front of it.

9. Several conditions contribute to the production of these
effects. In the first place, it is necessary that the mass ignited
at the commencement should not be too small, in order that
radiation and conduction may not be given time to deprive this
mass of an amount of heat, i.e. of energy, greater than that
which is indispensable for the propagation of the wave. In
fact, if the radius of the sphere ignited is equal to the thickness
of the radiating layer, the loss of heat is proportionately
greater than if the radiating layer is merely a fraction of this
radius.

Moreover, when the number of molecules surrounding the
point first ignited is too small they may not contain the com-
bustive and the combustible elements in the exact ratio that
corresponds to the average composition of the mixture; this
would lower the temperature of this section, and consequently
the energy of the molecules.

Another circumstance, no less important, is, that the ele-
mentary velocity of the chemical reactions, at the temperature
of the combustion, should be sufficiently great for the heat
given off in a given time to maintain the system at the point
required ; a condition which is all the more important when
the elementary velocity of the reactions increases rapidly with
the temperature. It can even be conceived that the explosive
wave is only propagated if its theoretical velocity (rate of
translation of the molecules) is below, or at the most equal to,
the elementary velocity of the reaction.

10. Thus there is a limit in the condition that* corresponds
to the propagation of the explosive wave ; this is the regime of
detonation.

But it is easy to conceive quite a different limit, in which
the excess of pressure of the ignited section upon the following
one tends to fall to zero, and consequently the excess of velocity
in the translation of the molecules, i.e. the excess of their
energy, or, what is the same thing, the excess of heat which
they contain, has the same tendency. In such a system the
heat will be almost entirely lost by radiation, conduction, the
contact of surrounding bodies and of inert gases, etc., with
the exception of the very small quantity that is required for
raising the adjacent portions to the temperature of combustion ;
this is the regime of ordinary combustion, to which the measure-

INTERMEDIATE VELOCITIES. 113

ments of Bunsen, Schlcesing, and Mallard and Le Chatelier
relate.

We may, moreover, imagine the existence of velocities that
are intermediate between these two limits ; but they do not
constitute a regular system. In fact, the passing from one
regime to the other is accompanied, as is generally the case with
transitions of this kind, by violent movements, and extensive
and irregular displacements of matter, during which the propa-
gation of the combustion takes place in virtue of a vibratory
movement increasing in amplitude and gaining in velocity.1
Thus the regime of combustion, developed under conditions of
continually increasing pressure, ends by arriving at the regime
of detonation. These two regimes, and the general conditions
that define the establishment of each of them, and the transition
from one to the other, apply not only to gaseous explosive com-
pounds, but also to solid and liquid explosive systems, seeing
that the latter are wholly or partially transformed into gas, at
the time of the detonation.

1 See Mallard and Le Chatelier, "Comptes Rendus des stances de
l'Acade*mie des Sciences," torn. xcv. pp. 599, 1352.

114 GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

BOOK II.

THERMO-CHEMISTRY OF EXPLOSIVE COMPOUNDS.

CHAPTEK I.

GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

THERMO-CHEMISTRY is based on the following three fundamental
principles : —

(1) MOLECULAR WORK. This furnishes the measure of chemical
affinity.

(2) THE CALORIFIC EQUIVALENCE OF CHEMICAL TRANSFORMA-
TIONS. The heat disengaged in a definite chemical transformation
remains constant, like the sum of the weights of the elements.

(3) MAXIMUM WORK, The forecast of chemical phenomena is,
in virtue of this principle, brought to the purely physical and
mechanical notion of the maximum work accomplished by the
molecular reactions.

FIRST PRINCIPLE — MOLECULAR WORK.

1. The quantity of heat liberated in any reaction measures the
sum of chemical and physical work accomplished in this reaction.

Now the heat liberated in chemical action may be attributed to
loss of energy, to changes of movement, and, lastly, to the relative
changes which take place at the moment when the different mole-
cules fly 'towards one another in order to form new compounds.

It follows from this principle that the heat liberated in a re-
action is precisely equal to the amount of work which would
have to be accomplished to restore the bodies to their primitive
state. This work is at once chemical (changes of composition)
and physical (changes of condition) ; the former alone can serve
as measure of the affinities. We further see that the heat
liberated in one and the same combination varies with the
changes of state (solid, liquid, gaseous, or dissolved), with the
external pressure, with the temperature, etc. Hence the neces-
sity of defining all these conditions for each of the bodies
experimented upon.

2. In general the heat of molecular combination which expresses

PRINCIPLE OF INITIAL AND FINAL STATE. 115

the real work of the chemical forces (affinities) must be referred to
the reaction of perfect gases taking place at constant volume ; that is
to say, that the components and the compounds must all be brought
to the state of perfect gases and react in an unvarying space.

In the cases in which the reaction of the gases with formation
of gaseous products gives rise to a change of volume at constant
pressure the heat liberated necessarily varies with the tempera-
ture ; but the variation is slight enough to be neglected, as long
as we consider intervals of temperature which are not very far
apart, and even up to 100° or 200°.

Table I. (p. 125) gives the principal data known on the subject.
It expresses the heat liberated in reactions between gaseous
bodies at constant pressure with formation of gaseous products.

3. In default of these conditions, which it is rarely possible
to realise, it is permissible to refer the reactions of the bodies to
the solid state; as has already been done in the case of the
specific heats, according to the law of Dulong. In this state
the influences of the external pressure and changes of tempera-
ture become only slightly sensible, and in consequence all
bodies are more comparable than in the other states. The
quantities of heat liberated hardly vary as long as the interval
between the temperatures at which the reactions are carried out
does not exceed 100° to 200°.

4. There remain the following definitions: — we shall term
exothermal every reaction which liberates, and endothermal every
reaction which absorbs heat.

SECOND PRINCIPLE— THE CALORIFIC EQUIVALENCE OF CHEMICAL
TRANSFORMATIONS; OTHERWISE TERMED PRINCIPLE OF THE
INITIAL AND FINAL STATE.

If a system of simple or compound bodies, under given con-
ditions, undergo physical or chemical changes capable of bringing
it to a new state without giving rise to any mechanical effect
exterior to the system, the quantity of heat liberated or absorbed by
the effect of these changes depends solely on the initial and final
state of the system. It is the same whatever the nature or the
sequence of the intermediate states may be. This principle is
demonstrated by the aid of the preceding, combined with the
principle of energy. From it there follow various very important
consequences, such as the following, which are simply stated,
those who wish to go more fully into this subject being referred
to the author's " Essai de Me'canique Chimique."

1°. General Theorems on Reactions.

Theorem I. — The heat absorbed in the decomposition of a body is
exactly equal to the heat at the time of the formation of the same
compound, since the initial and final states are identical.

This relation has been pointed out by Laplace and Lavoisier

116 GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

as far back as 1780. It enables us to measure the chemical
work of electricity, of light, of heat, etc.

Theorem II. — The quantity of heat liberated in a series of
chemical and physical transformations accomplished successively or
simultaneously, in one and the same operation, is the sum of the
quantities of heat liberated in each isolated transformation, (all the
bodies being brought to absolutely identical physical conditions.)

It is in this way that the heat liberated by reactions referred
to the solid state is calculated.

Theorem III. — If two series of transformations be carried out,
starting from two distinct initial states, and arriving at the same
final state, the difference between the quantities of heat liberated
in the two cases will be precisely the quantity liberated or
absorbed when the transformation is from one of the initial
states to the other.

In this way is calculated the heat liberated by the union of
water with acids, bases, anhydrous salts, by the synthesis of
alcohols, etc.

The same theorem is employed to calculate the heat liberated
by the transformation of an explosive substance, whenever this
transformation does not occasion a total combustion, but the
products are defined by analysis. In a word, it is sufficient to
know, first, the heat produced by the total combustion of this
substance, a heat which may be experimentally measured by
detonating the substance in pure oxygen; second, the heat
liberated by the total combustion of the products of explosion,
which may be calculated when these products are known and
well-defined. The difference between these two quantities
represents the value sought.

Theorem IY. — The same conclusion is arrived at when the
two initial states are identical, the two final states being different.

This relation serves as base to a number of calorimetric
methods introduced into thermo-chemistry during the last few
years, because it renders it unnecessary to define the inter-
mediate states in complex reactions.

It is specially applicable to explosive substances when com-
bustion is incomplete and gives rise to imperfectly known
products. In short, it is sufficient to detonate the substance,
first, in pure oxygen, which gives rise to total combustion ; then
in nitrogen, which yields incompletely burnt products. The
heat liberated in each of the explosions is measured, and the
difference between the two figures expresses the heat of com-
bustion of the products of the second explosion ; that is, the
energy capable of being utilised in total combustion.

Theorem V. — Substitutions. — If one body be substituted for
another in a combination, the heat liberated "by the substitution is
the difference between the heat liberated by the direct formation of
the new combination, and by that of the original combination.

FORMATION OF SALTS. 117

This theorem is applicable to reciprocal replacements among
the metals, the metalloids, bases, acids, etc.

Theorem VI. — Indirect reactions. — If a compound yield one of
its elements to another 'body, the heat liberated by this reaction is
the difference between the heat liberated by the formation of the first
compound, by means of the free element, and the heat liberated by the
formation of the new compound, by means of the same free element.

The theorem is applicable to indirect oxidations, hydrogena-
tions, and chlorinations, to metallurgical reactions, to the study
of explosive substances, etc.

In the latter study it gives the difference between the heat of
combustion by free oxygen, and the heat of combustion by
combined oxygen.

The oxidiser (nitrate, chlorate, bichromate, metallic oxide, etc.)
is not a simple magazine of oxygen, as was formerly said ; for
generally this oxygen has lost a portion of its energy, equivalent
to the heat of the first combination. In certain cases, on the
contrary, such as where potassium chlorate is employed, the
combined oxygen liberates more heat than the free would do.

Theorem VII. — Slow reactions. — The heat liberated in a slow
reaction is the difference between the quantities of heat liberated
when the system of the components and that of the products of the
slow reaction are brought by the aid of the same reagent to the same
final state.

This finds numerous applications in organic chemistry, in the
study of ethers, amides, etc.

2°. Theorems on the Formation of Salts.

Theorem I. — The heat of formation of a solid salt is obtained
by adding the heats liberated by the successive actions of the
acid on water (Dt at the temperature t\ of the base on water
(D7), and of the dissolved acid on the dissolved base (Q£), then
by subtracting from the sum the heat of solution of the salt
(A£), all being measured at the same temperature.

In general, calling S the heat liberated in the reaction of a
system of solid bodies, transformed into a new system of solid
bodies, by means of a solvent, we shall have —

S = 2Dt + Q* - SA£.

D£, D'£, Qt, A2, are obtained by experiment. They are quantities
such that all of them vary considerably with the temperature t ;
while the quantity S is almost independent of the temperature
— at least, within very wide limits, as will be presently shown.

Theorem II.— The heat of formation of saline, add, and
alkaline hydrates is the difference between the heat of solution of
the anhydrous body and that of the hydrated body, in the same
proportion of water and at the same temperature.

Theorem III. — The heat of formation of a double crystallised
salt is equal to the difference between the heat of solution of the

118 GENERAL PRINCIPLES OF THERMOCHEMISTRY.

double salt and the sum of the heats of solution of the component
salts, increased by the heat liberated ly the mixture of the
solutions of the separate salts, the whole at the same temperature
and in presence of the same quantity of water.

Theorem IV. — The heat of formation of acid salts is calculated
in a similar manner.

Theorem V. — Changes of state of precipitates. — The difference
between the quantities of heat liberated or absorbed during the re-
dissolving of a precipitate, under two different states, at the same
temperature, and in the same solvent is equal to the heat brought
into action when the precipitate passes from one state to another.

Theorem VI. — Influence of dilution. — The heat of formation of
dissolved salts varies in general with the dilution and temperature.
The variation of this quantity of heat with the dilution at a given
temperature is expressed by the formula —

M' - M = A - (8 -f S')>

M being the heat liberated by the reaction of an acid and a
base, taken at a certain degree of concentration at this tempera-
ture; M', the heat liberated by the same reaction, the two
bodies being taken at a different degree of concentration ;
A the heat liberated (or absorbed) when the solution of the salt
is brought from the degree of concentration corresponding to
the first reaction to the concentration corresponding to the
second. S and £' are the analogous values, which correspond to
the respective changes of concentration of the acid and of the
base, always at the given temperature.

From a suitable degree of dilution, such as 1QOH20 to 1
equiv. of an acid or of a base, the variation M' — M generally
reduces itself to negligable quantities, that is to say, within the
limits of experimental error. But it should be remarked that
the variation M' — M ceases to be negligable, even within these
limits, for salts formed by the union of bases with alcohols or
weak acids, or by the union of any acid with wedk bases, such
as the metallic oxides. For such salts, moreover, the variation
M' — M tends to reduce itself to A, because S and §' become
inappreciable. Thus —

Theorem VII. — Under these conditions the heat of dilution of
the salt represents the variation in the heat of combination.

This action of water constitutes a true characteristic of weak
acids and bases. The preceding theorems are applicable not
only to salts but to every compound, or system of compounds
solid or in solution.

Theorem VIII. — The reciprocal action of acids on the salts which
they form with the same base, in presence of the same quantity, of
water, may be expressed at a given temperature by the relation

K! - K = M - M1?
M, M! being the heats liberated by the separate union of the

FORMATION OF ORGANIC COMPOUNDS. 119

two acids with the base; K, K^ the heats disengaged by the
action of the salt formed by the other acid.

Theorem IX. — Similarly the reciprocal action of bases on the
salts which they form with the same acid

K'i - K' = M - ML

Theorem X. — The reciprocal action of the four salts formed by
two acids and two bases is expressed by the formula

K! - It = (M - M') - (ML - M'O,

K being the heat liberated, when the solutions of two salts with
different acids and bases (potassium sulphate and sodium nitrate)
are mixed, and Kj the heat liberated when the reciprocal pair
are mixed (sodium sulphate and potassium nitrate). This
theorem enables us to determine the double saline decompositions
which are effected in solutions, when two salts of the same acid
or the same base are unequally decomposed by the same quantity
of water, which happens in the case of weak acids and bases,
and the metallic oxides.

3°. Theorems on the formation of Organic Compounds.

The heat of formation of organic compounds, by means of their
elements, cannot be directly measured, but it may be calculated
by the aid of various theorems, which follow from the second
principle.

Theorem I. — Difference between the heats of formation from the
elements. — Let there be two distinct systems of compounds, formed
from their elements, carbon, hydrogen, oxytfen and nitrogen, or
from very simple binary compounds, such as water, carbonic acid,
carbonic oxide, ammonia j the difference between the heat of forma-
tion of the first system and that of the second is equal to the heat
liberated when one of the systems is transformed into the other.

It is in this way that the heat of formation of bodies belong-
ing to the cyanogen series has been measured.

Theorem II. — Difference between the heats of combustion. —
The heat of formation of an organic compound by its elements is
the difference between the sum of the heats of total combustion of
its elements by free oxygen and the heat of combustion of the com-
pound with formation of identical products.

It is in virtue of this principle that most of the heats liberated
by the formation of organic compounds and their reciprocal
transformations have been obtained.

Theorem III. — Conversely, the heat of combustion of a body
formed of carbon, hydrogen, oxygen and nitrogen, is calculated by
means of its heat of formation. It is sufficient to find the sum
of the quantities of heat liberated when the carbon and hydrogen
supposed free, which enter into the composition of this body, are
changed into water and carbonic acid, and to deduct from this
sum the heat of formation.

120 GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

Theorem IV. — Formation of alcohols. — The heat liberated
when an alcohol is formed by the union of water and of a
hydrocarbon is the difference between the quantities of heat liberated
ivhen the alcohol and the hydrocarbon form one and the same
combination with an acid such as sulphuric acid.

The formation and the decomposition of conjugate bodies
(ethers, amides, etc.) give rise to various other theorems, analogous
to those relative to the salts, but which are omitted in order not
to unduly extend this summary.

4°. Theorems relative to the Variation of the Heat of
Combination with the Temperature.

In general, the quantity of heat liberated in a chemical
reaction is not a constant quantity ; it varies with the changes
of state, as has been said above; but it also varies with the
temperature, even when each one of the reacting substances
preserves the same physical state during the interval considered.
This variation is calculated in the following manner for any
reaction whatever, according to the second principle.

The reaction may be determined at an initial temperature, t,
and the heat liberated, Qt, may be measured.

The component bodies may also be raised separately from the
temperature t to the temperature T : which absorbs a quantity
of heat, U, depending on the changes of state and of the specific
heats, then the reaction is determined, which liberates Q£;
lastly, the products are brought by a simple lowering of tempera-
ture from T to t, which liberates a quantity of heat, V, also
depending on the changes of state and of the specific heats. The
initial and final states being the same in both processes the
quantities of heat liberated are equal, that is to say : —

Theorem I. — The difference between the quantities of heat liber-
ated by the same reaction, at two distinct temperatures, is equal to the
difference between the quantities of heat absorbed by the components
and by their products, during the interval of the two temperatures.

QT = Q* + U - V.
U — V represents the variation in the heat of combustion.

Theorem II. — If, during the interval T — t, none of the original
or final bodies undergoes change of state, this expression reduces
itself to the sum of the mean specific heats of the first bodies
during this interval, minus the sum of the mean specific heats of
the second bodies, multiplied by the interval of the temperatures.
U - V = (Sc - SoO (T - t).

The heat of combination will go on increasing or diminishing
with the temperature, and may even change in sign, according
as the first sum is greater than the second, or vice versa.

Theorem III. — Gaseous combinations formed without condensa-
tion.— In order that the heat liberated may be independent of the
temperature, the two above sums must be equal. Now this

VARIATION IN HEATS OF COMBINATION. 121

equality exists in fact for compound gases formed without conden-
sation. It is admitted that it should exist in principle for per-
fect gases, if the combination were effected at constant volume,
hence the definition (p. 114) of the molecular heat of combination.

Theorem IV. — Combinations referred to the solid state. — The
same equality exists approximately for solid bodies ; the specific
heat of these compounds, referred to equivalent weights, being
nearly the same as the sum of those of their components. The
heats of combination can therefore be referred to the solid state,
as validly as the atomic specific heats already are by Dulong's
law; which shows the importance of the expression S given
above. The liquid or dissolved state does not present the same
advantages ; for instance, the heat liberated in the reaction of
dilute hydrochloric acid on dilute soda, these two bodies being
taken at a given degree of concentration, varies from -f- 14*7 Gal.
to 4- 10-4 CaL, between 0 and 100°, that is to say, nearly by
half the latter's value.

Theorem V. — The heat liberated or absorbed during solution of
an anhydrous salt changes continually in amount with the tempe-
rature of solution; since the specific heat of saline solutions
differs, generally speaking, from the sum of the specific heats of
the salt and water taken separately.

It is smaller with the majority of the dilute solutions formed
by the inorganic salts. But the contrary holds good with the
solutions of various organic salts. The heat of solution of
anhydrous salts changes as a rule even in sign, for an interval
in temperature not exceeding 100° to 200°; sometimes this
change of sign occurs near the surrounding temperature, and
can be determined by direct experiments.

Hence it follows that those of the inorganic salts which pro-
duce cold when dissolving in water, at the ordinary temperature,
produce on the contrary heat at a higher temperature, whence it
also follows that there exists a temperature for which no thermal
variation is produced during solution.

These results, of which the development and demonstration
will be found in the author's " Essai de Mecanique Chimique,"
torn. i. p. 123, et seq., prove that solution has hitherto errone-
ously been assimilated to fusion.

5°. Theorems relative to the Variation of the Heat of Combination
with the Pressure.

Theorem I. — In gaseous combinations and reactions the heat
liberated is independent of the pressure, operating at a constant
volume.

This statement is no other than Joule's law, and is only true
for slight pressures and on the assumption that there is no
appreciable internal work in the gases, this work being in fact
negligible for gases remote from the point of liquefaction and at
a low pressure.

122 GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

Theorem II. — In gaseous combinations and reactions effected
without condensation, the heat liberated is the same, whether at
constant volume or at constant pressure.

Such is the case with the combustion of cyanogen, whether
by free oxygen or nitric oxide.

Theorem III. — In reactions effected with condensation the heat
liberated at a constant pressure,^?, at the atmospheric pressure
for example, and at a given temperature, t, is connected with the
heat liberated at constant volume, v, and at the same tempera-
ture, by the following relation —

Qtv = Qtp + 0-542 (N' - N) + 0'002*.

N here expresses the quotient by 22'32 of the number of litres
occupied by the component gases reduced to 0° and 0760
metres, and N' the same quotient for the resulting gases.

This theorem is of great importance in calorimetric measure-
ments relative to explosive substances. It enables the difference
between the two quantities of heat at constant pressure and
volume to be calculated for every reaction of which the formula
is known (see p. 15). It is not only applicable to reactions
where all the bodies, components as well as products, are
gaseous, but also to those where some of them only possess the
solid or liquid state at the outset, or assume it at the end.

THIRD PRINCIPLE — MAXIMUM WORK.

Every chemical change, effected without the intervention of a
foreign energy, tends towards the production of the body or of the
system of bodies liberating the most heat.

The necessity of this principle may be seen by observing that
the system which has liberated the greatest possible amount of
neat, no longer possesses in itself the energy necessary for
effecting a new transformation. Every fresh change requires
work, which cannot be performed without the intervention of a
foreign energy. On the other hand, a system still capable of
liberating heat by a fresh change possesses the energy necessary
for effecting this change without any auxiliary intervention.

The foreign energies here in question are those of physical
agents: light, electricity, heat; the energy of disaggregation
developed by solution ; lastly, the energy of chemical reactions,
simultaneous to that under consideration. Now, the interven-
tion of electric or luminous energies in a chemical phenomenon
is ordinarily apparent, and it is the same with chemical energy,
borrowed from a simultaneous reaction. The only cases which
call for discussion are those in which calorific energy and the
energy of disaggregation by solution intervene. They are dis-
tinguished by the following general character, that these energies
are exercised solely to regulate the conditions of existence of each
compound regarded separately without in any other way inter-
vening in the place of the reciprocal chemical actions.

MAXIMUM WORK. 123

Thus they are manifested under the conditions where they
provoke either the change of physical state (liquefaction, vapori-
sation) of any one of the bodies experimented on, regarded
separately, or its isomeric modification, or its total or partial
decomposition. It is furthermore evident, generally speaking,
that a compound can only take part in a reaction, if it exist in
the isolated state under the conditions of the experiment, and
in the proportion in which it can exist. This remark rightly
understood, can, strictly speaking, be made to apply in practice,
for it is sufficient to regard each of the components and of the
products in a system, and to know its individual state of
stability or of dissociation, under given conditions, in order to
be able to apply the principle. It is, moreover, necessary to take
into account in calculations and reasonings all the compounds
capable of existing under the conditions of the experiment, such
as double salts, acid salts, perchlorides, hydrates, etc., and
secondary compounds of every kind, which are ordinarily
neglected in the general interpretation of reactions, but each
of which contributes its quota, and, so to speak, its weight to
the thermal balance of affinities.

Lastly, let us note that in the calculation of the quantities of
heat liberated by a transformation, we should consider, as far as
possible, the corresponding bodies in the initial and final system
taking them under the same physical state. This mode of pro-
ceeding offers the advantage of putting aside, without further
discussion, a whole class of foreign energies, such as the
energies consumed in changes of physical state.

We do not wish to enter here upon more extended develop-
ments ; it will suffice to refer the reader to the detailed discus-
sion which is to be found in " Essai de Mecanique Chimique." 1
There it will be seen how the third principle is deduced from
the experimental study of the phenomena of combination and
decomposition.2

The following theorems are given which are applicable to a
large number of phenomena : —

Theorem I. — No endothermal reaction is possible without the
intervention of foreign energies.

Theorem II. — A system is the more stable, everything else being
equal, the larger the fraction of its energy which it has lost.

Theorem III. — Every chemical equilibrium results from the
intervention of certain dissociated compounds, that is to sayt in
the state of partial and reversible decomposition, which act at
once by themselves, as compounds, and by their components.

Under these conditions, there always intervene, in opposition
to the chemical energies properly so called, foreign energies,
electric or calorific, the latter especially.8

1 Tom. ii. pp. 421-471. 2 Ibid. pp. 424-438.

3 " Essai de Mecanique," chap. ii. pp. 439 and following.

124: GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

Exothermal reactions are, as has just been said, the only ones
which can be effected without the aid of a foreign energy.
However, they often require, in order to start them, the inter-
vention of a certain preliminary work, analogous to ignition.

Theorem IV. — An exothermal reaction which does not take place
of itself at a certain temperature, can almost always take place of
itself at a higher temperature, that is to say, in virtue of the
work of heating.

Theorem V. — It can likewise take place at the ordinary tempe-
rature, with the aid of a suitable auxiliary work, and especially
with the aid of chemical work, due to a simultaneous and
correlative reaction.

Theorem VI. — Within the limits of temperature at which
exothermal reactions take place, they do so, generally speaking, more
rapidly the higher the temperature.

Theorem VII. — Successive transformations can only take place
directly without the intervention of foreign energies, if each of
the transformations, regarded separately, as well as their definite
sum, be accompanied by a liberation of heat.

In other words, the energy proper to a system may be ex-
pended either all at once, or little by little, and according to
several distinct cycles, but there cannot be a gain of energy, due
to the internal actions alone, in any of the intermediate changes.
We shall give lastly, a theorem of the greatest importance in
the study of saline and many other reactions.

Theorem VIII. — Every chemical reaction capable of "being
accomplished without the aid of preliminary work and indepen-
dent of the intervention of an energy foreign to that of the bodies
present in the system, is of necessity produced, if it liberate heat.

It is in virtue of the third principle that the forecast of chemical
phenomena is reduced to the purely physical and mechanical
notion of maximum work effected by the molecular actions.

5. Numerical Tables.

The following tables give the principal data relative to the
quantities of heat liberated by the formation of compounds used,
or capable of being used, as explosives.

In these tables the authorities for the different determinations
are indicated by their initials, viz. : —

Al = Alluard Gh = Graham Pf = Pfaundler

An = Andre* G = Grassi Rech = Rechenberg

A = Andrews Ha = Hammerl R = Regnault

B = Berthelot H = Hautefeuille Sab = Sabatier

Cal = Calderon Hs = Hess Sa = Sarrau

Ch = Chroutschoff Jo = Joannis S = Silbermann

Ds = Desains L = Louguinine T = Thomsen

Dv = Deville M = Mitscherlich Tr = Troost

Dt = Ditte Og = Ogier Vie = Vieille

D = Dulong P = Person Vi = Vielle

F = Favre Pett = Pettersen W = Woods

%* The authority preferred is bracketed ; F. & S. [T].

TABLES.

125

TABLE I. — FORMATION OP GASES BY THE UNION OF THE GASEOUS ELEMENTS, THE
COMPOUNDS BEING REFERRED TO THE SAME VOLUME, 22 LITRES (1 + a<),
UNDER NORMAL PRESSURE.

Names.

Elements.

Equivalent
of gaseous
components.

Heat.

Authorities.

Hydrochloric acid

H + C1

36-5

+ 22-0

T.B.

Hydrobromic acid

H + Br

81

+ 13-5

T. [B.]

Hydriodic acid . .

H + I

128

- 0-8

T.B.

Water

H2 + O

9x2

+ 29-5 X 2

O

Hydrogen sulphide

17x2

+ 3-6x2

T. H.

Ammonia ....

H8 + N

17

+ 12-2

B. T.

Nitrogen monoxide .
Nitric oxide

N2+O

22x2
30

- 10-3 x 2
- 21-6

F. & S. [B.]
[B.] T.

Nitrogen trioxide .

N2 + 0,

38x2

- 111x2

B.

Nitric peroxide

N + 02

46

- 2-6

B!

Nitrogen pentoxide ^

54x2

- 0-6x2

B.

Nitric acid . . . '7

N2+ 035+ H

63

+ 34-4

B.

Chlorine monoxide

C12 + 0

43-5 x 2

- 7-6x2

T.B.

Sulphur chloride .
„ dioxide . .

S2 + C12

67-5 X 2
32 X 2

+ 8-1x2
+ 35-8x2

[B.]F.'&S.

„ trioxide .

S + O,

40 X 2

+ 48-2 X 2

B.

„ trioxide . .

S02 + O

40 X 2

+ 12-4x2

B.

„ oxychloride .

S02 + 01,

67-2 X 2

+ 6-6x2

Og.

O+O2

24 X 2

- 14-8 X 2

B.

Carbon dioxide

co + o

22 X 2

+ 34-1 X 2

„ oxychloride .

CO + C12

49-5 x 2

+ 9-4x2

B.

„ oxysulphide .

co + s

SOX 2

— 1-8x2

B.

Hydrocyanic acid .

CN + H

27

+ 7-8

B.

Cyanogen chloride

CN + C1

61-5

+ 1-6

B.

Ethane

C.IL 4- H-

30

+ 21-1

B.

Propane ....

v/2-LJ-4 i^ A 2

44

n •* *

+ 22-8

B.

Dibromethane . .

C2H4 + Br8

188

+ 29-1

B.

Glycolic ether . . .

C2H2 + 0

44

+ 33-0

B.

Aldehyde ....

C2H2 + 0

44

+ 65-9

B.

Acetic acid ....

C2H4 + O2

60

+ 13-3

B.

C2H4 + H2O

46

+ 16-9

B.

Formic acid . . .

CO + H2O

46

+ 3-1

B.

»> »» ...

C02 + H2

46

- 5-8

B.

Chlorethane . . .

C2H4 + HC1

64-5

+ 31-9

B.

Bromethane

C2H4 + HBr

109

+ 32-9

B.

lodethane ....

C2H4 + HI

156

+ 39-0

B.

Ethyl acetate . . .

C/2H4 + O2H4O2

88

+ 13-2

B.

Ethylidene chloride .

C2H2 + 2HC1

97

+ 29x2

B. & Og.

Amyl chloride . .

C5H10 + HC1

106-5

+ 16-9

B.

,, bromide .

C5H10 + HBr

151

+ 13-2

B.

„ iodide . .

C5H10 + HI

198

+ 10-6

B.

Nitric acid ....
Acetic acid ....

N205 + H20
C4H60, + H20

63
60

+ 5-3
+ 10-0

B.
B.

Chloral hydrate . .
„ alcoholate .

C2HC18O + H2 O
C2HC180-|-C8H60

...

+ 2-0
+ 1-6

B.
B.

Diamylene ....

2C H

140

+ 15-4

B.

Benzene ....

3C2H8

78

+ 171

B.

•

78

+ 70-5

B.

Aldehyde ! ! !

CAO

44

+ 32-9

B.

1 D. Hs. F. & S. G. A. T. B.

D. F. & S. G. A. B. T.

126 GENERAL PRINCIPLES OF THERMOCHEMISTRY

TABLE II. — FORMATION OF SOLID SALTS FROM THE ANHYDROUS ACID AND
BASE, BOTH SOLID.

NITRATES.

N205 + H20 (solid) + 1-1

N205 + K20 „ + 64-2

N205 -f Na20 „ + 54-4

N20S + BaO „ + 40-7

N205 + SrO „ + 38-1

N20a + CaO „ + 29-6

N205 + PbO „ + 21-4

N205 + Ag80 „ + 19-2

IODATES.

I205 + K20 (solid) -I- 51-6]

I30a + BaO „ + 34-9

SULPHATES.

SO3 + H2O

(solid) + 9-9

S03 + K20

„ + 71-3

SO3 + Na2O

„ + 61-7

SO3 + BaO

„ + 51-0

SO3 + SrO

„ + 47-8

S03 + CaO

„ + 42-0

S03 + PbO

» + 30-4

SO, + ZnO
S03 + CuO
S03 + Ag2

(solid)

OTHER SALTS.
PHOSPHATES.

P205 + 3H20 (solid)
P2O5 + 3Na2O „
P2O5 + 3CaO „

+ 19-7
+ 19-5
+ 28-0

+ 4-9
+ 39-8
+ 26-7

SUCCINATES.

C4H4O3 + H2O (solid) + 4-1
C4H403 + Na20 „ +38-1

C02
C02
C02
C02

CARBONATES.

(solid) + K2O + 40-3

„ + Na2O + 34-9

„ + BaO + 25-0

4- CaO +18-7

TABLE III. — FORMATION OF SOLID SALTS FROM THE GASEOUS ANHYDROUS ACID
AND THE SOLID BASE.

Names.

Elements.

Heat
disengaged.

NnO. 4- K,O

+ 70-7

Nitrates ....

N2O5 + Na2O ....

-|- 60-9

N2OS + BaO . .

+ 47-8

Nitrites ...

N2O, + BaO

+ 33-8

Sulphates ....

§3 + K2O
33 + K20
3 + Na2O

+ 76-6
+ 95-7
+ 67-2

3 + BaO

+ 56-9

Sulphites ....
Acetates ....

(SU2-fK20
12S02 + K20
C4H608 + K20 ....
C4H603 +Na20 ....
C4H603 + BaO ....
CO2 + K2O

+ 53-1
+ 66-8
+ 55-1
+ 47-0
+ 35-5
+ 43-3

CO2 + Na2O
COo 4- BaO

+ 37-9
+ 28*0

C02 + SrO

+ 26-7

CO2 + CaO

+ 21-7

CO2 + PbO ....

+ 10-8

C02 + Ag20

+ 9-8

TABLES.

127

TABLE IV. — FORMATION OF SOLID SALTS PROM HYDRATED ACID AND BASE,

BOTH SOLID.

Acid 4- base = salt + water (solid).

The heat disengaged, S, has the property of not varying sensibly with the
temperature contrariwise to what happens in the reactions of dissolved bodies
(p. 117). ('* Annales de Chimie et de Physique," 5e serie, torn. iv. p. 74.)

Symbol
of the
metals.

Nitrates.
NO,M.

Formates.

Acetates.

Benzoates

Picrates.

Sulphates.

Oxalates.

Tartrates.

K

Na
Ba
Sr
Ca
Mn

+ 42-6
+ 36-1
+ 31-7
+ 29-2

+ 25-5
+ 22-3
+ 18-5
+ 16-7
+ 13-5
+ 7-6

+ 21.9
+ 18-3
+ 15-2
+ 14-7
+ 10-6
+ 4-5

+ 22-5
+ 17-4

+ "8-2

+ 30-5
+ 24-3

+ 407
+ 34-7
+ 33-0
+ 29-5
+ 24-7
+ 15-6

+ 29-4
+ 26-5
+ 20-81
+ 21-3
+ 18-9
+ 13-2

+ 27-1
+ 22-9

+ 16-71

Zn
Cu
Pb

Ag

+ T9-7
+ 18-0

+ 6-2
+ 5-4
+ 9-1

+ 3-3
+ 4-3
+ 5-1

+ 7-6

Phenate.
K+17-7

Succinates.
K+23-2
Na+20-0

+ 11-9
+ 10-5
+ 19-9
+ 17-9

+ 11-5

+ 13-1
+ 12-5

"lodates.
K + 31'5
Ba+25-6

1 This number refers to precipitated salts which contain combined water.
TABLE V. — FORMATION or SOLID AMMONIACAL SALTS.

Names.

Components.

Heat
disengaged.

(1) From the solid 1
Nitrate

lydrated acid and gaseous base
HNO8 + NH3
CH202 + NH3
C2H402 + NH8
C7H602 + NH8
C6H3(N02)303 + NH3
H2S04 + 2NH
C2H204 +2NH3
C4H604 + 2NH3 .
cid and base gaseous.
HC1 + NH3 .

l»

+ 34-0
+ 21-0
+ 18-5
+ 17-0
+ 22-9
+ 33-8
+ 24-4
+ 19-7

+ 42-5
+ 45-6
+ 44-2
+ 20-5
+ 23-0
+ 26-0
+ 29-0
+ 41-9
+ 39-8
ee gaseous.
+ 47-1
+ 33-7
+ 41-5
+ 53-5
+ 30-4
+ 31-6

+ 76-7
+ 71-2
+ 56-0
+ 42-4
+ 64-8
+ 87-9
+ 79-7
+ 142-4
+ 70-8

Formate

Benzoate

Sulphate

Oxalate

(2) A
Chloride

HBr + NH3 .

Iodide

HI + NH3

HCy + NH3 .

Sulphide

H2S + NH3

Acetate
Formate

C2H402 + NH3 ....
CH2O2 + NH3 ....
HN03 + NH3 . . . .
HC1 + C3H9N . . . .

id, water, and the base, all thr
N205 + H20 + 2NH8 . .
N20, + H20 + 2NH3 . .
C2H20 + H20 + NH8 . .
SO3 + H2O + 2NH3 . .
C02 + H20 + NH8 . . .
CO + H20 + NH3 . , .
beir gaseous elements.
C1 + H4 + N . . . .
Br (gas) + H4 + N . . .
I(gas)4-H4 + N . . .
S(gas) + Hs + N . . .
02 + H4 + N2 . . . .
08 + H4 + N2 . . . .
Cl + 04 + H4 + N . . .
8 + 04 + H8 + N2 . . .
Cl + H4 + N +0 . . .

Nitrate

Trimethylamine chloride
(3) From the anhydrous oxyac
Nitrate
Nitrite

Sulphate

Formate • • • •

(4) From t
Chloride . . ^ . . . .

Bromide

Iodide • • ....

Sulphide

Nitrite ... ...

Nitrate

Sulphate

Hydroxylamine Chloride

128

GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

1 i

l>00<£> OO

Oi CO COO-* Ai Oi

ciirH :<Ni-« : :<MT-I

cocpoo

^^00
J r-l r-t •*!

I~t l> 00 <-i

CO . *H|1<|>

1 + + + +

+ +1+ + 1 + + 1

II +++

CO CO tt> CO CO r-l -^

S5^66ci$

1 1 1 1 1 1 +

5 5 « ^

* 44 w- 'M"* L?" w

& H o o

N
N
N
N
N

N

TABLES.

129

!...T.T^*^

pq

e

EH

02

-> B

MM WM

PQQQ"

C5 00 *0 plO O

+

+++++ +1 +++++++ 1 1+ II 1 +++++

++

CO

GO •* OS 1C T* rH O: O CO t> CO CO O 00 CO O O

.g <jq cb •* t> »o rH o t> cs cb do 10 <N TH o r-i

SrH

1

+ •+++'"+++++++' "++" '++'++'

ii

p

(N O »O rH *? T1 »O *f CO

i> <>i cb cb !2 S »> o 3<i

: : t-«h :

-h

1 1

cogs cp pip qs TH co op THOO o

H cb o os :

+^" i ijtiiii i!

HJJ,

»p o o o »p

;(N 1-1 00 OS 1C 10 cVj rH O 00 CO 00 OS »O t- CO <M CN
OOt>OSO5i— lCOTttOOOrHCOOO(NCOCOI>OS <M

00 W

-- N N M

O 02 02

o o o 02 2

O O O O OQ

Q o

o o - .

O O O O 02

^3^-3 o S

rUrl S 3

Sf«§ § §

5 Q
||

*J£ A ^ ,5

P J?

gitl

03 'S

.2 a

,T3 03

"I

232^23 2S .2^0202

<D S *c •* • ^

ft t- fc J3 T*

11 ' ' I I

563

ll

K

130

GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

TABLE VII.— FORMATION OF METALLIC OXIDES ACCOBDING TO M. THOMSEN.J

Name.

Components.

Equiva-
lents.

Heat disengaged.

Solid state.

Dissolved
state.

K2 4- 0 (Beketoff)

47*1

+ 48-6

+ 82-3

Potash ......

K2 + O 4- H2O

56-1

4- 69-8

+ 82-3

K2 + H2 + 02

4- 104-3

+ 116-8

Na2 + O (Beketoff)

31

4- 50-1

+ 77-6

Soda . . .

tfa2 4- O + H2O

40

4- 67-8

4- 77-6

Na2 + H2 + 02

+ 102-3

+ 112-1

/N + H3 + H20
\N + H5 4- 0

35

...

4- 21-0
+ 90-0

Ca + O

28

4- 66-0

+ 75-05

Ca + O + H2O

37

+ 73-5

4- 75-05

Ca + H2 + O2

37

4- 108-0

+ 109-55

Sr + 0

51-8

+ 65-7

4- 79-1

Jgr + O + H2O

60-8

+ 74-3

+ 79-1

Sr + H2 + O2

60-8

4- 108-8

+ 113-6

Baryta

Ba + 0

76-5

X

x+ 14-0

Barium dioxide .

BaO + O

84-5

4- 6-05

Magnesia . . . . .

/Mg + 0 + H20
\Mg + H2 + 02

29
29

4- 74-9
4- 109-4

Alumina . . . • .

A12 + O3 + 3H2O

78-4

/+ 195-8 or
\+ 65-3x3

Manganese protoxide

Mn + O

35-5

4- 47-4

„ dioxide .

Mn + O2

43-5

+ 58-1

Permanganic acid

/Mn2 + O7 + H2O
\ (hydrated)

120

...

4- 89

Chromic acid ....

Cr2O3 4- O3

103

4- 3-1

/+ 4-2 or
\ + 1-4 X 3

Iron protoxide

Fe + O

36

4- 34-5

Iron peroxide ....

Fe2 4- 03

80

/+ 95-6 or
\+ 31-9 x 3

Iron (magnetic) oxide

Fe3 + O4

116

/ 4- 134-5 or
\+ 33-6 x 4

(FeO + Fe203

...

+ 4-5

Auric oxide ....

Au2 + 03

221

- 5-6

Zinc oxide (anhydrous) .

Zn + O

40-5

+ 43-2

„ (hydrous)

Zn 4- O 4- H2O

49-5

4- 41-8

Cadmium oxide .

Cd + 0

64-0

+ 33-2

Lead oxide (anhydrous) .

Pb + O

111-5

+ 25-5

„ (hydrous)

Pb 4- 0 + H20

120-5

+ 26-7

Cuprous oxide
Cupric oxide (anhydrous)

Cu2 + O
Cu + O

71-4
39<7

4- 21-0
+ 19-2

„ (hydrous) .

Cu 4- O 4- H2O

48-7

4- 19-0

Stannous oxide

Sn + O

67

4- 34-9

Stannic oxide .

Sn4-02

75

4- 67-9

Mercurous oxide

Hg2 + 0

208

+ 21-1

Mercuric oxide

Hg + 0

108

4- 15-5

Silver oxide .

Ag24-O

116

4- 3-5 2

„ sesquioxide

Ag4 + 08

240

4- 10-5

Bismuth oxide

Bi + 03

234

4- 68-3

Antimonious oxide

Sb4-03

146

+ 88-7

Antimonic oxide .

Sb4-05

162

+ 114-9

1 The heats of solution of the alkalis are by Berthelot. Without modifying the
experimental bases of Thomsen, his calculations in Tables X. and XL have been
corrected by the small amounts necessary to put them in accord with the other data
of the present Tables, such as the heat of formation of water, + 34-5 instead of + 3±-7.

2 Corrected by Berthelot.

TABLES.
TABLE VIII.— HALOID SALTS.

131

Names.

Components.

Equiva-
lents.

Heat disengaged.

Authorities.

Solid Salt.

Dissolved Salt.

Potassium

chloride .
Sodium

K + C1

74-6

+ 105-0

+ 100-8

T.

chloride .
Ammonium

Na + Cl

58-5

+ 97-3

+ 96-2

T.

chloride .
Strontium

N + H4 + C1

53-5

+ 76-7

+ 72-7

B.

chloride .
Barium

Sr + Cl2

79-3

+ 92-3

+ 97-8

T.

chloride .

Ba + C12

104-0

»+ 31-7

*+ 32-7

B.

Br

Br

Potassium

Gas. Liquid

Gas. Liquid.

bromide .

K + Br

119-1

+ 100-4 +96-4!

+ 95-0 +91-0

T.

I

I

Potassium

*-

iodide .

K + I

166-1

+ 85-4 j + 80-0

+ 80-1 +74-7

T.

TABLE IX. — FORMATION OF METALLIC SULPHIDES. '

Name.

Components.

Equiva-
lents.

Heat disengaged, the
body being

Authorities.

Solid.

Dissolved.

Potassium sulphide .

K2 + S

55-1

+ 51-1

+ 56-2

T. Sab.

Potassium polysulphide
Potassium sulphydrate
Sodium sulphide

K2S + S3
K2S + H2S
Na2 + S

103-1
72-1
39

+ 6-2
+ 9-2
+ 44-2

+ 2-6 2
+ 2-9 2
+ 51-6

Sab.
Sab.
T. Sab.

Sodium polysulphide

Na.S + S3

87

+ 5-1

+ 2-5 2

Sab.

Sodium sulphydrate

Na2S + H2S

56

+ 9-3

+ 3-9 «

Sab.

Ammonium sulphide

N2 + He + S

34

+ 28-4

B.

Ammonium sulphydrate
Ammonium sulphydrate

(NH4)2S + H2S
N2 + H10 + S2

51
51

+ 39-8

+ 3-02
+ 36-6

B.
B.

solid

Strontium sulphide

Sr + S

59-8

+ 47-6

+ 53-0

Sab.

Calcium sulphide

Ca + S

36

+ 46-0

+ 49-0

Sab.

Barium sulphide

Ba + S

84-5

a;-15-6

Sab.

Iron sulphide

Fe + S

44

+ 11-9

B.

Zinc sulphide

Zn + S

48-5

+ 21-5

B.

Lead sulphide .

Pb + S

119-5

+ 8-9

B.

Copper sulphide

Cu2 + S

79

+ 10-1

T.

Copper sulphide

Cu + S

47-5

+ 5-1

T.

Mercury sulphide
Silver sulphide .

Hg+S
Ag2 + S

116
124

+ 9-9
+ 1-5

B.
B.

1 These numbers refer to solid sulphur ; when the sulphur is gaseous, about
448°, it is necessary to add + 1-2. Towards 1000° the correction would be much
larger, but it is not known with certainty.

2 Components dissolved.

K 2

132 GENEKAL PRINCIPLES OF THERMO-CHEMISTRY.

c> t> •* rt

cb i> co eb o

CO ?O <M i-< I-H

O5 O
C000

1 ++++ 1 +++++++

<N »O CO tD "^ ^ O5 O O CO »H O "*"

00 O O t- O 1O rH (N (fq b- O5 OO CO

T*< co co CD <M I-H o i> :GO :«sr-i o

+++++!+++*+ ++ +

Op»0
' I +

1 +

cp»p»pqp
111 +

• *

I +

CO <N l>
<N »O <N

*P *? T1 T1 T1 T1 T1 ^

|> l> ,!H TH rH -f »O »O OS «O «O Al T^-I O5 t> i-i CO 00 QO r^
«O CO •* ^ ?O «O T^ (M (M 00 X »O OS 00 t> »O O 00

8

**&%

oooo

6 « «
+ +

1?

0

+
w
+

^

OOOQ

yanic acid

.

i

,,

I

g,8

•&

Cya
Hyd

H

TABLES.

133

TABLE XI. — FORMATION OP THE PRINCIPAL SOLID OXTSALTS, PROM THEIR
ELEMENTS TAKEN IN THEIR ACTUAL STATE.

Names.

Elements.

Equiva-
lents.

Heat disengaged.

Nitrates

/N 4- 03 + K
N + 03 4- Na
N2 + 03 4- H4

N2 4- O6 4- Sr

101-1
85
80
103-8

4- 118-7
4- 110-6
4- 87-9
4- 109-8

Sulphates

N2 4- O6 4- Ca
N2 4- 06 + Pb

/S4-04 + Na2
S + 04 4- H8 4- N2
S + O4 + Sr
S 4- 04 4- Ca
S 4- O4 4- Mg

82
165-5
170
87-1
71
66
91-8
68
60

+ 101-2
-f 52-8
4- 28-7
+ 171-1
4- 163-2
4- 141-1
4- 164-7
4- 160-0
4- 150-6

S 4- 04 4- Mn
S + 04 + Pb
S 4- O4 4- Zn
\S4-04 + Cu
\S + 04 4- Ag2
g2 + O7 + K2

75-5
151-5
80-5
79-7
156
127-1

4- 123-8
4- 107-0
+ 114-4
+ 90-2
+ 82-9
+ 236-0

S2 4- O6 4- K2

119-1

+ 205-7

Sulphite

g + 03 + K2

79-1

+ 136-3

g2 4. Q5 4- K2

111-1

4- 184-6

Hyposulphite . •

S2 4- O3 4- K2

95-1

4- 133-4

01 4- 03 + K
KC1 + O3
01 4- O3 4- Na

122-6
106-5

+ 94-6
- 11-0
4- 85-4

NaCl + O3
BaCl2 + 06
/Br. gas 4- 03 + K

< iZTt , f\

152-1
167-1

- 12-3
- 12-6
+ 87-6
11*1

lodate

\ KBr + O3
JI gas + 03 + K

214-1

4- 128-4
+ 44.1

Perchlorates
Phosphates .

/Cl 4>O4 4-3K
KC1 4- 04

\NaCl 4-4 04
BaCl2 + 08
NCI 4- 04 + N + H4

138-6
138-6
122-1

168-1
117-5
164

•±t A

4- 112-5
4- 7-5
4- 110-2
+ 3-0

+ I'l

4- 79-7
4- 451-6

i 4fiO'6

Carbonates (carbon dia-

\P2 4- O8 + Ca3
C + 03 4- K2
C 4- 03 4- Na2
C 4- O3 4- Sr
C + O3 4- Ca
( C 4- 03 4- Mg
\ C 4- O3 4- Mn

155
69-1
53
73-8
50
42
57-5

+ 138-9
4- 135-1
4- 139-4
4- 134-7
+ 133-8
4- 104-0

Bicarbonates

C + 03 4- Pb
\ C 4- O3 4- Zn
\C 4- 03 + Ag2
C + O3 + K + H
JC4- O3 + Na + H

133-5
62-5
138
100-1
84

4. 83-2
+ 97-1
4- 60-2
4- 232-8
4- 227-0

Formiates (carbon diamond)
Acetates (carbon diamond)

| C 4- O3 + N + H4
JC 4- H + K 4- 02

C2 + H3 4- K + 022
C2 + H3 + Na + 02
C2 + H7 + N + 02

79
84-1
68
98-1
82
77

4- 205-6
+ 154-8
4- 149-6
4- 184-9
4- 179-2
+ 159-6

134 GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

FORMATION OP THE PRINCIPAL SOLID OXYSALTS, ACCORDING TO THEIR ELEMENTS

TAKEN IN THEIR ACTUAL STATE. — (Continued.)

Names.

Elements.

Equiva-
lents.

leat disengaged.

(C2 + K2 + 04

166-2{

+ 323-6 or
161-8 x 2

Oxalates (carbon diamond) .

) C2 + Na2 + 04
j C2 + H8 + N2 + 04

134 |
124 |

+ 313-8 or
156-9 X 2
+ 272-4 or
136-5 X 2

r

I

+ 158-5 or

\ C2 -i- Ag2 + O4

304 ^

79-2 x 2

Chromates

/Cr203 pp. + 05 + K2
\Cr03 pp. + 0 + Na2

194-2
162

+ 206-7
+ 190-3

Bichromates

/2(Cr03) pp. + 0 + K2
\2(CrO8) pp. + O +N2 + H8

147-1
126

+ 115-5
+ 85-4

ACID SALTS.

Bisulphates

(S03 + K2SO4 = K2S2Oy
|H2SO4 sol + K2SO4 =
W>KHSO4

127-1
136-1

+ 13-1
+ 7-5

S04 sol + Na2S04 =

JNaHS04

120-0

+ 8-1

CrO8 + K2CrO4

147-1

+ 1-9

HYDRATES.

K20 + H20

56-1

+ 21-2

Hydrates

Na2O + H2O

40

+ 17-8

BaO + H2O

85-6

+ 8-8

SrO + H2O
CrO + H2O

60-8
37

+ 8-6
+ 7-55

TABLE XII. — HEAT DISENGAGED BY THE COMBUSTION OF ANY BODY WHAT-
EVER BY MEANS OF VARIOUS OXIDISING AGENTS.

Name of Oxidising Agent.

Formulae.

Equiva-
lents.

Heat disengaged.

Free oxygen . ...

£0

8

A Calories

Copper oxide . ...

£CuO

39-7

A - 19-2

Lead oxide . ...

iPbO

111-5

A - 25-5

Stannous oxide

• SnO

67

A - 34-9

Stannic oxide . ...

fSnO-

37-5

A - 34-0

Antimony oxide

£Sb02

38-1

A — 31-1

Mercury oxide

JHgO

108

A - 15-5

Bismuth oxide

Bi O

78

A - 23-0

Silver oxide . ...

£Ag20

116

A- 3-5

Nitrogen monoxide .

<N20

22

A + 10-3

Nitric oxide

iJ-NO

15

A + 10-8

Nitric peroxide (liquid) . .

£N2O4

11-5

A- 0-4

Nitric acid (liquid) .

£HNO3

12-6

A- 1-4

Potassium nitrate . . .

£KNO3

20-2

A- 5-4 or A -2-5

Sodium nitrate ....

£NaNO3

17

A — 4-4 or A — 2-3

Strontium nitrate . . .

"'Sr(NO )

21-2

A- 5-4

Barium nitrate ....

}5Ba(N03)2

26-1

A- 5-5

Lead nitrate

Ti5Pb(NO3)2

28-9

A- 8-8

Silver nitrate . .

i AffNO,

34

A— 4-8

Ammonium nitrate .

|NH4N03

40

A + 25-0

Potassium chlorate .

AKC1O3

20-4

A+ 1-8

Potassium perchlorate .

iKC104

17-3

A- 0-9

Manganese dioxide .

iMnO2

43-5

A — 10-7

Potassium bichromate .

£K2Cr207

49-1

A- 7-9

TABLES.

135

TABLE XIII. — MULTIPLE DECOMPOSITIONS OP AN EXPLOSIVE COMPOUND, BY

M. BERTHELOT.

NH4NO3 solid1 ==

N20 + 2H20

Water.

Liquid.

Gaseous.

+ 29-5
+ 50-1
+ 28-5

+ 48-8
+ 52-7

+ 10-2
+ 30-7
+ 9-2
+ 23-3
+ 29-5
+ 33-4
-41-3

N + O +2H O .

N + NO + 2H2O

KN2 + N2O3 + 6H2O

£(3N2 -f N2O4 + 8H2O)

i(2HNO + 4N + 9H2O)

HNO + NH (both gaseous) .

TABLE XIV.— FORMATION OP THE PRINCIPAL SALTS IN THE DISSOLVED OR
PRECIPITATED STATE BY MEANS OF DISSOLVED ACIDS (ONE EQUIVALENT
DISSOLVED IN TWO OR FOUR LITRES OF LIQUID) AT 15°, ACCORDING TO
BERTHELOT AND THOMSEN.

Bases.

id

6 S

Is"?

g»sr

fo-

s*C

^uS1

* *

S ««

a0-"

Sw .
£v$

4

Jo«

H'.

oSSf

lo-
•§,«?, ii

"333 d4

CO V

s a

Soj00

•Sw"

S s

M

Carbonates
C02
eq. = 15 lit.

I"H

rt

""*

^

Na200
K2O .
NH3 .
CaO («)
BaOC)
SrO(5)
MgO(")
MnO («)
FeO .
NiO .

13-7
13-7
1245
14-0
13-85
14-0
13-8 (")
11-8
10-7
11-3

13-7
13-8
12-5
13-9
13-9
13-9
13-8
11-7

13-3
13-3
12-0
13-4
13-4
13-3

11 -3

9-9

13-4
13-4
11-9
13-5
13-5
13-5

10-7

14-3
14-3
41-7
18-5
16-7
17-6

14-3

15-85
15-7
14-5
15-6
18-4
15-4
15-6
13-5
12-5
13-1

3-85
3-85
3-1
3-9

5*1
7-3

2-9
3-0
1-3
3-2
32
3-1

10-2
10-1
5-3
9-8
11-1
105
9-0
6-8
5-0

CoO

10-6

13-3

7-2

CdO
ZnO .
PbO .

10-1
9-8
7-7(8)
10-7(7>)

10-1

9-8

7-7

8-9
6-5

9-1
6-6

12:5
12-8

11-9
11-7
10-7

9*6
13-3

7-3

5-5

6-7

CuO
HgO .

Ag20 .
£A12O3

7-5

9-45(10)
20-1(";
9-3

7-5
5-2

6-2
3-0

4-7

6-6

7-0
12-9

9-2

7*2

10-5

15-8
24-35
27-9

15-5
20-9

2-4
6!9

*Fe203.
iCr,O, .

5-9
6-9

5-9

4-5

...

...

5-7

8-2

...

...

...

If the salt were fused the numbers should be increased by about + 4.
1 eq. = 2 litres. » 1 eq. = 25 litres.

1 eq. = 6 litres. 5 1 eq. = 10 litres.

Preci
ates as we

pitated, this applies to the earthy and metallic oxalates and carbon-
__ .. 3ll as to the metallic oxides and sulphides. 7 Crystallized.

1 eq. = 4 litres ; this applies to all salts formed by insoluble oxides.
Very dilute.

10 HgCl2 - solid ; -+ 11-0 ; HBr dilute : HgBr2 - dissolved ; + 13-7 ; solid +
15-4 ; HI dilute ; HgI2 red ; + 23-2.

11 HBr dilute and Ag2O; +22-5 to 25-5. HI dilute + Ag20 ; 26-5 at first,
afterwards + 32-1.

136 GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

o

§5

I

ft

g a

+ +

III

« ^99
+ + + +

i 10 O ^- US O US O

+ + 1

iOrt •* *- *s •<»« r-i p oo <£>
•+ I I + + + + I I +

to ta

00 O«OOOOOOO
« CO tH«- *~*~

oo

OOOO 0

8-S

.22

2§

g§

fl • • "§^ -ag •

•a 1 • • §§ .os«

^.242 g 2 go S » 5^a | *"3 £«« gja1^'

'Hit I

TABLES.

137

•s

1 i

«s as co as 4)i *- e*

O5 C<1 05 N \O OS 1-

+ + + + + + +

l

ss

.

0+ + +

+*£*

within about om
alue of the heat

12

°s

!!

Illllll

* as

l«

ill! "fraiijiirj

gzlg SS3===t|a&J

138

GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

TABLE XVI. — FORMATION OP ALDEHYDES AND ORGANIC ACIDS BY OXIDATION.

Names.

Components.

Compounds.

Heat
disengaged.

Physical state of
the Compounds.

1. WITH THE HYDROCARBONS.

Ethylic aldehyde . .

C2H4 + 0

C2H40

r+ 65-9

\ + 71-9

Gas
Liquid

Ortho-propylic aldehyde
Iso-propylic aldehyde .

}C3H6 + 0

C3H6O

/ + 87-3
\ + 83-3

Liquid
Liquid

( + 133-2

Gas, liquid

Acetic acid ....

C2H4 + Oz

C2H402

{ + 138-3

Solid

1 + 133-9

Solid

Propionic acid . . .

C3H6 + O2

C3H6O2

+ 153-3

Liquid

Oxalic acid ....

C2H2 + 04

+ 258-0

Solid

Acetic acid ....

C2H2 + O +H20

C2H402

/+ 118-7
\+ 121-2

Liquid
Solid

(+128-D

Gas

Formic acid ....

CH4 + 03

CH2O2 + H20

+ 143-5

Liquid

4- 147-3

Solid

2. WITH ALDEHYDES.

Both bodies

Acetic acid ....

C4H40 +O

C2H402

4- 67-3
4- 66-4

gaseous
Actual state

Propionic acid . . .

C3H60 + O

C3H602

4- 74-0

Actual state

3. WITH ALCOHOLS.

Liquid formic acid .
Liquid acetic acid .
Liquid valerian ic acid .
Solid margaric acid

CH40 + 02
C2H60 + 02
C5H120 + 02
C16H340 + 02

CH2O2 +H2O
C2H402 + H20
C5H1002 + H20
C16H3202 + H20

4- 100-0
4- 125-1
+ 131-0
+ 180-0

Actual state
Actual state
Actual state
Actual state

Solid oxalic acid

/C2H60 + 05
\C2H402 + 03

C2H204 + 2H20
C2H204 + H20

+ 261-0
+ 139-4

Actual state
Actual state

TABLE XVII. — VARIOUS ORGANIC COMPOUNDS.

Names.

Components.

Compounds.

Equivalents.

Heat
disengaged.

FORMATION OF AMIDES BY AMMONIACAL SALTS.

Formic amide

CH202, NH3 diss.

CH3NO (diss.)

45

- 1-0

Formic nitril or

hydro-cyanic"

• j
acid .

CH202, NH3 diss.

HCN (diss.)

27

- 10-4

Oxamide .

C2H204, 2NH3 cryst.

C2H4N202 (solid)

88

- 1-2x2

FORMATION OF ISOMERIC AND POLYMERIC BODIES.

(liquid )

( liquid

140

+ ll'S

Diamylene

2C5H10 gaseous

C10H20 liquid

140

+ 22-3

(gaseous)

( gaseous

140

+ 15-4

Benzene . .

3C2H2 (actual reaction)

C6H6 (gas)

78

+ 171-0

Dipropargyl .

J3C2H2 (theoretical action)
\C6He (Benzene idem.)

}C6H6 (gas)

78 {

+ 100-5
- 70-5

TABLES.

139

TABLE XVIII. — FORMATION OF NITRIC DERIVATIVES.
Organic compound + HN03 liquid = Nitric derivative + H20 (liquid).

Names.

Compounds.

Equivalents.

Heat
disengaged.

Nitric ether (B.) .
Nitroglycerin (B.)

C2H4(HN03)
C3H23(HN03)

91
227

+ 6-2
+ 4-7x3

Nitromannite (B.)
Gun-cotton (B ) .

C6H26(HN03)
C24H180911(HN08)
C6H5(N02)
C6H42(N02)

453
1143
123

168

+ 3-9x6
+ 11-4X 11
+ 36-6
+ 36-2 x 2

Nitrobenzene (B.)
Dinitrobenzene (B.)

Picric acid (Sa. & Vie.)
Chloronitrobenzene (B.)

06H33(N02)0
C6H4C1(N02)

229
157-5

+ 34-0 X 3
+ 36-4

Nitrobenzoic acid (B.)

C7H5(N02)02

167

+ 36-6

Nitronaphthaline (Tr. & H.)
Nitrotoluene (Tr. & H.) . .

C10H7(N02)
C7H7(N02)

173
137

+ 36-5
+ 38-0

TABLE XIX. — HEAT OP FUSION OF ELEMENTS AND SOME OF THEIR COMPOUNDS.

Names.

Formulae.

Equiva-
lents.

Temperature
of Fusion.

Heat of
Fusion.

Authority.

Br

80

degrees.
7.3

C.
— 0-13

I

127

+ 113-6

— 1-49

S

16

+ 113-6

-0-15

p

Phosphorus ....

P
Hff

31

100

+ 44-2
— 39-5

— 0-15
— 0-28

P.

p

Lead

Pb

103-5

+ 335-0

— 0-53

Bismuth
Tin

Bi

Sn

210
59

+ 265-0
+ 235'0

-2-6
— 0-84

P.
p

Ga

35

+ 30-0

— 0-66

B

Cadmium

Cd

56

+ 500-0

-0-65

p

Silver

A£

108

+ 954-0

-0-23

p

Platinum

Pt

98-6

+ 1775-0

-2-68

Vi.

Pd

53

+ 1500-0

— 1-9

Vi

Water

H,O

9

+ o-o

- 0-715

Ds.

Iodine chloride.
Nitrogen pentoxide . .

IC1
N205
HNO3

162-5
54
63

+ 25-0
+ 29-5
— 47-0

-2-3
-4-14
-0-6

B.
B.
B

Sulphuric acid . . .
Sulphuric acid (hydrate) .
Naphthalen ....

H2S04
H2SO4H2O

C,HoO,

49
58
128
92

+ 8-0
+ 8-8
+ 79-0
+ 17-0

-0-43
-1-84
-4-6
-3-9

B.
B.
Al.
B.

Formic acid
Acetic acid . .
Benzene . .
Nitrobenzene . . .
Phenol ... . .

OfiK'
C2H403
C6H6
C6H5N02
C6H6O

46
60
78
123
94

+ 8-2
+ 17-0
+ 4-5
+ 3-0
+ 42-0

— 2-43
-2-5
~2'27
-2-74
— 2-34

B.
B.

Pett.
Pett.
Pett.

Potassium nitrate . .
Sodium nitrate .

KN03

NaN03

101
85

+ 333-5
+ 306-0

-5-5
-4-9

P.
P.

140 GENERAL PRINCIPLES OF THERMOCHEMISTRY.

TABLE XX. — HEAT OF VOLATILIZATION (LATENT HEAT) OF THE ELEMENTS AND
THEIR PRINCIPAL COMPOUNDS, REFERRED TO THE SAME GASEOUS VOLUME
(22'32 LITRES) UNDER ATMOSPHERIC PRESSURE.

Names.

Formulas.

Molecular
Weights.

Latent
Heat.

Authorities.

Bromine (liquid) • • .

Br,

160

7-2

R.

I*

254

6'0

F.

Sulphur (liquid) . . . .

S.

64

4-6

F.

Hg

200

15-4

F.

Water .... ...

H»O

18

9'65

R.

NH3

17

4'4

R.

Hydrofluoric acid .
Nitrogen monoxide
Nitric peroxide
Nitrogen pentoxide (liquid)
Nitric acid ...

HF

N20

N02

HNO3

20
44
46
108
63

7-2
4-4
4-3
4-8
7-25

Guntz.
F.
B.
B.
B.

Sulphurous anhydride .
Sulphuric anhydride (solid)
Carbonic acid (solid) .
Carbon disulphide . . .

S02
S03
C02
CS2
HCN

64
80
44
76

27

6-2
11-8
6-1
6-4

5-7

F.
B.
F.
R.

B.

Cyanogen chloride

CNC1
CTT

61-5
70

8-3
5-25

B.
B.

Diamylene

r H

140

6-9

B.

elk20

78

7-2

R.

Terebenthene

CL H

136

9-4

R.

Methyl alcohol . . .

CH2(H20)

32

8-45

R.

Ethyl alcohol

CoH/HoO)

46

9-8

R.

Aldehyde . .

C2H4O

44

6-0

B

Acetone

C3H6O

58

7-5

R.

Formic acid

CH2Oo

46

4-8

B. & Og.

C2H4O2

60

5-1

B. & Og.

Ethyl acetate

CoH/CoH.Oo)

88

10-9

R,

Ordinary ether

CoH/CoH O)

74

6-7

R.

The calculation of the pressures exerted at the moment of
decomposition of an explosive substance, requires not only the
knowledge of the heat disengaged by the transformation, but
also that of the specific heat of the products of the explosion
and their volume. It is equally necessary to know the specific
heat of the component bodies, in order to know the effects of a
certain heating on these bodies. Finally, the volume which
they occupy for a given weight, and which is deduced from
their density, plays an essential part in the valuation of the
density of charge and specific pressure (p. 28). These con-
siderations have induced the author to give the following
tables : —

TABLES.

141

TABLE XXI. — SPECIFIC HEATS OF SUBSTANCES WHICH CAN BE OBSERVED IN
THE STUDY OP EXPLOSIVES. — GASES.

Specific Heat at constant pressure,

referred

Names.

Formulae.

Molecular
Weights.

to Igrm.

to the mole-
cular
weight
under

volume

22-32 litres.

Hydrogen

H2

2

3-410

6-82 1

o

32

0-217

6-96

Nitrogen

N2

28

0-244

6-82

01.

71

0-121 (0°-200°)

8'58

Carbonic oxide .

CO

28

0-245

6-86

Nitric oxide . . .

NO

30

0-232

6-96

Nitrogen monoxide
Carbonic acid .

N20
C02

44

44

0-226 (0°-200°)
0-215 (0°-200°)

9-94
9-50

Sulphurous anhydride
Water vapour .
Hydrochloric acid gas .

SO2
H20
HC1

64
18
36-5

0-154 (0°-200°)
0-480 (128°-220°)
0-185

9-86
8-64
6-75

Sulphuretted hydrogen gas

H2S

34

0-243

8-30

Ammonia gas ....

NH3

17

0-535 (0°-200°)

9-11

CH4

16

0-593 (0°-200°)

9-50

Ethylene ....

C2H4

28

0-404 (0°-200°)

11-30

1 These numbers represent small calories. The specific heats at constant
volume are deduced from these by subtracting the constant 2.

TABLE XXII. — SPECIFIC HEATS OF SUBSTANCES WHICH CAN BE OBSERVED IN
THE STUDY OF EXPLOSIVES. — SOLIDS AND LIQUIDS.

Specific Heats refen

ed

Names.

Formulaa.

Equiva-
lents.

to 1 gnn.

to the
equiva-
lent
weight.

ELEMENTS.

0-203 solid

q.O

Sulphur ....

S

16

0-234 liquid (120° to

O LA

3-7

150°)

Phosphorus

P

31

rO-19 solid
\0-20 liquid

5-9
6-3

Arsenic

As

75

0-081

6-1

Antimony ....
Bismuth ....

Sb
Bi

122
210

0-051
0-031

6-2
6-5

Tin

Sn

59

0-055

3-3

Carbon

C

12

(0-202 graphite, coke
\0-24 1 calcined wood

2-4
2-9

Fe

28

0114

3-2

Zinc

Zn

32-5

0-096

3-1

Copper .
Mercury
Lead .

Cu
Hg
Pb

31-5
100
103-5

0-095
0033
0-0314

3-0
3-3
3-3

Silver .

Ag

108

0-334

6-2

Platinum

Pt

98-5

0-324

3-2

Gold .

Au

98-5

0-324

3-2

14.2 GENERAL PRINCIPLES OF THERMO-CHEMISTRY.

SPECIFIC HEATS OP SUBSTANCES WHICH CAN BE OBSERVED IN THE STUDY OF
EXPLOSIVES — SOLIDS AND LIQUIDS. — (Continued.)

Names.

Formula.

Equiva-
lents.

Specific Heats referred

to 1 grm.

to the
equiva-
lent
weight.

OXIDES.

Magnesia ....
Chromium oxide .

MgO
Cr20,

20
76

0-244
0-19

5-9
14-5

Alumina ....

A1203

51

0-217

11 '2

Ferric oxide

Fe20,

80

0-16

13-1

Zinc oxide ....

ZnO

40-5

0-13

5-4

/CuO

39-5

0-14

5-7

Copper oxide . . .

71

0-11

7.7

Lead oxide ....
Stannic oxide .

Sn03

111-5
75

0-051
0-093

5-7
14-0

Silica

Si02

60

0-195

11-4

CHLORIDES AND SUL-

PHIDES.

Ammonium chloride .

NH4C1

53-5

0-373

20-0

Potassium chloride

KC1

74-6

0173

12-9

Sodium chloride . .

NaCl

58-5

0-214

12-5

Barium chloride .

BaCla

104-0

0-090

9-3

Calcium chloride .

CaCl2

55-5

0-164

9-2

Silver chloride .

AgCl

143-5

0-091

13-1

Potassium sulphide .
Sodium sulphide .

K2S
Na2S

55-1
38-0

V

8-9 l
8-9 l

Iron sulphide . . .

FeS

44-0

0-136

6-0

Potassium ferrocyanide

K4Fe2Cy6

212-0

0-28

59-0

NITRATES.

Potassium nitrate . .

Sodium nitrate .
Barium nitrate . .
Strontium nitrate .
Lead nitrate

KN03

NaN03
Ba(N03)2
Sr(N03)2
Pb(N03)2

1-01

8-5
130-5
105-8
165-5

JO-239 solid
\0-332 liquid
0-278
0-15
0-18
0-11

242
33-5
23-7
19-0
19-1
18-2

Silver nitrate .

AgN03

17-0

0-143

24-4

Ammonia nitrate . .

NH4NO

80-0

0-455

36-4

SULPHATES AND CHRO-

MATES.

Potassium sulphate

K2S04

87

0-190

16-6

Sodium sulphate . .

Na2SO4

71

0-229

16-2

Calcium sulphate .

CaSO4

68

0-18

12-7

Barium sulphate . .

BaS04

116-6

0-11

12-6

Strontium sulphate

SrS04

91-8

0-14

12-4

Magnesium sulphate .

MgS04

60-0

0-22

13-3

Copper sulphate
Potassium hyposulphite
Sodium hyposulphite .

CuS04
K2S203
Na2S203

80-5
95
79

0-134
0-20
0-221

14-1

18-7
17-5

Potassium chromate .

K2Cr04

97

0-19

27-6

Potassium bichromate

K2Cr207

147

0-187

18-2

Lead chromate . .

PbCrO4

161

0-09

14-5

CARBONATES.

Potassium carbonate .

K2C03

69-1

0-21

15-0

Sodium carbonate . .
Calcium carbonate .

Na2CO3
CaCO3

53-0
50-0

0-27
0-209

14-5
10-5

Barium carbonate .

BaCO3

98-5

0-11

10-7

Lead carbonate . .

PbCO8

134-0

0-145

10-7

1 Theoretical valuation.

TABLES.

143

SPECIFIC HEATS OF SUBSTANCES WHICH CAN BE OBSERVED IN THE STUDY OF
EXPLOSIVES — SOLIDS AND LIQUIDS. — (Continued.)

Specific Heats referred

Names.

Formulae.

Equiva-
lents.

to 1 grra.

to the
equiva-
lent

weight.

CHLORATES.

Potassium chlorate

KC1O3

122-6

0-21

25-7

Potassium perchlorate

KC104

138-6

0-19

26-3

WATER, ACIDS, ORGANIC

COMPOUNDS.

Water

H2O

9

(1-0 liquid
\0-50 solid

9-0
4-5

Nitric acid .

HN03

63

0-445 liquid

28-0

Sulphuric acid
Benzene

H2S04
C«H6

49

78

0-34 liquid
0-44 liquid

16-7
31-0

Alcohol

C2H6O

46

0-595 about 20°

27-3

Glycerin
Mannite
Cane sugar

O.H.OJ

C12H22On

92

182
342

0-591
0-324
0-301

54-4
59-1
103-0

The specific heats of solid compounds can be approximately
calculated from the sum of those of their elements ; the latter
being taken, not with the real values which they possess in
the free state, but with the values calculated by Kopp from the
mean of the observed values for their compounds. He has thus
obtained the following empirical values referred to the equiva-
lent weights : —

6-4 for K, Li, Na, Rb, Tl, Ag, As, Bi, Sb, Br, I, Cl.
5-4 for P.
5-0 for F.
3-8 for Si (28).

3-2 for Al, Au, Ba, Sr, Ca, Cd, Co, Cr, Cu, Fe, Hg, Ir, Mg, Mn, Ni, Os, Pb,
Pd, Pt, R, Sn, Ti, Mo, N, Zn, Se, Te.
2-7 for S (16), B (11).
2-3 for H.
2-0 for 0 (8).
1-8 for C (12).

144 GENERAL PRINCIPLES OF THERMOCHEMISTRY.

TABLE XXIII. — DENSITIES AND MOLECULAR VOLUMES or SOME BODIES.

Name.

Symbols.

Equiva-
lents.

Density.

Molecular
Volume.

Sulphur

S

16

2-04

cub. cms.
7-9

i'3-5 diamond

1-7

c

9*27 crrnnhifp

O.7

AAt grapiiii/c
1-57 amor ph.

w 1

3-8

Copper

Cu

31-6

8-94

3-5

Lead

Pb

•inq.K

n-4

9-1

Silver

Ag

il/O *J

108

-LA T

10-47

ft-3

Fe

00

7.0

O.f»

Tin

Sn

to

KQ

1- o

7-3

o o

Q.I

Mercury

Sg

O«7

100

13-59

O 1

7-35

Zn

09. K

8-Q

4-7

Magnetic iron oxide

Fe304

O£i O

116

D »7

5-9

A i
23

Lead oxide . .

PbO

111-5

9-36

11-9

Tin oxide . .

SnO2

75

6-71

11-2

Chromium oxide

Cr2O8

79

5-2

15

Alumina . .

A1208

51-5

3-5 to 4-1

15 to 12-5

Silica . . .

Si02

60

2-65

23

Potassium chloride

KCl

74-6

1-94

38

Sodium chloride

NaCl

58-5

2-15

27

Barium chloride .

BaCl2

104-6

3-70

28

Strontium chloride
Ammonium chloride

SrCl2
NH4C1

79-3
53-5

2-80
1-53

28
35

Potassium cyanide

KCN

65-1

1-52

43

Potassium nitrate .
Sodium nitrate .

KN03

NaN03

101-1

85

2-06
2-20

49
39

Barium nitrate .

Ba(N03)2

130-5

3-18

41

Lead nitrate

Pb(N03)2

165-5

4-40

38

Silver nitrate

AgN03

170

4-35

39

Ammonium nitrate

NH4NO3

80

1-71

41

Potassium carbonate

K2C03

69-1

2-26

31

Sodium carbonate .
Barium carbonate .

Na2C03
BaC03

53

98-5

2-46
4-30

21-5
23

Strontium carbonate
Calcium carbonate .

SrC03
CaC03

73-8
50

3-62
2-71

20
18

Potassium sulphate
Sodium sulphate .

K2SO4
Na2S04

87-1
71

2-66
2-63

33
27

Barium sulphate .
Strontium sulphate

BaSO4

SrSO4

116-5
91-8

4-45
3-59

26
26

Calcium sulphate .

CaS04

68

2-93

23

Potassium chlorate

KC1O3

122-6

2-33

52-6

Potassium bichromate

K2Cr2O7

147

2-69

55

Cane sugar ....

C12H22On

342

1-59

215

( 145 )

CHAPTER II.

CALORIMETRIC APPARATUS.

§ 1. GENERAL REMARKS.

1. THE author has carried out almost all the measurements of
the quantities of heat liberated or absorbed in his experiments
with the water calorimeter. It is very well adapted for deter-
minations concerning explosive substances. This instrument,
employed by Dulong and Regnault, and also by Thoinsen,
appears to offer the guarantees of the greatest accuracy. In
fact, the quantities determined by it approach as closely as
possible the theoretical definition of the " calorie " ; whilst
the ice calorimeter of Lavoisier and Laplace, as well as that of
Bunsen, and the mercury calorimeter of Favre and Silbermann,
determine different quantities, such as the weight of water
liquefied, or the expansions of certain liquids. The relation of
these quantities to the calorie must be found separately, by a
system of special experiments, and it is liable to incessant
variation, according to the conditions of the surrounding
medium. In the use of these instruments, therefore, all the
uncertainties of indirect measurements occur.

2. The conditions under which the calorimeter is employed
are very simple, and capable of being easily reproduced by all
chemists and physicists who desire to carry out similar experi-
ments. The measurements are, moreover, more promptly
executed, and the calculation easier than by any other method.
For the complete discussion of the process, the verification of
the thermometers, and the arrangements special to certain ex-
periments, the reader is referred to the author's "Essai de
Mecanique Chimique," where these subjects are more fully
treated.

§ 2. DESCRIPTION OF THE CALORIMETER.

1. The apparatus consists of three fundamental parts, viz. a
calorimeter; a thermometer; an envelope. The annexed

L

146

CALOEIMETRIC APPARATUS.

sketch will give a sufficient idea of the apparatus (scale one-
fifth).

2. The calorimeter, properly so called, consists of a platinum,
brass, or glass vessel, with very thin walls, goblet shaped, pro-
vided with various fittings, and placed on three cork points.
We shall now describe it in detail.

In the greater number of the experiments, a cylindrical
platinum vessel, capable of containing at least 600 cub. cms. of
liquid, was used. It is 0120 metre in height by 0'085 metre in

Fig. 19. — Berthelot's Calorimeter, with its Envelopes.

GG, calorimeter of platinum; C, the cover; 09, calorimetric thermometer; EE,
silver-plated envelope; C', cover of same; HH, double envelope of tin plate, filled
with water ; C", cover of same ; AAA, stlrrer ; tt, its thermometer ; <p<j>, jacket of
thick felt covering the tin plate envelope.

diameter, and weighs 63 '43 grms. It is provided with a
platinum cover, fixed with a bayonet joint on the edges of the
cylindrical vessel, and pierced with various holes for the passage
of the thermometer, stirrer, conducting tubes for the gases,
liquids, etc. This cover weighs 1218 grms.

It is only employed in certain experiments, the calorimeter
being for the most part uncovered.

In experiments in which the equilibrium of temperature is

STIRRER.

147

almost instantaneous, the cover and the stirrer may be omitted,
and the thermometer itself employed to agitate the liquid, which
simplifies operations.

Under these conditions the calorimeter is very simple, as will
be seen. Keduced to water it is equivalent to 3 grms. to 4 grms.,
according to the accessory pieces, that is to say, that its calori-
metric mass does not exceed the two-hundredth part of the
mass of the aqueous liquids which it contains, a circumstance
which is very favourable to accuracy in experiments.

The author has also used several other platinum calorimeters,
one with capacity of 1 litre, which has served for the greater
number of his experiments on the detonation of gases, another
of 2-5 litres.

In certain experiments where it was necessary that contact
with the air should be completely avoided, glass phials containing
700 cms. to 800 cms. have been
used as calorimeters, always plac-
ing them in the same protecting
envelope. These instruments
give measurements which are the
more exact the larger they are,
but on the condition of con-
suming larger weights of the
substances. This limits the use
of the large instruments. On
the contrary, the small ones are
more subject to corrections for
cooling which may be neglected
with calorimeters of half a litre
and upwards, for the duration of
an ordinary experiment (one to
two minutes) and whenever the
excesses of temperature remain
less than 2°.

3. Stirrer. — In the experiments
in which the stirring of the water
by means of the thermometer
was insufficient, or presented any
difficulty, a stirrer of special form
was employed, superior to those
hitherto used, because it more
completely mixes all the layers
of water, with less expenditure
of force.

This stirrer (Fig. 20) consists of four wide helicoidal blades,
A A' A" A/" very thin, inclined at about 45° to the vertical and
normal to the internal surface of the cylinder employed as a
calorimeter. They are mounted on a frame formed of two

L2

Fig. 20.

148 CALORIMETRIC APPARATUS.

horizontal rings, B B', which hold the frame together at its ends,
and of four strong vertical rods, the whole being in platinum or
brass, as may be required.

The blades, about O'OIO metre in width, and the rings of equal
breadth, are arranged so as to form a frame, concentric to an
internal cylindrical space, the whole being in its turn enveloped
and almost touched by the cylindrical vessel, V V, which con-
stitutes the calorimeter.

Two of the vertical rods are prolonged about 0'15 metre above
the calorimeter, and joined at their upper end by a half ring of
wood, C C, of suitable width and thickness. The lower ring is
provided with four small feet or prolongations, a few millimetres
in length, and arranged so that the stirrer rests on their rounded
ends, at the bottom of the calorimeter.

The whole may be seen, in the centre of the calorimeter
(Fig. 20). In the cylindrical space surrounded by the stirrer
are placed the thermometer and suitable apparatus.

In order to employ this stirrer, the half ring is held in the
hand, or by some mechanical appliance (turn-spit, hydraulic, or
electro-magnetic motor, etc.), the stirrer is lifted a few millimetres,
and a horizontal and rotary movement around its vertical axis
is imparted to it. This movement is alternating, and comprises
an arc of from 30° to 35°. In consequence the water in the
calorimeter is impelled towards the centre, and at all heights at
the same time, being sharply thrust forward by the heficoidal
blades, which strike the water at an angle of 45° with the
vertical.

The degree of perfection which is hereby attained in the
mixture of the layers, and the promptitude with which this
result is obtained, even with a slight effort and slow movement,
are surprising.

Besides, the stirrer not coming continually out of the liquid,
as happens with stirrers moved up and down, is not exposed to
the very sensible evaporation to which the latter give rise, nor
to the causes of error which result therefrom.

4. The calorimeter just described may be employed under
extremely varied conditions. A full account will be found in
the "Essai de Me'canique Chimique" and in the author's
"Memoires." Some of the special instruments employed for
effecting chemical reactions in the interior of this calorimeter
will be described in the following chapters, in connection with
the experiments for which they have been constructed.

§ 3. DETONATOR OR CALORIMETRIC BOMB.

1. A description will now be given of the apparatus used to
measure by detonation both the heat of combustion of hydro-
carbon gases, or by an inverse process the heat of formation of

HEAT OF COMBUSTION OP GASES. 149

the oxygen yielding gases such as nitrogen monoxide and
dioxide, the apparatus employed not having been described in
the work previously quoted.

2. The method consists in mixing in a suitable vessel the gas
or vapour with the proportion of oxygen strictly necessary to
burn it completely, or even with a slight excess of oxygen when
this excess is not detrimental ; then, in causing the explosion of
the mixture in a closed vessel, and at constant volume. The
detonator having been previously placed in a calorimeter, the
heat produced is measured. By proceeding in this manner,
the combustion lasts only a fraction of a second, and is always
total, at least for gases properly so called ; in short, the calori-
metric measurement is effected in the shortest possible time —
that is to say, under the conditions of the greatest accuracy.

3. From this measurement is deduced, by calculation, the heat
liberated by the total combustion of the gas, simple or com-
pound. If, further, the sum of the quantities of heat liberated
by the combustion of the elements, when the gas is compound,
be known, it is sufficient to deduct from this sum the heat of
combustion of the said compound gas to obtain the heat of
formation of this gas, by means of its elements.

For example, marsh gas, CH4, taken at the weight of 16 grms.,
liberates, when burning at constant pressure, 213*5 Cal. Now
its elements liberate respectively, for C = 12 grms., taken in the
diamond form, 94 Cal., and for H4 = 4 grms., 138 Cal. ; hence
we conclude that the formation of marsh gas from its elements
C (diamond) + H4 = CH4; liberates + 94 + 138 - 213-5 =
+ 18-5 Cal.

4. The same method has enabled the author to measure in an
inverse sense the heat of formation of nitric oxide employed
as oxygen yielding gas. This gas, mixed with hydrogen, does
not detonate under the influence of the electric spark ; but it
explodes violently when mixed with ethylene or cyanogen.
Such a mixture has, therefore, been made in the proportions
strictly necessary for total combustion, exploded in the apparatus,
and the heat liberated measured.

The same experiment has been made with the same com-
bustible gases and pure oxygen.

This having been done, it is sufficient to deduct the heat
liberated in the first case from that produced in the second, in
order to obtain the heat of formation of nitric oxide by its
elements without any other data than these two intervening in
this calculation. In this way we find a negative number, viz.
- 21-6 Cal. for N + 0 = NO (30 grms.), which means that the
combustion of an oxidisable body, effected by nitric oxide,
liberates more heat than the same combustion effected by pure
oxygen.

Thus nitric oxide is formed from its elements with absorp-

150

OALORIMETRIC APPARATUS.

tion of heat, and contains more energy than the oxygen and
nitrogen which constitute it. This circumstance is all-important,

for it explains the combustion
power of the oxygenated com-
pounds of nitrogen.

5. This being established, we
proceed to describe the apparatus
employed, and to give some types
of experiments in order to charac-
terise the method. As for the forms
of the apparatus, they belong to
two models — the ellipsoidal and
the semi-cylindrical bomb, the
method of closing these two models
being slightly different. But the
introduction of the gases, their
extraction, the ignition, and the
measurement of the heat liberated
are always effected in the same
manner.

Fig. 21 represents the calori-
rnetric bomb employed for the
author's first measurements. Its
capacity is 218 cms., and its value
in water 51 grms.

It is formed of a receiver, B'B',
and of a cover, BB (Fig. 22), held
together by a screw joint provided
with lugs, 00, both of steel plate,
2*5 mms. in thickness. They were
electro-plated internally with a very thick layer of gold, weighing
about 22 grms., which resisted all the explosions. At first the

bomb was plated internally with
platinum, but platinum thus de-
posited does not stand prolonged
use. After a certain number of
observations, the platinum is raised,
I or eliminated, during cleaning, and
- the exposed iron becomes oxidised
during the explosions, especially
when water is formed. Platinum
electro-plating was therefore com-
pletely abandoned. The weight
of the gold fixed on the interior
should be determined by special
weighings, so as to be able to find the value of it in water,
simultaneously with that of the steel. The exterior surface of
the bomb was also nickel-plated, to render it less oxidisable.

Fig. 21. — Calometric bomb
(section).

Fig. 22.— Cover.

CALORIMETRIC BOMB.

151

The cover carries laterally an insulating ivory fitting traversed
by a platinum wire,//, which is fitted with a small screwed
portion, which holds it in the ivory. By this wire the electric
spark is made to pass.

In every experiment, before closing the apparatus, a small
mica disk, pierced in the centre, is fitted to the surface of the
ivory to protect the latter from the flame of the explosion.
The gases are introduced at the outset, and extracted at the end

Fig. 23. — Bomb suspended in the calorimeter.

by the aid of a mercury pump, combined with an apparatus
similar to Eegnault's eudiometer, but greater in capacity (half a
litre) ; this introduction being effected through an orifice, p,
which can be stopped at will by the screw W, fitted with a
head 0 and a channel KK'.

Fig. 23 shows the calorimetric bomb in place inside the
calorimeter, with its supports and the glass three-way cocks for
operating it.

152

CALORIMETRIC APPARATUS.

6. M. Golaz also constructed for the author another apparatus
of a similar form, whoUy of platinum internally, Covered
externally with sheet steel. The screw and the tube which it
traverses are entirely of platinum, which allows of passin-
chlorine, and sulphuretted or acid gases through it. The con*

Fig. 24.

struction of this platinum

Fig. 25.

screw is a real masterpiece of

execution.

Fig. 24 is the drawing of this apparatus complete Fig 25
represents the receiver apart. Fig. 26, the cover fitted with the
closing screw. Fig. 27, the tightening piece, F F, of the cover

Fig 26.— Cover. Fig 27.— Tightening piece. Fig 28.— Auxiliary nut.

Lastly, Fig. 28, the auxiliary nut E, fitted with two pins a, a',
for screwing up the preceding piece. This nut does not form
part of the apparatus immersed in the calorimeter.

The second apparatus has an internal capacity equal to
247 cms., contains 662 grms. of platinum and 419 grms. of steel,
and is equivalent in water to 70*4 grms.

Its dimensions have been arranged so as to make it act in the
calorimeter of 1 litre, containing only 550 grms. of water.

7. By proceeding in this way, the elevation of temperature
may amount to 1'5° to 2'0°.

EXPERIMENTAL DETAILS. 153

The calorimetric measurements, carried out to within ^J0 of
a degree, involves a smaller error than by the old method,
seeing that the combustions are generally total, and the correc-
tions extremely reduced, through the short duration of the
experiment.

However, the accuracy is limited by the weight of the
substance on which we are obliged to operate ; the weight of
the carbonic acid formed generally not exceeding 0*200 grms. to
0*300 in the most favourable cases.

The quantity of gas burnt may be estimated either from its
initial volume, or from the weight of its products.

The estimation of the initial volume presents great difficulties,
owing to the necessity of taking into account the internal spaces
of the tubes joining the bomb to the receivers in which the
gases are measured. It has, however, been effected in the case
of hydrogen.

But in the majority of cases it is preferable to weigh, after
combustion, the gaseous products, which reduce themselves
ordinarily to carbonic acid. With this object the gases are
collected from the bomb, after explosion, by means of a mercury
pump, and passed through a tube filled with pumice-stone and
sulphuric acid, which dries them, then through a Liebig tube
filled with potash, followed by a TJ tube filled with solid potash,
in order to absorb the carbonic acid. The bomb is thrice
filled with air free from carbonic acid, in order to clear out the
gases completely, and each time the gases extracted from the
bomb are passed through the Liebig tube. The latter, and the
tube filled with solid potash, are finally weighed. It is necessary
to further make the following verifications.

In the first place, the combustion of each gas is effected in
the eudiometer, over mercury, in order to see that it is pure and
gives the theoretical figures.

Then a similar combustion is carried out in the calorimetric
bomb, the whole of the gases are extracted from it by the pump,
and collected over mercury. After the absorption of the car-
bonic acid and of the oxygen, it is ascertained whether there
remains any trace of combustible gas (carbonic oxide, hydrogen,
marsh gas, etc.).

This verification is made, first with the aid of acid cuprous
chloride, then by means of a fresh attempt at burning, by a
proper quantity of oxygen. If nothing burn, there is added
to the mixture the half of its volume of electrolytic gas, and the
attempt is repeated.

In this manner it has been ascertained that the combustions
are total with all hydrocarbon gases properly so called, such as
methane, methene, ethylene, ethene, ethane, dimethyl, pro-
pylene, etc.

8. The combustion of nitrated, chlorinated, brominated, iodated

154 CALORIMETRIC APPARATUS.

and sulphuretted gases can likewise be effected in the platinum
detonator which has just been described.

9. Not only are permanent gases burnt in the apparatus
above described, but it is easy to burn in them every vapour
the tension of which is sufficient for it to be completely trans-
formed into gas in the volume of oxygen capable of completely
burning it. In this case the liquid is weighed in a small sealed
glass bulb, and the bulb is placed in the bomb ; the latter is
closed and filled with oxygen, then by a few shocks the bulb is
broken. In a few moments after vaporisation has taken place
the bomb is placed in the calorimeter. After five or six minutes,
during which the thermometer is observed, the gas is exploded,
and the carbonic acid is collected and weighed as above.

By proceeding in this manner we have the advantage of
being able to control the weight of carbonic acid obtained, by
the weight of the original liquid.

In the case of aldehyde, glycolic ether, hydrocyanic acid,
hydrochloric and hydrobromic ethers, methylic and ethylic
alcohols, etc., the operations have been carried out in the above
way.

The combustions are total for every vapour having a consider-
able tension, such as that of bodies boiling below 50°.

But for the less volatile bodies, as benzene, there is no longer
the same certainty of total combustion, probably owing to the
condensation of some trace of matter on the walls and in the
grooves of the apparatus. In this exceptional case, the detona-
tion method loses some of its advantages and requires corrections
similar to the ordinary method by combustion.

10. The figures obtained by detonation have not exactly the
same significance as those obtained in the ordinary heats of
combustion ; the latter are carried out at constant pressure, the
former at constant volume. By this method numbers are
obtained which are better adapted to the majority of theoretical
discussions.

It is, moreover, easy to pass from the numbers obtained at
constant volume to those which would be obtained at constant
pressure. According to the formula given above

Qtp = Qtv + 0-5424 (N - N') + 0-002 (1ST - W)t.

Take, for example, the combustion of carbonic oxide at 15°.
CO 4- 0 = C02 liberates at constant volume + 68*0 Cal. In
order to pass from this to the heat liberated at constant pressure
we should note that on one hand CO occupies a unit of volume,
0 a half-unit. Therefore

N = l£.

On the other hand C02 occupies a unit of volume.

N' = l
N - N' = .

EXPLOSION OF ETHANE AND OXYGEN. 155

At 0° we should therefore have for the difference between the
heats of combustion at constant pressure and at constant
volume

+ 0-54 X J = + 0-27.

At 15° to this figure must be added + 0'03, which raises the
correction to + 0*30. The heat of combustion of carbonic oxide
at constant pressure and at 15° will therefore be + 68*3 Cal.
Take again the combustion of ethane —

C2H6 + 07 = 2C02 + 3H20

ff = l + 3J = 4J
N' = 2 (assuming water liquid)

N - N' = 2£.

The difference between the two heats of combustion is ex-
pressed at 15° by + 1425 Cal.

The correction relative to condensation should in principle be
reduced by a small quantity on account of the appreciable
tension of water vapour at 15°, but this quantity may be
neglected, owing to its smallness, in the present calculation.

11. We should, on the contrary, bear in mind that the
correction due to the formation of water vapour is very appreci-
able in the calculation of the heat of combustion at constant
volume, as well as at constant pressure, seeing that it represents
the formation of gaseous water, which liberates less heat than
the formation of liquid water. It has been verified in all these
experiments from the internal capacity of the bomb, and con-
formably to Kegnault's tables for the tension of water vapour
and the vaporisation heat of water, at the temperature of the
calorimeter.

12. More than three hundred explosions have been effected in
these instruments. No accident has occurred in the instruments
themselves, in spite of the magnitude of the sudden pressures
developed during the explosions. These pressures are estimated
at fifty atmospheres in certain cases where previously com-
pressed gaseous mixtures have been operated upon.

13. We have, however, twice observed the spontaneous
explosion of the gaseous mixtures while they were being shaken
in closed and very dry glass vessels, with mercury. This very
serious and singular accident appears due to internal electric
sparks, produced by the friction of the mercury on the glass of
the flasks, these being held in the hand and realising conditions
of condensation similar to those of a Ley den jar.

14. We shall now expound the data of a determination, with
the object of showing the method followed in the experiments,
verifications and calculations.

Ethane. — The gas was prepared by the electrolysis of
potassium acetate. It was freed from carbonic acid by potash,
from ethylene by bromine, and from carbonic oxide by a pro-

156 CALORIMETRIC APPARATUS.

longed shaking over mercury with its own volume of acid
cuprous chloride.

Its composition was verified ; 102 vols. of this gas, burnt in
the eudiometer by a slight excess of oxygen (360 vols.), produced
200 '5 vols. of carbonic acid. The total diminution of the
volume after explosion and absorption of carbonic acid, amounted
to 451 vols. ; the remainder, deprived of the excess of oxygen by
hydrosulphite, yielded two volumes of nitrogen.

According to the formula C2H6 -f 07 = 2C02 + 3H20,
100 vols. of combustible hydrocarbon should have produced
200 vols. of carbonic acid, occasioning a total diminution of 450
vols.

The gas analysed was therefore ethane, containing only two
hundredth parts of nitrogen, which have no appreciable influence
on the heat of combustion.

The foregoing results show that the gas employed is really
ethane, and that its combustion by a slight excess of oxygen is
total. However, it has appeared useful to prove that the com-
bustion is effected in the same manner in the calorimetric
bomb, that is to say, that the above equation is applicable to
the measurement itself.

With this object, the bomb was filled with the mixture of
ethane and oxygen in suitable proportions, placed in the water
of the calorimeter, and the gases exploded ; then the whole of
the gases contained in the bomb were extracted by the aid of
a mercury pump, passed into a large test-tube, in which the
carbonic acid was absorbed by potash and the excess of oxygen
by hydrosulphite.

It is known that this reagent does not act either on carbonic
oxide or on hydrocarbons. The residuum thus obtained under-
went no diminution of volume by cuprous chloride, it was not
combustible, and, mixed with half of its volume of oxygen, it
did not explode under the influence of the electric spark. In
another trial, for greater certainty, the analogous residuum was
mixed with its own volume of electrolytic gas after adding
oxygen to it, in order to burn the last traces of combustible
gases, if such existed. But this test showed that there remained
in the residuum nothing but nitrogen. The combustion of
the ethane- in the calorimetric bomb had therefore been total,
as well as in the eudiometer.

The following are the figures of a calorimetric experiment
performed on October 28, 1880.

200 cub. cms. of ethane and 720 cub. cms. of pure oxygen
were mixed over mercury, and the mixture was passed, with the
aid of a system of suitable tubes, into the calorimetric steel
bomb lined with platinum, shown on p. 152, a vacuum having
previously been created in the bomb with the aid of the mercury
pump. The cock of the bomb was closed, and the latter was

DETAILS AND RESULTS OF EXPERIMENT. 157

introduced into a platinum calorimeter of a capacity equal to
1 litre. Owing to the displacement produced by the bomb
550 cub. cms. of water sufficed to fill the calorimeter and
cover the bomb, with the exception of the screw-cock. The
thermometer, which served at the same time for stirrer, was
put in place. The value in water of the calorimeter, the
thermometer, and of the bomb amounted to 770-4 grms.

The whole was left at rest for some time, in order to allow
the equilibrium of temperatures to become established. This
accomplished, the following is the course of the thermometer : —

At the outset 13-295°

After 1 minute 13'295°

„ 2 minutes 13-295°

„ 3 „ 13-295°

„ 4 „ 13-295°

„ 5 „ 13-295°

The explosion is then caused by passing a single spark,
supplied by a very small induction coil and a bichromate cell.
The noise of this explosion is faint, but appreciable with ethane ;
this gas, and diallyl, have produced the greatest noise. Often,
in this kind of experiments, the noise of explosion is not even
heard, and its existence only known by the heating of the water
in the calorimeter.

The following is the continuation of this experiment : —

After 6 minutes (from the outset) 14-740°

„ 7 „ 14-745°

8 14-735°

„ 9 „ 14-725°

„ 10 „ 14-715°

11 14-705°

„ 12 „ ... 14-695°

The readings are suspended.

It will be noticed how short the combustion is, and how
sharply defined are the phases of the calorimetric measurements.
This done, the carbonic acid is extracted from the bomb with
the aid of the mercury pump, it is dried by passing it through
a curved tube of concentrated sulphuric acid, and a U tube
filled with pumice-stone and sulphuric acid, then it is slowly
passed through a Liebig tube with liquid potash, followed by a
small U tube with solid potash.

The extraction being carried out, and the vacuum established
down to a few millimetres of mercury, air (freed from carbonic
acid) is allowed to enter the bomb, then this air is extracted by
the pump and passed in its turn through the potash. This
operation is thrice repeated in order to extract the last traces of
carbonic acid formed by the combustion. The extraction lasts
altogether about a quarter of an hour. When it is accomplished,
the Liebig tube joined to the U tube is weighed, the increase of

158 CALORIMETRIC APPARATUS.

weight being equal to the weight of carbonic acid formed. It
has been found to be 0*2090 grm. in the above experiment.
This being established, let us calculate the heat produced. It
is equal to the product of the masses reduced to water and
multiplied by the variation of temperature ; Sft + At

Sju = 550 + 77'4 = 6274,

A* = 14745 - 13-295, or 1-45° + p,

p being the heat lost by cooling.

Now, in the initial period of five minutes which preceded
explosion there was neither gain nor loss. The maximum was
established one minute and a half after explosion. In the five
following minutes (final period) the loss was regular and equal
to 0*01° per minute. This being determined between the fifth
and the sixth minute the loss may be estimated at the half, or
0-005°; between the seventh and eighth it is 0-01°. The total
correction will therefore be 0'015, which makes t = 1465°.

The heat liberated is 6274 X 1465 = 919-14 cal. But
this figure does not correspond to a total transformation of
ethane into gaseous carbonic acid and liquid water. In fact,
a certain quantity of water retains the gaseous state in the
interior of the bomb. This quantity is easy to calculate, for it
corresponds to the maximum tension of water vapour at 1474°,
viz. 12*5 mms. according to Eegnault's tables. The capacity
of the bomb being 247 cub. cms., and the density of water
vapour at 14*7° being supposed theoretical (which is very near
the reality, according to Eegnault's experiments), the real
weight of the gaseous water remaining in the bomb may be
estimated at

- 1 2<F> 1

0-806 grm. x -^- X — X - - - = 0*0031 grm.
1000 760 1 + 0*00367 x 1474

Now, the vaporisation of this weight of water at 15°, still follow-
ing Eegnault, absorbs —1*85 cal., a quantity which must be
added with the contrary sign to 919*14 cal., which makes in
all + 920-99 cal. This is the heat liberated by the combustion
of the weight of ethane which gave 0-2090 grm. of carbonic
acid.

But 1 equiv. of ethane C2H6 = 30 grms., would have yielded
88 grms. of carbonic acid. It is therefore sufficient to calculate
the heat liberated by the formation of 88 grms. of carbonic
acid to obtain the heat of combustion of ethane at constant
volume, viz. 387*78 Cal.

At constant pressure this figure must be increased by 1425,
according to the formula on page 154, which makes 389*21 Cal.
This is the heat of combustion of ethane deduced from the above
experiment.

15. The method just described has been applied to the study

LIST OF GASES EXPLODED. 159

of the combustion of the following thirty gases and vapours, all
combustible bodies.1

Hydrogen.

Carbonic oxide.

Hydrocarbons, such as methane, ethane, ethylene, acetylene,
hydride of propylene, propylene, allylene, benzine, dipropargyl,
diallyl.

Oxygenated compounds, such as methylic ether, glycolic
ether, aldehyde, methylformic and ethylformic ethers, dimethylic
methylal.

Nitrated compounds, such as cyanogen, hydrocyanic acid,
trimethylamine, ethylamine.

Chlorated, bromated, iodated compounds, such as methyl-
chlorhydric, methylbromhydric, methyliodhydric, ethylchlor-
hydric and ethylbromhydric ethers, chlorides of methylene and
ethylidene.

Sulphuretted compounds, such as carbon disulphide.

It has also been used by an inverse process to measure the
heat of formation of the combustive gases, such as nitrogen
monoxide, and nitric oxide (p. 149), the heat of formation
of these gases being deduced from the difference between the
heats of combustion of the same carburetted (carbone) gas by
pure oxygen on the one hand, and by the oxygenated gas on
the other.

1 "Annales de Chimie et de Physique," 5e se*rie, torn, xxiji. p. 115 and
following.

( 160 )

CHAPTEE III.

HEAT OF FORMATION OF THE OXYGENATED COMPOUNDS OF NITROGEN.

§ 1. PRELIMINARIES.

1. POTASSIUM nitrate, otherwise termed nitre, or saltpetre, has
been employed for many centuries as an ingredient of gun-
powder. Its use was discovered by empirical means ; but theory
only commenced to throw a light upon it a century ago, when
the part played by oxygen in combustion was discovered by
Lavoisier, as well as the presence of a great quantity of oxygen
in potassium nitrate, but the difference between these two
substances as regards their explosive action has only become
clear within the last few years, as being due not to a difference
in chemical composition, but rather as explicable on thermo-
chemical grounds.1

The determinations in question presented extraordinary diffi-
culties, and the results were not realized at the first attempt.
They only reached their full accuracy after a series of experiments.

Attention has since been directed towards obtaining more
exact values, and the scope of the work has been extended to
the heat of formation of the various oxygenated compounds of
nitrogen, and iis theoretical importance has therefore consider-
ably increased. The following are the results obtained with
nitric oxide, which is the origin of most of the others.

§ .2. HEAT OF FORMATION OF NITRIC OXIDE.

1. The series of the five oxides of nitrogen, formed in propor-
tions varying according to simple ratios of weight and volume,

1 The measurement of the heat of formation of potassium nitrate involved
an elaborate series of experiments, based partly on the previous determinations
of Dulong, Hess, Graham, Favre and Silbermann, Andrews, Wood, Thomsen,
Deville and Hautefeuille, Bunsen and Schischkoff, etc., but largely on experi-
ments begun in 1870 by the author; and the following data, relating to the
heat of formation of the oxygen compounds of nitrogen, were an outcome of
this investigation. — EDS.

COMBUSTION OF CYANOGEN. 161

is one of the most important in chemistry. The knowledge of
the heats of formation of these oxides offers the more interest
as the two first are formed with absorption of heat from their
elements, while the three others are, on the contrary, formed
with liberation of heat from nitric oxide, which plays the part of
a true radical. A knowledge of the heats of formation of these
bodies is, moreover, indispensable to the theoretical study of
explosive substances, of which they go to form the greater part.
Unfortunately, the exact determination of these quantities pre-
sents great difficulties, as is the case with combinations which
cannot be brought about by direct synthesis.

The figure deduced by Favre and Thomsen, namely 267 Cal.,
from the action of chlorine on ammonia, was found to be wide
of the truth by the author, who estimated it at 12*2 Cal., and
this last figure has since been confirmed by Thomsen, so that all
values up to that date in which the formation of ammonia
intervenes have to be corrected by 14*5 Cal. But, before
applying such a correction to the heat of formation of the
oxides of nitrogen, the author endeavoured, with success, to
measure this more directly by comparing the heat of combustion
of methene and of cyanogen when burnt in oxygen and in
nitric oxide respectively. The results obtained were practically
identical, and the method admits of rapid and exact manipula-
tion, and the figures obtained are therefore incomparably more
valuable than the previous ones based on no less than nine
experimental data. The following are the details of the
experiment.

The combustible chosen was cyanogen or ethylene. It was
found that the slow combustion of a mixture of cyanogen,
ethylene or carbon disulphide with nitric oxide always pro-
duced an abundance of nitrous vapours; but this is avoided
by detonating the cyanogen and nitric oxide in the calorimetric
bomb.

1. Combustion of Cyanogen by free Oxygen.

The explosion of the following gaseous mixture: 26 grm.
CN + 02 = C02 + N liberated + 131'0, and + 1307, the mean
being : + 130*9 ; explosion at constant volume.

Hence we obtain the heat absorbed in the union of carbon
(diamond) and nitrogen.

C (diamond) + N" = CN absorbs - 36'9.

In another series of experiments made by burning a jet of
cyanogen in oxygen at constant pressure -I- 131 '6, for the heat
of combustion, and - 37'6, for the heat of formation, were the
figures obtained. The numbers found, whether at constant
pressure or at constant volume, can therefore be regarded as
identical, which one would expect, as the combustion of
cyanogen by free oxygen does not give rise to any change of

M

162 OXYGENATED COMPOUNDS OF NITROGEN.

volume. This remark applies likewise to the combustion
of cyanogen by nitric oxide.

2. Combustion of Cyanogen by Nitric Oxide.

The explosion of the following gaseous mixture, CN + 2NO
= C02 + N3 gave + 175-3 ; + 172-9 ; + 175-0 ; + 174-4 ; +
175 -3 ; the mean being 174-6 ; explosion at constant volume.

The difference between this number and the figure + 130-9
obtained with free oxygen under the same conditions, viz. the
value -|- 43*7, represents the heat liberated by the decomposition
of 2NO into its elements. According to these two data, the
union of nitrogen with oxygen to form nitric oxide (NO
30 grms.), N + 0 = NO absorbs - 21-8 Cal.

3. Combustion of Ethylene Try free Oxygen and Nitric Oxide.

Similar experiments were made with ethylene, and yielded
the same results. It is, therefore, unnecessary to enter into
details. It will be sufficient to state that the difference between
the numbers observed corresponding to the union of the elements
nitrogen and oxygen N -f 0 = NO, was - 21 -6 Cal.

§ 3. HEAT OF FORMATION OF NITROGEN MONOXIDE.

The heat of formation of nitrogen monoxide was measured by
exploding carbonic oxide mixed first with this gas and then
with oxygen and taking the difference of the two results.

1. Combustion of Carbonic Oxide by Oxygen.

CO (14 grms.) +0 = CO* liberated + 33'7 and + 34'4. The
mean, + 34'0, refers to the explosion at constant volume.

From this we pass to the heat of the combustion at constant
pressure 1 by adding 0'14, by reason of the condensation which
reduces 1 J vols. of the explosive mixture to 1 vol. ; we thus
obtain 3414 cals. This figure agrees almost exactly with that
previously obtained by the combustion of a jet of carbonic oxide
in oxygen, viz. + 34-09.2 It also agrees with the value obtained
by the wet process 3 with formic acid, by oxidizing on one hand
the formic acid, and on the other hand transforming it into
water and "carbonic oxide. By this method the combustion of
carbonic oxide gave -f 34'25.

2. Combustion of Carbonic Oxide by Nitrogen Monoxide.
CO + N20, 22 grms. = C02 + N2 liberated ; + 44'0 ; -f 451 ;
-f- 441, the mean being 44"4 ; explosion at constant volume.

1 " Essai de Me*canique Chimique," torn. i. p. 115.

2 " Annales de Chimie et de Physique," 5" seVie, torn. xiii. p. 13.

3 Same collection, torn. v. p. 316.

FORMATION OF NITRITES.

163

According to theory this number is the same at constant
pressure.

It follows from these figures that the formation of nitrogen
monoxide from nitrogen and free oxygen at constant pressure
N2 + 0 = N20 absorbs + 34'1 - 44'4 = - 10'3 ; or for N2 + 0
= N20 44 grms. - 20*6 Cal.

The heat absorbed in the formation of the monoxide (— 10*3)
is practically one half of the heat absorbed in the formation of
nitric oxide (— 21*6).

§ 4. HEAT OF FORMATION OF DISSOLVED AND ANHYDROUS
NITROGEN TRIOXIDE, AND THE NITRITES.

1. The heat of formation of nitric oxide being known, it
is easy to obtain from it those of the higher oxides ; for it is
easy to change nitric oxide, under conditions of calorimetric
experiments, into nitrogen pentoxide, tetroxide, and trioxide.

2. Conversion of nitric oxide into nitric acid by several
methods. One method consists in first forming a nitrite and
afterwards oxidizing it. In regard to the formation of nitrites,
nitric oxide and oxygen react very rapidly upon each other,
upon contact with an alkaline base, and are changed almost ex-
clusively into nitrites whatever be the relative proportions of
the two gases.1

This experiment was made in a closed vessel (Fig. 29) of a
capacity equal to 800 cub. cms.
almost filled with baryta water, the
strength and weight of which was
accurately measured.

This vessel served as a calori-
meter; it was surrounded by an
envelope, as in the annexed figure.

A calorimetric thermometer, 9,
was plunged into the vessel, pass-
ing through a large tube, K, at the
upper orifice of which it was fixed
by a small cork, b. The vessel itself
was closed by a large cork, pierced
with four holes, one for the passage
of the tube, K, another for that of
a tube, t, conducting the nitric
oxide and dipping into the liquid,
a third hole (hidden by the large

Fig. 29.

tube in the figure) received another tube for supplying the

oxygen, lastly a tube, s, for carrying away the excess of the gases.

Having introduced separately into the calorimetric apparatus

1 " Annales de Chimie et de Physique," 5e s£rie, torn. vi. p. 193.

M 2

OXYGENATED COMPOUNDS OF NITROGEN.

dry nitric oxide and oxygen, and shaken them incessantly, the
heat liberated and the increase of weight was measured and the
amount of trioxide and pentoxide formed was ascertained.
The weight of pentoxide formed is always very slight, it has
been taken account of in the calculations according to the data
on the following pages.

The following was found upon full calculation,

2NO + 0 4- BaO = (IST02)2Ba, dissolved : + 28'0 Cal.

3. Barium Nitrite. — In order to pass from barium nitrite in
solution to nitrous acid, it was necessary to make a special
study of barium nitrite itself, this salt being a perfectly
pure and well-defined body, and intended to serve as starting
point for other experiments on the respective transformation
of the nitrous acid and of the nitrites into nitric acid and
nitrates.

The barium nitrite was prepared by the reaction of nitrous
vapour (starch attacked by nitric acid) on a mixture of barium
carbonate and hydrate held in suspension in water. The barium
nitrite obtained was several times recrystallized, and its purity
verified by analysis.

This salt crystallizes in brilliant needle-shaped prisms,
gathered together without order. Very slow spontaneous
evaporation yields large, confused twin crystals, which have the
appearance of a rather acute, double hexagonal pyramid. This
is in reality a limiting form, belonging to the system of the
straight rhomboidal prism, and analogous to that of potassium
sulphate. The following are some thermal data relative to this
salt : —

One equivalent (N02)2BaH20, 123'5 grms. dissolved in 60
times its weight of water absorbs at 12°, — 4'3 Cal.

The dissolving of the anhydrous salt (N02)2Ba = 114*5 grms.
absorbs at 12°, - 2'84 Cal.

It follows from these figures that the reaction (N"02)2Ba solid
+ H20 liquid = (N02).2BaH20 solid, liberates -f T46.

The very weak solution of barium nitrite, decomposed by
dilute sulphuric acid, liberates for 1 equivalent, + 7*9 Cal.

Very dilute nitrous acid is set free under these conditions
without very sensible formation of nitric acid, as is proved by
adding potassium permanganate to it. Further, the formation
of barium sulphate, according to experiments made under the
same conditions of dilution and temperature, liberates 4- 18'50 ;
starting from sulphuric acid and the diluted base.

From these figures we may conclude that, N203 very diluted .
+ BaO diluted = Ba(N02)2 diluted, liberates + 10'6. This is
3 '2 Cal. less than nitric and hydrochloric acid, which shows
that nitrous acid must be ranked among the weak acids.

Dilute hydrochloric acid completely displaces nitrous acid

AMMONIUM NITRITE. 165

from alkaline nitrites, according to the thermal measurements,
while in presence of baryta, dilute acetic acid, weaker than
hydrochloric acid, gives rise to a division, variable according to
the relative propositions.

This division may be explained by partial dehydration, that
is to say, by the state of partial dissociation of the hydrate of
nitrous acid in its solution.

4. Ammonium nitrite. — This salt in the solid form has been
but little studied. The author having had occasion to prepare
it for thermo-chemical researches, several new facts have been
observed. It was prepared by decomposing pure barium nitrite
by ammonium sulphate in exactly equivalent proportions in
the cold. The filtered liquid was evaporated in vacuo over
caustic lime, at ordinary temperatures. The operation lasted
several weeks, and, owing to decomposition, even then the yield
was only thirty to forty per cent, of that required by theory.
It is necessary to evaporate, to complete dryness, and to pre-
serve the solid salt in vacuo, over caustic lime. The mass is
white, crystalline, somewhat elastic and tenacious, so that it
may be moulded between the fingers, and adheres to the sides
of the containing vessel. It is perfectly neutral, and corre-
sponds to the formula (NH4) N02. It is very diliquescent. At
the ordinary winter temperature, it decomposes very slowly ;
at that of summer more rapidly. Heated to 60° or 70° on the
water bath, it detonates violently after a few seconds. It
detonates also by a blow from a hammer. Its decomposition
disengages weight for weight about three quarters as much heat
as nitroglycerin, hence its activity. It cannot be kept in sealed
tubes, because they soon explode, from the pressure of the gases
generated. It is best kept as above.

If the decomposition take place with explosion, it yields only
nitrogen. When gradually decomposed by progressive heating,
the salt loses at first a little ammonia, and afterwards yields,
together with free nitrogen, a small amount of the monoxide,
nitric oxide, and trioxide vapour.

Its very slow decomposition, at the ordinary temperature,
yields nitrogen and water, without affecting its neutrality.

Aqueous concentrated solutions decompose more rapidly than
the dried salt in the cold, so that when agitated, they evolve
gas like champagne. Ammonium nitrite may be formed syn-
thetically by mixing together nitric oxide, ammonia, and oxygen.
The solid nitrite condenses on the walls of the tube in crystal-
line masses, apparently cubical in form.

The three gases immediately react on each other, but as they
do not contain the water necessary for the constitution of
ammonium nitrite, nitrogen is formed at the same moment.
0 + 2NH3 = 2N2 + 3H20.

2NO + 0 + 2NH3 + H20 = 2NH4N02.

166 OXYGENATED COMPOUNDS OF NITROGEN.

Both reactions, in fact, are simultaneously developed, but
the volume of the nitrogen collected is much greater than
that which should be produced if the whole of the available
water were changed into ammonium nitrite. In the above
experiments it represented more than double the theoretical
quantity, which is easily explained by the simultaneous de-
composition of a portion of the nitrite. An analysis showed
that the products did not contain any sensible proportion of
nitrate.

The following are various thermal data relative to ammonium
nitrite. NH4N02 (64 grms.) dry + 120 times its weight
of water at 12'5° absorbs in dissolving 475 Cal. The heat
liberated when dilute nitrous acid unites with ammonia may
be deduced from the heat liberated when ammonium sulphate is
precipitated by barium nitrite, N203 dilute + NH3 dilute + H2O
liberates -f 91.

The heat liberated by its decomposition into nitrogen and
water, NH4N02 solid = N2 4- 2H20 liquid, amounts, according to
the formula, to -f 73 '2 ; the water being gaseous, we should
have + 54 Cal.

5. Silver nitrite. — By double decomposition with solutions of
different degrees of concentration N203 dissolved, + Ag2O preci-
pitated = 2AgN02 dissolved, liberates -}- 3*36. N203 dissolved*
4- Ag20 precipitated = 2AgN02 crystallised, liberates + 121.

The heat absorbed in the solution of an equivalent of silver
nitrite is equal to — 874 Cal.

It is worthy of remark that the thermal formation of solid
silver nitrite 4- 821 exceeds that of silver nitrate -f 10*9, both
formations being reckoned from the diluted base and acids :
while the formation of the alkaline nitrates, such as solid
barium nitrate, liberates + 18*6, and that of ammonium nitrate
calculated from the same components 4- 187, figures which are
on the contrary higher than the heat of formation of the corre-
sponding nitrites. In fact, the formation of solid barium nitrite
liberates only + 13 '4, and that of solid ammonium nitrite
+ 13*8.

These relations deserve some attention, for they tend to
connect the nitrites with the chlorides and halogen salts, in the
case of which the thermal formation of the salts of silver,
calculated in a similar manner, exceeds even that of the alkaline
salts. Such relations are in conformity with the known
analogies between the group (N02) which plays the part of
a radical compound in the nitrites, and the simple radical
halogens, such as chlorine and its congeners; in other words,
Ba(N02)2 is here compared to BaCl2 and AgN02 to AgCl.

6. Formation of nitrogen trioxide. — The preceding numbers
concerning barium nitrite being known, the heat liberated by
the transformation of nitric oxide into dilute nitrous acid, is

NITRIC PEROXIDE. 167

deduced from them, thus : 2NX3 + O 4- water = N203 dilute
+ 28-0 - 10-6 = 174 Cal.1

From this figure, and from the heat of formation of nitric
oxide, is deduced the formation of dilute nitrogen trioxide from
its elements, nitrogen and oxygen. N2 4- O3 + water = N203
dilute, absorbs — 4*2 Cal.

The experiments relative to the formation of nitrogen trioxide
might be quoted here, but these experiments will be more con-
veniently described after those relating to nitric peroxide. The
numerical result will suffice at present : —

i(Na + 08) = i(NaOa) gas absorbs - 111 Cal., or for N203
- 22-2 Cal.

7. Formation of the nitrites from their elements. — According
to the above numbers, the thermal formation of the nitrites from
their elements liberates —

Salt Salt

dissolved. anhydrous.

Potassium nitrite, N + 02 + K = KN02 ... + 887 ...

Sodium nitrite, N + 02 + Na = NaN02 ... +84-0 ...

Ammonium nitrite, N2 + 02 + H4 = NH4N02 +60-0 ... +64-8

Barium nitrite,2 N2 + 03 + BaO = Ba(N02)2 + 26*8 ... + 29'6

Silver nitrite, N2 + 02 + Ag = AgN02 ... +2-7 ... +11-4

§ 5. HEAT OF FORMATION OF NITRIC PEROXIDE.

1. The heat of formation of this body was measured by two
inverse methods, and according to three distinct processes,
intended to control one another, viz.

(1) By synthesis or by the direct reaction of nitric oxide on
oxygen, both gases being employed in equivalent ratios.

(2) By the transformation of the already formed nitrogen
tetroxide into nitric acid by means of chlorine and water.
By the transformation of the same nitric peroxide into barium
nitrate by means of barium dioxide, whence we pass by calcula-
tion to the transformation effected by means of free oxygen.
The heat liberated by the direct metamorphosis of nitric oxide
and oxygen into dilute nitric acid, being known from former

1 Favre had estimated this quantity at -6'6 Cal. from erroneous data.
Thomsen calculated + 18'2, relying upon the union of three more exact
thermal data, one derived from the reaction of nitric oxide, and oxygen form-
ing nitric peroxide (+ 19'57), another from the dissolving of the latter body
in water (+ 7*75), a solution which he supposes to give rise to nitric acid and
nitrous acid in equal equivalents, the last datum being deduced from the
reaction of chlorine on the same solution, which it changes entirely into nitric
acid. This method is much more complicated than the one applied above,
and is founded on less sure reactions. However, the results coincide
sufficiently.

2 The heat of formation of this salt has been given from baryta only, the
heat of oxidation of barium being unknown. In the transformation of barium
nitrite as well as in that of the nitrate, this datum moreover suffices for all
calculations relative to explosive substances, as these calculations can always
be established from the baryta itself.

168

OXYGENATED COMPOUNDS OF NITROGEN.

experiments, we deduce from the latter trials by difference the
heat which would be liberated by the metamorphosis of nitric
oxide and of oxygen into nitric peroxide.

The first method, though simpler, is less exact than the
others from a consideration of the weight of the gases employed,
and of the quantity of heat produced.

2. First process. Nitric oxide and oxygen. — Two concentric
glass bulbs are enclosed one inside the other and sealed
separately, each containing one of the dry gases in the exact
ratio of 2 volumes nitric oxide (250 to 280 cub. cms.) to 1
volume of oxygen (see Fig. 30).

The system is plunged into the water of the calorimeter, then
by a jerk of the hand the internal bulb is broken, leaving its
envelope intact. Both gases react at once, and the action is
allowed to complete itself. The nitric peroxide remains gaseous
even up to 10° or 15°, because its tension in the bulb is less by
a third than the atmospheric pressure. The
latter circumstance slightly lowers the figures
which would be observed at the normal pres-
sure, viz. by 0-3 Cal. (p. 155).

Operating in this way and calculating the
reaction at constant pressure1 the following
was observed:— 2NO + 02 = N"204 gas + 19'6 ;
+ 19-9 + 18-3 + 19-8 : mean + 194 Cal.

3. Second process. Nitric peroxide, gaseous
chlorine and water. — In principle, this reaction
is the following :— N02 gas + Clgas -f H20 +
water = HN03 dilute + HC1 dilute.

The heat of formation of water (34*5 for
H2 + 0) and that of dilute hydrochloric
acid + 39-3 for H + Cl + water = dilute
HC1 being taken as known.
In practice it has been found preferable to operate on liquid
nitric peroxide, and this led the author to determine its heat
of vaporisation,2 viz. 4'33 Cal. for N02 = 46 grms.

The weighing of liquid nitric peroxide may be performed
very accurately in a hermetically sealed bulb.

In order to weigh chlorine directly in the same way, recourse
was had to the following artifice. Instead of allowing the nitric
peroxide and the chlorine to act directly on the water, the
chlorine was absorbed by a dilute solution of potash, the latter
being in excess, the heat liberated and the weight of chlorine
absorbed was determined by means of the vessel shown on
page 163.

1 Thomsen obtained + 19-57 by introducing both gases simultaneously into
a chamber placed in a calorimeter.

2 See the method employed (" Annales de Chimie et de Physique," 5e seVie,
torn. v. p. 154).

Fig. 30.

NITRIC PEROXIDE AND BARIUM DIOXIDE. 169

A volume of the alkaline solution containing a weight of chlorine
precisely equal to that of the nitric peroxide was then taken,
and the bulb containing the acid placed in it. The bulb was
then opened by the breakage of one of its points, taking care
that the mixture of the two solutions should be gradual, and the
heat liberated during the reaction measured.

Lastly, an excess of dilute hydrochloric acid is added to the
solution, the heat liberated by this addition being also measured.
Thus the whole is brought to a very simple final state, that of a
weak aqueous solution, formed by an equivalent of potash, an
equivalent of nitric acid, and a known proportion of hydro-
chloric acid somewhat greater than an equivalent.

In an independent experiment the heat liberated by the
mixture of the three components taken directly in the same
proportions and degree of dilution as in the above experiment
was measured.

This being known it is easy to deduce from the data obtained
the heat liberated by the following transformation : —
N02 gas + Cl gas + H20 -f water = HN03 dilute + HC1 dilute.

The weight of N02 being 2-281 grms 17-9 Cal.

„ 1-125 grms 17'7 „

Mean 17-8 „

deducting from this value the difference in the heats of forma-
tion of dilute hydrochloric acid, viz. : —

39-3 - 34-5 = + 4-8,

we find, 2N02 liquid + 0 gas + water = 2HN03 dilute + 13*0.
Adding now the heat of vaporisation of nitric peroxide, we obtain
2N02 + 0 gas + water = 2HN03 dilute + IT'3.1

The heat liberated by the transformation of nitric oxide and
oxgyen into dilute nitric acid, + 35*9 Cal., being taken as
known, we shall definitely have for the heat liberated by the
formation of gaseous nitric peroxide, from its immediate com-
ponents, 2NO + 02 = 2N02 gas -f'35'9 - 17'3 = 18'6 according
to the experiments of the second process.

4. Third Process. Nitric peroxide and barium dioxide. — This
process is based on the following reactions, 2N02 -J- Ba02
= (N03)2Ba. But this reaction does not take place with pure
and anhydrous bodies under the conditions adapted for calori-
metric measurements, and the following method was used.

The liquid nitric peroxide is weighed in a bulb, then the

1 Thomson obtained for this reaction the value + 16-9. In order to
measure it, he adopted the following process, which is less certain than that
indicated in the test. He allowed the nitric peroxide gas to act upon water,
so as to dissolve it, which liberates + 7-75 ; then he introduced chlorine into
the liquor, which liberates + 14-28 more, and he derives from these data the
heat of oxidation of nitric peroxide gas, forming dilute nitric acid.

170 OXYGENATED COMPOUNDS OF NITKOGEN.

equivalent barium dioxide is weighed, the latter is dissolved in
dilute hydrochloric acid and the heat liberated measured. Next,
nitric peroxide is allowed to react gradually upon the solution
which escapes from one of the broken points of the bulb which
is completely immersed in the solution. The heat liberated
during this second reaction is also measured.

The sum of the two results gives the total heat corresponding
to the following transformation —

2N02 liquid -f Ba02 anhydrous + 2HC1 dissolved = 2HN03
dissolved + BaCl2 dissolved.

It is immaterial whether we suppose the baryta united with
the hydrochloric acid, or with the nitric acid, or shared between
both, since the heat liberated by the union of this base with
either acid is the same.

The foregoing experiment was made with pure and anhydrous
barium dioxide.

The heat of formation from anhydrous baryta and free oxygen —

BaO + 0 = Ba02 liberates + 6'05 Cal.1

This quantity being known, as well as the heat of solution of
anhydrous baryta in dilute hydrochloric acid (+ 27*8) ; lastly,
the heat liberated by the reaction of nitric acid (formed from
nitric peroxide) upon dilute barium chloride being sensibly nil ;
the calculation of the experiments made with nitric peroxide
enables the heat liberated in the following reaction to be
arrived at —

2N02 liquid + 0 gas + H20 = 2HN03 dilute.
Thus—

Weight. Found.

N02 2-279 + 12-4Cal.

N02 1-358 +12-2 „

N02 0-951 +12-6 „

Mean + 12'4 „

In order to pass to gaseous N02 we must add the heat of
vaporisation of this body, viz. -f 4*33 ; which makes altogether
+ 16-73.

Finally, this number deducted from the heat of formation of
dilute nitric acid by nitric oxide and oxygen (viz. + 3 5 '9)
gives for the formation of nitric peroxide gas from nitric oxide
and oxygen, + 35'9 - 16*73 = + 19-17.

5. To sum up, the reaction NO + 0 = N02 gas liberates —

According to direct experiment + 19'4

„ „ the reaction of nitric peroxide gas on chlorine ... + 18-57
„ „ the reaction of nitric peroxide on barium dioxide +19-17

Mean + 19-04

1 " Annales de Chimie et de Physique," 5e se"rie, torn. vi. p. 209.

NITROGEN TRIOXIDE. 171

The heat of formation of nitric oxide itself from the elements
being — 21*6, from it is deduced that of nitric peroxide gas.
N 4. 02 = N02 gas, absorbs — 2'6. The formation of liquid
nitric peroxide on the contrary liberates heat, viz. -f 17.

6. The figures which express the heat of formation of nitric
peroxide gas, whether from nitrogen and oxygen or from nitric
oxide and oxygen, were obtained near the ordinary temperature.
Their value, however, becomes notably altered when referred to
a higher temperature. In fact, the specific heat of nitric per-
oxide gas varies very considerably with the temperature.1 This
gas undergoes, especially between 26° and 150°, a kind of mole-
cular transformation of an exceptional order, which nearly
doubles its density, in order to bring it to a value corresponding
to the molecular weight N02 = 46 grms. This transformation
may be estimated by supposing the theoretical specific heat of the
gas constant and equal to the sum of those of its components.

We have thus found that the transformation absorbs

From 27 to 67° 2'65 Cal.

„ 67 to 103° 1-75 „

„ 103 to 150° 0-85 „

„ 150 to 200° ... 0-03 „

Total 5-28 „

This number, added to the heat of vaporisation properly so
called, viz. 4*3, brings the heat absorbed to nearly 9 '6. Hence it
follows that the reaction of nitric oxide on oxygen, NO + 0 =
N02 gas, liberates quantities of heat decreasing with tempera-
ture, at least from 26° up to about 200°. It produces only 4-
13 '7 Cal. towards 200°. Similarly the formation of the com-
pound from the elements,

N + 02 = N02,

absorbs quantities of heat continually increasing in absolute
value, or — 7'9 towards 200°.

These figures hardly vary from 200° to 250°. Beyond this
they seem to decrease again, though much less rapidly, and
in conformity with what happens in the case of carbonic acid
and gases formed with condensation.2

7. Formation of nitrogen trioxide. — The calculation of the
heat of formation of this acid has not been given above, because
it is inseparably connected with the experiments relative to
nitric peroxide. Take now the reaction

2NO + 0 = N203.

If this reaction could take place separately, it would suffice to
bring together four volumes of nitric oxide and one volume of
oxygen, and to measure the heat liberated.

1 " Bulletin de la Socie*te* Chimique," 2e se*rie, torn, xxxvii. p. 434. 1882.

2 " Essai de Me"canique Chimique," torn. i. p. 334.

172 OXYGENATED COMPOUNDS OF NITROGEN.

But under these conditions a portion of the two gases is
always changed into nitric peroxide, and it does not seem
possible to obtain nitrogen trioxide without having at the same
time the products of its transformation, N02 and NO, the
whole constituting a system in equilibrium.

However, by increasing the proportion of nitric oxide, that of
the nitrogen trioxide is increased. But we are limited in this
respect by the necessity of operating upon a volume of oxygen
sufficient to give notable calorimetric effects.

By carrying out the reaction by the aid of a system ot
concentric bulbs (see p. 168), containing a known volume of the
two dry gases (about 400 cub. cms. of NO), the heat liberated was
measured, the proportion of NO and N02 formed was deter-
mined by absorbing the products in a weak alkaline solution,
the weight of oxygen employed affording a verifying equation.

The products being known as well as the heat of formation of
NOa, we can calculate the heat of formation of the nitrogen
trioxide. The datum which results from these measurements,
though less accurate than that of the other oxides of nitrogen, is
nevertheless useful.

From the mean of three experiments, 2NO -f 0 = N203 gas
liberates -f- 10*5 Cal. Hence the fixation of a second equivalent
of oxygen, that is, the transformation of the nitrogen trioxide
into nitric peroxide in the gaseous state, viz. N203 + 0 = 2N02
gas liberates + 19'0 - 10*5 = 8'5.

8. The fixation of a third equivalent of oxygen, transforming
the peroxide into nitric anhydride. 2N02 + 0 = N205 gas
liberates + 2;0.

The heat liberated by the same weight of oxygen, at the
ordinary temperature, decreases according as the oxygen in-
creases in the compound of nitrogen, starting with nitric oxide,
a fact which is demonstrated by the series of numbers -f 10 f5
+ 8-5 + 2-0.

The latter figure is in keeping with the slight stability of
nitric anhydride, a compound which cannot be formed from
nitric oxide by direct synthesis.

9. Direct measurements, independent of all analysis, show that
the same volume of oxygen, in the presence of an equal or more
than double volume of nitric oxide, in sealed bulbs, liberates
the more heat the greater the volume of nitric oxide, and conse-
quently that of the nitrogen trioxide formed. This result con-
tributes with the preceding ones to prove that the heat liberated
in the formation of nitric peroxide gas is not double that
liberated by the nitrogen trioxide gas, contrary to the relation
existing between the weights of oxygen successively fixed.

Finally the formation of a nitrous gas from its elements at
the ordinary temperature, calculated from the above data —

N2 + 03= N203 gas, absorbs -*22'2.

NITRIC ACID. 173

§ 6.— HEAT OF FORMATION OF DILUTE NITRIC ACID, OF
MONOHYDRATED NlTRIC ACID AND OF ANHYDROUS
NITRIC ACID.

1. The heat of formation of nitric acid from its elements is
a datum of the first importance, and may be deduced from that
of nitric oxide. Several methods were employed. We shall
first indicate them in principle and then give them in detail.

(1) BY THE NITRITES. Having first formed a nitrite in
presence of a base, such as baryta, the barium nitrite is trans-
formed into nitrate ; and this by four distinct processes.

(2) BY NITRIC ACID AND BARIUM DIOXIDE. The nitric oxide
is dissolved in concentrated nitric acid, which changes it into
nitrogen trioxide, the latter being further oxidised after dilution
by barium dioxide.

(3) BY NITRIC PEROXIDE. This body is further oxidised by
various agents, which have already been indicated. They may
be employed either to measure the heat of formation of nitric
peroxide, deduced from that of nitric acid, which is known, or
to measure the heat of formation of nitric acid, that of nitric
peroxide being known.

2. First method. Transformation of barium nitrite into
nitrate. In this reaction barium nitrite is oxidised, and the
heat liberated is measured by four processes distinct and inde-
pendent of one another.

First process. Gaseous chlorine. — Initial system: Ba(N02)2
dissolved ; C14 gas ; H4 gas ; 02 gas ; nBaO dissolved ; nHCl dis-
solved, these bodies being all separate from one another. Final
system : Ba(N03)2 dissolved ; 4HC1 dissolved ; n(BaC!2 + H20)
dissolved, these bodies being mixed.

It is first of all supposed that H2 is allowed to react on 0,
which forms water, liberating -f 69 Cal. ; then the following
experiments are carefully carried out. Dry chlorine is agitated
with baryta water, of known strength and weight, in a calori-
metric flask (p. 163) ; the heat liberated, Q, is measured, and the
chlorine absorbed, p, is directly weighed to within O'OOl grm.
Care is taken that there shall remain a considerable excess of
free baryta, and agitation is kept up incessantly during the
operation, in order to prevent the formation of any other oxide
of chlorine than hypochlorous acid; the measurement of the
heat liberated supplies a verification in this respect.1

A quantity of barium nitrite strictly equivalent to the weight
of chlorine absorbed (Ba(N02)2 for C12) is then taken and
dissolved separately in water, the solution is then mixed with
that of the hypochlorite, an operation which liberates a quantity
of heat, £, which can almost be neglected.

1 " Annales de Chimie et de Physique," 5e sdrie, torn. v. pp. 335-338.

174 OXYGENATED COMPOUNDS OF NITROGEN.

Dilute hydrochloric acid is at once added in considerable
excess, which liberates a fresh quantity of heat, Qx. Under these
conditions the whole of the chlorine introduced at the outset,
is at the end, and in a moment, changed into hydrochloric acid,
as can be easily proved. The final as well as the initial state
is therefore completely definite. The sum

Q + q + Qi

represents the total heat liberated during the passage from the
initial to the final state ; or, for C12 = 71 grms.

* x 71 = S.

P

The heat liberated, from the initial to the final system, is
therefore 69 + S.

But we could have passed from the same initial to the same
final state according to the following succession — unite 2H with
2C1, forming 2HC1 dilute, liberating + 78-6 Cal. ; then unite
nHCl dilute with nBaO dissolved, forming nBaC!2 dissolved,
which liberates + 13'85n ; lastly unite 02 gaseous -f Ba (N02)2
dissolved, which produces Ba(N03)2 dissolved, liberating x. The
thermal sum being the same in both processes, we have the
equation —

S - 13-85n - (78-6 - 69) = x.

Three concordant experiments, made according to this process,
each on about 2 grms. of nitrite, yielded

x = 221 Cal.,

a value which corresponds to the following reaction, Ba(N02)2
dissolved -f- 02 gas = Ba(N03)2 dissolved. The precautions em-
ployed in these experiments to avoid the use of gaseous chlorine,
either in a neutral or acid medium, and also against the sudden
transformation of the chlorine into chloride, or the variable form-
ation of oxides of chlorine, should be observed;1 free hypochlorous
acid has also been included, because this acid is difficult to
obtain quite free from chlorine or the higher oxides of chlorine,
besides, it decomposes spontaneously, and also in presence of
bodies which it oxidises, especially in an acid medium.2

Second process. Barium dioxide. — The barium nitrite is
changed into nitrate by barium dioxide dissolved in hydrochloric
acid. Ba(N02)2 dilute + 2BaO2 + 4HC1 dilute = Ba(N03)2
dilute + 2BaCl2 dilute + 2H20.

The initial system is the following : —

Ba(K02)2 dilute, 2BaO anhydrous ; 02 gas ; 4HC1 dilute,
all these bodies being separate.

The final system is the following : —

Ba(N03)2 dilute, 2BaCl2 dilute + 2H20.

1 " Annales de Chimie et de Physique," 5e serie, torn. v. p. 322.

2 Ibid., p. 342.

BARIUM DIOXIDE PROCESS. 175

We may pass from one to the other by following two different
cycles.

FIRST CYCLE.

Ba(N02)2 dilute + 02 = Ba(N03)2 dilute ...... x

2Ba02 anhydrous + 4HC1 dilute = 2BaCl2 dilute + 2H20 + 55'58

Sum ...... + 55-58 + x

SECOND CYCLE.
2(BaO + 0) = 2Ba02 anhydrous ............ + 12-1

2Ba02 + 4HC1 dilute, then reaction of this solution on dis-

solved Ba(N02)2 .................. R

Sum ...... ... R -f 12-1

K being determined by experiment it is easy to calculate x.
The experiment is carried out as follows : —

A known weight, p, of anhydrous barium dioxide, say 8 '5 grms.
for example, is dissolved in the calorimeter by dilute hydro-
chloric acid; the quantity of heat liberated, Q, is measured.

With 8'50 grms. it is practically equal to -^- Cal. To the

zo

liquid is then added a quantity of barium nitrite strictly
equivalent to the barium dioxide employed, or 6'20 grms.
The nitrite should be dissolved beforehand in twenty-five times
its weight of water, and the temperature of the solution
accurately measured a moment before it is mixed in the calori-
meter with the hydrochloric acid and solution of barium dioxide.
Immediately upon this mixture being effected, several pheno-
mena are produced and succeed each other rapidly, the solution
becomes yellow, then for a moment it becomes turbid, as if a
precipitate were forming ; a few excessively fine gaseous bubbles
appear for an instant without giving rise to the production
of an appreciable volume of gas; then the liquor becomes
perfectly clear. A minute, and even less, suffices for the
accomplishment of all these effects.

At this point the liberation of heat is at an end, and the nitrite
entirely changed into nitrate.

From the data observed is calculated the heat liberated during
the last metamorphosis, say Q'.

Hence we have, calling E the equivalent of barium dioxide,
Ba02 = 84' 5 grms. as the expression for the heat liberated.

therefore the heat liberated in the transformation of nitrite of
baryta into nitrate will be

x = B + 121 - 55-58 = B - 43'5.

176 OXYGENATED COMPOUNDS OF NITROGEN.

Experimental results. First experiment at 12° —

2Ba02 dissolved in 4HC1 dilute +22-02

Reaction on Ba(N02)2 dissolved + 43-23

R = + 65-25
whence may be deduced

x = -f 22-25.

Second experiment at 12° —

In this experiment crystallised barium nitrite was directly
dissolved in the hydrochloric solution of barium dioxide.

2Ba02 + 4HC1 dilute +21-84

Reaction on crystallised Ba(N02)2H20 ... + 38'75

Sum +60-59

This experiment was purposely made with crystallised nitrite
of baryta in order to vary the conditions. To make it com-
parable with the proceeding it is necessary to add to the number
obtained the dissolving heat of the salt at the same temperature,
taken with the contrary sign, viz. -{- 4*30.

Hence we have

E = + 64-89

whence may be deduced

x = 21-49

The two experiments have therefore given

22-25 and 2149
or a mean of

21-87 or 21-9 Cal.

This is therefore the quantity of heat liberated in the following
reaction :

Ba(N02)2 dissolved + 02 gas = Ba(N03)2 dissolved.

Third process. Liquid bromine. — The theoretical reaction is
the following : —

Ba(N02)2 dissolved + Br4 + 2H20 = Ba(N03)2 dissolved +
4HBr dissolved.

Pure liquid bromine is weighed in an hermetically sealed tube,
say, for example, 2 '254 grms. A strictly equivalent weight of
pure barium nitrite is also weighed. The water is placed in the
calorimeter, the salt is dissolved in it, and the tube introduced.
When equilibrium of temperature is established the bromine
tube is crushed, and the whole is quickly stirred.

The reaction of bromine on barium nitrite does not, however,
take place so rapidly as in the case of chlorine ; it does not
completely dissolve until after some time, and the solution retains,
even after twenty minutes, a strong odour of bromine. In a

POTASSIUM PERMANGANATE PROCESS. 177

word, the nitrite and the bromine are not entirely changed into
nitrate and hydrobromic acid under these conditions, but there
exists in the solution a bromonitric compound analogous to
aqua regia, and the continued presence of which interferes with
the application of the calorimetric calculation by means of the
equation given above.

As a matter of fact the calculation applied to the results of
these experiments has given results falling below the theoretical
reaction Ba(N02)2 dissolved -f O2 = Ba(N03)2 dissolved ; the
value obtained fluctuating about 4- 18 Cal. instead of -J- 22 Gal.
These results are, therefore, not included in the averages on
account of the incomplete nature of the transformation. But it
has been thought proper to point out, from the purely chemical
point of view, the true character of the reaction of bromine on
the nitrites.

Fourth process. Potassium permanganate. — It is well known
with what accuracy this reagent may be employed to convert
nitrites into nitrates.

The experiments were performed with a solution of absolutely
pure potassium permanganate of known strength,1 mixed with
a large excess of dilute sulphuric acid ; 2 for instance, 192 cms. of
the permanganate solution (20 grms. = 1 litre) and 1800 cms. of
a solution of dilute sulphuric acid, mixed in a large platinum
calorimeter and 2*470 grms. of crystallised barium nitrite,
Ba(N02)2H20, added. The reaction is instantaneous. The heat
liberated is measured as soon as the reaction is accomplished,
the excess of the permanganate is reduced by a standard solution
of oxalic acid (160 cms. for instance), the whole of the carbonic
acid formed remains dissolved. The quantities of heat liberated
in this second reaction are also measured.

The sum of the quantities of heat which result from the fore-
going experiments represents the heat liberated.

As a check, the excess of oxalic acid remaining in the liquid
is ascertained. These measurements, combined with the data
contained in the author's Memoir e sur la chaleur de combustion
de I'acide oxalique? and with the figures obtained in the reduction
of potassium permanganate by oxalic acid,4 enable us to calculate
the heat liberated in the transformation of barium nitrite into
nitrate.

Finally, by this method it was found that the reaction
Ba(N02)2 dissolved -f 02 gas = Ba(N03)2 dissolved liberates—

First trial + 21-7 Cal.

Second trial +20-5 „

Mean + 2M „

These results are rather less reliable than those of the two

1 " Annales de Chimie et de Physique," 5e sdrie, torn. v. p. 306.

2 Ibid., p. 308. 3 Ibid., p. 305. * Ibid., p. 309.

N

178 OXYGENATED COMPOUNDS OF NITROGEN.

first methods, owing to the complex nature of the thermal
reactions of permanganate. Their mean, however, is sufficiently
concordant.

To sum up, the reaction

Ba(N02)2 dissolved + 02 = Ba(N03)2 dissolved liberates—

According to the results obtained with gaseous chlorine ... + 22*1
„ „ „ „ „ barium dioxide ... + 21 '9

„ „ „ „ „ potassium permanganate +21-1

Mean +21-7

This value is applicable not only to the oxidation of barium
nitrite, but also to the oxidation of all dissolved alkaline
nitrites.

Hence from the knowledge of the heats of neutralisation
of nitrogen trioxide (4- 10*6) and nitric acid (-f 13*8) by
baryta :

N203 very diluted + 02 + H20 = 2HN03 dilute liberates
+ 18-5 Cal.

For both the bodies gaseous we have, according to the data
given further on :

NA gas + 02 = N205 gas liberates + 10'5 Cal.

Hence the change of the nitrogen trioxide into the pentoxide
liberates, when the action takes place, upon the gases, + 10 '5 ;
on the dissolved bodies, -f- 18'5 ; lastly, in presence of alkalis,
-f- 247. The great difference between the quantities of heat
liberated by one and the same transformation, according to
the state of the bodies, deserves attention, being due to the
difference in the heats of hydration and neutralisation of the
two acids.

We will also give here the heats of oxidation of the solid and
anhydrous nitrites, which are easily calculated if their dissolving
heat be known.

Dissolved salts. Solid salts.

Ba(N02)2 + 02 = Ba(N03)2 liberates ... +21-7 ... +23-5

NH4N02 + 0 = NH4N03 liberates +21-8 ... +23-3

AgN02 + 0 = AgN03 liberates +20-3 ... + 17'2

3. Second method. By nitric acid and barium dioxide. A
known quantity of concentrated and pure nitric acid is allowed
to absorb dry nitric oxide, the weight of which is determined
by a fresh weighing after having measured the heat liberated,
say Q for NO = 30 grms. The concentrated acid is contained
in a small tube, which is weighed and closed and plunged
into the calorimeter throughout the whole duration of the
absorption. A thermometer sensitive to 20^ of a degree gives
the temperature of the calorimeter; a smaller thermometer,

NITRIC ACID AND BARIUM DIOXIDE. 179

sensitive to ^ of a degree, gives that of the acid, the latter being
kept as near as possible to that of the calorimeter. The cor-
rections for cooling are made according to the ordinary processes.1
It is with the aid of all these data that the quantity, Q, is
calculated. Further, a weight of anhydrous barium dioxide
equivalent to the above weight of nitric oxide absorbed (3Ba02
for 2NO) is dissolved in the excess of dilute hydrochloric
acid, which liberates QL. The concentrated nitric acid, which
has dissolved the nitric oxide, is then mixed in the calorimeter
with the hydrochloric solution of barium dioxide. The whole is
thus brought to the final state of dilute nitric acid and dilute
barium chloride, liberating a measured quantity of heat equal
to Q2. Lastly is dissolved the same weight of the same pure
nitric acid as used at the outset, in the same volume of dilute
hydrochloric acid, which liberates a quantity of heat, P.

As a check, a weighed quantity of barium dioxide is added to
the solution, which should, and in fact does, produce the same
quantity of heat as if it were dissolved in a solution containing
only hydrochloric acid ; which proves that the nitric acid
employed is very free from nitrogen trioxide.

This being established we have all the data for the calcula-
tion. Let the initial system be 03; 2NO; 3BaO ; m(2HN03
+ fiAq) ; 6 HC1 dilute ; these bodies being taken separately ;
and let the final system be 3(BaCl2 + H20 dissolved + (m -f 1)
2HN03 dilute). We can pass from one to the other according
to the two following cycles : —

FIRST CYCLE.

3BaO + 30 = 3Ba02 anhydrous liberates + 18-0 Cal.

6HC1 dilute + 3Ba02 .. > ... ... Ql

Reaction of NO on the concentrated nitric acid ... Q

Reaction between the two mixtures Q2

Sum ... + 18-0 Cal. + Q + Qi + Q2

SECOND CYCLE.

3BaO anhydrous + 6HC1 dilute liberates + 27'8 x 3

(according to the author's experiments) or ... + 83' 4 Cal.

2NO + 03 + water = 2HN03 dilute x

m(HN03 + wAq) + water

Sum + 83-4 Cal. + x + P

whence we deduce the value of x :

+ 18-0 Cal. + Q + Q! + Q2 = 4- 834 Cal. + x + P.

The experiments gave x = 344 Cal., a value which is
slightly too small.
4. The third method is based on the use of nitric peroxide.

1 " Essai de Me*canique Chimique," torn. i. p. 208.

N 2

180 OXYGENATED COMPOUNDS OF NITROGEN.

The results have been given above (pp. 168 to 171). They may
be summed up as follows : —

NO + 0 = N02gas ...............

Liquefaction of N0a ...............

2N02 liquid + 0 + H20 + water = 2HN08 dilute ...

+19-4
+ 4'33
+ 12-7

We have therefore :

2NO -f 03 + H20 + water = 2HN03 dilute -f 364.1

These results are worthy of attention, but they do not appeal-
capable of great accuracy, owing to the uncertainty of the
reactions.

5. To sum up the results of the different methods —

NO + 08 + water = HN03 dilute : by the nitrites ... + 35'9
„ „ „ „ by nitric acid ... + 34-4

„ „ „ „ by nitric peroxide ... +36-3

Mean ............ +35-5

However, the accuracy of the three methods is very unequal, so
the value 3 5 '9 will be adopted, the method by which it was
obtained being most reliable, taking the two other figures as
mere verifications.

6. Heat of formation of dilute nitric add from its elements.
This heat, calculated from nitrogen and oxygen, is easily deduced
from the preceding results, for it is sufficient to deduct from the
latter the heat absorbed in the formation of nitric oxide.

N2 4. Q6 -f- HaO + water = 2 HN03 dilute liberates 4- 35'9
- 21-6 = -f 14-3 Gal.

If we consider the integral formation of nitric acid from its
three elements, HN03, we must add the heat of formation of
water, viz. + 34'5. Thus N2 -f 06 + water = 2HN03 dilute
liberates -f- 48*8 Cal. Such a reaction, therefore, liberates heat.
Hence it can take place directly, and, in fact, it is observed
in the combustion of the hydrogen in the air, but it only affects
a small quantity.

7. Heat of formation of nitric acid. It is sufficient to measure
the heat liberated by the solution of the liquid acid in a large
quantity of water, viz. + 7 '2. We have therefore —

N2 + 08 + H20 = 2HNOS liquid liberates + 7-1

N2 + 06 + H2 = 2HNO, „ + 41-6

To pass from the gaseous to the solid state, it is sufficient to
measure the heat of vaporisation and the heat of fusion at a low
temperature of nitric acid, HN03 = 63 grms.

The result found for the heat of vaporisation is + 7'35, and
for fusion— 0*6.

1 Thomsen, following an analogous though not identical cycle, found
+ 36-47.

NITROGEN PENTOXIDE. 181

We have, therefore, neglecting the differences between the
specific heats of the body under its various states —

Na -f 05 + H20 solid = 2HN08 solid (at low temperature) ... + 7-0

„ =2HN08 „ ... F ...... + 42-2

gas =2HN08gas ......... - 0-1

„ = 2HNO, „ ...... +34-4

8. Add at different degrees of concentration. To pass to nitric
acid regarded under different states of concentration, it is
sufficient to know its heat of dilution, under its various states,
and to add it to the foregoing figures. It has been measured
for the whole scale of dilutions, and the results will be found
in the "Annales de Chimie et de Physique," 5e se*rie, torn. iv.
p. 446. We confine ourselves to giving here the heat of
formation of the hydrate HN03, 2H20, which represents
approximately the acid of commerce.

HN03 + 2H20 = HN03'2H20 liquid, liberates + 5-0 Cal.
Engravers' aqua fortis approaches the formula

HN03 + 6-5H20.
The formation of such a hydrate,

HN08 + 6'5H20 = HN03 6'5H20, liberates + 7'0.

9. Nitrogen pentoxide, N205. Preparation. It is well known
that this body was discovered by Sainte-Claire Deville in the
reaction of chlorine on silver nitrate. A more simple method
was devised by the author. The principle of the process
was discovered by Weber in 1872,1 and consists in dehydrating
nitric acid by phosphoric anhydride. Nitric acid, cooled by a
freezing mixture, is mixed with pulverulent phosphoric oxide in
small portions at a time, taking care to avoid any elevation of
temperature. The temperature of the mass should never
exceed 0°. When a little more than its weight of phosphoric
oxide has been added to the nitric acid, the mass becomes of
the consistency of a jelly. It is then placed in a roomy
tubulated retort and distilled with extreme slowness. The
products are condensed in receivers with ground stoppers,
immersed in ice. In this way large, brilliant, colourless crystals
of nitrogen pentoxide are obtained, which are perfectly pure.
From 150 grms. of nitric acid nearly 80 grms. of the crystals
were obtained. This substance is non-explosive either as a
solid or in vapour. It decomposes, however, very easily, and
this at the ordinary temperature, as Deville has observed, into
nitric peroxide and oxygen. It should not be preserved in
hermetically sealed vessels. It keeps well in good glass-
stoppered bottles placed under a bell glass with sulphuric acid.
In the air the crystals evaporate slowly, evolving abundant
vapours, but not liquefying. Hence they can be weighed

1 " Ann. Pogg.," torn, cxlvii. p. 113.

182 OXYGENATED COMPOUNDS OF NITROGEN.

without difficulty. Light accelerates its decomposition, as does
also heat, though even at 43° it is not very rapid. This
change into nitric peroxide and oxygen is endothermic, and is
not reversible.

The following is the analysis of these crystals : — 5 '5 55 grms.
of crystals weighed, and dissolved in water, yielded a solution
which, according to alkalimetric test, contained 5 '54 grms. of
nitrogen pentoxide ; no appreciable quantity of nitrogen trioxide
was present (reaction of potassium permanganate). A large
quantity having been prepared, its action on water in the
calorimeter was studied, taking it successively under the three
states, solid, liquid, and gaseous.

Solid state —

crystallised + H20 -f water at 10° = 2HN03 dilute
+ 8-34 Cal.

Now

N205H20 pure + water at 10° = 2HN03 dilute liberates +
718 Cal. ;

therefore

N205 solid + H20 liquid = 2HN03 pure and liquid liberates
+ 116 Cal.

This quantity of heat is very small, as might be expected,
owing to the contrary thermal effect produced by the liquefaction
of the anhydrides.

Hence, the action of the solid anhydride on water is not very
violent, which confirms in another way the above result. The
union of the solid anhydride with atmospheric aqueous vapour
is likewise slower than that of bodies said to be very hygroscopic.
In fact, at the ordinary temperature, the anhydride evaporates
without leaving any appreciable quantity of dilute acid. The
following reaction refers to the solid state —

N205 solid + BaO solid = Ba(N03)2 solid liberates + 407.

This quantity of heat is less by + 10*3 than that liberated by
the formation of barium sulphate starting from the anhydrous
acid and the anhydrous base, both being solid, viz. + 51*0.

Liquid state. — Heat of fusion was determined by two
methods :

(1) By the solidification of the dissolved acid contained in
a tube, immersed in the calorimeter.

(2) By dissolving directly the solid acid in water.

The following figures result from determination of the above
methods : —

N205 ( = 54 grms.) liquid, in becoming solid, liberates + 414,
or, for N208 ( = 108 grms.), + 8'28.

HEAT OF VAPORISATION OF NITROGEN PENTOXIDE. 183

This value is very high and equal to about six times the heat
of solidification of water (+ 072 for H20 = 9 grms., according
to Desains).

Therefore

N206 liq. + H20 liq. = 2HNO3 liq. and pure liberates + 5-3 Cal.
N206 liq. + H20 liq. + water = dilute acid -f 12'48 Cal.

The first value is nearly but not quite equal to that of the
heat of hydration of acetic anhydride (C4H603 liquid + H20
liquid = C4H804 liquid liberates + 6'9), but the second is much
greater than the heat of hydration of anhydrous acetic acid
referred to the dilute acid ( -f 7'3). Hence the action of the
liquid nitrogen pentoxide on water is extremely violent in con-
trast with the very much weaker reaction which water exercises
on the solid acid.

Gaseous state. — Heat of vaporisation. N205 gas, changed into
liquid, liberates + 242, and into solid, -f 6'56.

This quantity was determined by introducing dry air charged
with nitrogen pentoxide vapour into the water of the calori-
meter, at a temperature of 43°. The preliminary vaporisation
of the pentoxide in the current of air was produced by means
of a small air bath.

The decomposition of the pentoxide into the tetroxide and
oxygen is not appreciable under these conditions of vaporisation.

Known weights of pentoxide, previously weighed in a sealed
tube, have been operated on, the result being checked by the
acidimetric test of the aqueous solution.

By this means is obtained the heat liberated when nitrogen
pentoxide is changed into dilute acid, viz.

N205 gas + water = dilute acid, at -f- 10° liberates + 14-9.

The heat of vaporisation of the liquid acid is therefore for
the weight —

N205 = 54 grms.
14-9 -12-48 = 2-42,

or, for (N205 =108), 4'84 Cal.

That of the solid, for (N205 = 54 grms.)—

14-90 -8-34 = 6-56,
or, for (N205 = 108 grms.), 13'12.

According to the above figures, the heat of vaporisation of the
liquid nitrogen pentoxide (admitting £N203 = 2 vols.) will be for
N205, 4-84. It is nearly the same as that of nitric peroxide
at the same volume, or 4*3 for NO2. It is also nearly the same
as the heat of vaporisation of nitrogen monoxide, viz. 4'42 for
N20, according to Favre.

The thermal formation of nitrogen pentoxide from the elements

184 OXYGENATED COMPOUNDS OF NITROGEN.

is deduced from the foregoing data. Under the three states we
have —

N2 + 06 = N205 gas +0-6

N2 -f 06 = N205 liquid + 1-8

N2 + Ofi = N206 solid +5-9

10. The following table shows the thermal formation of the
oxides of nitrogen under the gaseous form, referred to the
ordinary temperature : —

N2 + 0 = (2v.)N20 - 103.

} - 11-3
N + 0 = (4v.)NO - 21-6{

} +10-5

N2 + 03 = (2v.)N203-H-l,

} + 8-5
N + 02 = (4v.)N02 - frtf

\ + 2-0

N2 + 06 = (2v.)N206 - 06'

It will be seen that the progressive formation of the oxides of
nitrogen follows a peculiar course. It first absorbs a quantity
of heat nearly proportional for the first two, then liberates
quantities which go on decreasing. These bodies are here
given under the gaseous form, the only one which is really
comparable. The most stable compound, nitric peroxide,
corresponds neither to the maximum nor to the minimum of
the heat absorbed. In short, there exists no simple numerical
relation between the quantities of heat brought into action.

The most general fact, following from the foregoing table, is
that the formation of all the oxides of nitrogen from their
gaseous elements absorbs heat, their decomposition must there-
fore liberate it. Nevertheless not one of them is explosive by
simple heating. But nitric oxide, formed with the greatest
absorption of heat, is decomposed into its elements with facility,
as will be established further on (see p. 191). The heat absorbed
in its formation renders it comparable to cyanogen ( — 37*3 for
C2N2) or to acetylene ( — 30-5 for C2H2). These three bodies can,
moreover, undergo a true explosion under the influence of the
sudden and violent shock of mercury fulminate (p. 66). These
three bodies indicate an aptitude for combination altogether
comparable, to that of the simple radicals. Hence from a
knowledge of these relations may be understood why the
formation of the oxides of nitrogen never takes place directly,
and why it requires the aid of a foreign energy, such as electri-
city, or of a simultaneous chemical action.

It also explains the great energy of explosive mixtures and
compounds formed by the oxygen compounds of nitrogen.

SILVER HYPONITRITE. 185

§ 7. HYPONITROUS ACID AND HYPONITRITES.

1. In studying the products of the reduction of the nitrates
by sodium amalgam, Divers1 discovered in 1871 a new salt,
which he called silver hyponitrite, and of which he determined
the composition and the properties. This salt and its derivatives
have since been the object of researches by Van der Plaats2
and Zorn.8 These chemists have attributed to silver hyponitrite
the formula AgNO, which would suppose it derived from
nitrogen monoxide, associated with silver oxide. But the
recent researches which Ogier and the author have made upon
this salt from a chemical and thermal point of view have led
them to prefer the formula Ag4N405, that is to say 2Ag20,N203,
which makes of the hyponitrous acid a sesquioxide of nitrogen.

The alkaline hyponitrites are also formed in the electrolysis
of the nitrites, and they are formed, though to a very small
amount, in the decomposition of the nitrites by heat, especially
in presence of iron. It is by means of silver hyponitrite that
hyponitrous acid and its salts are prepared ; we shall speak,
therefore, first of all, of this compound.

2. Silver hyponitrite is a yellow amorphous very insoluble body,
which is precipitated when silver nitrate is poured into a
neutral solution of alkaline nitrite. In order to obtain it pure,
it must be re-dissolved in very dilute nitric acid, and re-precipi-
tated, by neutralising by ammonia.

This body undergoes a very sensible decomposition when
heated to 100° or a little over.

Hence the hyponitrite should be dried in vacuo at the ordi-
nary temperature, and in the dark.

Its analysis has supplied the following figures : —

Ag 76-2 76-1

N 9-7 9-8

O 14-1 14-1

These results lead to the formula —

Calculated from Found.

AgNO(138) iAg4N4Ofl(284)

Ag 78-3 ... 76-1 ... 76-1

N 10-1 ... 9-9 ... 9-8

0 10-6 ... 14-1 ... 14-1

Hyponitrous acid has therefore as formula N4032H20, which
constitutes it a sesquioxide of nitrogen, corresponding in the
anhydrous state to the formula N403.

1 « Journal of the Chemical Society," vol. xxiv. p. 484 ; " Proceedings of
the Royal Society," vol. xxii. p. 425 ; " Bulletin de la Socie'te' Chimique,
torn. xv. p. 176.

2 " Berichte der Deutsch. Chem. Ges. Ber.," torn. x. p. 1508.

3 Ibid., torn. x. p. 1306, and torn. xv. p. 1258.

186 OXYGENATED COMPOUNDS OF NITEOGEX.

This formula accounts for the existence of the acid salts,
observed by Zorn.

Silver hyponitrite is decomposed by heat, with formation of
nitric oxide, nitrogen trioxide, and metallic silver —

Ag4NA = 2X0 + NA + Ag4.

But the nitrogen trioxide reacts partially upon the silver, so
as to reproduce a certain quantity of nitrite, and even nitrate
of silver.

3. By decomposing silver hyponitrite by a dilute acid,
hyponitrous acid is obtained in an aqueous solution. This acid
is not at all stable. Its solutions raised to boiling point are
decomposed, yielding nitrogen monoxide mixed with nitrogen,
retaining at the same time a certain quantity of dilute nitric
acid —

4NA dilute + H20 = 7£T20 + 2HN03.

On contact with the air they absorb oxygen slowly, becoming
changed into nitric acid.

4. We have, with a view to calorimetric tests, methodically
subjected hyponitrous acid to the action of the three following
oxidising bodies — iodine, bromine, and potassium permanganate.

(1) A solution of iodine in potassium iodide did not exert
any appreciable action on the hyponitrous acid combined with
the silver1 or previously liberated by an equivalent quantity
of dilute hydrochloric acid.

(2) The oxidation by bromine is very characteristic. A
known weight of silver hyponitrite, 2 grms., was mixed with
hydrochloric acid in excess, and an aqueous solution of bromine,
also slightly in excess, the strength of which was determined ;
the reaction was allowed to go on for some time, when the
excess of bromine was determined. This method tends to give
rather high figures, owing to the evaporation of some traces of
bromine.

Or the hydrochloric acid may be mixed beforehand with
bromine water in which salts of silver have been dissolved
(series I.) ; or the salt dissolved in the acid and the bromine
added (series II., p. 187).

The equivalent ratio between the silver and bromine employed
has been found to be very nearly 1 : 3 '5, which agrees with the
formula.

Ag4N A + 7H20 + 14Br = 4HX03 + 1 OHBr + 4 AgBr.

The formula AgNO would require the ratio 1 : 4, which is
greatly higher than all the quantities observed.

(3) The oxidation by potassium permanganate gives rather
irregular results, the oxygen absorbed varying from 4'6 to 8'9
per cent., and the action going on almost indefinitely. However,

1 Except the conversion of the silver into iodide.

HEAT OF FORMATION OF SILVER HYPON1TRITE. 187

by operating in presence of a very great excess of sulphuric
acid more concordant results may be obtained, such as 8'3 ;
7-5; 8-4; 8;9.

These figures correspond sensibly to three equivalents of
oxygen absorbed.

The solutions do not contain ammonia, but liberate by
ebullition a considerable quantity of nitrogen monoxide. In
another experiment, the oxygen absorbed and the nitrogen
monoxide were ascertained by analysis. The following results
were obtained : —

0 fixed : 8-3 : : N20 liberated : 8'0 per 100 parts of salt.

These figures correspond very sensibly to the following
transformation : —

Ag4N405 + 03 + H20 = ]ST20 + 2HN03- 2 Ag20 combined with

the acid.

This is therefore a fresh confirmation of the formula. The
analysis by the permanganate must be made by introducing
the salt of silver in a body into the mixture of perman-
ganate and sulphuric acid made beforehand and in excess,
as the hyponitrous acid set free slowly absorbs the oxygen
of the air.

These facts being established, we have proceeded to the
calorimetric measurements, and successively determined the
heat of formation of the salt of silver, that of the acid itself, as
well as the heat liberated by its union with silver and potassium
oxides.

5. Heat of formation of silver hyponitrite. We determined
the heat of formation of silver hyponitrite by oxidising it with
bromine water in accordance with one of the foregoing
experiments.

The figures obtained are sufficiently close. The following is
a list of them : —

First Series. — Action effected by a single operation for Ag= 108 grms.

First ' ... 29-83 Cal.\ m.ftn o0.fio

Second ... 31-54 „ /

Second Series. — Successive actions of HC1 and Br.

Third ... 28-00 Cal.)

Fourth 29-85 „ V mean 28'62.

Fifth ... 28-00 „ )

The general mean of both series is equal to 29'65 Cal.

The experimental ratio between the silver and the bromine
absorbed in equivalents was found to have a mean value of
3'71 ; a figure which is rather too high, owing, as before stated,
to the loss of bromine by evaporation. The theoretical ratio is
3-50. Let therefore the initial system be Ag4N405 + 7H20 +

188 OXYGENATED COMPOUNDS OF NITROGEN.

14Br (gas) -f water, the final state is arrived at by the following
cycle : —

N4 + 05 + Ag4 = N405Ag4.

7(H2 + 0) = 7H20 liberates + 34-5 x 7 ... =241-5

4Br gaseous + water = 14Br dissolved ... + 29'0

Beaction (for Ag4) + 59*3

x + 329-8
the final state being —

2HN03 dilute + lOHBr dilute -f 4AgBr.
The same final state may be arrived at by the following cycle : —

2(H + N + 08) + water = 2(HN03) dilute ... + 97-6
10(H + Br gas) + water = lOHBr dilute ... +167-5
4(AgBr gas) = 4AgBr + 55'4

+ 320-5
Both thermal sums being equal, it follows that

x = -9-3 Cal.
This is the heat absorbed in the reunion of the elements

Ag4 + N4 + 05.

We have further, starting with nitrogen, oxygen, and silver
oxide —

2Ag20 + N4 + 03, - 16-3 Cal.

6. Heat of formation of hyponitrous acid. To pass to the acid
itself, we measured the heat liberated in the reaction of dilute
hydrochloric acid on silver hyponitrite, viz. for one equivalent
of silver, Ag, contained in this compound.

+ 8''

which makes for Ag2 + 17 '88.

The hyponitrous acid exists, moreover, after this operation, or
at least throughout the duration of the experiment, as is shown
by the agreement of the estimations of bromine effected before
and after the action of the hydrochloric acid.

This being established, the reaction

2HC1 -f Ag20 = 2AgCl 4- H20 liberates + 201 ;
whence it follows that
N403 dilute 4- 2 Ag20 = Ag4N405 liberates + 40'2-17'9 = +22-3,

or 4- 1215 for each equivalent of oxide combined. Hence we
have N4 4- 03 4- water = N403 dilute — 38*6 Cal. Hyponitrous
acid is therefore formed from its elements with absorption of
heat, as would be supposed from the instability of the acid itself.
Its transformation into nitric acid, by oxidation (by means of
bromine), liberates N403 dilute 4- 07 4- 2H20 = 4HM)3 dilute

OXIDATION BY PERMANGANATE. 189

•f 67*2 or -f 9'6 Cal. per equivalent of oxygen fixed. This
figure is hardly higher than that obtained for the transformation
of dissolved nitrous acid into dilute nitric acid. N203 dilute
+ 02 + H20 = 2HN03 dilute + 1815, or 9-25 cal. for each 0
fixed. However, if we regard the two successive oxidations, the
calculation shows that the oxidation of the nitrogen trioxide
forming nitrous acid liberates a little more heat, viz. 101 per 0
fixed, than that of the nitrous acid changed into nitric acid, viz.
9 '25. The change even of one of the salts into the other would
liberate, for solid salts of silver —

Hyponitrite changed into nitrite per 0 fixed -f 10'4
Nitrite into nitrate -f 8*8.

For the dissolved salts of potassium the difference is increased,
owing to the difference in the heats of neutralisation.

Hyponitrite changed into nitrite per 0 fixed + 13 '6.
Nitrite into nitrate 4- 10'8.

The relations are always of the same kind.

7. The oxidation by permanganate, with formation of nitrogen
monoxide (deducting the heat due to the reduction of the
permanganate) —

N403 dil. + 03 + H20 = 2HN03dil. + N20 gas liberates + 42 -3.

The slow decomposition of the hyponitrous acid in contact with
the air, and at the expense both of the free oxygen and that
dissolved in the water, liberates exactly the same quantity of
heat with formation of nitrogen monoxide. The pure and
simple separation N403 dil. = 2NO + N20 gas would liberate
-f- 6 '4. The nitrogen monoxide can moreover be formed with-
out nitric acid by other methods, which liberate much more
heat, and are therefore preferable —

7N403 + water = 7N20 gas -f 2HN03 dil. liberates + 96'6,
or -{- 241 for N403. Combinations of hyponitrous acid present
a mobility and complexity of reactions which are explained by
their endothermal formation. Many analogous phenomena are
known in the series of the lower oxides of sulphur and
phosphorus, not to speak of hydroxylamine, which also very
easily yields nitrogen and nitrogen monoxide.

8. The heat of neutralisation of dilute hyponitrous acid by
silver oxide has been given above, viz.

N403 dilute + 2Ag204 = Ag4N405 -f 1115 X 4.
We have tried also to estimate the heat of neutralisation of
hyponitrous acid by the alkalis, by decomposing salts of silver
by the alkaline chlorides. The reaction is almost instantaneous.
We obtained i(Ag4N405 -f 4HC1) dilute at about 14° + 5'50
Cal. With barium chloride, BaCl2, the liberation of heat has
been more considerable, but it seems to be complicated by the

190 OXYGENATED COMPOUNDS OF NITROGEN.

partial precipitation of barium hyponitrite. With ammonium
chloride there is produced a special decomposition, setting free
ammonia, which has already been observed by Divers. Accord-
ing to the above figures, we have, for potash and hyponitrous
acid at 14°—

]ST403 dilute + 2K20 dilute liberates 2(+ 8'9 + 13'8 + 2'75
- 201) = + 2 x 5-35 Cal.

Let us now compare these results with the analogous numbers
relating to the two other acids of nitrogen —

i(2HN03 dilute + Ag20, forming 2AgN03) solid... + 10-7 Cal.
i(N203 dilute + Ag20, forming 2AgN02) solid ... + 12-1 „
}(N408 dilute + 2 Ag20, forming Ag4N406) + 11-1 „

These are nearly the same values as for silver oxide forming
solid salts. For potash, on the contrary, forming soluble salts —

i(2HN03 dilute + K20 dilute) + 13-8 Cal.

i(N203 dilute + K20 dilute) + 10'6 „

i(N403 dilute + 2K20 dilute) + 5'4 „

The relative weakness of the latter acids, a weakness which is
correlative with their decreasing percentage of oxygen, is here
more and more marked.1

§ 8. STABILITY AND EECIPROCAL KEACTIONS OF THE OXYGEN
COMPOUNDS OF NITROGEN.

1. The carrying out of so many thermal determinations has
led to the study of the formation and decomposition of the
various oxides of nitrogen, a subject which had not been re-
considered since the time of Gay-Lussac,2 Dulong,3 Dalton,4
and Priestley. Some of Peligot's5 famous experiments on
nitric peroxide and nitrogen trioxide have also been repeated.

The results obtained were unexpected, and contrary to the
received opinions on the stability of nitric oxide.

2. Nitrogen monoxide, according to Priestley, is decomposed at

1 We think it well to give here the calculation of the heats of formation of
the hyponitrites according to the old formula. The calculation can only be
etfected upon the supposition that the oxidation by the bromine should not be
quite complete, 3*71 equivalents of oxygen having been fixed instead of 4,
which is equivalent to admitting that the action of the bromine should have
liberated + 30-65 Cal. per equivalent of silver (taking into account the
formation of AgBr, which is not changed). We thus find —

J(N, + 02 + AgBr = 2AgNO) - g-25

N2 + 0 + water = N20 dissolved - 22-90

N20 dissolved + Ag20 = 2AgNG precip +11-15

N20 dissolved + K20 = 2KNO dissolved ... + 5-35

The deductions and general points of similarity remain moreover the same.

2 " Annales de Chimie et de Physique," torn. i. p. 394. 1816.

3 Ibid., torn. ii. p. 517. 1816.

4 Ibid., torn. vii. p. 36. 1817.

5 Ibid., 3e serie, torn. ii. p. 58. 1841.

NITROGEN MONOXIDE. 191

a red heat, or by the electric spark, into nitrogen and oxygen.
This decomposition is the easier, as it liberates heat.

N20 = N2 + 0 + 10-3 Cal.

In this way, it is not accompanied by dissociation, and is not,
therefore, reversible.

Experiments were made with a view of determining at about
what temperature this decomposition commences, and if nitric
oxide were produced. The monoxide resists the action of a
moderate heat better than is generally supposed. By heating
it to a dull red, about 520°, for half an hour, in a tube of
Bohemian glass hermetically sealed, hardly 1-5 per cent, is
decomposed into nitrogen and oxygen. The decomposition is,
therefore, extremely slow. Let us note here that the trans-
formation of nitrogen monoxide into nitric oxide at the ordinary
temperature,

N20 = N + NO, would absorb - I'O Cal.

The sudden compression of nitrogen monoxide in an apparatus
analogous to the gas tinder box (briquet a gaz) and under con-
ditions capable of causing the explosion of a mixture of
hydrogen and oxygen only produces traces of decomposition.

Nitrogen monoxide, mixed with oxygen and brought to a dull
red heat in a sealed tube, does not yield nitric oxide, which is
intelligible, its formation absorbing heat : i(N20 -f- 0 = 2NO)
would absorb — 11 '3. Finally, it must be remembered that
nitrogen monoxide does not exert an oxidising action in the
cold upon any known body, and that it is neither absorbed nor
decomposed by alcoholic or aqueous potash.

The action of the electric spark on nitrogen monoxide was
examined principally in order to study its first phases, for the
general products have already been noted by Priestley, Grove,
Andrews and Tait, as well as by Buff and Hoffmann. The
experiment was made in a sealed tube in order to avoid any
secondary action, from water or mercury.

Decomposition takes place rapidly, and nitrous vapour is
immediately formed. One-third of the gas was decomposed
within a minute. The decomposed part was divided in nearly
equal proportions between the two following reactions : —

N20 = N2 +0.
4N20 = N204 + 6JST.

The first action may be regarded as especially due to the
action of the heat of the spark; the second to the heat and
electricity combined. Further, both reactions are exothermal :
the first liberating + 10'3 Cal., and the second -f 38 Cal., that
is to say + 9*5 Cal. for every equivalent of nitrogen monoxide
decomposed. At the end of three minutes, with stronger sparks
(six Bunsen elements), nearly three-quarters of the gas was

192 OXYGENATED COMPOUNDS OF NITROGEN.

decomposed ; always in the same manner, the second reaction
slightly prevailing. Hence it will be seen that nitric oxide does
not and cannot appear in the electric decomposition of the
monoxide, since the latter always gives rise to an excess of free
oxygen.

The proportion of nitric peroxide, formed in these experi-
ments, represented nearly one-seventh of the final volume,
a proportion which cannot be very far removed from that corre-
sponding to the final equilibrium produced by the spark in an
equivalent mixture of free nitrogen and oxygen, according to
experiments detailed further on.

3. Nitric oxide. This gas is reputed one of the most stable.
It has, however, been taught that the spark (Priestley) or the
action of a red heat (Gay-Lussac) slowly decomposes nitric
oxide into nitrogen or nitric peroxide, and that in the presence
of mercury or iron there remains nothing but nitrogen (Buff
and Hoffmann, 1860).

These opinions do not appear to be well founded. Mtric
oxide l contained in a sealed glass tube and brought to a dull
red heat, about 520°, commences to decompose. At the end of
half an hour, the volume of the gas decomposed amounts to
nearly the quarter of the initial volume. The decomposed
portion was partly broken up into nitrogen monoxide and
oxygen —

2NO = N20 + 0, a reaction liberating + H'3 Cal.,
and partly into free nitrogen and oxygen —

2NO = N2 + 02, a reaction liberating -f- 21'6.

The first reaction, that is the formation of nitrogen monoxide,
was predominant ; but the oxygen, gradually regenerated in
presence of an excess of non-decomposed nitric oxide, had
partially transformed it, at first, into nitrogen trioxide —

2NO -f 0 = Na03 liberates + 10-5;
the total reaction,

4NO = N20 + N203, liberating + 217.

Then, the oxygen increasing owing to a more advanced decom-
position, nitric peroxide is formed —

2NO + 02 = 2N02 liberates -f 19'0 ;
the total reaction, that is to say

4NO = N2 + 2N02, liberating + 40'6 Cal.

1 This gas was prepared by the reaction of nitric acid on a boiling solution
of ferrous sulphate ; it is the only reaction which yields it quite pure. The
use of copper and nitric acid, even very dilute and cold, always gives rise to
monoxide of which the proportion, variable with the length of duration ot
the reaction, may amount to more than a tenth of the volume of the gas
disengaged.

ACTION OF ELECTRICITY ON NITRIC OXIDE. 193

Another experiment, lasting six hours, under the same con-
ditions, gave sensibly the same results, the proportion of nitric
oxide decomposed being the same, and of monoxide rather less,
but always very considerable. The action of the electric spark
confirms and extends these results. It commences to exert
itself with extreme promptitude, and presents several successive
terms which deserve attention.

Operating upon the gas enclosed in sealed tubes with rather
weak sparks (two Bunsen elements) a sixth of the gas was
already decomposed at the end of one minute. The proportion
would certainly have been larger, if the platinum electrodes
were situated in the centre of the mass instead of being at the
extremity, which retarded the mixture of the gases. About
a third of the decomposed product consisted of nitrogen
monoxide —

4NO = N20 + NA,

the other two-thirds producing nitrogen and nitric peroxide —

N2 + 2N02.

At the end of five minutes three-quarters of the nitric oxide
was decomposed with formation of nitrogen monoxide and
nitrogen trioxide and nitric peroxide. The ratio between the
nitrogen monoxide arid the nitrogen, that is, between the two
modes of decomposition, was nearly the same as above. It is
further necessary to distinguish between the calorific action of
the spark, which causes the formation of monoxide (a body not
formed by the spark acting on the elements), as well as of a
portion of free nitrogen, and the action peculiar to electricity,
as shown by an experiment of longer duration.

In fact, the flow of sparks prolonged for an hour leaves
nothing but a mixture of non-decomposed nitric oxide (thirteen
per cent, of the initial volume), nitrous vapour (more than forty
per cent.), and nitrogen. But no appreciable quantity of
monoxide was discovered. This gas therefore disappears before
the nitric oxide, doubtless under the influence of the high
temperature of the spark.

This fact, in apparent contradiction to the initial transforma-
tion of a part of the nitric oxide into monoxide, seems to show
that the nitric oxide commences to undergo decomposition at
a lower temperature than the monoxide, and that it nevertheless
lasts, in part, longer, or at a higher temperature, in presence of
the products of its decomposition.

However, the still more prolonged action of electricity causes
it to disappear in its turn, at the same time that it diminishes
the volume of the nitrous vapour produced in the first period.
After eighteen minutes only twelve per cent, of nitrous vapour
formed, solely of nitric peroxide. The gaseous mixture con-

o

OXYGENATED COMPOUNDS OF NITROGEN.

tained N = 44, 0 = 37, N02 = 13 for 100 volumes of the
original gas.

On account of the duration of the reaction, and of the
antagonistic influence tending to the formation of nitric peroxide,
in a mixture of pure nitrogen and oxygen traversed by the
spark, the above system must be regarded as nearly in a state
of equilibrium.

But to return to the nitric oxide. On the whole this compound
is less stable under ordinary conditions than the monoxide,
since it is capable of producing it by decomposition under the
influence of heat or the spark. Here an apparent contradiction
between the known properties of the two gases manifests itself.
Why do carbon, sulphur, phosphorus, when once ignited,
continue to burn more easily in the monoxide than in the
nitric oxide, a circumstance which has caused until now a
greater stability to be attributed to the latter gas ? The
explanation is the following (see pp. 62 and 63) : on the one
hand, nitric oxide does not contain more oxygen at equal
volumes than the monoxide, and, on the other hand, this
oxygen only becomes available in totality for combustion at a
much higher temperature, the nitric oxide being at first changed
to a great extent into nitric peroxide, a body really more stable
than nitrogen monoxide. The combustive energy of the nitric
oxide, at the temperature of incipient red heat, must therefore
be less than that of the monoxide, which is immediately resolved
into nitrogen and free oxygen.

We have explained in the same way the impossibility of
exploding a mixture of nitric oxide and hydrogen, or carbonic
oxide. The combustion produced at the point of contact
with the incandescent body, or on the path of the spark, does
not raise the temperature to the degree requisite for the de-
composition of nitric peroxide, whilst explosive mixtures
liberating far more heat, as is the case with cyanogen and
ethylene, explode with extreme violence.

The want of stability of nitric oxide is equally manifested in
a number of slow reactions, carried out with the pure gas at the
ordinary temperature, whether it be resolved into nitrite and
monoxide under the influence of potash (Gay-Lussac) —

4NO + K20 dilute + water = 2KN02 dissolved + N20
liberates + 39'2 Cal.,

or whether it gradually oxidise various mineral bodies in the
cold, according to the early observers, or certain organic com-
pounds, according to the author's own experiments.1

The latter reactions take place in various ways. Sometimes
the whole of the nitrogen of the nitric oxide is set free, libe-
rating -f 21 '6 more than the heat produced with free oxygen.

1 " Chimie organique fondee sur la synthese," torn. ii. p. 485.

DECOMPOSITIONS OF N1TKIC OXIDE. 195

Sometimes half the nitrogen only is set free, a slow reaction
observable with essence of turpentine or benzene, which leave
a residuum of nitrogen equal to the fourth of the volume of
the nitric oxide. Sometimes nitrogen monoxide is set free,
another slow reaction observable with sodium sulphide or
stannous chloride, which leave nitrogen monoxide and nitrogen
in equal volumes.

Sometimes even ammonia is set free, with the aid of the
hydrogen of water, or various organic compounds.

Nitrogen monoxide, nitrogen, and ammonia are formed from
the same causes in the greater number of reactions where an
oxidisable body tends to bring nitric acid to the state of
nitric oxide. Hence the latter gas, prepared by the reaction of
the metals on dilute nitric acid, is seldom pure.

A similar tendency to slow and multiple decompositions
is the distinctive character of unstable compounds formed
with absorption of heat. Nitric oxide is comparable, under
this head, with cyanogen and acetylene. Now, all these
endothermal compounds have a capacity for entering into
reaction, a sort of chemical plasticity very superior to that of
their elements, and comparable to that of the most active
radicals, a circumstance which may be explained by the excess
of energy stored up in the act of their synthesis.

The potential energy of the elements generally diminishes in
the act of combination ; acetylene, cyanogen, and nitric oxide,
however, form exceptions. There is no doubt some relation
between this increase of energy and the capacity possessed by
these compound radicals for entering directly into new com-
binations with the elements.

Under the influence of electricity we obtain the direct,
though always endothermal reunion of the elements which
form either acetylene itself or the hydrogenated combination
of cyanogen, or the super-oxidised combination of nitric oxide.

4. Nitrogen trioxide. Let us first note the following thermal
relations concerning anhydrous nitrous acid :

N203 = 2NO + 0 would absorb - 10-5 Cal.
N203 = 2N02 liberates + 8-5.

Hence it follows that the breaking up of nitrous acid into
nitric oxide and peroxide,

N203 = NO + NO* would absorb - 2'0 Cal.

In fact, the three bodies contained in the last equation con-
stitute a system in the state of dissociation, a system of which
the equilibrium varies with the relative proportions, temperature,
condensation, etc.

Gay-Lussac observed that oxygen and nitrogen, mixed in
volumes in the ratio of 1 : 4 in the presence of a concentrated
solution of potash, yield only nitrite.

o 2

196 OXYGENATED COMPOUNDS OF NITROGEN.

The same reaction occurs, whatever ~be the relative proportions
of the two gases and the order of the mixture, in presence of con-
centrated alkaline solutions, and even of baryta water. Not
only do the ratios between the volumes of the gases establish
this fact, but analyses made on several grammes of matter have
shown that the proportion of nitrogen trioxide formed corre-
sponds to 96 or 98 per cent, of the nitric oxide employed.

If the reaction occur without proper precautions being taken
to absorb the nitrogen trioxide, and particularly if it be
executed with anhydrous bodies, nitric peroxide is formed.1
Nitrogen trioxide acid cannot exist for any length of time
except in the presence of the products of its decomposition. It
is this complex mixture, variable according to circumstances,
which constitutes the body called nitrous vapour, whenever
oxygen is not in excess. The same remark applies, moreover, to
the liquid acid, the purest nitrogen trioxide which has been
obtained (Fritzche ; Hasenbach), containing about one-eighth
of nitric peroxide, according to the analyses. Peligot has
for long insisted on this circumstance.

In presence of an excess of oxygen, there is formed, or
rather there exists, only nitric peroxide, as is known from the
labours of Gay-Lussac, Dulong and Peligot, who obtained in
this way the crystallised acid. We will not dwell further
on this point, except to observe that, nitrogen trioxide being the
initial product of the reaction, even in presence of an excess of
oxygen, we are forced to admit that nitric peroxide results
from this nitrogen trioxide, combined afterwards with a second
equivalent of oxygen —

N203 + 0 = 2N02.

In a dry gaseous mixture, as well as in presence of water,
the formation of the two oxides takes place almost in-
stantaneously. Admitting, according to analogy, and in con-
formity with an approximate gaseous density given by
Hasenbach, that the formula N203 represents two volumes, the
second reaction would offer this remarkable character, hitherto
unique in the study of direct actions, of a real gaseous com-
bination accompanied by increase of volume, three volumes of
the component gases furnishing four volumes.

1 The experiments were made with a system of two concentric bulbs (see
p. 168) of known capacity, hermetically sealed, one containing dry oxygen,
the other dry nitric oxide, about 300 to 400 cms. The inner bulb is broken,
by a jerk, and the two gases are allowed to react. When the reaction is
complete, the point of the outer bulb is broken in a solution of potash of
known strength ; the nitrogen trioxide and nitric peroxide are absorbed with-
out affecting the nitric oxide. The nitrogen trioxide is absorbed without
change, as proved by the foregoing tests. Nitric peroxide in the state of
vapour is likewise completely absorbed, being changed according to a well-
known reaction into nitrogen trioxide and nitric acid.

ACTION OF WATER ON NITROGEN TRIOXIDE. 197

It would be the same with the metamorphosis of nitrogen
monoxide into nitric oxide —

N20 + 0 = 2NO,

if it could occur. In reality, this reaction does not take place
directly, being endothermal. But (pp. 192 and 194) the real
existence of the inverse decomposition, which presents an
anomaly of the same order and correlative, has been established,
viz. a simple gaseous decomposition effected with contraction:
four volumes being changed into three. The relation is more
clearly defined than the first, if not in principle, at least in fact,
seeing that it occurs between three gases of which the density
is known. If nitric peroxide is the final stage of oxidation of
anhydrous nitrogen trioxide by free oxygen, it is not the same
with nitrogen trioxide dissolved in water; for dilute solutions of
nitrogen trioxide gradually absorb free oxygen, and become
gradually changed into nitric acid: N203 + H20 -h 02 = 2HN03
dilute liberates -j-18'5. If ozone be substituted for oxygen the
oxidation of the nitrogen trioxide is instantaneous.

We now return to the action of water on nitrogen trioxide.
In presence of water the anhydrous acid becomes wholly or in
part hydrated nitrogen trioxide; it also shows a tendency to
decompose into nitric acid and nitric oxide. The reaction
3NA gas + water = 2HN03 dilute + 4NO liberates + 44.
But this last reaction only takes place to any appreciable
extent if water be present in sufficient quantity. In this case
it is partially decomposed into nitric oxide and oxygen, which
gradually transforms another portion of nitrogen trioxide into
nitric acid. This may be observed by treating solutions of
barium nitrite of various degrees of concentration with dilute
sulphuric acid. The immediate reaction here attributed to
nascent oxygen is the same as the slow reaction of free oxygen
on dissolved nitrogen trioxide.

From the well-known reaction of water on anhydrous nitrogen
trioxide, and from experiments on the distribution of baryta
among dilute hydrochloric and acetic acid and nitrogen trioxide,
the author is of opinion that a double dissociation is observed
when nitrogen trioxide is in presence of an insufficient quantity
of water, viz. the dissociation of the hydrated nitrogen trioxide,
which is partly changed into water and anhydrous acid, and the
dissociation of the anhydrous nitrogen trioxide, which is partly
changed into oxygen and nitric oxide. The effects are moreover
complicated by the ulterior action of the oxygen which dis-
appears in transforming another portion of the nitrogen trioxide
into nitric acid.

Under these conditions, the nitric oxide being eliminated
as produced, it would seem as if its formation should be in-
definitely reproduced.

198 OXYGENATED COMPOUNDS OF NITROGEN.

But the progressive dilution of the portion of hydrated
nitrogen trioxide which remains undecomposed (a dilution
resulting from the reaction itself) limits more and more the
relative proportion of anhydrous acid up to the point at which
the small quantity of nitric oxide remaining dissolved suffices
to ensure the stability of the system. Perhaps dilution, carried
out to a certain degree, completely arrests the decomposition
of the hydrated nitrogen trioxide, no longer permitting any
portion of the anhydrous acid to subsist.

In practice it is certain that a final system is realised contain-
ing at one and the same time water, dilute nitric acid, and
hydrated and diluted nitrogen trioxide. By diminishing the
relative proportion of water, the equilibrium would be destroyed ;
it would also be destroyed by raising the temperature, which
gives rise to a liberation of nitric oxide. Conversely, the
diminution of water may be compensated for by the lowering of
the temperature.

5. Nitric peroxide. We shall now examine the degree of
stability of nitric peroxide. This body is rightly regarded as
the most stable of the oxides of nitrogen ; in fact, it may be
heated in a sealed glass tube to about 500° for an hour, without
showing the least sign of decomposition. It moreover exerts no
reaction, either on oxygen in a cold state, or on free nitrogen at
a dull red heat under the same conditions. However, under
the influence of the electric current the mixture of oxygen
and nitric peroxide becomes discoloured, and gives rise to
a new compound, pemitric acid,1 about which very little is
known.

Nitric peroxide is decomposed into its elements by the electric
spark —

2N02 = £T2 + 04.

^ After an hour, as much as a quarter was decomposed. After
eighteen hours, a mixture was obtained containing in volume —

1ST = 28 ; 0 = 56 ; N02 = 14

We should note that the decomposition stops at a certain
point, as in all cases where the electric spark develops an
inverse action. It has, indeed, been known since the time of
Cavendish that the spark effects the combination of nitrogen
with oxygen. But this combination, effected with dry gases,
cannot yield anything but nitric peroxide, seeing that free
oxygen always remains, as will now be shown. Operating upon
atmospheric air it was found that after an hour 7'5 per cent.,
that is, a third by volume, had yielded nitric peroxide. Eighteen
hours of electric action did not sensibly alter this ratio.

This numerical value is not absolute. An exact measurement
would call for more numerous experiments, made under more
1 " Annales de Chimie et de Physique," 5e s^rie, torn. xxii. p. 439.

NITRIC PEROXIDE.

varied conditions, both with regard to electric energy, and
pressure and the relative proportions of the gases. The
important point is the existence of the limits, as a necessary
consequence of the two antagonistic reactions.

The action of water on nitric peroxide deserves attention.

If the water be in small quantity and the nitric peroxide
liquid, we obtain, as is well known, at a low temperature,
anhydrous nitrogen trioxide —

4N02 + H20 = N203 + 2HX03.

In the presence of a large quantity of water, nitric peroxide
gas, acting gradually, is completely absorbed with the formation
of hydrated nitric acid and nitrogen trioxide —

4N 02 + NH20 = 2HN03 dilute + N203 dilute.

This reaction liberates 7*7 Cal. for N02 = 46 grms.

But liquid nitric peroxide, in presence of the same quantity
of water, gives rise, generally speaking, to some nitric oxide,
according to the following reactions, which refer to quantities
of substances of which the proportion is variable with the
conditions of contact :

3N02 + ^H20 = 2HN03 dilute + NO.

This reaction, which may be limited almost to nil when contact
is gradually effected, liberates, after it takes place, + 4*8 for
N02.

The following experiment is easy of repetition, and clearly
shows both modes of decomposition of nitric peroxide under the
influence of water. Into a rather large tube, closed at one end
and formed at the other into a funnel, is poured a little liquid
nitric peroxide, which, in order to drive out the air, is brought
into a state of ebullition, leaving only an insignificant quantity
of liquid. The tube is then hermetically sealed. Liquid peroxide
is then poured into a similar but much smaller tube ; the air is
expelled in the same way by boiling, and the tube is closed.
After cooling, the large tube, being opened over water, fills
completely, owing to the total decomposition of the peroxide
into nitrogen trioxide and nitric acid. On the other hand, the
small tube is only partly filled, owing to the formation of nitric
oxide.

The difference between these two reactions appears to be
due to the slight stability of hydrated nitrogen trioxide above
defined (p. 197). If the peroxide has at the outset enough water
to form hydrated nitrogen trioxide without decomposition the
absorption is complete. This is the case with gaseous peroxide
and water gradually reacting over a large surface. But if it
comes into contact at one point with too small a quantity of
water the acid will be partly decomposed with formation of

200

OXYGENATED COMPOUNDS OF NITROGEN.

nitric oxide which will not be redissolved. Lastly, the contact
of the same quantities of substances, effected by degrees, will
not give rise to nitric oxide, or if so, only to a very small
extent.

6. Nitric acid. We have said that anhydrous nitric acid
manifests a certain tendency to be spontaneously decomposed at
the ordinary temperature, and this appears to be due to the
action of light. A few rays of sunlight are sufficient to cause
an abundant liberation of oxygen and nitric peroxide. Sponta-
neous decomposition also takes place in diffused light, but very
slowly. This decomposition is accelerated with rise in tempera-
ture, without, however, being very rapid up to 43°. It is
endothermal, for it absorbs — 2*0 ; for N205 gas = 2N02 + 0,
and is not reversible, dry nitric peroxide not absorbing oxygen
at any temperature, as has been proved by exact analysis. It is
well known that light also decomposes monohydrated nitric
acid.

7. Heat liberated in the various oxidations effected by nitric
acid. The oxidation of the metals and other oxidisable bodies
by nitric acid gives rise, according to circumstances, to the four
lower oxides of nitrogen, to nitric peroxide, nitrogen itself, to
hydroxylamine, ammonium nitrate, and ammonia, the ultimate
term of the reduction of nitric acid by hydrogenated bodies.
The following is the method of calculating the heat liberated. Q
being the heat supposed to be produced by the union of an
equivalent of free oxygen (0 = 8 grms.) with the oxidisable
body, the latter being changed, further, either into an oxide or
soluble salt, we shall have —

The products being

With HN03 by

With HN03 + 4H02
ordinary acid.

With HN03
dilute.

N2O4 gas + O yielded.

Q-9-7

Q - 16-1

Q - 16-9

N2O, gas + O2 yielded

(Q - 9-1) x 2

(Q - 12-3) x 2

(Q-127)x2

N20, diss. + 02 yielded .

» >»

(Q - 9-3) x 2

N2O2 gas + O3 yielded
§N408 diss. + 03.5 yielded
N20 gas + 04 yielded.
N2 gas + O5 yielded .
2NH,O diss. + Ofl yielded

(Q - 9-6) x 3

>» »» »>
(Q - 4-3) X 4
(Q - 1-4) x 5
2H2O in excess.

» »

(Q -'5-9) x 4
(Q - 2-6) x 5
(Q - 16-3) x 6

(Q- 12-0) x »
(Q-9-6)x3-5
(Q - 6-1) x 4
(Q - 2-8) x 5
(Q-16-4)x 6

2NH3 + 08 yielded . .

>» »•

(Q - 12-0) x 8

(Q-12-l)x8

2HNO3NH3 diss. + 08 yielded

fHN03 + H20^

tinthisreaction/

(Q - 10-4) x 8

(Q-10-5)x8

It will be seen that the heat liberated constantly increases from
nitric peroxide to nitrogen according as the reduction becomes
more complete, without, however, attaining to the heat which
free oxygen would produce. When hydrogen comes into play,
the formation of hydroxylamine and ammonia diminishes, on the
other hand, the heat liberated.

FORMATION OF AMMONIA FROM NITRIC ACID. 201

8. We give also the figures relating to nitrogen trioxide.

N203 dilute =

N202 + 0 yielded, .liberates Q-17'4

N20 + 02 „ „ (Q- 3-0) x 2

N2 + 03 „ „ (Q + 1-4) x3

N2H602 + 04 yielded /3H20 supplemented! „ (Q-20-1) x 4
2NH3 + O.J „ \ in the reaction) / „ (Q-13'0) x 6

It is well known that nitrogen trioxide oxidises bodies more
easily than nitric acid. This difference is accounted for by the
state of dissociation characteristic of nitrogen trioxide (pp. 196
and 197).

The formation of ammonia in oxidations effected at the
expense of nitric acid is equally deserving of our attention.

It is a secondary reaction, for it seems to be produced only
by the action of free hydrogen (spongy platinum) or by a metal
capable of liberating the hydrogen of water by dissolving in
more or less diluted acids, which requires the subsidiary relation
Q > 34-5.1

In order to form a proper idea of the conditions of this
formation, it is well to distinguish the general function of dilute
acids, the water in these compounds tending to be destroyed by
the metals with liberation of hydrogen, from the special function
in virtue of which nitric acid produces ammonia. Take dilute
sulphuric or hydrochloric acid in presence of a metal capable of
setting free its hydrogen, and a small quantity of nitric acid
to intervene, we shall provoke the following reaction : —

HN03 dil. + 8H = NH3 dil. + 4H20, which liberates + 248*2,
or 41*4 Cal. for every equivalent of oxygen (0 = 8 grms.)
eliminated. The ammonia combining with the excess of
sulphuric acid, the heat liberated will be raised by -f 12*4, which
makes altogether for each equivalent of oxygen + 43*5.

1 Or rather Q > 34'5 - S, S being the heat of solidification of hydrogen, for
it would be necessary to compare the metal and hydrogen under the same
physical state.

( 202 )

CHAPTEE IV.

HEAT OF FORMATION OF THE NITRATES.

1. THIS chapter will treat of the heat of formation of potassium
nitrate and the other nitrates, used in the manufacture of a multi-
tude of explosive mixtures.

The heat of formation of potassium nitrate from its elements
is easy to calculate provided we know, at a temperature of
about 15°—

(1) The heat of formation of dilute nitric acid from nitrogen
and oxygen.

N2 + 05 + H20 + water = 2HN03 dil. liberates + 14'3.

(2) The heat of formation of dilute potash from potassium
and oxygen.

K2 + 0 + H20 + water = 2KHO dil. liberates + 82'3.

(3) The heat liberated in the combination of dilute nitric acid
and dilute potash.

KHO dil. + HN03 dil. = KN03 dil. + H20 liberates + 13-8.

(4) Lastly, the heat which would be liberated if the solid
potassium nitrate separated itself from its dilute solution, a
heat which is precisely equal in absolute value to the heat
absorbed in the act of dissolving the same salt, but with the
opposite sign.

KN03 dilute = KN"03 crystallised -f water would liberate -f 8-3.
The sum of these four quantities, viz.

14-3 + 82-3 + 13-8 + 8'3 = + 1187 Cal.,
exactly expresses the heat liberated by the union of the elements
of crystallised saltpetre, taken at the weight of 101 grms.

N2 + 06 4- K2 = 2KN03 solid liberates + 118'7.
The formation of dissolved saltpetre from the same elements
would liberate + 110*4.

From anhydrous potash, nitrogen and oxygen, N2 -f 05 + K20
= 2KN03 solid liberates 701.

HEATS OF FORMATION. 203

From dissolved potash, the formation of dissolved saltpetre, N,
+ 0, + K20 dilute = 2KN03 dilute liberates + 281 only.

2. Similarly we have for sodium nitrate —

N2 + 06 +Na2 = 2NaN03 crystallised + 110'6,

and for the dissolved salt -f 105*9.
From anhydrous soda, oxygen and nitrogen —

N2 4- 05 + Na^O = 2NaN03 crystallised + 60-5.

From dilute soda the formation of dissolved sodium nitrate
liberates + 28 '0.

3. The formation of ammonium nitrate —

N2 + 03 -f H4 = NH4lsr03 crystallised + 87-9.

The dissolved salt + 817.

If we suppose that the equivalent of water necessary to the
constitution of the ammoniacal salts is formed beforehand, we
have liberated for the salt supposed solid + 53 '4 CaL, for the
salt supposed dissolved + 47'2 Gal., or + 23'6 Cal. for each
equivalent of nitrogen entering into combination, in presence
of an excess of water. From ammonia gas and pre-existing
water —

N2 + 05 + 2NH3 + H20 = 2NH4N03 crystallised + 41-2.
From dilute ammonia, the dissolved salt + 26'8.

4. The formation of calcium nitrate —

N2 + 06 + Ca = Ca(N03)2 anhydrous + 101-3.

For the dissolved salt + 103-3.
From anhydrous calcium oxide —

N2 + 05 + CaO = Ca(N03)2 anhydrous + 35-3.

From dissolved calcium oxide, the salt being likewise dis-
solved, + 28-2.

5. The formation of strontium nitrate —

N2 + 06 + Sr = Sr(N03)2 anhydrous + 109-8.

For the dissolved salt -f 107'3.
From the anhydrous base —

N2 + 05 +SrO = Sr(N03)2 anhydrous + 41-1.

From dissolved strontium oxide, the salt being likewise dis-
solved, + 28-2.

6. The formation of barium nitrate cannot be calculated
from the elements, because the heat of oxidation of barium is
unknown. Fortunately, this total heat of formation never
intervenes in calculations relative to explosive substances. To
calculate the thermal effects which barium nitrate produces in
combustions it is sufficient to know its heat of formation
starting from anhydrous baryta.

204

HEAT OF FORMATION OF THE NITRATES.

]sr2 + 05 4- BaO = Ba(N03)2 liberates 4- 47'2.
From the dissolved base, the salt also being dissolved, 4- 28*2.

7. It may be remarked that the heat of formation of the
alkaline and alkaline-earthy nitrates, by means of gaseous
nitrogen, gaseous oxygen, and the dissolved base, is* sensibly the
same for all. The same figure (4- 28 1) applies equally to
magnesium nitrate, as it is formed from solid magnesium
hydrate.

8. The formation of the anhydrous nitrates from the
anhydrous base and anhydrous nitric acid, whether gaseous or
solid, is given in the tables on p. 126. Similarly, the forma-
tion of the solid nitrates, formed by solid hydrated nitric acid
and basic hydrates also solid, is given in the table on p. 127.

9. We should further note that the metamorphosis of the
alkaline nitrites into nitrates M(N02)2 dissolved 4- 02 =
M(N03)2 dissolved, liberates a quantity of heat nearly equal
to 4- 21f7, and sensibly the same whatever be the base of the
salt (p. 178).

10. The heat of formation of the anhydrous magnesium, iron,
cobalt, nickel, and manganese nitrates, cannot be calculated,
these salts being only known in the hydrated state. In the
dissolved state we have, from the metals and the metallic
oxides —

N2 + 06 + Mg
N2 + 06 + Mn
N2 + 06 -f Fe
N2 + 06 + Zn
N2 + 06 + Co
N2 + 06 + Ni
N2 + 06 + Cd
N2 + 06 + Cu

11. The formation of lead nitrate from the elements

N3 4- 06 4- Pb = Pb(N03)2 anhydrous liberates 4- 52'8.

That of the dissolved salt 4- 48'7.

The formation of the same salt from the anhydrous oxide,
N2 4- 05 4- PbO = Pb(N03)2, liberates 4- 27'3.

The dissolved salt 4- 23'2.
The formation of silver nitrate from the elements

N2 4- 06 4- Ag2 = 2AgN03 anhydrous liberates 4- 28'7.

That of the dissolved salt 4- 23'0.
The formation of the same salt from the oxide,

N2 4- 05 4- Ag20 = 2AgN03, 4- 25-2.
The dissolved salt 4- 19'5.

12. We will add the following general remarks. Between
the formation of two salts obtained by the union of the

rates 103-0
73-5
59-5
67-3
56-9
56-3
57-6
39-8

N2 + 05 + MgO libe
N2 + 05 + MnO
N2 + 05 + FeO
N2 + 05 + ZnO
N2 + 05 + CoO
N2 + 05 + NiO
N2 + 06 + CdO
N2 + 06 + CuO

;rates + 58-1
. + 26-1
+ 25-0
+ 24-1
9 + 24-9
+ 25-6
+ 24-4
+ 21-8

IMPORTANCE OF THEORETICAL CONSIDERATIONS. 205

same alkaline base with two distinct acids, these salts being
considered under the solid and anhydrous form, we find a
nearly constant thermal difference, whatever be the base, when
we reckon the quantities of heat liberated from the elements up
to the anhydrous salts. For example, the formation of the
anhydrous potassium, sodium, ammonia, calcium, strontium,
lead and silver sulphates, liberates a mean value of 54 Cal.
more than the formation of the corresponding nitrates.

A similar difference exists between the nitrates and the
majority of the oxygen salts. It exists even between the
alkaline chlorides, bromides, and iodides, without, however,
extending itself to the anhydrous metallic chlorides.

13. These numbers permit, as will be shown later, of estima-
ting the heat liberated by any decomposition or definite com-
bustion of service powder or other powders, inflammable
materials or explosive mixtures constituted by the nitrates.
It is with the aid of analogous data, derived from the heat of
formation of nitric acid, that we can calculate the heat of
formation of nitroglycerin, and of organic compounds derived
from nitric acid. The figures thus calculated agree moreover
with the experiments of Sarrau and Vieille, as far as can be
expected in verifications of this nature.

14. If this agreement is dwelt upon, it is because, in the
author's opinion, the applications of explosive substances, as
well as the applications of human industry, need to be guided
by theoretical notions. We must raise ourselves above em-
piricism if we wish to obtain the most favourable results. It
is thus that blasting powder, so long exclusively employed in
practical applications, tends to-day to be replaced by dynamite
in the majority of its uses. Now this substitution is encouraged
and regulated by theory. Indeed, the latter teaches us that
blasting powder, as well as service powder, is far from utilising
in the best manner the combustive energy of nitric acid.

In the combustion of ordinary powder, the products formed
are neither the most oxidised, nor those which would liberate
the most heat for a suitable proportion of the various ingredients,
seeing that the maximum of heat which would be developed by
a known weight of saltpetre acting on the sulphur and the
carbon does not correspond to the maximum volume of the
gases liberated. Between these two data of the problem,
empiricism has led to a sort of compromise being adopted,
which is our traditional powder. But it would be far preferable
to arrange in such a manner that the maximum of the two
effects should occur in it for the same proportions.

This is not all. The formation of potassium nitrate itself,
reckoned starting either from nitric acid or the elements, corre-
sponds to very powerful affinities and gives rise to a greater
liberation of heat, and consequently to a greater expenditure of

206 HEAT OF FORMATION OF THE NITRATES.

energy than most of the other combinations derived from nitric
acid.

Theory therefore shows that saltpetre is not a favourable
agent of combustion ; and in this way it explains the superiority
of the organic compounds derived from nitric acid, and especially
the nitric ethers, such as nitroglycerin. As a matter of fact,
the author's experiments show a much inferior liberation of
heat, that is to say, a greater preservation of energy in the
formation of these substances. The energy introduced into an
explosive compound, formed by the same weight of nitric acid,
is in nitroglycerin double that which is found in service powder.
Hence it is easy to understand how the abandonment of blast-
ing powder for industrial purposes is gradually extending.
Perhaps it will be soon the same with service powder, if practice,
guided by the new theories, succeeds in discovering more active
nitrogenated compounds than powder, which will satisfy the
manifold conditions called for in the use of explosive substances
in firearms.

( 207 )

CHAPTEE V.

ORIGIN OF THE NITRATES.

§ 1. NATURAL NITRIFICATION.

1. THE formation of nitre in nature has long been regarded as a
most obscure phenomenon.

It has long been known that the alkalis and the alkaline
carbonates, when exposed for some time to the air, yield the
reactions of nitric acid. Stahl had already observed this two
hundred years ago. At all times and in all places, under the
action of natural forces, there are produced small quantities of
nitrates.

2. There also exist certain plants which appear to produce salt-
petre, at the expense of the nitrated combinations contained in
the soil or in manures. Such are borage, pellitory, beetroot,
tobacco, and especially plants of the family of the amarantaceae.1
Nevertheless, the conditions of natural nitrification are still
imperfectly known.

3. It is not proposed to refer here to the sodium nitrate
mines in Chili, formed under the influence of geological con-
ditions with which we are unacquainted, but only to the
nitrification going on every day under our eyes.

4. In the first place, we know that nitric acid is formed in
the atmosphere in small quantities under the influence of
storms, simultaneously with a little ammonium nitrate, and
introduced into the soil by rain and there united to the bases.
This formation is of great interest. But a searching examination
has shown that such an origin does not suffice -to account for
the production of the nitrates in nature and their concentration
in a soil impregnated with animal matter.

5. As a matter of fact natural nitrification results principally
from the slow oxidation of the nitrogenous organic compounds,
or even of ammonia, effected by the oxygen of the air, with the
aid of water and of an alkaline or earthy carbonate.

1 Compare note sur Tattraction du salpetre, par Faucher (" Memorial des
potidres et salpetres," p. 162. 1883).

208 ORIGIN OF THE NITRATES.

Too strong a light checks it. Clayey substances and
porous matters appear to favour it, but it does not appear
that free nitrogen intervenes in this mode of formation of
saltpetre.

6. Various questions here present themselves. Thus it has
been asked whether this slow oxidation is simply provoked by
the presence of clay and porous bodies, as occurs in Kuhlmann's
experiments, where the ammonia is changed into nitrous vapour
and nitric acid on contact with spongy platinum and oxygen at
about 300°.

Are the humus principles, the sulphuretted and ferruginous
compounds, and the other oxidisable bodies which are decom-
posed in the soil, at the same time that nitre is formed, the
medium of some special reaction ?

Do they provoke the oxidation of the ammonia, becoming
oxidised themselves, as occurs with copper in presence of the
air ? Phosphorus does in fact exert an analogous reaction, and
this influence has also been attributed to humus.

Does an oxidising body properly so called intervene, after the
manner of potassium bichromate and sulphuric acid, or of
manganese dioxide at a red heat, when the latter agent changes
the ammonia into nitrous vapour ?

Does ozone play some such part, as held by Schonbein,
according to whom certain plants emit ozone, a substance, in
fact, capable of oxidising ammonia at ordinary temperatures,
with formation of nitrite.

Lastly, do the mycoderms and microbes cause this oxidation
after the manner of a fermentation ?

Such are the principal hypotheses which have been brought
forward since the eighteenth century up to our time to explain
the apparently spontaneous formation of nitre in nature.

At the present day these questions, which have been for so
long a time the object of controversy, appear to have made a
decisive step forward in consequence of the recent experiments
of Schloesing and Mimtz.1

7. These investigators have found that the nitrification of
ammonia and the nitrogenous organic compounds takes place
under the influence of pointed, rounded, or slightly elongated
organised corpuscles, sometimes adhering in pairs of very
small dimensions, and very similar in appearance to the
corpuscular germs of bacteria. These corpuscles occur in all
arable soils and in sewage water, which they aid in purifying.
They cause the fixation of oxygen upon ammonia and nitrogenous
substances, generally forming nitrates, sometimes nitrites, when
the temperature is below 20° or the aeration insufficient. The
nitrites also result from the reduction of the original nitrates

). 301, 1877; torn. Ixxxv. p. 1018;
torn. "

1 "Comptes rendus," torn. Ixxxiv. p. 301, 1877; torn,
m. Ixxxvi. p. 892 ; torn. Ixxxix. pp. 891 and 1074 : 1879.

THE NITRIC FERMENT. 209

by the intervention of the butyric ferment and of analogous
secondary ferments.1

Their action is exerted between determinate limits of
temperature. Below 5° it is inappreciable, becoming appreciable
at 12°. It becomes more and more active as the temperature
rises to about 37°, at which temperature the nitrification is ten
times more rapid than at 14°, though still rather slow, all the
other conditions moreover being the same. Beyond this it
grows slower; at about 45° it is less active than at 15°, and
ceases completely at 55°.

According as the temperature rises, and especially if it be
brought to 100°, the vitality of the corpuscles diminishes, so
that mould or water in course of nitrification loses this property
without recovering it after cooling. They also perish under the
influence of the vapours of chloroform and antiseptics.

Moisture is indispensable to them. It is even sufficient to
dry in the air a fertile piece of mould for it to become sterile
after a time. The corpuscles do not support a prolonged
privation of oxygen, at least when operated upon in a liquid.

They act equally well in the dark or under the influence of
a moderate light, but a strong light is prejudicial to them.

Their action requires the aid of a slight alkalinity, due
either to the presence of calcium carbonate, or to that of two to
three thousandth parts of alkaline carbonates. Beyond this
degree alkalinity injures them, which accounts for the un-
favourable influence exerted by liming upon nitrification. The
development of the nitric ferment in water requires the
simultaneous presence of an organic substance and a nitrogenous
compound. But the ratio between the carbonic acid and the
nitric acid produced is in no way constant. It is the same with
the absorption of oxygen, which is continually going on at the
expense of a soil which has been rendered sterile by a tempera-
ture of 100° or by the action of chloroform vapours.

The nitric ferment is multiplied by sowing a nourishing
liquid, or earth, with a small piece of arable soil or a few
cub. cms. of sewage. It does not generally exist in the dust in
the air. Its multiplication is slow, and seems to be effected by
budding. The existence or absence of porous bodies appears to
have very little to do with nitrification, contrary to the views
formerly held.

Ordinary mould and mycoderms are quite distinct from this
ferment, and even contrary to its action. In fact, they destroy

1 Dehe rainet Maquenne, " Comptes rendus," torn. xcv. p. 691 ; Gayen,
same collection, torn. xcv. p. 1365. These auxiliary ferments, or rather
perturbators, reduce inversely the nitrates with production of nitrites,
nitrogen monoxide, free oxygen, and even of ammonia, according to their
nature and the greater or less intensity of their action. The hyponitrites
must also intervene.

P

210 ORIGIN OF THE NITRATES.

the nitrites, and change them into organic nitrogenous com-
pounds during the development of their mycelium. They act
in the same way upon ammonia or the ammoniacal salts, and
even by preference. Later on, during fructification, a portion
even of the nitrogen is eliminated in the gaseous form, some-
times with intermediate reproduction of ammonia.

These observations, as a whole, show the existence of par-
ticular organised beings, analogous to the acetic ferment, which
cause the fixation of oxygen upon ammonia and nitrogenous
organic compounds, and consequently the change of these
substances into nitrates. They go far to resolve the problem of
nitrification, effected in nature at the expense of the nitrogenous
or ammoniacal compounds ; a problem, moreover, which is quite
distinct from the fixation of free nitrogen taken from the
atmosphere. It is, however, allied to it ; for natural nitrifica-
tion is effected upon already formed and pre-existing nitrogenous
compounds.

§ 2. CHEMICAL AND THERMAL CONDITIONS OF NITRIFICATION.

1. These facts being admitted, it will be useful to show that
the study of the quantities of heat liberated during the act of
natural nitrification throw a fresh light upon the latter. In
order to render the discussion clearer, it will be best to attempt
at the outset to define the chemical conditions of this oxidation,
as far as can be done in the present state of our knowledge.

2. The most developed experiments which have been per-
formed on the chemical conditions of nitrification are, even at
the present day, those of Thouvenel, although they date from
nearly a century ago.1 They show that nitrification takes
place principally in connection with the gaseous compounds
produced in putrefaction, mixed with an excess of atmospheric
air. We know at the present day that the most important of
these compounds are ammonia, ammonium carbonate, hydro-
sulphide, hydrocyanide, and perhaps hydrocyanic acid. That
it requires the aid of moisture. That it is more easily effected
in the presence of the alkaline or earthy salts than in their
absence. Lastly, it hardly occurs save with carbonates, to the
exclusion of sulphates. For example, a basket pierced with
holes, and containing well-washed chalk, being placed over
blood in a state of putrefaction, the chalk was found after some
months to contain 2 -5 per cent, of nitrate. A plate, containing
washed mortar and placed in the atmosphere of a stable, con-
tains nitrates at the end of three weeks, etc. These conditions
agree with the biological conditions which preside at the

1 "Me'moires de 1'Acade'mie des Sciences" (Savants Strangers), torn. xi.
1787.

NECESSITY FOR ALKALINE MEDIA. 211

development of the nitric ferment, as they have been defined
above.

3. These various circumstances may also be accounted for
from the chemical point of view. We proceed to enter into
detail upon this subject. Ammonia and oxygen are, we have
said, the generators of the nitrates. Take, first, ammonia. The
liberation of gaseous ammonia, supplied by the slow transfor-
mation of nitrogenous organic principles, takes place only in an
alkaline medium. In an acid liquid it is clear that this
liberation cannot take place.

Neither can it take place in a liquor capable of forming only
neutral and fixed ammoniacal salts by double decomposition,
such as the sulphate.

On the other hand, it is facilitated when the liquor can give
rise by double decomposition to a volatile and partly dissociated
ammoniacal salt,1 such as the carbonate. /The presence of a")
fixed alkali, or of an alkaline carbonate, is not only useful for
setting free the pre-existing ammonia of the ammoniacal salts ;
it further causes the generation of ammonia, at the expense of
the principal organic nitrates, in virtue of a sort of predisposing
affinity, owing to the intervention of the excess of energy
resulting from the saturation of the bases by the acids produced
during oxidation. Let us now turn to the latter phenomenon.

Air, or rather its oxygen, is indispensable, because we are
here dealing with a phenomenon of oxidation incapable of
taking place in a reducing medium, such as a substance under-
going putrefaction.

From the same point of view, the presence of an alkali, or of
a salt having an alkaline reaction, is very efficacious in accele-
rating the oxidation of organic principles by the oxygen of the
air, and at the ordinary temperature, while they offer much more
resistance in an acid medium. The mode itself in which the
oxidation of ammonia takes place during nitrification helps to
account for the efficacy of the fixed alkalies and their carbonates.
Now, the slow oxidation of ammonia develops nitrous, then
nitric acid, which must gradually combine with the portions
of free and non-oxidised ammonia. Hence, finally, results
ammonium nitrate, that is, a salt fixed at the ordinary tempera-
ture and devoid of alkaline reaction. If a nitrogenous principle,
taken by itself, were operated upon, half the ammonia would
thus be withdrawn from the oxidising action, and at the same
time the liquor would constantly tend to lose the alkaline
reaction due to the existence of free ammonia, a reaction which
facilitates oxidation. But the alkaline carbonate retains the
alkaline character, because it gradually transforms the nitrate
of ammonia into fixed alkaline nitrate and ammonium carbonate,

1 "Essai de Mfoanique Chimique," torn. ii. p. 717.

P2

212 ORIGIN OF THE NITRATES.

which is partly dissociated, with formation of free ammonia.
Now the latter is capable of ulterior oxidation.

Further, the author has established, by direct and accurate
experiments, that dissolved ammonium nitrate in presence of
potassium or sodium carbonate is instantly transformed into
potassium or sodium nitrate and ammonium carbonate, the
strong acid taking by preference the strong base and leaving to
the weak acid the weak base.1 Calcium carbonate produces the
same reaction. "We shall return to the consideration of this
reaction on account of the part which it plays in natural
nitrification.

If we now consider the thermal phenomena which
accompany these various chemical reactions, we shall be able
to understand more fully the part played by them in
nitrification.

4. Take first the transformation of ammonia into nitrous
acid, nitric acid, and ammonium nitrate 2 —

Nitrous acid, NH3 + 03 = HN02 + H20.

Nitric acid, NH3 + 04 = HN03 + H20.

Nitrate ammonium, 2NH3 + 04 = NH4N03 + H20.

The formation of gaseous ammonia by its elements
N + H3 = NH3

liberates, according to the author's measurements, + 12*2 Cal. ;
that of dissolved ammonia liberates + 21 '06 Cal.
Lastly, the formation of water,

H2+ 0 = H20,

liberates + 34'5 or + 29'5 according as the water is produced
in the liquid or the gaseous state. It follows from the above
that the oxidation of ammonia, whether rapid or slow, liberates
the following quantities of heat according to the nature and the
state of the products to which it gives rise.

(1) Formation of nitrogen.

2NH3 + 03 = N + 3H2(X

Gaseous ammonia and gaseous water + 88-5 - 12-2 = + 76*3.
Dissolved ammonia and liquid water + 103-5 - 21-0 = -f- 82-5.
Gaseous ammonia and liquid water + 103-5 - 12*2 = + 91-3.

(2) Formation of nitrous acid.

-f 03 = HN02 + H20.

Gaseous ammonia, water, and dilute nitrous acid ... + 87'1.
Dissolved ammonia, water, and dilute nitrous acid ... 4- 78 -3.

1 " Essai de Me*canique Chimique," torn. ii. p. 717,

2 It would be well, no doubt, also to establish analogous calculations for the
hyponitrites (see p. 188).

NITRA

AMMONIUM NITRATE CHANGED INTO NITRATE. 213

(3) Formation of nitric acid.

NH3 + 04 = HN03 + H20.

Gaseous ammonia, water, and gaseous nitric acid + 81*2.
Gaseous ammonia, liquid water, dilute nitric acid + 105' 6.
Dissolved ammonia, dilute nitric acid .. +96-8.

(4) Formation of dissolved ammonium nitrate.

2NH3 + 04 = NH4N03 + H20.
Gaseous ammonia, dissolved nitrate ... + 125*3.
Or, for KE3 -f 02, + 62-6.

(5) Transformation of dissolved ammonium nitrite into nitrate
ly fixation of oxygen.

This transformation, and more generally that of a dissolved
nitrite into a nitrate of the same base, liberates + 21*8 ; a value
which is sensibly the same for the various dissolved alkaline
nitrites. This value offers the more interest, as the change of
the nitrites into nitrate and the inverse transformation take
place in nature, as shown by the very curious experiments of
Chabrier1 and the recent researches of Gayon, Deherain, and
Maquenne.

The presence of the nitrites has been remarked in stables, aa
co-existing with the nitrates, by Goppelsroder. They also exist
in rainstorms. The hyponitrites should also be searched for.

5. All the foregoing figures are applicable to the oxidation
of ammonia by free oxygen, whether this oxidation take place
by sudden combustion, or whether it be excited at a lower
temperature by spongy platinum, or whether it take place
slowly and in the cold state, as in nitrification.

They show that the formation of the oxygenated compounds
of nitrogen by the oxidation of ammonia always takes place
with liberation of heat. It can, therefore, always take place
without the aid of any foreign energy; the microbes con-
fining themselves, as in all cases where their action is exerted,
to cause a formation, to which they contribute no energy of
their own.

Conversely, the formation of ammonia by the action of
hydrogen on the various oxides of nitrogen liberates more heat
than the same formation effected by means of free nitrogen;
which accounts for the greater facility of the first reaction.
But it is not necessary to go at length into this subject, which
is foreign to the question of nitrification, though it plays a
certain part in the reduction of the nitrates to the state of
ammonia by natural agents.

6. Various experiments have been made with a view to
discovering whether free ammonia could be directly oxidised by

1 " Oomptes rendus des stances de 1' Academic des Sciences," 1871.

214 ORIGIN OP THE NITRATES.

the oxygen of the air, at the ordinary temperature, with the aid
of time, and without that of the microbes.

Large flasks full of air, well closed, and exposed to a
moderate light in presence of potash and its dissolved carbonate,
were employed. There was also introduced simultaneously
with the alkalies a small quantity of oxidisable substances,
naturally indicated for the purpose, such as glucose, and
essence of turpentine. But no nitre was obtained even after
several months (March to June, 1871). In spite of these
negative trials, the oxidation of ammonia during nitrification
cannot be questioned, but the conditions attendant upon it are
only known since the already cited experiments of Schloesing
and Miintz.

7. It will be interesting to further examine the integral trans-
formation of ammonium nitrate into potassium nitrate. It has
been stated, in fact, that ammonia could yield at first, in becom-
ing oxidised, ammonium nitrate. It can further be shown that
the whole of the nitrogen contained in this salt passes to the
state of potassium nitrate.

Two phases manifest themselves during this change.
The first transformation produces potassium nitrate and
ammonia, finally oxidisable. This transformation is effected,
both in nature and in the laboratory, by dissolved potassium
carbonate. The double decomposition between the two salts,
separately dissolved in equivalent proportions, gives rise, accord-
ing to the author's experiments, to a noteworthy thermal
phenomenon; that is, to an absorption of 3 Calories per
equivalent. This phenomenon shows that the potassium
carbonate is changed into ammonium carbonate in the liquor ;
since the formation of the latter salt by means of the dissolved
acid and the dissolved base, liberates far less heat than that of
the potassium carbonate.1

Now the ammonium carbonate thus formed in the solution
disappears by reason of the evaporation of the liquor, or even
by the mere fact of the diffusion of carbonic acid and ammonia
into the atmosphere ; so that there remains nothing at the end
but potassium nitrate, either in the liquor concentrated by
evaporation, or in the efflorescent residuum which this liquor
yields by spontaneous evaporation.

The ammonia, on the other hand, after having been brought
to the gaseous state, is separated from the carbonic acid, owing to
the diffusion of the two gases into the atmosphere ; it is oxidised
afresh under the 'influence of the same causes, whichever they
may be, that have already changed the half of this base into
nitric acid. The other half becomes in its turn ammonium
nitrate, and the latter body again reproduces ammonia by the
same mechanism, but it does not reproduce more than a quarter

1 " Essai de Mdcanique Chimique," torn. ii. p. 717.

FOREIGN ENERGY UNNECESSARY IN NITRIFICATION. 215

of the original quantity. The sequence of reactions goes on in
this way and the whole of the ammonia is finally changed into
potassium nitrate, provided the liquor contains an excess of
potash.

The transformation of ammonium nitrate into calcium or
magnesium nitrates takes place in virtue of similar reactions,
with this difference, however, that the double decompositions
can take place between ammonium nitrate and the earthy
carbonates, especially when the latter are dissolved by carbonic
acid (bicarbonates). Magnesium carbonate can also be dissolved
in another way, forming a double salt with ammonium carbonate.
Notwithstanding these diversities of detail, the general mechan-
isms remain the same whether in the case of potassium, calcium,
or magnesium nitrates.

8. Let us now refer nitrification to gaseous ammonia, and
dissolved potassium nitrate, without concerning ourselves with
the media, and calculate the heat liberated.

2NH3 gas + 402 + K2C03 dilute = 2KN03 dilute + 3H20 +

C02 dissolved. This reaction liberates 109*2, and hardly differs
from the formation of dilute nitric acid.

9. In cases where nitrification is not effected at the expense
of free nitrogen and oxygen, but at the expense of free oxygen
and of a pre-existing nitrogenous compound, such as ammonia,
the cyanides, etc., the heat liberated varies with the nature of
the said compound; but it is almost independent of the
particular nature of the dissolved alkali which takes part in
the reaction (potash, soda, lime) ; it is also the same with the
various carbonates compared with one another. This results
from an observed fact, viz. that the union of the same acid with
the various fixed alkalis liberates nearly the same quantities
of heat.

It will be seen from these data that natural nitrification once
excited and under the conditions in which it occurs, that is, in
presence of alkaline or earthy carbonates, can be effected without
the aid of any foreign energy.

10. It is effected all the easier, however, when this aid is not
wanting, seeing that the oxidation of the nitrated or non-nitrated
organic principle is developed at the same time as that of the
ammonia yielded by those principles, and liberates an additional
quantity of heat. This point deserves to be developed.

The presence of an alkali, free or carbonated, facilitates, as
has been said, the absorption of oxygen by the organic principle.
Here is another fact which may be accounted for by thermal
considerations; for the oxygen of the said principles forms
acids, the formation and the simultaneous combination of which
with the alkali liberate more heat than the pure and simple
formation of the same free acid would do. For example, the

216 ORIGIN OF THE NITRATES.

change of alcohol into potassium acetate, when in contact with
dilute potash, liberates 13 Calories more than its change into
free acetic acid.

The oxidation itself often becomes more thorough under the
influence of this additional work, which further increases the
liberation of heat. This is the case with alcohol. It is well
known how difficult it is to oxidise alcohol by free oxygen at
a low temperature and without a medium. It is necessary to
raise the alcohol, taken by itself, to a very high temperature in
order to cause it to absorb oxygen, forming at first aldehyde
and acetic acid. But it is otherwise if alcohol be placed in
presence of oxygen and of an alkali simultaneously ; then the
alcohol is gradually oxidised at the ordinary temperature, and
it forms not only acetic acid, but even oxalic acid, or rather an
oxalate. Now the transformation of alcohol into dissolved
potassium oxalate liberates a quantity of heat (288) nearly
double that produced by the transformation of alcohol into
acetate (136).

Phenomena of the same kind are very common in organic
chemistry. They certainly play a part in natural nitrification.
In the author's opinion their interpretation should be sought in
thermo-chemical considerations, seeing that chemical reactions
are the easier, cceteris paribus, the greater the amount of heat
liberated by them.

11. We shall show, lastly, how an analogous concurrence may
be brought about, under the hypothesis that the nitrates result
directly from the oxidation of nitrogenous organic principles.
It will be sufficient, to take an exact instance, to calculate
approximately the heat liberated in the nitrification of hydro-
cyanic acid, or rather of potassium cyanide, a calculation not
without interest in itself, the cyanides often existing in bricks
and other materials capable of nitrification. Take, therefore,

CNK dissolved + 50 = KN03 -h C02 gas.

The heat liberated amounts to + 177 Cal. It is nearly
double the heat liberated in the nitrification of ammonia, at
the expense of dissolved potassium carbonate. This excess is
due in a great measure to the oxidation of the carbon ; it is
probably to be met with in the oxidation of the other nitro-
genous organic substances. Gaseous hydrocyanic acid and
dilute potash would liberate + 186 Cal. in yielding an equivalent
of potassium nitrate.

Lastly, dissolved ammonium cyanide and potash absorb nine
equivalents of oxygen in being transformed into potassium
nitrate —

CNH.NHs dilute + K20 dilute + 90 = 2KN03 dilute + C02

gas + 2H20,

FIXATION OF NITROGEN IN NATURE. 217

and liberate + 2791 Cal. ; + 139'5 Cal. per equivalent of
nitrogen.

All these numbers exceed that corresponding to the oxidation
of ammonia alone (-f 109), there is therefore ground for sup-
posing that nitrification is facilitated by the simultaneous
oxidation of the carbon contained in the organic principle.

§ 3. ON THE TRANSFORMATION OF FREE NITROGEN INTO
NITROGENOUS COMPOUNDS.

First Section. — Problem of ike Fixation of Nitrogen in Nature.

1. The problem of the fixation of the nitrogen of the air
and its transformation into nitrogenous compounds, such as the
nitrates or aminoniaeal salts in the mineral kingdom, the
alkalis, amides, and albumenoid compounds in the vegetable
and animal kingdom, has long formed a subject of controversy.
A nitrogenous compound of any class being formed, it is easier
afterwards to change it into a compound of another class, and
it is precisely of this transformation that we have been treating
in the foregoing paragraphs. But there still remains the
problem of the formation of this initial compound, for nitrogen
does not combine directly with any body at the ordinary
temperature and in the absence of the conditions which will
presently be indicated. On the other hand, the natural nitro-
genous compounds tend constantly to be destroyed, under the
diverse influences of slow or rapid combustion, fermentation,
putrefaction, and even of the normal nutrition of animals,
influences which all tend to set free nitrogen. Hence it follows
that natural nitrogenous compounds being constantly destroyed
and never reproduced, the actual supply of them should con-
tinually diminish. Thus it is that the methodical researches
made on the use of manures in agriculture have not done much
more than reveal causes of destruction, without establishing
with certainty any general cause of regeneration, that is to say,
any cause sufficiently powerful to explain the reproduction of
the nitrogenous compounds. Nevertheless, vegetation is in-
definitely prolonged, and without languishing, on the same spot
of ground, whenever it is not over stimulated and rendered
exhaustive by human industry, a fact which seems to show
that there exist slowly acting causes of reproduction of nitro-
genous compounds, sufficiently efficacious to support spontaneous
vegetation. It is these causes which we are about to consider.

2. Slow oxidations. From the purely chemical point of view,
and under natural conditions, free nitrogen may be united to
oxygen in certain slow oxidations. It is beyond question, for
instance, that air kept for some time in contact with phosphorus
contains several thousandth parts of oxynitric compounds, it
being sufficient to agitate this air with lime or baryta water, and

218 ORIGIN OF THE NITRATES.

to evaporate the latter to obtain small quantities of nitrates.
Even in sudden oxidations, hydrogen, and the hydrocarbon
gases, burning in oxygen mixed with nitrogen, yield some traces
of the oxygen compounds of nitrogen.

3. Ozone. Schonbein attributed the first formation to the
action of ozone, formed by phosphorus, on free nitrogen. Ozone,
he said, oxidises nitrogen in the cold, especially in presence of
water or alkalis ; its formation in the atmosphere would account
for the natural formation of nitric acid, which would reduce the
problem of the formation of the latter to that of ozone.

But this theory has fallen in face of the experiments
separately by Carius and the author,1 experiments from which
it results that pure ozone does not oxidise nitrogen in any way.
The assertions of Schonbein, according to which the evaporation
of water in presence of nitrogen is sufficient to cause the
combination of these two bodies and the formation of ammonium
nitrate, have likewise been found erroneous, since he seems to
have neglected the pre-existence of traces of nitrates in the
waters upon which he operated.

It is none the less certain that the slow oxidation of phos-
phorus and the rapid combustion of hydrogen and the hydro-
carbon bodies develop nitrous compounds. But these are
exceptional reactions not sufficiently widespread nor efficacious
to account for the whole of the natural phenomena.

4. Function of porous bodies. The same may be said of
Longchamp's theory, according to which nitrogen is absorbed in
presence of alkalis and porous bodies. The sole experiments
which have been cited in confirmation up to the present, are
those of M. Cloez, according to which a million litres of air,
directed during a period of time amounting to six months across
pumice-stone impregnated with potassium carbonate, yielded a
few milligrammes of nitrates. This quantity is too small for
its origin to be attributed with certainty to free nitrogen. The
least trace of nitrated compounds of mineral or organic origin,
not arrested by the purifying agents (acid and alkaline) in
passing across, perhaps even a trace of neutral and volatile
compounds, would be sufficient to account for such small
quantities of nitrates. Whatever be the interest of these
observations, there is therefore no certain conclusion to be
derived from them, so long as the conditions involve the
formation of traces of nitrates only.

5. Nascent hydrogen. It has, in like manner, been supposed
that free nitrogen can be united to hydrogen, especially under
the conditions in which the latter is formed at the expense of
hydrogenated bodies. The formation of rust by the slow oxida-
tion of iron is especially cited with reference to this point. In
this formation traces of ammonia have been found. But these

1 " Annales de Chimie et de Physique," 5e se'rie, torn. xii. p. 440.

ACTION OF ELECTRICITY. 219

traces are attributed by the majority of authors to the presence
of nitric acid 1 or other nitrogenous compounds in the atmosphere.
The appearance of ammonia in the reaction of the metals (iron,
zinc, arsenic, lead, tin) upon dissolved alkaline hydrates, appears
in the same way due to the existence of a trace of cyanides or
nitrates in these alkalis.

6. Earthy substances. Mulder has asserted that during the
slow alteration of earthy substances, small quantities of ammonia
are formed. But quantitative measurements have not shown
that these quantities are capable of compensating the incessant
loss of nitrogen produced during vegetation.

7. Hence the purely chemical reactions which take place in
nature seem insufficient to explain the incessant reproduction of
the nitrogenous combinations.

Nevertheless, the latter does take place, but it results, in the
opinion of the author, from an energy foreign to purely chemical
actions.

It is electricity which causes the fixation of free nitrogen, and
principally at the ordinary temperature and at the low tensions
which electricity possesses at the surface of the earth every-
where and at all times, even during the finest weather.

Second Section. — Actions of Electricity in general.

1. Electricity can be employed under various forms to excite
chemical reactions, viz. voltaic current, electric arc, electric
spark, or silent discharge. The last-named mode of action may
itself be effected in several ways ; for instance, by suddenly
varying the potential, by the effect of rapid discharges, some-
times all in one direction, sometimes in alternate directions, or
again by maintaining the potential constant throughout the
whole duration of the experiment. Now it is certain, and this
is a fundamental fact, that all the modes of action of electricity,
with the exception perhaps of the voltaic current traversing
liquid electrolytes, bring about the chemical activity of nitrogen,
but in very different ways. Before reviewing them, let us decide
a preliminary question.

2. Does there exist a special isomeric modification of nitrogen
analogous to ozone, which is the origin of the nitrated
compounds ? This is the point which the author has set him-
self to clear up. He has observed that the activity of nitrogen
is only called into play at the moment when this element is
submitted to the action of electricity. Pure nitrogen, however,
does not undergo appreciable permanent modifications either by
the action of the arc, or by that of the spark, or of the silent
discharge. In fact, nitrogen brought into immediate contact
with hydrogen at a distance of a few centimetres, by silent
discharge tubes, or by spaces in which it undergoes the action

1 Cloez, " Comptes rendus," torn. lii. p. 527.

220

ORIGIN OF THE NITRATES.

of the arc or that of a series of strong sparks, never shows any
sign of combination. It is the same with nitrogen brought
afterwards into contact with oxygen, and also with organic
substances. In all known cases it is necessary that nitrogen
and the organic substance, or hydrogen, or oxygen should
simultaneously undergo the electric action for the combination
to take place.

3. The appliances in which the arc or spark is first caused to

act on nitrogen can be easily
imagined.

For the silent discharge the
apparatus shown in the annexed
figure is employed.

The apparatus consists of a
glass tube, c, provided with two
tubular passages, a and ~b. An-
other tube, d, penetrates into the
first tube which surrounds it, and
is ground into it at c. It is filled
with a conducting liquid (water
acidulated with sulphuric acid),
the whole being placed in a
test glass filled with the same
liquid.

The electrodes of a powerful
Euhmkorff machine communi-
cate with the liquid in the in-
ternal tube and with the external
liquid.

The silent discharge takes
place in the annular space com-
prised between the tubes c and d.
It acts upon the gases which
enter at a and escape at b. The
3piSita=_ nitrogen which issues from this
H apparatus has acquired no fresh
property.

4. The same negative results
Fig. 31.— Berthelot's silent discharge were obtained by the author with
apparatus for the modification of hydrogen in presence of organic

substances, either nitrogen or

oxygen, immediately after the hydrogen had undergone the
action of the sparks, or of the silent discharge, results which are
very different from those observed with oxygen. There does
not therefore appear to exist for nitrogen or hydrogen any
permanent electrical modification, analogous to that of oxygen
forming ozone.

ACTION OF THE ELECTRIC SPARK.

221

Third Section. — Action of the Voltaic Arc and the Electric Spark.

1. We shall now study the action of electricity under its
various forms in bringing about nitrogenous combinations, by
acting upon nitrogen in presence of the other elements.

Under the form of the voltaic arc, or the spark, electricity
produces in fact the union of nitrogen with oxygen (synthesis
of the nitric compounds), the union of nitrogen with hydrogen
(synthesis of ammonia), the union of nitrogen with acetylene
(synthesis of hydro-
cyanic acid).

2. These reactions can
easily be produced with
the following apparatus,
which does not require
either the use of plati-
num wires fused into
the glass or special con-
ductors. Bent glass
tubes and free platinum
wires suffice.

The following is the

arrangement. The gas (measured or not) is placed in an ordinary
test-tube, in a mercury trough ; then into this test-tube are intro-
duced two gas- tubes, twice bent to slightly obtuse angles (Fig. 32),
but still keeping the same direction. The tubes being open at
both ends, their introduction
is effected without difficulty
and without establishing com-
munications with the atmo-
sphere. This done, a thick
and long platinum wire is
taken, of which the length
considerably exceeds that of
the bent tube, and it is intro-
duced by the external orifice
of one of the tubes, by push-
ing it gently through the
mercury which fills the tube ;
it is thus got past the bends Fig. 33.— Action of the electric spark
until its end passes out of the on gases*

internal orifice of the tube. The same operation is performed
with a second platinum wire slipped through the second tube.

Two insulated conductors are thus obtained, which are put
into communication with the two poles of a Kuhmkorff coil,
or any other generator of high tension electricity. The spark
passes between the two points situated in the interior of the
test-tube, the distance and relative positionwJich can be

222 ORIGIN OP THE NITRATES.

regulated at will. Fig. 33 shows the tubes in place and the
experiment ready.

3. Now, if mixed dry nitrogen and oxygen, or even atmospheric
air, are subjected to the action of a series of electric sparks,
after a few minutes the test-tube is filled with nitrous vapour,
but it would need several hours to arrive at the limit of the
reaction. This is, moreover, never complete, the spark inversely
decomposing nitric peroxide (see p. 198).

4. If the operation take place in presence of a solution of
potash, the acid gases are gradually absorbed and potassium
nitrate is finally obtained. This is Cavendish's celebrated
experiment (1785).

5. The combination of nitrogen with oxygen requires the
intervention of a foreign energy represented by — 21*6 CaL,
when the union of nitrogen with oxygen takes place, forming
nitric oxide —

N + 0 = NO.

The latter compound afterwards unites with an excess of
oxygen, forming nitric peroxide.
The definitive formation,

N -f 02 = N02 gaseous,

only corresponds to an absorption of — 2 '6 Cal. at the ordinary
temperature, a quantity which increases to about — 7 Cal.
towards 200°.

6. It is precisely in virtue of analogous reactions developed
in the atmosphere during the passage of forked and sheet
lightning that nitric and nitrous acids are formed. These acids
appear in rainstorms, partly in the free state and partly as
ammonium nitrate or alkaline nitrates, the latter being derived
from the dust of the air. For example, Filhol, at Toulouse,
obtained per cubic metre of rain, T09 grms. of nitric acid.
From the analyses of M. Barral, one hectare of ground at Paris
would have received in November, 1852, from the rain, 659
grms. of nitrogen in the form of nitric acid. These quantities
are considerable, nevertheless the analysis of cultivated plants
has shown that they do not suffice to make good the losses of
nitrogen taken from the soil by vegetation.

Fourth Section. — Actions of the Silent Discharge at High Tension.

1. The combination of nitrogen and oxygen with the formation
of nitrous compounds is not only produced by the electric
spark, but also by the action of the silent discharge, when the
electric tension is very great (see the instruments, pp. 226
and 230).

2. This is, again, a condition which occurs in the atmosphere.
During the interval of time which precedes the instant when
the discharges of lightning, properly so called, trace a certain

NITROGEN AND WATER. 223

line in the atmosphere, there are very widespread surfaces
which gradually become electrified by influence, then suddenly
discharge themselves at the moment of the explosions (return
shock). Over these electrified surfaces there are exerted certain
chemical reactions analogous to those developed by the silent
discharge at a high tension and with a suddenly varying
potential. These are, moreover, accidental, local, and momentary
effects, as well as those of lightning properly so called. It is
probable that they are especially produced on mountains and
isolated peaks.

3. The electric influence thus causes the formation of hypo-
nitric and nitric acids, and even that of pernitric acid,1 an un-
stable compound produced by the reaction of the silent discharge
at a high tension on a mixture of hyponitric acid and oxygen.

4. Nitrogen and water. Under the influence of high electric
tensions, free nitrogen and water combine to form ammonium
nitrite, according to the author's experiments 2 —

N2 + 2H20 = NH4N02,

the energy necessary for this reaction (— 73 '2 Cal.) being
supplied by electricity.

5. The effects just described are produced under the influence
of external discharges of the Euhmkorff coil, the potential of
the electrified bodies thus passing in a very short interval of
time through all values, from zero to a limit amounting to
several thousand volts.

6. The same effects also take place, each pole being alternately
charged with positive and negative electricity, as with the
Euhmkorff coil, or each pole being constantly charged with the
same electricity, as may be obtained by the Holtz machine.

7. But these reactions gradually become weakened if the
potential be lowered, and finally cease entirely, when it fails
below a certain limit, relatively very high, that is, reaching to
several hundred volts. Below this limit nitrogen and oxygen
cease to combine, although ozone is still formed.

8. It should be noted that this limit of potential is far higher
than the ordinary tensions which atmospheric electricity can
assume, except in stormy weather. The direct formation of the
oxygenated compounds of nitrogen in nature is, therefore, limited
to the conditions of very great electric tension and the influence
of storms.

9. We will examine, from the same point of view, the com-
bination of nitrogen with hydrogen ; that is to say, the formation
of ammonia by the action of electricity.

1 "Annals de Chimie et de Physique," 5" se"rie, torn. xxii. p. 432. The
author had noticed the formation of the last combination ; but it has been
demonstrated in a more complete manner and studied more particularly by
Chappuis and Hautefeuille.

2 Same collection, 5' se*rie, torn. xii. p. 455.

224

ORIGIN OF THE NITRATES.

Take, first, the action of the spark. It is well known that
ammonia is decomposed by a series of sparks into its elements,
the volume of the gas being practically doubled after a rather
short period of time. Nevertheless, there remains a trace of
ammonia, not capable of measurement, though capable of being
manifested, as will be presently shown. Now, nitrogen and
hydrogen undergo reciprocally a commencement of combination,
by the action of a series of electric sparks. However, the pro-
portion of ammonia formed is so slight as not to be shown by a
change in volume. But it is sufficient to introduce into the
gases a bubble of hydrochloric acid gas to produce abundant
fumes. (In order that the experiment may be reliable, it is
necessary to operate with gases thoroughly dried before the
experiment and over dry mercury, the least trace of water
vapour being indicated in the same way by hydrochloric acid
gas.) This reaction is so delicate that it reveals the thousandth
part of a mgrm. in a small volume of gas.

To accumulate the effects of this reaction, it is sufficient to
operate in presence of dilute sulphuric acid, so as to gradually
absorb the ammonia. It is then easy to collect a considerable
quantity of it at the end of a sufficient time. The author has
not been able to discover the inventor of this experiment, but
it appears as already classic in the first edition
of Kegnault's "Traite de Chimie," printed in
1846, and dates from still further back.

10. The action of the silent discharge is far
more efficacious than that of the spark in
causing the union of nitrogen with hydrogen.

The silent discharge has also the double
property of decomposing ammonia into its
elements and of combining elementary nitrogen
and hydrogen. These two gases being mixed
in the ratio of three volumes of hydrogen to
one volume of nitrogen, if the silent discharge
be made to act upon the mixture, after a few
hours as much as three per cent, of the mixture
will be found to have been transformed into
ammonia. The latter may then be measured
by volume, and manifested by all its reactions.

11. The apparatus which was most commonly
employed for making the silent discharge act upon
the gases is formed of two distinct glass tubes —

(1) A very thin stoppered tube, enlarged at the lower part,
and forming a test-tube, so arranged as to permit of the intro-
duction, the extraction, and the rigorously exact measurement
of the gases over mercury, all as clearly and easily as with
ordinary gas test-tubes.

This tube is surrounded by a thin strip of platinum, arranged

Fig. 34. — Silent
discharge test-
tube.

APPARATUS FOR SILENT DISCHARGE.

225

spirally on its external surface (Fig. 34), this strip being fixed
with gum. The whole glass surface in contact with the
atmosphere is carefully coated with shellac, in order to insulate
it more fully.

(2) A V tube (Fig. 35), slightly less in diameter than the
test-tube, so arranged as to be able to be introduced into it,
almost without friction.

This tube is closed at one of its ends (Fig. 35), and filled with
dilute sulphuric acid.

The test-tube being placed over a large mercury trough, the
gases on which it is desired to operate are introduced into it
after having been measured in a graduated test-tube with the
usual precautions. The volume is regulated according to the
capacity of the test-tube, diminished by that of the vertical
portion of the V tube. It is also necessary to take account of
the increase of volume pro-
duced by decomposition, if
there be occasion to do so.
The closed part of the V tube
is then introduced into the
interior of the test-tube, first
having been filled with water
acidulated with sulphuric
acid.

Then, the test-tube being
held in the left hand, a small
porcelain basin, like those
usually employed for measur-
ing nitrogen in organic com-
pounds, is introduced by the
right hand under the mercury,
and passed under the test-
tube, held vertically, when Fig- 35>
the whole is taken away, so as to isolate the test-tube arranged
over the basin, as in Fig. 36.

It is held in place with the aid of the wooden jaw of a
Gay-Lussac support, which, for the sake of simplicity, has not
been shown. This support, at the same time, applied against
the platinum strip in Fig. 36, keeps in place a thin sheet of
platinum, fixed at the end of a wire communicating with one
of the poles of a very large Euhmkorff coil, whilst the other
pole is attached to a second wire which dips into the acidulated
water of the V tube.

12. The combination of nitrogen with hydrogen, as well as
that of oxygen and nitrogen, ceases below a certain potential of
the electric apparatus, which produces the silent discharge. It
does not take place at all at the low tensions.

13. The combination of free nitrogen with the hydrocarbon

Q

226

ORIGIN OF THE NITRATES.

compounds is of great importance. Before the author's experi-
ments it was entirely unknown. It is a remarkable circum-
stance that this combination takes place equally well with the
highest and even the lowest electric tensions, contrary to what
happens in the case of oxygen and hydrogen. The products,
moreover, vary according to the greatness of the electric tensions.
14. Hydrocyanic acid. In allowing the voltaic arc or the
electric spark to act directly upon gases, the author has observed
that acetylene and nitrogen combine directly at equal gaseous
volumes, forming hydrocyanic acid. The same reaction takes
place with every hydrocarbon gas or vapour capable of forming
acetylene under the influence of the spark. This formation of

Fig. 36. — Action of the silent discharge on mercury.

hydrocyanic acid constitutes the best defined positive character
of nitrogen and is the easiest to show.

If a series of strong sparks be passed into a mixture formed
by the two pure gases, the gases assume almost immediately
the characteristic odour of hydrocyanic acid.

After a quarter of an hour, or even less, if the sparks are
long and strong the reaction is already well advanced. It is
then sufficient to agitate the gas with potash to change the acid
into alkaline cyanide and to manifest the reactions which are
characteristic of it (Prussian blue, etc.)

Under the circumstances just described the formation of
hydrocyanic acid is accompanied by that of carbon and hydro-
gen, formed in virtue of a distinct but simultaneous decomposi-

NITKOGEN AND HYDROCARBONS. 227

tion of the acetylene. But this complication may easily be
avoided by adding beforehand to the mixture a suitable volume
of hydrogen, for instance, ten times the volume of the acetylene ;
no further deposit of carbon is then observed, and the reaction
absolutely corresponds to the following equation : —

C2H2 + N2 = 2CNH.

The presence of the hydrocyanic acid formed is not, however,
completely accomplished under the conditions just described,
and the reaction ceases at a certain limit, because the hydro-
cyanic acid is inversely decomposed by the spark, into nitrogen
and acetylene. But if the hydrocyanic acid be gradually
removed by potash, care being taken to dry the gases each time,
before renewing the action of the spark, a given volume of
nitrogen may be completely transformed into the acid, as has
been expressly verified. Hydrocyanic acid is formed solely by
the action of the spark or arc, and not of the silent discharge.

15. Nitrogen and organic compounds. Nevertheless nitrogen
is also absorbed by organic matters, when operating with the
silent discharge by means of a powerful Euhmkorff coil and
the test-tube just described. It is easy to observe (at an
ordinary temperature) the absorption of a measurable volume of
nitrogen either by hydrocarbons (benzene, essence of turpentine,
etc.), or by ternary substances, such as ether, moist dextrine, or
paper;

16. Nitrogen and hydrocarbons. The experiment is very well
defined with benzene, a compound devoid of oxygen, 1 grm.
of benzene absorbing in a few hours 4 to 5 cub. cms. of
nitrogen, the greater part remaining unaltered. The reaction
is effected principally between electrified benzene, in vapour, or
under the form of very thin liquid layers, and nitrogen gas. It
gives rise to a polymeric and condensed compound, a sort of
solid resin, which collects on the surface of the glass tubes
through which the discharge is effected. This compound, when
highly heated, is decomposed, with liberation of ammonia. But
free ammonia does not pre-exist, nor is it formed by the silent
discharge, either in the dissolved state in the excess of benzene,
or in the gases. The latter, moreover, contain a little acetylene,
which appears constantly in the reaction of the silent discharge
on the hydrocarbons. Essence of turpentine also gave rise to an
absorption of nitrogen, in reality slower under the same con-
ditions. There was also produced a condensed resinous body,
which liberates ammonia on ignition.

The vapour of ether also x absorbs nitrogen. Methane behaves
in the same manner. It yields at once (in a small quantity) a
very condensed solid nitrogenous product, which liberates
ammonia, by heat, and free ammonia, which remains mixed
with the non-condensed gases. With acetylene, the principal

Q2

228 ORIGIN OF THE NITRATES.

product is a polymeric substance, discovered by Thenard.
Nitrogen and acetylene, moreover, do not form hydrocyanic
acid under the influence of the silent discharge, a result which
contrasts with the abundant formation of this compound under
the influence of the spark. However, the condensed product
formed by acetylene modified in presence of nitrogen, when
subsequently destroyed by heat, liberates towards the close
some traces of ammonia.

17. Nitrogen and carbohydrates. The following are various
experiments relative to the absorption of nitrogen by the action
of the silent discharge at high tension, which are calculated to
show that this absorption really takes place, when operating
with the principal constituents of vegetable tissues, either with
pure nitrogen or in presence of oxygen, that is, by bringing
atmospheric air into action.

White filter paper (cellulose or ligneous principle) slightly
moistened and submitted to the influence of the silent discharge,
in presence of pure nitrogen absorbs a very marked quantity of
it in the space of eight to ten hours. It is sufficient to heat the
paper strongly afterwards with soda-lime, to liberate from it a
great quantity of ammonia. The original paper did not appre-
ciably yield any under the same conditions. Ammonia, besides,
is only produced towards a dull red heat by the destruction of
a particular and fixed nitrogenous compound, as with the
hydrocarbons.

18. The presence of oxygen does not prevent this absorption
of nitrogen. The following experiment shows this. The glass
tubes through which the electric influence is exerted having
been covered with a thin coat of a syrup-like solution of
dextrine (a few decigrammes in all), a certain volume of
atmospheric air was introduced into them over mercury.

After having made the silent discharge act for about eight
hours, an absorption of 2'9 per cent, of nitrogen and 7'0 of
oxygen in 100 volumes of the original air was observed. It
will be seen that the absorption of the oxygen was not total
under these conditions. As a check the organic matter remain-
ing on the surface of the tubes was collected and heated with
soda-lime, it liberated ammonia in great abundance and only
towards a dull red heat, which completes the demonstration.
For the rest, it was not found that free ammonia, nitric or
nitrous acids were formed in any appreciable amount, at least
under these conditions.

19. The principal phenomenon is therefore the production of
a complex nitrogenous compound by the direct union of free
nitrogen with the carbohydrate experimented upon, a reaction
perfectly comparable to those which must be produced in
nature, by the contact of vegetable matter with the electrified
atmospheric air.

NITEOGEN AND CELLULOSE. 229

20. The absorption of nitrogen by organic compounds takes
place likewise under the influence of loth kinds of electricity.
It takes place in just as well defined a manner with the lowest
as with the highest tensions, but in a time which is the longer,
the lower is the electric tension. It is very marked even
with the low tensions which no longer yield the oxides of
nitrogen. This absorption has been verified, both by insulating
the silver or platinum1 armatures held in contact with the
paper and the gases, and also by insulating the paper itself
from all metallic contact between two glass surfaces. At the
same time as the fixed nitric compounds already referred to, and
under these conditions, no trace of ammonia was formedt and
no trace of nitric or nitrous acid, or of hydrocyanic acid.

21. Working under similar conditions, and with very low
tensions, it was found that the fixation of the nitrogen was
especially abundant with paper, less with ether, and still less
with benzene, a diversity corresponding to the unequal stability
of these principles and to the different nature of the nitrogenous
principles derived from them. With paper especially, there
are produced at the same time insoluble nitrogenous com-
pounds, very slightly coloured, which remain fixed upon the
woody fibre, and nitrogenous bodies which are soluble in water
and almost colourless, which are condensed upon the sheet of
platinum ; the latter contain such large quantities of nitrogen
that they yield free ammonia which turns litmus paper blue,
even without any addition of soda-lime.

22. The experiments just described define the general con-
ditions of the chemical reactions produced by the silent dis-
charge, but they do not indicate clearly the effects of the
electrical tension, free from all complications. In fact, in the
experiments made with the help of the Kuhmkorff apparatus, or
the Holtz machine, the tension changes continually during the
interval between the outer sparks, and this between limits that
vary by several thousand volts.

What is the influence of these .incessant variations and the
sudden alternations accompanying them? Are the chemical
reactions determined by the very fact of these alternations and
the molecular shocks and vibrations resulting from them, or can
the chemical reactions be produced by a simple difference of
potential, or a simple determination of the gaseous molecules,

1 The metallic armatures had been brought to a red heat in the open air
before each experiment in order to destroy every trace of organic matter on
their surfaces. Care must be taken not to touch them with the fingers. The
Swedish paper and the dextrine employed did not contain more than a ten-
thousandth part of nitrogen according to a special analysis, a proportion which
is of no account when a few centigrammes of paper are operated upon. This
verification must be made each time upon strips taken from the same sheet of
paper and in an alternate manner, the paper sometimes accidentally contain-
ing nitrogenous substances.

230

ORIGIN OF THE NITRATES.

without there being either any voltaic current properly so called,
as with a closed battery, or elevation of temperature, as with
the spark, or sudden and incessant variations of tension, as with
the silent discharge developed under influence of the Holtz or
Kuhmkorff machines ? The following experiments were made
in order to solve these questions.

Fifth Section. — Action of Electricity at very Low Tension.
1. These fresh trials were made with a battery, without closing

Fig. 37. — Apparatus open.

Fig. 38.— Apparatus arranged
for the experiment.

the circuit, and under such conditions that the entire experi-
ment resolved itself into the establishment of a constant
difference of potential between the two armatures. This
difference was measured by the electro-motive force of five
Leclanche cells (a force equivalent to about seven Daniell cells)
in the greater number of the experiments about to be described.
Each experiment lasted from eight to nine consecutive months.
2. Metallic armatures had to be given up on account of the
special reactions they bring about, and it was necessary to

SLOW FIXATION OF NITKOGEN.

231

place the gases in the annular space separating two concentric
glass tubes fused together at the top.

The apparatus is shown on the preceding page. The inner
tube is open and filled with dilute sulphuric acid ; the outer one
is closed at the blowpipe, and plunged into a test-glass contain-
ing the same acid. The gases and other bodies are introduced
beforehand into the annular space, by means of small tubes,
which are then closed at the blowpipe.1 The positive pole of
the battery is put in communication with the acid liquid of the
inner tube, which acts as armature ; and the negative pole with
the acid liquid of the test-tube, which acts as a second arma-
ture, separated from the first by a dielectric formed of two
thicknesses of glass and the gaseous stratum between. The
gases are thus contained in a space completely closed by fusion
of the glass without any metallic contact.

3. The following results were observed under these conditions :
the formation of ozone, into which it is not necessary to enter
here; the absorption of the free nitrogen by the paper and
by the dextrine ; and the formation of special nitrogenous com-
pounds, exactly as in the experiments on p. 229.

4. Some of the experiments were made under quantitative
conditions, so as to measure the

weight of nitrogen absorbed in a
given time. For this purpose over
half the outer surface of a large
cylinder of thin glass, A, termi-
nated by a spherical cap, a sheet
of Swedish paper, weighed before-
hand and damped with pure water,
was laid. The other half of the
same outer surface was coated with
a syrupy solution of dextrine tested
and weighed under conditions that
enabled us to know exactly the
weight of dry dextrine employed.
The inner surface of the cylinder
had been covered beforehand with
a sheet of tinfoil (internal armature). This cylinder was placed
upon a glass plate, and then covered over with a concentric
cylinder of thin glass, B, as closely as possible, the inner surface
of this cylinder being left uncovered, and the outer surface
covered with a sheet of tinfoil (external armature).

The system of two cylinders was covered with a bell-glass, C,
to keep out dust, and placed upon a glass plate, arranged so as
to keep the apparatus airtight.

The internal armature was put in communication with the
positive pole of a battery formed of five Leclanche cells, arranged
1 " Annales de Chimie et de Physique," 5e se>ie, torn. xii. p. 463.

Fig. 39.— Slow fixation of the
nitrogen.

232 ORIGIN OF THE NITRATES.

in series, the external armature with the negative pole of the
same battery. In this way there was a constant difference of
potential between the two armatures of tinfoil, separated by
the two thicknesses of glass, by the stratum of air between, and
lastly by the paper or dextrine applied to one of the cylinders.
Before the experiment the nitrogen was estimated in the paper
and in the dextrine (working upon two grammes of dry material),
and was found to be, in 1000 parts —

Paper -10, dextrine 12.

At the end of a month (November), having worked at first with
a single Leclanche element,

Paper 10, dextrine 17,

mould had formed. There being no variation in the paper and
very little in the dextrine, the experiment was continued with
five Leclanche cells for seven months, the outside temperature
being raised little by little until at times it reached 30°. Again
mould was observed. At the end of this period, in 1000 parts,
the nitrogen was found to be —

Paper *45, dextrine 1*92.

The space between the two cylinders was from three to four
millimetres. Another trial, made at the same time, with nearly
treble the space between two other concentric cylinders, similar
to the first, gave, in nitrogen in 1000 parts —

Paper '30, dextrine 114.

All these analyses go to establish the fact that there is a fixation
of nitrogen upon paper and upon dextrine, i.e. upon vegetable
substances that are not directly nitrogenous, under the influence
of excessively low electrical tensions.

The effects are here provoked by the difference of potential
existing between the two poles of a battery formed of five
Leclanche cells, a difference that may be compared to atmospheric
electricity acting at short distances from the earth.

5. The influence of the mould, observed in the course of the
experiments, cannot be taken into account, for Boussingault has
proved, by very careful analysis,1 that this vegetable substance
does not possess the power of fixing atmospheric nitrogen.

6. The influence of light did not enter into the above ex-
periments, in which the fixation of the nitrogen was effected in
total darkness. Other experiments, however, performed in the
light, showed that light does not impede the electrical fixing of
the nitrogen.

7. The reactions just described are determined by very low
electrical tensions, the value of which is quite comparable to
those of atmospheric electricity, as is shown by the measure-

1 " Annales de Chimie et de Physique," 3e se*rie, torn. Ixi. p. 363.

NITKOGEN FIXED BY ATMOSPHERIC ELECTRICITY. 233

ments published by Thoinsen, Mascart, and various other
experimentalists.

8. In order to complete this demonstration, it was thought
expedient to operate upon atmospheric electricity itself. For this
purpose, the author worked by means of the difference of
potential existing between the earth and a stratum of air about
two metres above it in the garden of the observatory at
Montsouris.

The results obtained, during experiments which lasted from
July 29 to October 5, 1876, i.e. rather more than two months,
will now be given, the mean electrical tension having been
about that of three and a half Daniell cells, and having
fluctuated in absolute value from + 60 Daniell to about — 180
Daniell, in the apparatus.

In all the tubes, without exception, whether they contained
pure nitrogen or ordinary air, whether they were hermetically
sealed or in free communication with the atmosphere, the
nitrogen fixed itself upon the organic substance (paper or
dextrine), forming an amide compound, which was decomposed
by soda-lime at about 300° to 400°, with regeneration of
ammonia.

The same substances, left freely exposed to the atmosphere
of a room apart from the laboratory, did not give the least sign
of the fixation of nitrogen.

The quantity of nitrogen thus fixed under the influence of
atmospheric electricity is, moreover, very small in each tube.
This may be explained by the smallness of the weight of
organic matter (a few centigrammes), by the slowness of the
reactions, and lastly by the limited extent of the surfaces
influenced.1 As, however, the number of tubes capable of
being arranged in the same circuit might certainly be very
much increased, without affecting the electrical effects any more
than the chemical effects derived from them, we see that the
quantity of nitrogen capable of being deposited on a surface
covered with organic matter at the end of a suitable time may
be rendered considerable without any other depositing influence
being brought to bear upon it than the natural difference of
potential between the earth and the strata of air two metres
above it. We thus find ourselves in conditions similar to those
of vegetation increased in the relation existing between the
distance from the outflow tube in the Thomsen apparatus to
the earth and the distance between the two armatures of the
author's tubes.

1 No trace of nitric acid was found either in the water which had been in
contact with the organic substances, or in special tubes containing only air
and water and subjected simultaneously to atmospheric electricity. The
silent discharge under these conditions of feeble tension does not, therefore,
seem to determine the union of the nitrogen with oxygen, so as to form nitric
acid.

234 ORIGIN OF THE NITRATES.

9. Two of the experiments enable the demonstration to be
carried even further. In fact, the damp paper contained in two
tubes (nitrogen with an armature of silver in the inner tube, air
with an armature of platinum in the annular space) was found
to be covered with greenish stains, formed of microscopic algae,
with fine filaments interlaced and covered with fructifications.
They derived their origin, no doubt, from some germs introduced
accidentally before the closing of the tubes. Now, in these two
tubes there was much more nitrogen fixed than in tubes deprived
of vegetable matter. In the nitrogen tube especially, the gases
emitted a sourish and slightly foetid odour, similar to that of
certain fermentations, and the deposition of nitrogen was much
greater than in any of the others.

10. From these facts it follows that the deposition of nitrogen
in nature, which is indispensable for the formation of nitrates,
and also for the development of vegetable life, may take place
directly and under normal atmospheric conditions, without
necessarily being correlative either with the formation of ozone
or with the previous production of ammonia or nitrous com-
pounds ; this last-named production only taking place with the
help of stormy and exceptional tensions.

We know, however, that working in a closed space, Boussing-
ault, whose ability is well known, did not succeed in proving
the absorption of free nitrogen. But atmospheric electricity at
a low tension did not act in these experiments in vitro, in which
the potential is the same at all the internal points of the
apparatus, and its intervention is apparently of a nature to
modify the conclusions of this eminent authority.

11. The result of the author's experiments is to show clearly
the influence of a new natural cause, an influence of great
importance to vegetation. Up to the present, whenever the
question of atmospheric electricity has been studied from an
agricultural point of view, only its luminous and violent
manifestations have been considered, such as thunder and
lightning. Even the action in nature of those high tensions
which determine the formation of nitrous compounds by
influence had scarcely been taken into consideration before the
author's experiments (p. 215).

In all cases, only the formation of nitric and nitrous acids
and of ammonium nitrate was studied. The author considers
that up to the present there has been no other suggestion made
with regard to the influence of atmospheric electricity being
capable of constituting the distant and indirect source of the
fixing of nitrogen on vegetable substances. Before the experi-
ments just described, there was no idea of the direct reactions
that can take place between vegetable matter and atmospheric
nitrogen under the influence of feeble electrical tensions.
The starting into activity of the nitrogen under these feeble

LAWES AND GILBERT'S EXPERIMENTS. 235

tensions is, however, of great interest, and it is these feeble
tensions that seem to be the most efficacious, the slightness of
the effects being compensated by their duration and by the vast
extent of the surfaces influenced. We have to do with quite
a new kind of action, until now completely unknown, which is
working incessantly under the most unclouded sky, to deter-
mine a direct fixing of nitrogen upon vegetable tissues. In
studying the natural causes capable of acting upon the fertility
of the soil, and upon vegetation, causes which it has been sought
to define by meteorological observations, we must for the future
take into consideration not merely luminous or calorific
influences, but also the electrical condition of the atmosphere.

12. We will now specify more particularly the character of
these reactions in nature. When studied at a given spot, and
over a small surface, they can certainly be only very limited,
otherwise the humic substances in the soil would rapidly
become rich in nitrogen ; whereas the regeneration of naturally
nitrogenous substances, when exhausted by cultivation, is, on
the contrary, as we know, excessively slow.

But this regeneration is indisputable, for in no other way
can we account for the unlimited fertility of soils that receive
no manure, such as the meadows on high mountains, as studied
by Truchot, in Auvergne.1

It will be remembered that Messrs. Lawes and Gilbert, in
their celebrated agricultural experiments at Eothampstead, came
to the conclusion that the nitrogen in certain crops of leguminous
plants exceeds the sum of the nitrogen contained in the seed,
the soil, and in the manure, even adding the nitrogen supplied
by the atmosphere under the known form of nitrates and
ammoniacal salts ; a result which is all the more remarkable,
seeing that a portion of the nitrogen combined is eliminated in
a free state during the natural transformations of vegetable
products. We observe, therefore, only the difference between
these two effects, i.e. that the actual fixing of nitrogen is much
greater than the apparent. In most cases it is concealed by the
causes of loss. The above-mentioned writers concluded from
their observations that there must exist in vegetation some
source of nitrogen sufficient to account for the great mass of
combined nitrogen in existence on the surface of the globe.
But the source of this was until now quite unknown. Now, it
is precisely this hitherto unknown source of nitrogen that would
seem to be established in the author's experiments on the
chemical reactions provoked by electricity at low tensions, and
especially atmospheric electricity.

13. To complete this explanation, we will compare the
quantitative data of the experiments with the richness in
nitrogen of the vegetable tissues and organs that are renewed

1 " Annales agronomiques," torn. i. pp. 549 and 550. 1875.

236 ORIGIN OF THE NITRATES.

each year. The leaves of trees contain about *008 of nitrogen,
wheat straw about '003.

Now, the nitrogen fixed upon the dextrine, in the experiments,
at the end of eight months, amounted to about '002 (p. 232), i.e.
a nitrogenous substance was formed of a richness almost com-
parable to that of herbaceous tissues, produced in vegetation in
the same space of time, with the help of the influences exercised
by natural electrical tensions, which may be compared to those
of the foregoing experiments.

14. This new cause of the fixing of the atmospheric nitrogen
in nature is of the highest importance. It engenders condensed
nitrogenous products of the humic order, so widely diffused over
the surface of the glode. However limited the effects may be
at each moment and at each point of the terrestrial superficies,
they may, however, become very considerable, on account of
the extent and continuity of a reaction working universally and
perpetually.

( 237 )

CHAPTER VI.

THE HEAT OF FORMATION OF HYDROGENATED COMPOUNDS OF

NITROGEN.

§ 1. HEAT OF FORMATION OF AMMONIA.

1. THE heat of formation of ammonia, of nitric oxide, of water,
of carbonic acid, and of hydrochloric acid, constitute, perhaps,
the most important data of thermo-chemistry. The three last
have been, for the last forty years, the subject of numerous
direct measurements on the part of the most skilled experi-
mentalists ; they may therefore be looked upon as known within
one or two per cent, of their absolute value. In the foregoing
chapter the heat of formation of nitric oxide has been given, and
we may now proceed to study that of ammonia. Before the
author's last researches it was only known in a somewhat
unsatisfactory manner; two measurements only had been
taken of it, and these by an indirect process without control.

2. It is by making chlorine act upon diluted ammonia, and
then weighing the chlorine absorbed, that Favre and Silbermann,
and afterwards Thomsen, endeavoured to estimate the heat of
formation of ammonia. They assumed that the reaction worked
upon the whole of the chlorine according to the following
formula, which is admitted in the elementary treatises, but in
none of these works is the quantitative realisation of this
equation verified by the calorimeter —

4NH3 dilute + 301 gas = N gas + 3NH4C1 dilute.

Favre and Silbermann obtained results which, for fourteen
grammes of nitrogen, gave —

N + H3 = NH3 gas + 2273 Cal.
U + H3 + water = NH3 dissolved + 3147.

Thomsen, having repeated the same experiment, obtained
different results—

238 HYDROGENATED COMPOUNDS OP NITROGEN.

N + H3 = NH3 gas, -|- 26-71 Cal.
]ST -f H3 + water = NH3 dissolved, + 3515 Cal.

The difference is considerable, amounting to 4 Cal., or nearly
20 per cent. Thomsen tried to reconcile these figures by re-
calculating the figures of Favre and Silbermann, according to
his own data regarding the heat of formation of hydrochloric
acid and ammonium chloride. But corrections of this kind are
very problematical,1 seeing that the figures of the above-men-
tioned writers form a complete whole : the cause of the divergence
is apparently quite a different one.

3. In fact, some years ago, the author began to doubt the
accuracy of all these figures, in the course of his studies of the
heat of formation of the oxygen acids of the halogen elements.2
Having measured that of the hypobromites, he thought it might
serve to determine that of urea, in accordance with the process
of analysis generally followed for that substance. But it was
desirable first to verify the reaction of the hypobromites upon
ammonia itself, and it was then found that extraordinary losses
of heat took place, quite irreconcilable with those that could be
calculated from the data that have been accepted with regard to
ammonia. The experiments were made, starting with pure
liquid bromine of a determined weight. It was dissolved in a
weak solution of soda, and the heat liberated was measured ;
then weak ammonia was also added in considerable excess, and
the second escape of heat was measured. The total result must
represent the transformation of the bromine, ammonia, and soda
into sodium bromide, water, and nitrogen —

6Br + 2NH3 dilute + 3^0 dilute = 6NaBr + 3H20 + N2.

This is the thermal result observed, as obtained from the effect
of the two operations, performed one after the other —

£(6Br acting on 3Na20) dilute ......... +18-0

NH3 dilute acting on the hypobromite ... ... + 88'8

Total ... + 106-8

1 It would at least be as reasonable to correct Favre and Silberrnann's
results by the following considerations. Their data were almost all obtained
with the mercury calorimeter; now the unit employed by them in this
instrument was apparently too high by about one-tenth, according to the
error that they committed in the estimation of the heat of neutralisation of
nitric acid, hydrochloric acid, etc. All the quantities that enter into the calcu-
lation of the heat of formation of ammonia, and consequently this heat of
formation itself, should therefore be reduced in the same proportion.

2 "Annales de Chimie et de Physique," 5' se*rie, torn. v. p. 333, hypo-
chlorites; torn. x. p. 377, chlorates; torn. xiii. pp. 18 and 19, bromates et
hypobromites ; p. 20, iodates. See Book II. chap. XII. of the present work.

ACTION OF CHLORINE ON AMMONIA. 239

If we admit the preceding reaction, we shall take —

As the initial condition J(6Br + 6H + 2N + 3Na20 dilute)
Final condition ... |(6NaBr dissolved + 3H20 + N2)

FIRST CYCLE.

£[6(H + Br) + water = 6HBr dilute] + 88- 5 (B)

J[6HBr dilute + 3Na20 dilute] + 41-1 (B)

SECOND CYCLE.
N + H3 + water = NH3 dilute x

Successive reactions of the bromine upon the soda and of the
hypobromite upon the ammonia, -f 106'8, whence we get
x = + 22-8 in place of +3515 or 31'5. The same experiment,
repeated with potash and with baryta, gave similar results. It
was proved, moreover, by collecting over mercury the nitrogen
set free, that the reaction differs little from the above equation ;
in fact, the volume of nitrogen given off amounted to about
nine- tenths of the theoretical value, a secondary phenomenon1
having abstracted from the fundamental transformation a por-
tion of the bromine employed.

Whatever hypothesis may be formed as to the missing tenth,
we cannot explain the difference between 3515 and 22'8.

In other words, these experiments, which are very simple and
easily executed with the calorimeter, gave 12*35 Cal. more than
were indicated by the received numbers ; an excess which is too
great to be explained by any error in the experiments. How-
ever, even the heat of formation of the ammonia does not come
out with sufficient accuracy in these trials ; fearing, therefore,
some mistake in such an important question, the further study
of this subject was postponed. It was, however, recently
resumed, with the following results.

4. It was first attempted to determine whether chlorine, in
the presence of dilute ammonia, really decomposes it without
heat, with the immediate liberation of a quantity of nitrogen
equal to the chlorine employed. The experiment is easily
made. We require merely to pass a known volume of chlorine
(displaced in a gasometer by a flow of concentrated sulphuric
acid) through diluted ammonia, taken at the surrounding-
temperature and enclosed in a small receiver, so as to collect
the gases given off. It was found in two experiments made
with an excess of ammonia (which is necessary in order to
avoid the formation of nitrogen chloride) —

Chlorine 140 cc., nitrogen 20'5 cc., instead of 467 cc.
„ 243 „ „ 32 „ „ 81 „

1 The formation of a small quantity of bromate ?

240 HYDROGENATED COMPOUNDS OF NITROGEN.

Moreover, these figures vary considerably, according to the
conditions of the experiments, as might be expected. It would
be easy to reduce them still further, and perhaps even to annul
them altogether, by taking precautions to diminish the elevation
of temperature developed upon the first contact of the chlorine
with the ammonia, a diminution which was not attempted by
any special contrivance. As they are, these numbers are in
relation to the same conditions in the calorimetric measure-
ments, and they are sufficient to establish the incomplete
character of the reaction.

The liquids thus subjected to the action of chlorine contain
ammonium hypochlorite, a compound previously mentioned by
Balard and by Soubeyran, who had prepared it, the one with
hypochlorous acid,, the other with chloride of lime. The
presence of hypochlorous acid may, in fact, be manifested in
it. Perhaps there are also some chloro-substitution bases, inter-
mediate between nitrogen chloride and ammonia.

The above liquids are in an unstable condition; they are
continually giving off nitrogen. We have merely to pour
them off into another vessel or stir them with a rod in order
to make them pass into the gaseous form. They are well
adapted to the repetition of Gernez's elegant experiments.
Even after a day or two, the slow liberation of the nitrogen
continues.

The author tried whether he could obtain at one stroke the
nitrogen in solution, by adding to the liquid an excess of
hydrochloric acid. The liquid, which had at first furnished
32 cms. of nitrogen, gave off upon this second operation
38*6 cms. ; in all, 70'6 cms. instead of 81 cms. This last deficit
results either from the solution of a small quantity of nitrogen,
owing to the great volume of the final liquid, or to some
quantity of chlorine being employed in a secondary reaction,
such as the formation of a little chlorate or perchlorate. How-
ever this may be, the facts above mentioned show the causes of
the errors of the first experimentalists. The action of chlorine
upon ammonia could not, at any rate under the conditions
with which they worked, be employed for measuring the heat
of formation of this substance.

The action of the hypobromites would seem to be preferable,
judging from the measurement of the volume of nitrogen
liberated. This reaction, however, was not wholly satisfactory.
The object in view was arrived at by quite another method,
which is very simple and apparently faultless, as regards the
completeness of the reaction, the direct combustion of the
ammoniacal gas was effected by means of free oxygen.

5. Combustion of ammonia. The combustion of ammoniacal
gas in free oxygen is effected with the same facility as that of
hydrogen. It may easily be performed in the glass combustion

COMBUSTION OF AMMONIA.

241

vessel described elsewhere,1 and which has already been used
by M. Ogier and the author for burning pure carbonic oxide,
acetylene, olefiant gas, benzene, cyanogen, phosphuretted,
arseniuretted and silicated hydrogen, for forming hydrochloric
acid gas, etc. It is shown in the subjoined figure.

This reaction, when effected satisfactorily, produces only
nitrogen and water, in accordance with the equation

2NH + 0

3H20.

The greater part of the water is condensed in the combustion
tube, and the surplus upon the solid potash in two consecutive
U-shaped tubes. This surplus re-
presents a very small proportion
of the water formed, a proportion
corresponding to the normal
saturation with the vapour of
water of the gases set free. Its
gaseous form has been taken into
consideration in the calculations.
The weight of the water is fur-
nished by the variation in the
weight of the vessel (filled with
pure oxygen) and of the U-shaped
tubes. From this we deduct the
weight of the ammonia consumed,
27 grms. of water being furnished
by 17 grms. of ammonia.

The combustion should take
place all at once and without
relighting, an operation which
necessitates the opening of the
vessel and involves losses of
watery vapour. If the condensed
water shows any signs of the
presence of the oxygen compounds
of nitrogen, the quantity does
not exceed some ten thousandths,
that is to say, it may be dis-
regarded.

The combustion of the ammonia, moreover, is complete, for
no appreciable trace of it was found in the condensed water, and
a tube of pumice-stone and sulphuric acid placed as a test at the
end of the U-shaped tubes of solid potash, never increased in
weight in the experiments.

These facts being stated, the following results were obtained
under constant pressure, at about 12° :—

1 " Essai de M&anique Chimique," torn. i. p. 246.

Fig. 40. — Combustion of ammoniacal

242 HYDROGENATED COMPOUNDS OF NITROGEN.

Weight of water obtaiaed. "^T^SSS*

0-880 grms.
0-819 „
1-004 „
1-110 „
1-006

+ 91-1 Cal.
+ 90-7 „
+ 91-7 ,,
+ 91-4 „
+ 91-4 ,

Mean + 91-3 „

The heat of combustion of ammonia in solution will thus be
+ 82*5.

6. It is easy to deduce from this the heat of formation of
ammonia by its elements, without resting on any other basis
than the heat of formation of water. This being admitted,
according to the following data —

H2 + 0 = H20 liquid gives off + 34'5,
we deduce —

N + H3 = NH3 gas liberates + 103*5 - 91*3 = + 12*2.

The author found l that the solution of the ammoniacal gas
in a large quantity of water gives off + 8*82. Thus —

N 4. H3 -f water = NH3 dilute gives off 4- 21 Cal.

The value obtained with the hypobromite (-f 22*8) differs little
from this ; but it is necessarily less exact on account of the
complication of the reactions.

The author therefore adopts the respective values of +21
and + 12-2 for the formation of ammonia in solution and in the
gaseous form.

Between the result + 12*2 and the figures +26*7 previously
adopted, there is a discrepancy of 14*5 ; this is the greatest
experimental error that has up to the present been committed
in thermo-chemistry. Its source has been shown, and it has
been rectified accordingly.

7. Some months after the first publication of the results of
the author's researches, Thomsen repeated the experiments, and
he obtained for the heat of combustion of ammonia + 90 '65, a
value agreeing with +91*3 within the limits of error allowed
in experiments of this order. This is an important confir-
mation of the experiments. The heat of formation of ammonia
seems, therefore, to be definitely fixed at + 12*2, or very
near this.

1 " Annales de Chimie et de Physique," 5e serie, torn. iv. p. 526.

VOLATILITY OF AMMONIUM NITRATE.

243

§ 2. HEAT OF FORMATION OF AMMONIACAL SALTS FROM THEIR

ELEMENTS.

1. The table of the heat of formation of the principal
ammoniacal salts from their elements follows :

Sal

to.

Chloride • •

<51 4- H4 -f N

solid.
4- 76-7

dissolved.
4- 72 '7

Bromide

Br gas4- H4 + N

4- 71'2

4- 66'7

«

Br liquid
Br solid

4-67-2
4- 67-1

4-62-7
4- 62'6

Iodide • •

I SAB -4- IL 4- N

4- 56-0

4- 52-4

I solid

4- 49-6

+ 46-0

Sulphide . ....

S gas 4- H4 4- N

4-424

4- 39-2

S solid

4- 39-8

4- 36-6

C diamond 4- H4 + N-

4- 3-2

— 1-2

Nitrite ....

N, + H, 4- O,

4- 64-8

4- 60-1

Nitrate

No 4- H. 4- O

4- 87*9

4- 81-7

Perchlorate ....

C1 + N + H4 + O4
S solid 4- N2 4- H8 4- 04

4- 79-7
4- 141-1

4- 73-3
4- 140-5

Bicarbonate ....
Formate
Acetate . . .

C (diam.) 4- N 4- H4 4- O3
0 ( „ ) + N + H5 + 02
Ga( S 4. N 4- H 4- O«

4-205-6
4-129-4
4- 159-6

4- 199-H
4- 126-5
4- 159-8

Oxalate

CL ( i) 4- No 4- Ha 4- O,

4- 272-4

4- 264-4

2. The heat of formation of the same salts in .a solid form,
from ammoniacal gas and anhydrous or hydrated acids, taken
in the gaseous form and in the solid form, has been given
(p. 127).

3. We may also observe that the difference between the heats
of formation from the elements, of anhydrous salts of potash
and ammonia formed by strong acids, such as the nitrates,
sulphates, perchlorates, is almost constant, being about + 30
Cal. But this difference decreases for weaker acids ; it falls to
25 Cal. with the formates, oxalates, acetates, bicarbonates, etc.

§ 3. ON THE VOLATILITY OF AMMONIUM NITRATE.

1. It has been shown how ammonium nitrate may be decom-
posed in seven different ways, according to the process of heat-
ing (p. 5). Here certain experiments may be mentioned that
indicate an eighth mode of action of heat, viz. the volatilization
pure and simple of this salt.

2. Ammonium nitrate melts at about 152°, a temperature
which the water, previously existing or formed by the decom-
position of the salt, does not allow us to fix very accurately.
It is only at 210° that it begins to decompose, that is, sufficiently
to furnish an appreciable volume of gas in a few minutes, for
the decomposition begins really at a lower temperature. This
decomposition becomes more and more active, in proportion as

R2

244

HYDROGENATED COMPOUNDS OF NITROGEN.

the temperature of the salt melted is raised by some source of
heat, without, however, the temperature being arrested at any
fixed point between 200° and 300°. Pure nitrogen monoxide is
thus given off.

But if we go on raising the temperature, the reaction becomes
explosive at the time that the multiple products appear that
are due to the many distinct modes of simultaneous decom-
position, such as are shown on p. 5 of this work. All these
phenomena are of the same order as those manifested generally
by exothermal reactions, and their variety is a characteristic of
explosive substances.

3. However, according to the author's experiments on the
decomposition of ammonium nitrate, even with the greatest
care, the quantity of nitrogen monoxide
collected remains always considerably less
than the theoretical quantity. This is
on account of the volatility, real or ap-
parent, of the ammonium nitrate. The
difference is very great, even if we work
with the lowest possible temperature, and
in such a way as to prevent, as far as
possible, the portions sublimed in the
cold parts of the apparatus from gradually
falling into the heated parts at the same
time as the condensed water.

4. We can, in fact, sublime ammonium
nitrate without destroying it to any
extent (Fig. 41) . by placing this salt,
previously melted, in a capsule, E, which
is closed by means of a sheet of blotting
paper fastened over the top and sur-
mounted by a cardboard cylinder, CC',
the latter being filled with large pieces of glass. This is heated
over a sand bath, S, by means of a Bunsen burner, B, properly
regulated, care being taken that the temperature of the melted
salt (which is shown by a thermometer, 0, plunged obliquely
into it) does not exceed 190° to 200°. A very considerable
proportion of the salt is then sublimed in beautiful brilliant
crystals, adhering to the sides of the capsule and to the lower
surface of the paper. A portion of the salt even passes through,
and condenses above the capsule, in the form of a white smoke
very finely divided and very difficult to collect.

At first, the existence of some special compound in this
smoke, such as nitric amide, was suspected; but its identity
with ammonium nitrate was proved by a complete analysis.
The temperature of the paper thus traversed by the vapour
may rise above 120 and even 130 degrees (as shown by a
horizontal thermometer, t, laid upon the upper surface of the

Fig. 41.— Sublimation of
ammonium nitrate.

HYDROXYLAMINE. 245

paper) without the paper being affected to any considerable
extent. This experiment has some importance, as it shows that
ammonium nitrate may be volatilised as it is without being
at first resolved into ammonia and gaseous nitric acid —

NH4N03 = HN03 + NH3,

which would afterwards re-combine, the mixture when dis-
sociated possessing all the energy of the simple components.
In fact, we cannot understand how the vapour of nitric acid
could be in contact with the paper, at a temperature which
necessarily ranges between 130° and 190°, without oxidising it
or destroying it instantaneously.

5. Ammonium nitrate, from the point of view of its volatility,
and on account of many considerations, may be regarded as a
typical explosive substance. In fact, pure nitroglycerin may
also be evaporated without decomposition. Picric acid itself
gives off very appreciable vapours, which sublime, and are
condensed without alteration when the substance is heated
with great care.

§ 4. THERMAL FORMATION OF HYDROXYLAMINE OR OXYAMMONIA.

1. We know that hydroxylamine is a product of reduction
intermediate between hyponitrous acid and ammonia. It may be
formed in a number of oxidations. It was thought expedient
to determine its heat of formation, and this was done by
decomposing its hydrochloride by means of a saturated aqueous
solution of potash, very fine and very pure crystals of the salt
being employed.

2. Hydroxylamine, exposed under these conditions, is im-
mediately resolved into nitrogen and ammonia, according to
M. Lossen's observations.

After having ascertained that no other product was formed
(with the exception of a few hundred ths of nitrogen monoxide)
during the first moments of a sudden reaction, and that the
proportion of hydroxylamine thus destroyed at the ordinary
temperature and in a few minutes may amount to four-fifths of
its total weight, the reaction was reproduced in the calorimeter,
working with a known weight of hydrochloride, and collecting
the gases given off over the water in the calorimeter itself,
so as to measure them exactly.

3. We will now describe the apparatus employed in the
experiments (Fig. 42).

(1) At the bottom of a large tube, TT, closed at one end, is
placed a known weight of aqueous solution of potash, saturated
at the temperature of the experiment.

(2) In this large tube is suspended above the potash a smaller
tube, tt, containing exactly one grm. of hydroxylamine hydro-
chloride.

246

HYDROGENATKD COMPOUNDS OF NITROGEN.

(3) The small tube is wound round with a thick and heavy
spiral of platinum, gg, intended later on to plunge the system
below the level of the potash, and thus to determine the contact
and the reaction between the alkaline solution and the solid salt.

(4) The upper end of this spiral is hooked on to a platinum
wire 2^ mm*, in diameter, stretched between the two copper
wires of a small electric cable of gutta-percha, KK. This
cable is intended to convey the current, which is to heat to
redness and finally melt the little platinum wire, allowing the
small tube to fall into the solution of potash, where the salt
will react after its submersion.

(5) The large glass tube, TT, is closed with a cork, through
which on one side passes the cable which winds in and out
until outside the apparatus, and through the other side is passed
a tube, dd, used for the liberation of gases.

(6) This large glass tube, TT^ and the tube dd, including the

curved extremity of this latter,
through which the gases are to
escape, are contained together
in a small bell glass rather wide
and capable of containing 200 to
250 cms. of gas, a volume con-
siderably larger than that of the
gases given off in the reaction.

(7) This bell glass is in its
turn placed upside down with
its tubes and appurtenances,
in an ordinary platinum calori-
meter, CC, of a capacity of
1050 cms., but containing only
850 grms. of distilled water.

Thick copper wires, uu, ar-

Fig. 42.

ranged beforehand in the form- of a star round a central point
on the upper surface and on the axis of the bell glass, support
it and keep it in a fixed position under the water. These
wires are connected with a centra! rod, S, which rises verti-
cally above the apparatus, and enables it to be attended to
without any special instrument being introduced into the
calorimeter. "

It need not be said that the weight of .each portion of this
complicated apparatus was determined beforehand, so as to
enable us to reduce the submerged masses to units of water.
Moreover, special measurements of the specific heat of the cable
and that of the cork were taken ; these measurements may be
made somewhat roughly, since the weight of the cable sub-
merged does not exceed a few grammes ; that of the cork is
still less. As to the' glass, copper, and platinum, their specific
heat is known.

EXPERIMENTAL DETAILS.

247

(8) The parts being all adjusted, the air is exhausted in the
bell glass by means of an inverted syphon.

(9) Then we have merely to follow the progress of the
thermometer, 0, for ten minutes.

(10) We then heat and finally melt the little platinum wire
by means of a current of four Bunsen elements, the hydroxy-
lamine hydrochloride falls into the potash and is immediately
destroyed. The gases produced by its destruction are given
off under the bell glass. We give this glass a rotatory move-
ment for a few minutes by means of the rod S, taking care to
keep it completely submerged. Headings of the
thermometer are taken every minute. ±

(11) This done, we break the bottom of the
large glass tube by means of a- platinum crusher
introduced from outside and fixed at the ex-
tremity of a long rod of the same metal (Fig. 43) ;
the liquids and other substances contained in the
tubes spread out into the calorimeter and remain
in it completely intermingled, this being effected
by a suitable agitation which is easily performed
by means of the rod S.

(12) During this interval, and for a little while
after, the progress of the thermometer is followed ;
all the thermal data are thus determined.

(13) This being done, all that remains is to
know the volume of nitrogen developed by the
decomposition. For this purpose we put the
platinum calorimeter with the bell glass into
the water contained in a very large earthen pan,
so as completely to submerge them. The bell
glass is then raised, so as to render it independent
of the calorimeter, and the gases are transferred
to a graduated testing apparatus.

These gases contain the nitrogen given off Fig. 43.— Plati-
(mixed with three or four per cent, of nitrogen num crusher,
monoxide, according to the analyses), plus the
air contained at first in the large tube and in the liberating
tube. The volume of this air is known by the previous gauging
of the tubes, if we deduct the liquid volumes of the potash
and the various other objects introduced into the tube for the
experiment. These volumes having been each measured sepa-
rately, we succeed, finally, in ascertaining within about half
a cubic centimetre the volume of nitrogen given off by the
destruction of the hydroxylamine.

In the author's experiments this volume corresponded to
78 and 79 per cent, of the weight of the salt subjected to the
reaction. The surplus of the salt, or more correctly the surplus
of the hydroxylamine derived from it, is found unaltered in the

248 HYDROGENATED COMPOUNDS OF NITROGEN.

water of the calorimeter, where it is mixed with the excess of
potash.

The apparatus just described is very complicated, but the
experiment is in itself very simple ; it admits of a very accurate
measurement of the heat given off, and the conditions at the
commencement being ascertained with exactness, it is possible
to arrive at a strictly definite final condition in one operation.

4. In order to calculate the decomposition of pure hydroxy-
lamine, it is necessary to measure —

(1) The total heat given off in the reaction just described.

(2) The heat given off by an equal weight of the same potash
reacting upon the weight of pure water contained in the
calorimeter.

(3) The heat absorbed by the solution of an equal weight
of pure hydroxylamine hydrochloride in the same quantity of
water.

(4) The heat given off when the hydroxylamine hydrochloride
in a weak solution is decomposed by the diluted potash ; in this
case the hydroxylamine is set at liberty at first without being
destroyed.

All these data being obtained by special experiments, it is
easy to calculate the heat given off by the simple destruction* of
an equivalent of hydroxylamine.

5. The results deduced from the experiments are as follows : —

3NH30 dissolved = N2 + NH3 + 3H20 disengaged1 + 57'3 & + 56'7; mean

+ 57-0 Gal

Other distinct experiments have given : —

NH30 diss. + HC1 dilute at 24° liberates 2 + 9'2

NHgOHCl cryst, (1 p. of salt + 90 p. of water) in dis-
solving at 24° - 3-31

(NH30)2S04H2 cryst. + 100 parts of water at 12-5 ... - 2-90

(NH30)2 dilute + H2S04 dilute, at 12-5 + 10-8

6. Formation from the elements : —

N + H3 + 0 = NH30 dissolved liberates ... . + 19-0

N + H3 + 0 + HC1 dilute = NH3OHC1 diss + 28-2

N + H4 + 0 + Cl gaseous = NH3OHC1 cryst. ... , + 70-8

N2 + H6 + 02 + S04H2 dilute = (NH30)2H2S04 diss.. + 29'8

N2 + H8 + 06 + S ~ (NH30)2H2S04 cryst + 138-8

7. Different modes of formation : —

OXIDATION OF AMMONIA.

NH3 diss. + 0 = NH30 diss. will absorb ... - 2-0

NH3HC1 diss. + 0 = NH3OHC1 diss - 7-2

NH3HC1 cryst. + 0 = NH3OHC1 cryst. ... - 5'9

1 In the calculation of the experiments the formation of a little nitrogen
monoxide was taken into account, say 3 to 4 per cent., under the conditions in
which I was working.

2 According to the decomposition of pure hydrochloride dissolved in
water, by dilute potash.

NASCENT HYDKOGEN AND HYDKOXYLAMINE. 249

SIMILAR OXIDATION OP THE SULPHATE.

(NH3)2H2S04 diss. + 0 = (NH30)2H2S04 diss - 5-7

(NH3)2H2S04 cryst. + 0 = (NH30)2H2S04 cryst - 4-1

We see that a fixed oxidation would absorb quantities of heat
varying from — 2 '6 to — 7 '2; according to whether it takes
place on free hydroxylamine or on its salts in solution. It is
essential to note that this quantity is negative, unlike what
takes place for oxides of nitrogen.1 Moreover, the three above-
mentioned reactions are purely theoretical ; they are, however,
worthy of mention, as by their endothermal nature they may be
compared to the formation of oxygenated water and to that of
nitrogen monoxide.

We get for the formation of hydroxylamine by the hydrogena-
tion of nitric oxide —

NO + H3 + water = NH30 diss. + 40-6.

This last reaction is effected, in fact, by means of nascent
hydrogen, that is to say, in reactions which furnish, in addition,
the heat which would have been given off at the time of the
formation of the free hydrogen, under the same conditions.

8. Eeactions of hydroxylamine. Action of hydrogen —

NH30 dissolved + H2 = NH3H20 dissolved + 71*0.

We see by this that the hydroxylamine will be easily changed
into ammonia by the nascent hydrogen. This is why the
production of the first body, in the reduction of the oxides of
nitrogen, requires very special conditions. Among all the
formations of nitric compounds that nitric acid can effect by
producing oxidation, that of hydroxylamine gives off the least
heat. In fact, each equivalent of oxygen imparted by the
dilute nitric acid to the body to be oxidised with the formation
of hydroxylamine gives off — 16*4 Gal. less than free oxygen,
whereas the free formation of ammonia gives off only — 12'1 Gal.
less, that of nitric oxide - 12 Gal., that of nitric peroxide
- 9-6 Gal, that of nitrogen -14 Gal., etc.2

9. Action of oxygen. Heat of combustion —

2NH30 dilute + 0 = ¥2 + 3H20 liquid + 84'5.

The combustion of dilute ammonia gives off a little less, or
-f 82*5 ; but it requires three times as much oxygen for the
same weight of nitrogen contained in the compound.

We get—

2NH30 dilute + 02 = N20 gas + 3H20 liquid +74-2

4NH30 „ + 05 = N403 dilute + 6H20 liquid + 65'4

2NH30 „ +04 = N203 „ + 3H20 „ +80-3

2NH30 ,; +05 = 2HN03:: + 2H20 „ +98-8

1 p. 171, nitric oxide ; pp. 178 and 179, nitrogen trioxide ; p. 189, nitric
peroxide.

2 See p. 200.

250 HYDROGENATED COMPOUNDS OF NITROGEN.

or for each fixed equivalent of oxygen (8 grms.), 371, 261,
201, 19-8.

10. Action of dilute alkalis.

The reaction of the alkalis upon the salts of hydroxylamine
is worthy of notice. Dilute alkalis confine themselves to dis-
placing the hydroxylamine, at least in an operation of short
duration. The measurement of the heat given off shows that
hydroxylamine is a much weaker base than baryta, potash, and
even ammonia. In fact, with dilute potash and the hydro-
chloride it was —

2NH3OHC1 dissolved -f K2O dilute at 23°, + 444 ;
with dilute baryta and the sulphate at 12'5°—

2(KE30)H2SOi dilute -f BaO dilute, 4- 7'8 ;
likewise with ammonia and the chloride —

]STH3OHC1 dissolved + NH, dilute at 12-5°, + 3'35.

These thermal measurements show that the displacement of
the hydroxylamine by the ammonia is complete, i.e. in pro-
portion to the weight of this base. It is the same even when
we employ only half the ammonia necessary for a complete
decomposition.

Hydroxylamine is, therefore, one of the weakest of bases,
hence its salts offer a very pronounced acid reaction.

It was found that the sulphuric acid, which is combined with
it, might be accurately estimated by an alkalimetric test;
almost like the soda in borax, but by an opposite test.

11. The concentrated alkalis act very differently, for they
determine the decomposition of hydroxylamine itself. Thus
with concentrated potash we get destruction of the hydroxy-
lamine.

12. Ammonia. 1. With a saturated aqueous solution of
ammonia at about zero, the hydroxylamine is displaced in its
salts without undergoing decomposition* even at the end of
several days. 2, With ammoniacal gas and solid hydro-
xylamine hydrochloride there is slow decomposition of the
hydroxylamine. Theory indicates that the displacement pro-
perly so called —

NH3OHC1 solid -|- NH3 gas = NH4C1 solid + 1STH3O, liberates

-f 12-6 - a,

a being the heat of dissolution of NH3O, a compound which
appears to be liquid.

In fact, it was observed that the dry hydrochloride absorbs the
ammoniacal gas immediately, in the proportion of one equivalent,
and even a little more. If we employ a considerable excess of
ammoniacal gas, working over mercury, and immediately remove

CONDITIONS OF STABILITY OF HYDROXYLAMINE. 251

this excess by means of a gas pipette, the gas separated contains
barely a few hundredths of a gas almost insoluble in water
(nitrogen or nitrogen monoxide), which shows that the decom-
position of the hydroxylamine is almost inappreciable under
these conditions. However, the gas so separated contains a few
hundredths of the vapour of hydroxylamine. We may prove
this by the following process. This gas is heated with a few
drops of water, which dissolve the vapour at the same time as
the ammonia ; the gas not dissolved is taken away by means of
a gas pipette, then we add to the water a large piece of potash
(with its surface previously damped, so as to eliminate the
gases adhering to it) ; under these conditions the hydroxylamine
which existed in the water, and consequently in the ammoniacal
gas which this water had dissolved, is immediately destroyed
with formation of nitrogen, which is really produced and which
may then easily be observed.

Hydroxylamine may then be regarded, according to these
facts, as existing in a free state and in a liquid form, in the
testing apparatus, where it impregnates the ammonium chloride.

Its vapour tension, as deduced from the preceding experi-
ments, would indicate a boiling point near that of water.

But hydroxylamine so formed does not exist long in a state
of purity ; it is destroyed little by little, giving rise especially
to nitrogen monoxide and ammonia —

4KE30 = £T20 4- 2NH3 + 3H20.

At the end of forty-eight hours^ nearly two-thirds had under-
gone this transformation, as found by an exact analysis made of
the products derived from a known weight of the hydrochloride ;
about a seventh had in the same time changed into nitrogen and
ammonia.

The fundamental reaction, which* in this case produces nitrogen
monoxide, gives off, according to calculation, -f- 48*4 Gal., a
result relating to the following conditions —

4NH30 dilute = N2Q'gas -f 2NH3 dilute -J- 3H20 liquid.

The real reaction, N"H3 being supposed to be gaseous, and a
being the heat of solution of NH30> gives- off 4- 39*6 — a.

We see that all these quantities are far below the heat given
off in the reaction, engendered by nitrogen, viz. + 57. This ex-
plains why this last reaction preponderates under the influence
of concentrated potash.

13. From these facts it follows that hydroxylamine is only
stable in presence of acids, but its union with these agents
deprives it of part of its energy. This is, moreover, generally
the case in chemistry; a system is the more stable, all else
being equal, in proportion as the fraction of its energy which it
loses is greater (see p. 123).

252 HYDROGENATED COMPOUNDS OF NITROGEN.

In the same way, it was found that hydrochloric acid gas in
excess, and also boron fluoride, do not determine the decompo-
sition of hydroxylamine, notwithstanding their avidity for the
water which might be formed at its expense. But this relative
stability is explained by the preceding considerations, i.e. in
proportion to the formation of the saline compounds.

But on the contrary, hydroxylamine, when free or dissolved
in a very small quantity of water, i.e. possessed of all its energy,
manifests a strong tendency to spontaneous destruction, and
this destruction works in a way that gives off the more heat the
more suddenly it is effected.

14. To recapitulate these various processes of decomposition.

(1) In the most simple decomposition

NH30 dissolved = N -f H + H20 -f water would liberate

+ 50-0.

But this sudden reaction has not been observed ; the nascent
hydrogen remained completely associated with the nitrogen
in these conditions, and it forms ammonia, a formation ac-
companied by a second liberation of heat.

(2) In a sudden reaction we see the transformation of a third
of the nitrogen into ammonia, as follows —

3NH30 dilute = JSTH3 dilute + 2N + 3H2O,

a reaction which gives off in addition -f- 7, or altogether +57.

We observe also the absence of the compound NH, which one
would think ought to appear under these conditions. Though
sought for particularly, no trace of it was obtained.

The formation of water itself, which it would seem a priori
ought to be effected in preference, preponderates only in the
sudden reaction brought about by potash; probably by reason
of the tendency of this alkali to form hydrates with liberation
of heat. Thus the slightest influence determines the manner
in which this unstable compound is destroyed.

(3) On the contrary, in the spontaneous decomposition of
hydroxylamine, such as takes place in the presence of am-
moniacal gas, we see chiefly nitrogen monoxide appear, with less
liberation of heat ( + 484 X 2 instead of + 57 X 2, all the
substances being supposed to be in solution).

15. Constitution. This last decomposition, effected upon two
molecules of hydroxylamine, one of which abstracts the hydro-
gen from the other, recalls the resolution of an aldehyde into
the corresponding alcohol (or rather carburet) and acid. We
may here remark that the slow decomposition of hydroxylamine
is at the same time that which develops least heat and which is
produced in preference, under conditions in which most care is
taken. Moreover, it takes place at exactly the same temperature
as the decomposition that gives off the most heat. But these

AMIDES AND SOME ORGANIC ALKALIS. 253

various relations are not necessary, and we might quote contrary
examples in which a slow decomposition gives off more heat than a
rapid one effected at the same temperature (decomposition of
barium dioxide by a diluted acid, with the rapid formation of
oxygenated water, which is itself slowly resolved into water and
free oxygen ; the decomposition of a hypochlorite by a dilute
acid, &c.).

The initial temperature of the reactions is not connected except
in a general manner with their unequal thermal value, as is shown
by the comparison of the reactions of potassium chlorate and
iodate. In short, the conditions of more or less rapid action or
higher or lower initial temperature are not those that regulate
the phenomena.

On the contrary, the phenomena are determined, on the one
hand, l>y the general tendency towards the conservation of the
initial molecular condition, arid, on the other hand, ly the tendency
of any system towards the condition that corresponds to the
maximum of heat given off. This last condition is realised fully
whenever the corresponding bodies can begin to be produced in
the conditions of the experiments. It is in order to avoid, so
far as possible, the realisation of conditions favourable to the
production of these bodies that we avoid raising the temperature
and hurrying the reactions. We thus keep as closely as
possible to the primitive molecular type.

Without dwelling longer on considerations of this order, it
may be said in conclusion that the thermal observations confirm
and specify the unstable properties of hydroxylamine, and this
instability is due to the exothermal character of its various
decompositions.

§ 5. HEAT OF FORMATION OF SOME ORGANIC ALKALIS.

First Section — General Remarks.

1. Ammonia, on uniting with organic compounds, such as
hydrocarbons, alcohols, aldehydes, acids, forms compounds of
various natures, alkalis and amides in particular.1

The thermal study of these compounds has been very little
worked. It would be of great interest in the study of the force
of explosive substances derived from ammoniacal salts, cyanides,
diazo compounds, etc. The author measured the heat of for-
mation of the cyanide compounds, of several diazo compounds,
and of some alkalis and amides. As special chapters are
devoted to the cyanide series and the diazo compounds, the
alkalis and amides will only be discussed here.

i " Traite Ele'mentaire de Chimie Organique," torn. xi. pp. 224 and 313.
Second edition (1881), with the collaboration of M. Jungfleisch, published by
Dunod.

254 HYDROGENATED COMPOUNDS OF NITROGEN.

Second Section — Ethylamine.

1. This alkali is gaseous in summer ; it boils at + 18'1, it is
extremely soluble in water, and forms well-defined salts.

2. Analysis. Its purity was proved by eudiometric analysis,
a more reliable process than analysis by weight for such com-
pounds. These are the results in volume : —

ETHYLAMINE.

Volume of the Gas.

C02 produced.

Nitrogen.

Total diminution after
combustion and
absorption of C02.

Found 100

201

50-5

428

Calculated 100

200

50-0

425

3. Heat of combustion of ethylamine. Four detonations made
with weights of this base ranging between '11 and '12 of a grm.
gave, at about 20 '5° with gaseous ethylamine (C2H7N = 45
grms.), the volume being constant —

2C2H7N gas + As = 4C02 gas + 7H20 liquid + Na.

According to the initial
weight of the alkali.

416-3 Cal.
409-3 „
400-7 „
402-7 „

Mean 407-2 ,

According to the final
weight of the carbonic acid.

413-0 Cal.
403-3 „
406-4 „
416-4 „

Mean 409-3 ,

The general mean 4- 408*5 must be increased by 1*2 to pass
to the ordinary heat of combustion under constant pressure,
which makes 4- 409*7 Cal. This number entails a limit of
error of about ± 4 Cal., an uncertainty that also occurs in the
following deductions.

4. Heat of formation. The heat of combustion of the
elements being + 42 9 '5, we get for the heat of formation —

From the elements —

C2 (diamond) -f H7 + N = C2H7N gas + 19-8

C2 (charcoal) „ „ +25-8

From ammonia —

C2 (diamond) + H4 + NH3 = C2H7N gas +7-6

C2 (charcoal) „ „ + 13-6

From ethylene —

C2H4 + NHS = C2H7N +23-0

From alcohol —

C2H5(HO) gas + NH3 gas = C2H7N + H20 +6-1

5. Solution in water. Two experiments made at 190° on

HEAT OF COMBUSTION OF TKIMETHYLAMINE.

255

weights of gaseous ethylamine, equal respectively to 2*555
grms. and 2*415 grms., and dissolved in 400 grms. of water
gave for C2H7]Sr (45 grms.); + 12*92 and + 12*90 ; mean 12*91
Cal. These results exceed those of ammonia by one-half.
6. Formation of salts in solution at 190°.

C2H7N (1 eq. = 7 litres) + HC1 (1 eq. = 2 litres) liberates -f 13-2
+ C2HA „ „ 4-12-9

+ H2S04 „ „ 4-15-2

figures that are intermediate between those given .by potash and
ammonia.

Third Section. — Trimethylamine.

1. This is a liquid that boils at 9°; it is consequently
gaseous at the ordinary temperature. It is ^ery soluble in
water and forms well-defined salts.

2. Analysis. These are the results obtained -by eudiometric
analysis : —

TRIMETHYLAMINE.

Volume of the Gas.

C02 produced.

Nitrogen.

Total diminution after
combustion and
absorption of <!02.

Found 100
Calculated ..... 100

302
300

50
50

580
575

3. Heat of combustion of trimethylamine. Three detonations
made with weights of the base ranging between 112 grms. and
•186 grms. gave for C3H9N (59 grms.), the volume being
constant —

2C3H9lSr + 021 = 6C02 gas + 9H2O liquid -f N2.

According to the initial weight 586*2, 583*5, 601*1 ; mean
+ 590*3.

According to the final weight of the carbonic acid, on an
average + 591-7.

The general mean is + 590*5, which gives for the heat of
combustion at a constant pressure -f- 592, with a limit of error
of about + 6 Cal., an uncertainty that applies to the following
deductions.

4. Heat of formation.

From the elements —

C3 (diamond) + H9 + N = C3H9N gas - 9-5

C3 (charcoal) „ „ - 0-5

From ammonia —

C3 (diamond) + H6 + NH3 = C3H9N gas +2-7

C3 (charcoal) „ „ +6-3

From methylic alcohol —

3[CH3(HO)] + NH3 = (CH2)3NH3 -1- 3H20 gas ... -7-3x3

256 HYDROGENATED COMPOUNDS OP NITROGEN.

The exact deductions that can be drawn from the heats of
combustion are really only valid for low heats of combustion or
for considerable differences, and attention must be called to the
limits of error involved in calculations of this kind, in order to
prevent any misapprehensions.

5. Solution in water. Three experiments, made at about 20°,
on weights of the base equal respectively to 4*753, 4*994, and
4-633 grms., and dissolved separately in 400 grms. of water,
gave for C3H9N (59 grms.) gaseous + 270 H20 about + 12*82,
+ 12-76, -f 13-2 ; on an average + 12-90 Cal.

This figure is equal to the heat of solution of ethylamine, and
it shows in both bases a special affinity for water.

6. Dilution. This affinity may be shown still more clearly
as regards trimethylamine, by experiments on dilution.

A liquid saturated at about 19° contained 409*6 grms. of the
base per litre, or 478 grms. per kilog. Its density was *858
at 16°. It corresponded to C3H9N + 7*17H20.

On being diluted with thirty times its volume of water, it
gave off + 3*89 Cal. at 19°.

Thus C3H9N, on combining with 7*17H20, gives off only +
9 Cal., and that its subsequent dilution gives off about half as
much heat.

For purposes of comparison we may repeat here some of the
figures obtained with ammonia.

NH3 + 7H20, by its subsequent dilution, gives off + -32.
JSTH3 -j- 19H20, gives off only -f *02.

These figures show that ammonia has much less tendency
than trimethylamine to form hydrates.

The heat of dilution of the latter base when concentrated is
very considerable, and its value amounts to even double that of
potash and soda, taken at a corresponding degree of concentra-
tion. The heat of dilution of concentrated trimethylamine is
quite comparable to that of the hydracids. Now, such values
express the formation of certain successive hydrates,1 a very
important circumstance in the study of the reactions of
hydracids, as well as those of trimethylamine.

7. Formation of salts in solution. At 21° it was found —

C3H9N (1 eq. = 5 litres) + HC1 (1 eq. = 2 litres) ... + 8*9

„ „ + C2H402 „ ... + 8-3

+ H2S04 „ ... +10-9

As a check to these results, we get by a double reciprocal
decomposition —

C3H9N (1 eq. = 2 litres) + KC1 (1 eq. = 2 litres) + 4-40\,,, ,T
K20 (1 eq. = 2 litres) + C3H9NHC1 (1 eq. = 2 litres) - O28/M

1 " Essai de Me'canique Chimique," torn. ii. pp. 151, 167.

TRIMETHYLAMINTE HYDROCHLORIDE. 257

From these data it follows that the heat given off by the
union of potash with hydrochloric acid exceeds by -f 47 that
given off by trimethylamine (p. 118), which gives for the com-
bination of this base in solution with dilute hydrochloric acid,
the value + 9, which results agree with that given above.

We also see, by the above numerical experiments, that the
potash entirely, or almost entirely, displaces the trimethylamine
in its acid compounds. It seems, however, that there are some
indications of division —

C3H9N (1 eq. = 2 litres) + NH3HC1 (1 eq. = 2 litres) - 2'33\M M,_ , „ .
NH3 (1 eq. = 2 litres) + CSH9NHC1 (1 eq. = 2 litres) + M7/M

The division of the acid between the two bases is here
evident. It is no doubt due to the formation of dissolved
hydrates of trimethylamine as mentioned above, and also of its
hydrate, which is discussed further on. Without entering
further into this point, we will content ourselves with saying
that we deduce from these figures, for the heat of neutralisation
of trimethylamine by hydrochloric acid, + 8 -95.

The three values found agree, viz. 8 '9, 9, 8*95. They are
lower by about a third than the heats of neutralisation of potash
by the corresponding acids ; they are even lower than the results
obtained with ammonia. Their numerical values approximate,
on the contrary, to the heats of neutralisation of the same acids
by hydroxylamine and by aniline, bases which are much weaker
than ammonia.

Again, we find —

C3H9N (1 eq. = 8 litres) + C02 (44grms. in 26 litres) liberates + 4-4
C3H9NHC1 (1 eq. = 2 litres) + Na2C03 (1 eq. = 2 litres) „ - M7

The last value indicates the transformation of trimethylamine
chloride into sodium chloride, the strong base, i.e. the soda,
taking the strong acid, i.e. the hydrochloric acid, as it happens
also between ammonium chloride and sodium chloride, and for
the same reasons.1 If we suppose the reaction to be total, we
deduce from it that C02 in solution, + C3H9N in solution,
would give off -f- 4'1 in the presence of 4 litres of water.

In the presence of 17 litres of water the experiment gave a
lower value, which seems to indicate the gradual dissociation
of the carbonate by dilution, always as with ammonia.2

8. Trimethylamine hydrochloride. The heat of formation of
this salt has already been given in a state of solution. In order
to estimate it in a solid state the heat of solution was determined
upon a fine specimen, supplied by M. Vincent, carefully dried
upon blotting paper, under a bell glass over sulphuric acid.

1 For discussions of reactions of this order, see "Essai de Me'canique
Chimique," torn. ii. pp. 712 and 717.

2 « Annales de Chimie et de Physique," 4a seVie, torn, xxxix. pp. 477-485.

3

258 HYDKOGENATED COMPOUNDS OF NITROGEN.

The analysis of it agreed pretty closely with the formula —
C3H9NHC1.

10 gnns. of this salt were dissolved in 500 grms. of water
at 180°.

A slight absorption of heat was produced, answering to
-•5 Gal., for

C3H9NHCL = 95-5 grms.

According to this result,
C3H9NCgas + HClgas = C3H9NHC1 solid liberates + 39*8 Gal.

This value is lower than the heat of formation of solid
ammonium chloride starting from its gaseous components, or
+ 45-5 Gal.

But the value deduced probably does not represent the actual
heat of formation of trimethylamine chloride as it exists in
diluted solutions. In fact, this salt attracts the atmospheric
moisture with such avidity that it falls almost immediately
into a liquid state, which indicates the formation of a definite
hydrate in its solutions, whereas ammonium chloride seems to
exist in its solutions in an anhydrous state. The heat of
formation of anhydrous trimethylamine chloride must there-
fore be increased in its solutions by the heat of formation of
its hydrate, if we wish to calculate the energy really called
into action in the formation of the chloride in solution, i.e. the
true energy put forth in the reactions of this substance.1

§ 6. THE HEAT OF FOKMATION OF SOME AMIDES.

1. The amides are derived, in general, from the union of the
acids and ammonia, with separation of water, that is, they are
ammoniacal salts deprived of the elements of water. This class
comprises a number of very important compounds ; it extends
even as far as the albumenoid principles which form the basis
of animal tissues and organs. Many explosive substances are
also included in it. But their thermal study is not as yet far
advanced, with the exception of that of the cyanide series,
which will be discussed in a subsequent chapter. Besides
these, the .author has only, up to the present, examined two
amides, viz. oxamide and formamide.

2. Oxamide. Oxamide is a solid body, almost insoluble,
differing from ammonium oxalate by the elements of water —

C2H204(NH3)2 = C2H4N202 + 2H20.
It may be obtained, either by the decomposition of the salt,

1 We must also take into account its own state of dissociation as a
hydrate and as an anhydrous salt. " Essai de Mecanique Chimique,"
torn. ii. p. 445.

OXAMIDE. 259

or by the reaction of ammonia upon oxalic ether, a reaction
more accessible to measurement.

A + 2CAO,

In fact, the reaction of ammonia upon oxalic ether is
immediate. This circumstance was taken advantage of to
determine the heat of formation of oxamide. For example,
1-9495 grm. of oxalic ether and 10 cms. of a very concentrated
solution of ammonia were enclosed in a phial,

(NH33H20 about),

the two bodies being brought together in a little receiver
immersed in the water of the calorimeter. The reaction is
complete at the end of three or four minutes. The products are
then mixed with the water of the calorimeter, so as to bring the
whole into a state of dilute aqueous solution.
All the calculations being made,1

(C2H4)2C2H204 pure + 2NH3 dilute
= C2H4lsr202 solid + 2C2H60 dilute

gave off + 26-2 and + 26'6, on an average -f 26'4, or 13'2 x 2.
Now the formation of ammonium oxalate, by means of
oxalic ether and ammonia, in the presence of a large quantity
of water —

(C2H4)2C2H204 pure + 2NH3 dilute
= C2H2042NH3 dissolved + 2C2H60 dilute,

would give off + 16'0 X 2.

By subtracting from the difference (16'0 — 13'2) x 2, the heat
of solution of the ammonium oxalate, or — 4 x 2, we find that
the formation of oxamide from the solid salt,

C2H2042ISrH3 cryst. = C2H4lSr202 + 2H20 liquid, absorbs - 2'4
or - 1-2 x 2.

In the conditions of direct metamorphosis, by the action of
heat on ammomium oxalate, the water takes a gaseous form.
Hence

C2H2042NH3 cryst.
= C2H4£T20 + 2H202 gas absorbs - 217 or - 10'8 x 2.

From the above measurements we deduce the heat of forma-
tion of oxamide from the elements

C2 (diamond) + H4 + N2 -f 02 . . . . + 140'0.

3. Formamide. It was found that the transformation of
formic amide into formic acid and ammonia (or rather into

1 The author also studied the action of dilute ammonia upon oxalic
ether, dissolved beforehand in a large quantity of water. This reaction gave
off + 8-2 x 2. It did not produce oxamide, all the bodies remaining in
solution, even after several days, no doubt in the form of oxamic ether.

s2

260 HYDROGENATED COMPOUNDS OF NITROGEN.

ammonium chloride) is effected by means of concentrated
hydrochloric acid.

According to the figures obtained, the reaction CH3NO diss.
+ H20 = CH202NH3 in solution, gives off + 1/0.

The opposite reaction, the two conditions being similarly
comparable, absorbs — 1, a result very near — 1/2 observed
with oxamide. It is also very near the absorption of heat
produced in the formation of ethers.

Eeciprocally, the fixing of the water on the oxamide (as upon
formamide) with the production of ammoniacal salts, gives off
in heat -j- 2*4 for oxamide, always like the fixing of water on
the ethers.

4 We see by this that the hydration of organic compounds
generally gives off heat, whether we are considering the decom-
position of ethers dissolved in acids and dilute alcohols, the
transformation of amides into ammoniacal salts, the transforma-
tion of anhydrous acids into hydrated acids, or of the acid
chlorides into hydrochloric acid and dilute organic acids.
This is a very general result, to which attention was drawn in
1865, and which is confirmed and put in a definite form by the
present experiments. Its importance in the theory of animal
heat may be easily understood. From a more technical point
of view, this relation, and especially the values found for the
hydration of oxamide and formamide, may be useful in the
approximate calculation of the heat of formation of amidated
compounds capable of being employed in the manufacture of
explosive substances.

( 261 )

CHAPTER VII.

HEAT OF FORMATION OF NITROGEN SULPHIDE.1

§ 1. NITROGEN SULPHIDE.

1. THIS body is a solid, crystallised, yellow, explosive substance,
expressed by the formula NS, and by the equivalent 46. It is
prepared by the action of ammoniacal gas upon sulphur
chloride, dissolved in carbon disulphide.2

The specimen used in these experiments gave upon analysis —

Found. Calculated.

N 69-64 69-56

S 30-41 30-44

H ... ... 0-01 —

2. Nitrogen sulphide is stable at the ordinary temperature.
It is preserved without alteration both in dry and in damp air.
It may be moistened and then dried at 50° without any appreci-
able alteration, even should these operations be repeated several
times.

Its density at 15° was found to be equal to 2'22.

Nitrogen sulphide detonates with violence upon being struck
with a hammer, but its sensitiveness to this shock is less than
that of mercury fulminate.

On being heated it explodes at 207° and above this heat. Its
decomposition is, however, much slower than that of mercury
fulminate or diazobenzol nitrate. We may remark that this
temperature of conflagration is near that of the combustion
of sulphur freely exposed to air.

3. Heat of detonation. The decomposition of the nitrogen
sulphide was provoked in a pure dry atmosphere of nitrogen, in
a bomb lined with platinum.

It was ignited by means of a very fine metallic wire, plunged
nto the substance and heated to incandescence by means of ail

1 This study was made jointly with M. Vieille.

2 Fordoz et Gelis, " Annales de Chimie et de Physique," 3° serie, torn,
xxxii. p. 385.

262 HEAT OF FORMATION OF NITROGEN SULPHIDE.

electric current. Two experiments gave, the volume being
constant —

Weight of the Heat given off Per equivalent

substance. per grm. (46 grms.).

2-997 grms 701-1

2-979 „ 700-4

Mean 700-7

At a constant pressure, we should have had + 31 '9. The
experiment gave for 1 grm. 243*1 cms. of gas (the volume being
reduced to 0° and 76 metre).1

Theory gives 242*2 cub. cms.

These gases consisted of pure nitrogen, within about 25o^n-
Thus the decomposition was produced according to the equation
NS = N -|- S, i.e. the nitrogen sulphide is resolved purely and
simply into its elements.

4. Heat of formation. From these results we conclude that
the formation of nitrogen sulphide from its elements,

N + S = NS, absorbs - 32*2 Cal,

the volume being constant, or 31*9 Cal. at a constant pressure.

This formation is, therefore, endothermal, which explains why
it does not take place directly. But it is effected by making
ammoniacal gas act upon sulphur chloride. The chlorine in
this latter compound unites with the hydrogen of the ammonia
to form hydrochloric acid, and consequently ammonium chloride,
while the nitrogen sulphide is forming. This transformation
finally gives off 4- 123*0 Cal. The energy consumed in the
association of the sulphur and the nitrogen (—31*9) is thus
furnished by the formation of the hydrochloric acid, or rather
by that of the ammonium chloride, at the expense of the sulphur
chloride and the ammoniacal gas (+ 230*1 — 75*2).

5. It will be observed that the combination of the nitrogen
with the sulphur absorbs heat (— 31*1 Cal.), exactly like the
combination of nitrogen with oxygen (— 21*6 Cal.). The
nitrogen sulphide is, therefore, analogous to nitric oxide as
regards its endothermal character, as well as its formula. This
is a fresh proof of the general analogy existing between the
conditions of formation of oxygenated compounds and those of
sulphuretted compounds. It is difficult to carry further these
points of resemblance in the heats of formation, seeing that the
conditions of the two compounds are not comparable, any more
than the conditions of the elements, although a certain com-
pensation may be allowed between the solid form of the sulphur
and that of nitrogen sulphide.

6. Heat of combustion. If we are working in air or in oxygen,
nitrogen sulphide burns —

NS + 02 = N + S02,
and gives off -f 101*1 at a constant pressure.

NITROGEN SELENIDE. 263

§ 2. NITROGEN SELENIDE.

1. This compound is similar to nitrogen sulphide; it has
recently been the subject of careful study on the part of M.
Verneuil,1 who has fixed its formula at NSe. He kindly
furnished M. Vieille and the author with a specimen for the
experiments which they were making upon explosive substances.

It is an amorphous powder, of a deep orange colour, very
dangerous to handle. It explodes at about 230°, according to
M. Verneuil. It also explodes either by friction or by a very
slight percussion of iron upon iron, or a more violent percussion
of wood upon iron. The contact of a drop of sulphuric acid
also makes it explode.

2. Heat of detonation. We effected the explosion in our
usual apparatus by the same process as we adopted for nitrogen
sulphide. Working with 3 grms. of the substance, two trials
gave, for the reaction —

NSe (93 grms.) = 1ST + Se

+ 42-9 Cal. and -f 42'4 Gal., on an average + 42'6 Gal. with
a constant volume, or + 42 '3 Cal. at constant pressure, a value
which is only approximate, owing to the difficulty of obtaining
this substance quite pure.

3. Heat of formation. We conclude from these observations
that the nitrogen selenide is formed from its elements, with an
absorption of heat equal to — 42*3 Cal. at a constant pressure.

4 The heat of combustion.

NSe + 02 = N + Se02

is equal to + 99'9 Cal.

5. Thus nitrogen selenide is an endothermal combination
( — 42'3). It may, therefore, in this respect be classed with
nitrogen oxide (— 21*6 Cal.) and nitrogen sulphide (— 31'9
Cal.), the condition of these bodies being almost comparable as
regards the nitrogen sulphide and selenide, and the heats
absorbed forming a sort of arithmetical progression at the rate
of about 10 '5. In all cases they increase in absolute value
with the equivalent, in accordance with a relation that is pretty
general among the series of similar compounds,2 such as the
series of chlorine, bromine, and iodine ; the series of nitrogen,
sulphur, and selenium, etc. It follows that in such series the
explosive character of the endothermal compounds becomes more
and more pronounced in proportion as their atomic weight is
greater.

1 " Bulletin de la Socie'te' Chimique," 2" se*rie, torn, xxxviii. p. 548.

2 " Annales de Chimie et de Physique," 5" seVie, torn, xxxii. p. 391.

( 264 )

CHAPTEE VIII.

HEAT OF FORMATION OF COMPOUNDS DERIVED BY THE ACTION OF
NITRIC ACID UPON ORGANIC SUBSTANCES.

§ 1. GENERAL KEMARKS.

1. A LARGE number of artificial compounds result from the
association of organic principles with nitric acid. These com-
pounds are generally explosive, and they play an important
part both in warfare and in mining industry. In order to
estimate their explosive force, it is necessary to know the heat
disengaged in their decomposition. In fact, the explosive force
of nitro-carbon compounds results from a kind of internal com-
bustion analogous to that of ordinary gunpowder, from which,
however, it is distinguished by the fact that the nitric acid and
combustible principle are intimately combined, instead of being
simply mixed together, as in the case of ordinary gunpowder.
This force is greater in proportion as the combustion develops
more gas and more heat. Now, if all else be equal, the heat
disengaged by the combustion will be inversely proportional
to that disengaged by the previous union of the nitric acid
with the organic principle.

2. The heat disengaged in the formation of the following
more important nitrated compounds by means of nitric acid
was determined — nitric ether, nitroglycerin, nitro-mannite, gun-
cotton, nitro-cellulose or xyloidin, the nitro, dinitro, and chloro-
nitrobenzene, and nitrobenzoic acid. The heat of formation of
trinitrophenol, otherwise called picric acid, and of its salts,
was deduced from calculations based on certain analogies
which have just been confirmed by some experimental de-
terminations of Sarrau and Vieille. In 1871 Troost and
Hautefcuille had published, a few days after the author's
communication, some measurements relating to the heat of
formation of various nitrated derivatives, the results agreeing
very closely.

HEAT OF FORMATION AND TOTAL COMBUSTION. 265

3. We may, then, arrive at the heat disengaged in the forma-
tion of nitrated compounds from pure nitric acid and organic
principles, such as alcohol, benzene, phenol, glycerin, mannite,
cellulose, etc. But this quantity does not enable us to calculate
the heat given off by their explosive decomposition, even if we
know exactly the products of this decomposition. It is neces-
sary, in addition, to have the heat of formation of these
products from their elements, together with that of nitric acid,
water, and the original compound that gave rise to the nitrated
body.

The products of the explosive decomposition of nitrated com-
pounds are generally simple, e.g. water, carbonic acid, and
nitrogen ; these three being the only substances produced in a
complete combustion, such as that of nitroglycerin or nitro-
mannite. But in incomplete combustion, where the oxygen is
deficient, as in that of gun-cotton, we get also carbonic oxide,
hydrocyanic acid, hydrogen, marsh gas, occasionally oxides of
nitrogen, etc. The heat of formation of all these substances
should be known beforehand.

In fact, the heat of formation of all these compounds has
already been given (pp. 128, et seq.), together with that of nitric
acid from its elements. With regard to the original generator
of the nitrogenous body, its heat of formation may be determined
by its total combustion in oxygen, or by various other processes.
For all the substances enumerated above, the heat of formation
will be found in the thermo-chemical tables (pp. 136, 137). We
will give an example, in order to make this clear. Let us
estimate the heat disengaged in the combination of the elements
of nitric ether. For this purpose we add the heat disengaged in
the formation of alcohol, C2H60 (+70 Cal.), to that disengaged
in the formation of nitric acid, HN03 -f 41*6 Cal., and then to
the sum we add that disengaged in the reciprocal reaction of
these two bodies (+ 6*2 Cal.), which reaction produces nitric
ether. The sum of these three quantities, minus the heat of
formation of the water eliminated in the reaction (H20), i.e.
69 Cal., gives the quantity, + 49*& Cal., which represents the
heat disengaged by the combination of the elements of nitric
ether. On subtracting this quantity from the heat disengaged
by the pure and simple combustion of the said elements by
means of free oxygen, we get the heat of total combustion of
nitric ether in free oxygen, or + 311'2 Cal.

4. In this way were calculated both the heat of formation
from the elements, and the heat of total combustion, of nitric
ether, nitroglycerin, gun-cotton, and nitrobenzene.

The heat liberated "by their explosive decomposition can be at
once deduced, provided that we know exactly the real equation
representing this decomposition. It is also necessary to take
into account, in the calculations, the conditions of the decom-

266 COMPOUNDS DERIVED FROM NITRIC ACID.

position ; for the figures are not the same when we are working
at constant pressure, as in the open air, as when we are working
at constant volume, as in a bomb shell or other closed vessel.
The rule for determining corrections of this nature has already
been given (p. 15).

5. Conversely, if we know the heat disengaged by the decom-
position of an explosive substance, in a closed vessel, as well
as the exact nature of the products, it is easy to deduce the
heat of formation of the nitrated compound from its elements.
Sarrau and Vieille have followed this method. It furnishes us
with a check on the results obtained by the direct method, as
the two series of data should at least agree within the limits of
error allowed in experiments of this kind.

6. Sometimes the products of the decomposition are either
not well known, or are too complicated, or imperfectly defined
as to their physical condition — as in the case in which are
formed carbonaceous substances still retaining nitrogen, hydrogen,
oxygen, etc. This is what happens, for instance, with diazo-
benzene nitrate and with the picrates.

In cases of this kind, the heat developed by the explosion is
always a useful quantity to measure, but it cannot be calculated
a priori.

7. For a great number of applications it is necessary to
measure the heat of formation of such explosive compounds
from their elements. We then have recourse to a general
method, which consists in causing the body to explode in an
atmosphere of pure oxygen. This converts it entirely into water,
nitrogen, and carbonic acid. Calculation then becomes easy.
This method was employed for diazobenzene nitrate; Sarrau
and Vieille also adopted it for the picrates.

8. Instead of oxidising the body by free oxygen, we may do
so by means of an oxidising compound. This is frequently
done in practice, such as when gun-cotton or the picrates are
mixed with potassium nitrate, ammonium nitrate, potassium
chlorate, or even sometimes certain metallic oxides.

Under these circumstances it is convenient to calculate the
heat of combustion of the hydrocarbon compound, taking into
account the heat of formation of the oxidising body, according
to the table on page 134.

The calculation is easy if the oxidising substance be potassium
chlorate, each equivalent of oxygen supplied entailing a supple-
mentary disengagement of T83 Cal.

With ammonium nitrate, the additional energy is enormous,
amounting to + 25'05 Cal. per equivalent of oxygen.

With potassium nitrate the calculation is somewhat more
complicated, on account of the alkali present, which may change
into carbonate or sulphate, according to circumstances. Let
us take, for example, a compound containing carbon in sufficient

COMPLETE COMBUSTION BY AN OXIDISING AGENT. 267

quantity to convert all the potash into potassium carbonate.
The five equivalents of oxygen supplied by the potassium nitrate
will give off, then, 27 Cal. less than if they were free, and
generated with the carbon, free carbonic acid; or 54 Cal.
for each equivalent of oxygen. We may even add that this
estimate is not quite exact whenever cooling takes place in an
atmosphere of carbonic acid and aqueous vapour, because these
convert the neutral carbonate, K20, C02, into the bicarbonate ;
K20, H20, 2C02, causing a complementary disengagement of
248 Cal. (beginning from liquid water). Consequently, the excess
of heat developed during the combustion of a hydrocarbon
compound, in free oxygen, over that developed by the same
combustion by means of potassium nitrate, is reduced to 14*6
Cal. only for KN03 (= 101 grins.), i.e. to 2*9 Cal. for each
equivalent of oxygen employed.

9. We will add that, in cases where the combustion of the
explosive body is rendered complete by the addition of an
oxidising agent, we must not forget that the weight of the latter
is added to that of the explosive substance ; so that a gramme
of the mixture, subjected to total combustion, may give off less
heat than a gramme of the explosive body decomposing
separately in pursuance of a less complete oxidation. Various
compensations may be made with respect to this. For instance,
when it is required to make up one kilogramme of an explosive
mixture, copper oxide is the most efficacious of the oxides in
use, on account of the smallness of its equivalent (79*2 grms.)
and the comparatively low value (38*4 Cal.) of its heat of
formation. Lead oxide presents the double inconvenience of an
equivalent three times as high (223 grms.) and a greater heat
of formation (51*0 Cal.), which diminishes in a corresponding
degree the heat given off in combustion in which it is the
agent.

The oxides of mercury and silver present, on the contrary,
smaller heats of formation (31-0 Cal. and 7'0 Cal.). But the
thermal increase resulting from this is counterbalanced with
the unit of weight, by the magnitude of their equivalents (216
grms. and 232 grms.).

The oxides of tin and antimony, which are, for a given
weight, somewhat richer in oxygen than copper oxide, have
heats of formation that are, for each equivalent of oxygen,
almost double that of the latter.

It has been thought advisable to give these numbers, because
they render definite and correct many of the current ideas on
combustion by means of metallic oxides. We see that prefer-
ence should be given to copper oxide on account of the smallness
of its equivalent If lead oxide, and particularly the oxides of
mercury and silver, seem to be more powerful, it is no doubt
because they react and decompose at a lower temperature ; a

268 COMPOUNDS DERIVED FROM NITRIC ACID.

circumstance which enables the reaction to commence and
continue with greater vigour. Thus the explosives which they
form act with greater violence. But their useful effects, as
regards both work and pressure, are much less, even in the case
of silver oxide, which is so easily decomposed, and of mercuric
oxide, which, on the other hand, furnishes a gaseous metal.

10. We will conclude with one remark. When the explosive
substance is an acid, such as picric acid, its salts, already
existing as such, will produce a less useful effect than
simple mixtures of picric acid and metallic oxide; for their
formation involves, at the moment of the union of the acid with
the oxide, a liberation of heat, i.e. a loss of energy. But, on the
other hand, simple mixtures will be more dangerous, less stable,
and also subject to spontaneous explosions, owing to the possible
combination of the acid with the metallic oxide.

11. We have just calculated the heat of formation of a
nitrated derivative, supposing that of the generator to be known.
Conversely, if we know the heat of formation of a nitrogenous
body from its elements, together with that of nitric acid and
water, and also the heat disengaged in the reaction of the
nitric acid on the original generator of the nitrogenous body,
the heat of formation of this original generator can itself be
calculated. We may observe that this method is less direct
than the immediate combustion of the last compound ; therefore
the results are less exact They are, however, useful as
checks.

12. Such are the general conclusions that can be deduced
from the measurement of the heat disengaged by the combina-
tion of nitric acid with organic compounds. These having been
given, the experiments of the author, dating from 1871, will
now be described.

First of all, it will be remembered that the action of nitric
acid on organic substances gives rise to compounds of two dis-
tinct kinds, formed according to a similar equation and with a
similar separation of the elements of water ; the one kind con-
sists of true ethers, capable of being decomposed by alkalis
with regeneration of nitric acid and alcohol, whereas the other
kind, designated specially by the name of nitro- compounds, can
no longer be split up by distinct reactions, so as to reproduce
the generating substances, which are, in the most simple cases,
nitric acid and a hydrocarbon. The cause for this difference of
reactions will be explained later on. The ethers themselves
are divided into two groups, according to whether they are
formed from true alcohols, simple in their function, or from
alcohols of mixed function, such as cellulose, or condensed ether,
derived from several molecules of glucose, which is itself an
aldehydic alcohol.

The heat of formation of several bodies belonging to these

NITROBENZENE. 269

three groups was measured and found to differ according to the
diversity of functions of the bodies experimented on.

§ 2. NiTRO-COMPOUNDS IN GENERAL.

Nitro-compounds result from the action of nitric acid upon
organic substances, with separation of water. For instance,
C6H6 4- HN03 - H20 ; one, two, three, or four equivalents of
nitric acid may thus enter into combination with either a hydro-
carbon, an acid, an alcohol, an alkali, etc. Compounds of this
class are formed principally in the aromatic series, i.e. in the
series of compounds derived from benzene, or rather, condensed
acetylene. Up to the present the terms of this series are the
only ones which have been employed in connection with ex-
plosive substances ; they are also the only ones that have been
studied by the author.

It will be recollected that when nitro-compounds are treated
with alkalis, they do not reproduce nitric acid, but various
bodies of a special character, and nitrogenous, like the generators
themselves. Treated with reducing agents, nitro-compounds do
not reproduce the original body, but an amide.

1. Nitrobenzene, C6H5N02.

1. The reaction for the production of this compound is as
follows :— C6H6 + HN03 = C6H6N02 + H20.

This was performed in a little platinum cylinder, floating in
a platinum calorimeter containg 500 grms. of water. The same
conditions were observed as in the calorimetric experiments.
The density of the acid used was 1*5, and its composition corre-
sponded to the formula, HN03 + -335H20.

Fifteen grms. of this acid were poured into the little platinum
cylinder, which was then closed with a cork coated with
paraffin. The temperature of the water in the calorimeter was
taken by means of a thermometer sensitive to ^J<j of a degree ;
and the temperature of the nitric acid with a smaller thermo-
meter sensitive to ^ of a degree.

The two temperatures being made to agree, the cork was then
removed, and the benzene allowed to drop into the nitric acid
through a pipette having a very tapering mouth, and only allow-
ing exceedingly small drops to pass through. During this
procedure the acid was continually stirred, so as to mix it
gradually with the benzene ; the water in the calorimeter was
also stirred. In this way a known weight of benzene was intro-
duced— 1-835 grm. and 3*670 grms. respectively in two different
experiments — the operation of pouring in lasting, in all, two
minutes.

The cylinder was then corked up and worked through the

270 COMPOUNDS DERIVED FROM NITRIC ACID.

water of the calorimeter, being pushed along by means of the
large calorimetric thermometer, which also served to agitate the
water at the same time. The progress of this thermometer was
followed, also that of the small thermometer immersed in the
acid. At the end of six minutes, the two thermometers gave
readings agreeing within about one-tenth of a degree, which
difference represented the excess of the temperature of the acid
over the water in the calorimeter ; the variation of temperature
in the two experiments being 170° and 3-45° respectively.
Lastly, the rate of cooling was noted.

The following data were then known. On the one hand, the
weights of the water, the platinum, and thermometer reduced to
units of water, and also their variation of temperature ; on the
other hand, the weights of the acid and benzene, and also the
thermal variation involved by their combination, which had
converted the benzene into nitrobenzene, with the simultaneous
production of water.

The heat communicated to the water, platinum, and thermo-
meters may easily be calculated. But an exact calculation of
the heat communicated to the mixture of acid and nitrobenzene
would require a knowledge of its specific heat. Now, it is
sufficient to know that this specific heat approximates pretty
closely to '47, which is the same as that of the acid employed.
Thus, in the two experiments in question, the mass of acid and
nitrobenzene, reduced to units of water, will be from about 8*5
to 9 '5 grms., amounting to about one-sixtieth of the entire
heated mass. This fraction is so small as to be of slight im-
portance in the calculation of the heat disengaged. Thus the
latter can be estimated within the limits of experimental
error without its being necessary to measure more exactly
the specific heat of the mixture.

We may thus make a complete calculation of the heat dis-
engaged in the reaction that has taken place in the calorimeter.
It is brought by calculation to an equivalent of nitrobenzene,
i.e. Q, for the weight, C6H5N02 = 123 grms.

The compound formed under these conditions is really nitro-
benzene. To make sure of this, it was precipitated, after the
experiment, by means of water, and its density taken, which
was found to be equal to 1194 at 14°. Now, Kopp has
given the value 1187 at the same temperature. The differ-
ence, therefore, is so slight that the reaction may be accepted as
true. This reaction, however, under the conditions of the
author's experiment, is complicated by two circumstances,
which must be taken into consideration. On the one hand, the
nitrobenzene remains dissolved in the excess of acid ; and on
the other, the reaction itself gives rise to water which must give
off a certain amount of heat, owing to its combination with the
excess of acid. In order to be able to bring in this last factor,

HEAT OF FOKMATION OF NITROBENZENE. 271

a special series of experiments was made for the purpose of
measuring the heat disengaged by the same acid, when treated
with certain proportions of water, which are increased from a
limit below that produced in the experiments given to one a
little above it. These experiments were carried out at the
same temperature, under the same conditions, and on the same
day — each trial being twice repeated. In this way were
obtained two pairs of results, from which it was easy to trace
the curve representing the heats of hydration of the acid, be-
tween limits which comprised the hydration in the preparation
of the nitrobenzene. The heat, q, corresponding to the propor-
tion of water formed at the same time as the nitrobenzene may
thus be calculated.

2. Lastly, there was selected from amongst these mixtures the
one that agreed best with the final data of the experiment re-
lating to the formation of the nitrobenzene. In it was dissolved
pure nitrobenzene, in the same relative proportions ; the heat of
solution was very small. It was brought, by calculation, to the
data of the experiment relating to the formation of nitrobenzene,
which gave a value, qlt for the weight, C6H5N02. In short, the
number,

Q - 2 - <?i>
represents the heat disengaged in the following reaction : —

C6H6 + (HN03 + -335H20)
= C9H5N02 + H20 + -335H20.

The numbers found in the two experiments were, + 35 and
4- 35*2, average 35*10. In order to make this number apply to
true monohydrated acid, HN03, we must add to it the heat
given off in the reaction of '335H20 upon this last acid ; or
-h 1*5, according to the author's experiments.1

3. We get then, finally—

C6H6 (pure) + HN03 (pure) + H20 disengages + 36'6 Cal.

4. It is easy to deduce from this the heat of formation of
nitrobenzene from its elements —

C6 (diamond) + H5 + N + 02 disengages + 4'2.
In short, it was found that —

Benzene, C6 (diamond) + H6 = C6H6 (liquid) ... +5-0
Nitric acid, H + N + 03 = HN03 (liquid) ... + 41-6

Reaction ... ... ... ... ... -|- 36-6

Sum + 73-2

1 " Annales de Chimie et de Physique," 5e seVie, torn. iv. p. 448.

272 COMPOUNDS DERIVED FROM NITRIC ACID.

On the other hand —

C6 + H6 + N + 02 = C6H6N02 (liquid) x

H2 + 0 = H2 0 (liquid) +69

Sum ... + 69 + x

whence, x = + 4*2 for 123 grms.

5. Decomposition ly heat. We know that nitrobenzene
is not, properly speaking, an explosive substance. It may be
distilled at a certain temperature. If, however, it is subjected
to great heat, a powerful reaction is effected between the oxygen
of the nitrous molecule and the hydrocarbon elements of the
benzene molecule. But the products of this reaction are im-
perfectly known.

6. The heat of complete combustion of nitrobenzene is cal-
culated from the above data ; that of the elements being —

12C + 1202 = 12C02 4-564-0

i(5H2 - 50 = 5H20) + 172-5

+ 736-5

On subtracting the heat of formation of nitrobenzene + 4'2 we
get—

J (2C6H5N02 liquid + 250 = 12C02 + 5H20 liquid + ET2) gives
off + 732-3 Cal.

This weight relates to 123 grms. For 1 grm. we should get
5952 cal.

2. Dinitrobenzene, C6H4(N02)2.

This substance was prepared by dissolving a known weight
of nitrobenzene in nitrosulphuric acid. The apparatus was the
same as for nitrobenzene, and the experiment was performed in
exactly the same manner. In the platinum cylinder were placed
35 grms. of a mixture previously prepared from 1500 grms. of
nitric acid similar to that already described, and 2944 of boiled
sulphuric acid.

In these 35 grms. of nitro-sulphuric acid were dissolved : in
one experiment, T262 grm., and in another 2'534 grms. of
nitrobenzene. The elevations of temperature were 73° and T44°
respectively.

It was proved that the nitrobenzene was entirely converted
into dinitrobenzene. The calculations and corrections for
obtaining the quantity Q were made as previously (p. 270).
'The calculation of q (p. 271) is somewhat complicated. In
fact, the formation of the dinitrobenzene, in this case, produces
two phenomena : it changes the hydration of the nitrosulphuric
acid and also alters the relation between the nitric and sulphuric
acids, causing the latter to predominate, as a portion of the
nitric acid disappears, owing to the fact of the combination. In

HEAT OF FORMATION OF DINITROBENZENE. 273

order to estimate correctly the influence of these two effects
which would enter into subsequent operations, it was necessary
to make several series of experiments. In the first place, it
was convenient to measure directly, two experiments being
made in each case, the heat disengaged by the mixing of the
nitric acid

(HN03 - -335 H20)

with the boiled sulphuric acid, in four different proportions,
chosen so as to comprise within their limits all the cases
possible in the experiments which were performed. In this
way the curve was obtained for the quantities of heat produced
for the whole series of intermediate mixtures.

Then considerable quantities of each of these liquids were
prepared and proportions of water added to them, increasing
according to distinct ratios, which also comprised within their
limits all the cases possible in the experiments. Each time the
heat disengaged was measured, and curves constructed for the
heats of hydration of these various systems of mixtures.

Thus were obtained the elements necessary for calculating by
interpolation the quantity q, in all cases included within the
limits of the experiments.

This method is somewhat tedious, but it seemed to be the
most suitable for the object in view, viz. the study of a series
of analogous formations. If, however, there were only one
experiment of this kind to make, it would be preferable to
measure the heat given off in three cases only, viz. the mixing
of the two acids in their initial proportions; the mixing of
the two acids in their final proportions, in which they
exist after the performing of the experiments; and lastly,
by the addition of water (in the proportion furnished by this
experiment) to the mixture of the two acids corresponding to
the final proportions.

Lastly, the quantity, q1 (p. 271), was measured directly, by
dissolving a known weight of crystallised dinitrobenzene in a
mixture of the two acids and water of proportions similar to
those of the final condition of the liquid, that gives rise to the
dinitrobenzene. This quantity is negative, as generally happens
when solid bodies are dissolved. It was found equal to - 2 -6 9
for C6H4(N02)2.

We thus arrive definitely at the quantity

Q - q. - ft-

But this quantity relates to the formation of dinitrobenzene by
means of the nitrosulphuric acid of the experiments. In order
to apply the reaction to pure nitric acid, we must, in addition,
take into account the heat given off by the previous combina-
tion of the two acids, and also that by the union of HN03 with
•335 H20.

T

274 COMPOUNDS DERIVED FROM NITRIC ACID.

2. It was found, after making all the necessary calculations,
that the theoretical reaction —

C6H5N02 (pure) + HN03 (pure) = C6H4(N02)2 (crystal.) + H20,
disengages -J- 36*45 and + 36*35 ; average, 4- 36*4.

This result is to all intents the same as in the formation of
mononitrobenzene ; + 36'6 ; or, in other words, the heat dis-
engaged is proportioned to the number of equivalents of acid linked
on to the hydrocarbon.

The complete formation of dinitrobenzene, starting from
benzene —

C6H6 + 2HN03 = C6H4(N02)2 + 2H20,

would give off + 73.

3. These numerical values show that the formation of nitro-
compounds involves a considerable loss of energy ; it is much
greater than that entailed by the formation of nitric ethers, as
we shall presently show.

We can therefore understand why the explosive energy of the
latter compounds is greater, and their stability less. We can
also understand why nitro-compounds do not act like ethers,
the latter being capable of decomposition by potash with re-
formation of an alcohol and acid. Potash, which, when
combined with dilute nitric acid, gives off only 137 Cal.,
cannot furnish, by a simple reaction, the energy required for the
reproduction of the acid and benzene, the union of which, in
order to form nitrobenzene, has. disengaged 36*5 Cal. This
energy, on the contrary, is available in the case of nitric ether
and nitroglycerin, which require only 4 to 6 Cal. for the re-
generation of each equivalent of acid.

4 Moreover, the figures -f 36 '5, relating to nitrobenzene, are
worthy of notice from another point of view. In fact, this
quantity is approximately three-quarters of the heat disengaged
in the action of hydrogen on dilute nitric acid, with the
formation of nitrous acid, which remains in solution.

H2 + HN03 (dissolved) = H20 + HN02 (dissolved) gives off

+ 50-5.

In this reaction, the action of the hydrogen is, in certain respects,
similar to that of benzene in the formation of nitrobenzene.

5. This shows that the formation of nitrobenzene and similar
substances may be compared to oxidation.

On the other hand, the formation of nitric ether and nitro-
glycerin, which causes the liberation of much less heat,
represents a simple substitution of the elements of the acid for
the elements of water.

6. The decomposition of nitrobenzene may be effected by
sudden heating ; but the products have not been studied.

NITROBENZOIC ACID. 275

7. The formation of dinitrobenzene from its elements —

C6 (diamond) + H4 + N2 + 04 = C6H2(N02)2 (dissolved)
gives off 12-7 for 168 grms.

8. The heat of complete combustion of dinitrobenzene
(= 168 grms.)—-

C6H4N204 + 010 = 6C02 + 2H20 + N2,

gives off + 689 '3 Cal. for 168 grms., which amounts, for
1 grm., to 4103 cal. All these calculations were made for
dinitrobenzene obtained without heat ; in the action of dinitro-
benzene on nitric acid, and without having regard to the mixture
of isomeric substances produced in this case. The observations
which have been published1 tend, moreover, to show that
various isomeric substances of the same chemical function are
formed, causing disengagements of heat almost identical.

3. Chloronitrdbenzene, C6H4C1(N02).

The formation of this compound takes place according to the
following equation : —

C6H5C1 + HN03 = C6H4C1(N02) + H20.

It was found that this reaction gives off + 36*4.

We know that several isomeric substances are formed. The
heat of solution in a mixture similar to that formed in the
reaction was determined.

The details of these experiments may be omitted, as they are
similar to those already described.

The heat of chlorination of benzene being unknown, it is not
possible to calculate the heat of formation of the above sub-
stance from its elements.

4. Nitrobenzoic Acid, C7H5(N02)02.

The formation of this compound takes place according to the
following equation : —

C7H602 + HN03 = C7H5(N02)02 + H20.
This reaction gives off -f 36*4.

We see that this value is nearly constant for the nitration
of benzene and all its immediate derivatives. The formation of
nitrobenzoic acid from its elements is easily calculated if we
admit for the heat of formation of benzoic acid the value -f 54
(Kechenberg). We then get —

C7 (diamond) + H5 + N + O4 = C7H5(Isr02)02 - + H2 (liquid)
gives off + 63 Cal. for 167 grms.

The "heat of complete combustion of the same substance
= 761-5 Cal. for 167 grms., or 3772 cal. for 1 grm.

1 " Bulletin de la Soci&e* Chimique," 2e sfrie, torn, xxviii. p. 530.

T 2

276 COMPOUNDS DERIVED FROM NITRIC ACID.

5. Nitro-derivatives of the Aromatic Series in general.

1. It has just been shown that the formation of a nitro-
compound, belonging to the aromatic series, is generally
accompanied by a liberation of heat approximately = +36 Cal. ;
this number was also obtained by Troost and Hautefeuille for
the derivatives of toluene and naphthalene. It will be shown
presently that it also holds for the formation of trinitrophenol,
otherwise called picric acid.

2. This being admitted, it is easy to give general formulae for
calculating d, priori the heat of formation of a nitro-compound
from its elements, and also its heat of combustion, provided
that we possess these data for the original hydrocarbon.

Let A be the heat of formation of the generating substance ;
4- 41'6 Cal. being that of nitric acid ; -f 364 the heat of
nitration ; and lastly + 69 the heat of formation of water ;
the equation representing nitration is as follows : —

K + HN03 = X + H20,

and from it we arrive at the expression for the heat of formation
of the nitro-compound, X, or

A + 41-6 + 36-4 - 69 = A + 9 Cal.
For a binitrated, trinitrated, etc., compound, we shall get —
A + 18 Cal. ; A + 27 Cal ; and generally, A + 9rc.

3. In the same way, the heat of complete combustion of a
nitro-compound is deduced from that of the original hydro-
carbon. The latter being supposed = Q ; that of the mononitro-
compound, which contains one equivalent less of hydrogen, will
be Q - 34-5 - 9 = Q - 43'5 ; for a dinitro-compound, Q - 87 ;
for a trinitro-compound, Q — 130 '5. These formulae must only
be regarded as approximate, as the effect of the nitration is
often complicated by the change of physical condition, which
should be taken into consideration separately.

4. The large quantity of heat liberated in the formation of
nitro-compounds, when using pure nitric acid, enables us to
understand the formation of the same compounds when using a
mixture of nitric and sulphuric acids. We know for a fact that
this mixture is employed, in preference to pure nitric acid, for
the preparation of nitro-derivatives ; but this is an empirical fact.

The theoretical explanation of it may be given; it results
from the difference between the heat of formation of sulphuric
derivatives and that of nitro-derivatives, joined to the tendency
of sulphuric acid to form a secondary hydrate with the water
resulting from the formation of the nitro-compound. For
instance, the formation of benzene-sulphonic acid—

C6H6 + H2S04 = C6H6S03 + H20, gives off + 14'3 - a;1

1 a represents the heat of solution of benzene-sulphonic acid in water ; a
positive quantity amounting to a few Calories.

PICRIC ACID AND PICRATES. 277

whereas that of nitrobenzene —

C6H6 + HN03 = C6HJlsr02 + H20, gives off + 36'6.
The difference between these two quantities, -f 22 '2 4- a, is
enormous and cannot be compensated, either by the difference
in the quantities of heat disengaged by the union of H2O with
the excess of nitro-sulphuric acid, in the two experiments, or
by the difference in the respective heats of solution, in the same
liquid, of nitrobenzene and benzene-sulphonic acid. The differ-
ence is further increased by the heat of formation of the
secondary sulphuric acid hydrate. Thus the formation of nitro-
benzene gives off much more heat than that of benzene-sul-
phonic acid ; the formation of the nitro-derivative, in preference
to a sulphuric derivative, is therefore a natural consequence
of the general principles of thenno-chemistry.

6. Trinitrophenol, or Picric Acid and its Salts.

1. Let us apply these formulae to picric acid. This acid is
derived from phenol, by the replacement of three atoms of
hydrogen —

C6H60 + 3HN03 = C6H3(N02)30 +3H20.

Now the heat of formation of phenol may be estimated either
at + 34 Gal., or at -f 28 Cal., according to whether we adopt
the heat of combustion of Favre and Silbermann (737) or that
of M. Eechenberg (743), the difference between which values
does not amount to quite one-hundredth.

We will take the mean, 31 Cal., for an equivalent, 229 grms.
This being allowed, the heat of formation of picric acid from its
elements, C6 (diamond) -f H3 + N3+07, will be + 31 -f 27 =
+ 58 Cal. for 229 grms. ; the heat of combustion being + 609'5 Cal.,
according to our formulae.

2. It is easy to proceed from this to the heat of formation of
picrates. Let ammonium picrate be

C6H2(N02)30]SrH4 = 246 grms.

According to the calculations,1 the formation of this body by
means of pure acid and ammonia gas —

C6H3(N02)30 (solid) + NH3 (gas),

disengages -f 22*9 Cal., which gives, for the heat of formation of
the salt from its elements, for 246 grms. —

C« + H6 + 2N2 + 07 ; + 58 + 12'2 + 22-9 = + 831 Cal.

Messrs. Sarrau and Vieille 2 found + 80*1 Cal. for combustion in
oxygen, a value agreeing with the former within the limit

1 Table v. p. 127.

2 "Comptes rendus des stances de 1'Acad^mie des Sciences," torn, xciii.
p. 270.

278 COMPOUNDS DERIVED FROM NITRIC ACID.

of experimental errors ; as the difference does not amount to
half per cent, of the heat of combustion. In fact, the heat
of total combustion of this salt is, according to calculation 4-
688 Gal., according to experiment + 691, for 246 grms., or,
for 1 grm., 2797 cal.

3. We now come to potassium picrate —

C6H2(N02)3KO = 267 grms.
According to table iv., p. 127, the reaction of the acid and base —

C6H3(lSr02)30 (crystal.)^ KHO (solid)
= C6H2K(N02)30 (solid) + H20 (solid), gives off + 30-5 Cal.

Admitting that K + H + 0 = KHO gives off + 104'3, we get
for the heat of formation of potassium picrate from its elements,
for 267 grms.—

C6 + H2 + K + N3 + 07 ; + 58 + 104-3 + 30'5 - 704 =
+ 1224 cal.

Sarrau and Vieille gave, for combustion in oxygen, 4- 117*5
Cal. The difference in these values amounts to less than
one-hundredth of the total heat of combustion, thus being
within the limits of error ; more so when we take into account
that the action of the water, formed in the combustion, on the
the potassium bicarbonate has been disregarded by these writers
in their calculation, as well as the partial dissociation of the
last-named salt.

4. The heat of total combustion of potassium picrate, with
formation of potassium bicarbonate, amounts to 61947 Cal., or,
for 1 grm., 2321 cal.

5. The explosive decomposition of potassium picrate gives rise
to complex products : carbonic acid, carbonic oxide, hydrocyanic
acid, free hydrogen, nitrogen, marsh gas. The relative propor-
tion of these bodies varies with the conditions.

Thus carbonic acid and marsh gas increase with the pressure,
at the expense of the carbonic oxide and hydrogen.

As to the solid residue, it is composed of potassium carbonate
and cyanide containing, according to Sarrau and Vieille, the
third of the alkaline metal, with a small quantity of carbon.
With a density of charge of '5, the results observed by these
writers are represented approximately by the following empiric
equation : —

16C6H2K(N02)30 = 4KCN + 6K2C03 + 21C02 + 52CO +
6CH4 + 22N2 + 4H2 + 70.

According to this equation, an equivalent of potassium picrate
(267 grms.) would disengage, in decomposing, + 2084 Cal., or,
for 1 grm., 780 cal.

NITRIC ETHER. 279

§ 3. NITRIC ETHERS FROM ALCOHOLS PROPERLY so CALLED.

General Remarks.

1. Nitric ethers are obtained by the action of nitric acid upon
alcohols, accompanied by the substitution of the elements of
water for those of acid ; 1, 2, 3 to 6, and even more equivalents
of acid may take the place of H20, 2H20, 3H20 to 6H20, etc., in
the alcoholic molecule. For instance —

Nitric ether, C2H4(H20) + HN03 = C2H4(HN03) + H20.
Nitroglycerin, C3H2(H20)3 + 3HN03 = C3H2(HN03)3 + 3H20.

2. The equations representing the formation of nitric ethers
are analogous to those for nitro-compounds. But there is a
fundamental reaction that characterises the nitric ethers ;
namely, that they reproduce the acid and original alcohol, under
the prolonged influence of water and dilute alkalis, which does
not happen in the case of nitro-compounds. Eeducing agents
also decompose the nitric ethers with reproduction of the
original alcohol, whereas, in the case of nitro-compounds, the
same agents form compound ammonias.

3. These differences in reactions are correlative with the un-
equal quantity of heat given off in the action of nitric acid on
various organic compounds. If it gives rise to a nitro-derivative
(p. 276), it disengages on an average 4- 36 Cal., or, in the case of
an alcohol, properly so called, to an ether, it disengages 4- 5 to
4- 6 Cal., and 4- 1 1 Cal. at the most, in the case of complex
bodies with analogous functions, such as cellulose. It is this
that causes the greater instability of nitric ethers. The presence
of alkalis, or even moisture, is sufficient to cause a change in
them after a little while.

But this circumstance gives greater energy to nitric ethers in
their use as explosives ; the combustive energy of the nitric acid
being much less weakened at the time of its first combination
with the organic compound.

This being understood, we will now examine the thermal
formation of nitric ethers, beginning with those derived from
ordinary alcohol.

1. Nitric Ether, C2H4(HN03) = 91 grms.

The formation of this ether was effected in a calorimeter, in
a direct manner, by means of pure alcohol and nitric acid,
sp. gr. 1'5, and without the addition of any other auxiliary
body. The product is approximately the same as would be
expected from theory. The experiment, as has been said, can
be performed directly, but it is a very delicate operation.

It is effected in the apparatus already described (p. 269), by

280 COMPOUNDS DERIVED FROM NITRIC ACID.

letting pure alcohol fall, in exceedingly minute drops, into nitric
acid, which is pure and free from nitrous compounds. With
each addition the acid is stirred vigorously, in order to avoid
any local elevation of temperature. At the same time the
vessel containing the acid is moved about in the water of the
calorimeter, so as to cause the gradual absorption of the heat
disengaged.

These are essential conditions. When they are very scrupu-
lously observed, we succeed in avoiding all secondary reactions,
as well as any disengagement of nitrous vapours, and in con-
verting the alcohol entirely, or almost entirely, into nitric ether,
as we can prove by precipitating the mixture, immediately it is
formed, by means of water, and collecting and weighing the
ether produced.

The addition of urea to the pure nitric acid does not render
the experiment more successful ; but it is different when a less
concentrated acid is used, as in the usual method of preparation
of nitric ether.

The only essential condition is that the drops of alcohol
should be excessively small, and very rapidly mixed with the
mass, so as to avoid any local elevation of temperature, which
would promote secondary reactions.

The experiment does not always succeed, and it is better
only to take into consideration the calorimetric measurements
got by means of a successful reaction. On some occasions 7'6,
on others 15 grins, of nitric acid and "84 grm. of alcohol were
experimented upon by the author. After the reaction, the
products should immediately be poured into water, otherwise
a secondary reaction begins to manifest itself. The latter
reaction is also quickly developed when pure nitric ether, pre-
pared beforehand, is dissolved in pure nitric acid, an operation
which the author was compelled to perform in the calorimeter,
in order to complete the data of the calculations relating to
the formation of nitric ether.

2. After all calculations, it is found that the formation of
nitric ether —

C2H60 (liquid) + HN03 (liquid) = C2H4(HN03) (liquid)
+ H20 (liquid),

gives off -f 6'2 Cal. ; the bodies being supposed pure, separated
from each other, and taken at the ordinary temperature.

The heat of solution of nitric ether in water was also
measured :

C2H4(H]Sr03) (1 part) + 180 parts of water gives off + '99 ;
whence we get

C2H60 (in solution) + HN03 (in solution) = C2H4(HN03) (in
solution) + H20 + water absorbs - 3 '2 Cal.

NITROGLYCERIN. 281

We see that the thermal effect varies inversely with the dilu-
tion, just as in the case of ethyl-sulphuric acid, and those acids
allied to it.

The formation of nitric ether is, in this respect, analogous
to that of those from organic acids, in which case their pro-
duction causes absorption of heat, whether the bodies in ques-
tion be in solution or in a pure state.1

But, on the contrary, the formation of nitric ether from con-
centrated acid gives rise to disengagement of heat. This
opposition results from the great difference of energy existing
between nitric acid in the pure state and that diluted with
water.

3. The formation of nitric ether from its elements —

C2 (diamond) + H5 + N + 03 = C2H4(HN03) (liquid),

gives off + 49-3 Cal. for 91 grms., or, for 1 grm., 542 cal.

4. Decomposition. — Nitric ether may be distilled with great
regularity, but care must be taken to avoid all local overheating.
The approach of a flame, or even a temperature of about 300°,
causes the ether to explode with violence. A terrible accident,
which happened at a chemical works at St. Denis, has shown the
dangers attendant upon the handling of large quantities of this
ether. The products of this explosion have not been analysed.
The oxygen contained in the compound is, moreover, insufficient
to oxidise the carbon and hydrogen, even supposing the first
body to be converted only into carbon monoxide. Admitting the
following reaction —

C2H4(HN03) = 2CO + H20 + 3H + N,

the composition of the liquid ether, with the formation of liquid
water, would give off + 71 '3 Cal. for 91 grms. If the ether
and water were in the gaseous form, the figures would be
slightly different, amounting, for 1 grm., to 787 cal.

5. The heat of total combustion of nitric ether by means of
pure oxygen —

£[2C2H4(HN03) + 70 = 4C02 - 5H20 - N2],
gives off + 311-2 Cal. for 91 grms., or 3420 cal. for 1 grm.

2. Nitroglycerin, C3H2(HN03)3 = 227 grms.

1. Nitroglycerin was prepared in a calorimeter, by means of
nitrosulphuric acid, and under conditions similar to those
recently described by M. Champion ; conditions under which
the product amounts to only four-fifths of the theoretical value,
owing to unavailable secondary oxidations. Quantities of 1*201
grm. and T934 grm. of glycerin were experimented upon. It
1 " Annales de Chimie et de Physique," 5e serie, torn. ix. p. 344.

282 COMPOUNDS DEBITED FROM NITRIC ACID.

was contained in a little capsule, accurately weighed, and poured
drop by drop into the middle of the nitrosulphuric mixture.
When a sufficient quantity of glycerin had been poured out
the capsule was re-weighed ; the loss in weight showed the
quantity of glycerin introduced.

2. All necessary calculations having been made, it was found
that the ordinary reaction, i.e. the case in which the substances
are taken in their actual condition —

C3H803 + 3HN03 = C3H2(HN03)3 + 3H20,

gives off + 14'7 ; or -f 4*9 for each equivalent of acid that has
entered into combination.

These figures, which are rather below those obtained for nitric
ether, show that both the acid and the glycerin have preserved
almost all their reciprocal energy throughout the reaction, a
circumstance which explains the remarkably easy decomposition
of nitroglycerin and the formidable effects thereof.

3. Again, we find that

C3H803 (in solution) + 3H£T03 (diluted) = C3H2(HN03)3 (pure)
-f 3H20 (liquid), absorbs - 8'8, or - 2-9 X 3.

Therefore we have thermal inversion, arising from the solution
of the substances ; exactly as in the case of nitric ether. This is
another point of resemblance between nitroglycerin and ethers
formed from organic oxy-acids.

4. The heat of formation of nitroglycerin from its elements may
be calculated from its heat of formation, as deduced from the
heat of combustion which was observed by M. Louguinine. We
thus find

Ce (diamond) + H5 + N3 + 09 gives off + 98 Cal. for 227 grms.,
or 432 cal. for 1 grm.

5. The heat of total combustion and the heat of complete decom-
position are, in this case, interchangeable terms, since nitro-
glycerin contains an excess of oxygen —

i[2C3H2(HN03)3 = 6C02 + 5H20 + 3N2 + 0].

Sarrau and Vieille have verified the reality of this reaction.

From the preceding data, we find that the heat of combustion
is equal to + 356'5 Gals., or, for 1 grm., 1570 cal.

Sarrau and Vieille obtained -f 360 '5 Cal. ; a value agreeing
as nearly as could be expected.

Nitroglycerin is decomposed differently if it is ignited as
dynamite, i.e. an intimate mixture of silica and nitroglycerin,
and if the gases which are formed are allowed to escape freely,
under a pressure nearly equal to that of the atmosphere. Sarrau

1 "Comptes rendus des stances de I'Acad&nie des Sciences" torn, xciii.
p. 270.

NITKOMANNITE.

283

and Vieille obtained under these conditions, for 100 volumes of

gas—

NO 48-2

CO 35-9

C02 12-7
H 1-6

N 1-3

CH4

0-3

These conditions are similar to those under which a mining
charge, simply ignited by the cap, burns away slowly under a
low pressure ; this is called a miss-fire.

3. Nitromannite, C6H2(HN03)6 = 452 grins.

1. This substance was prepared by means of nitrosulphuric
acid. The reaction is slow and somewhat prolonged. One grm.
of mannite and 30 grms. of acid liquid were operated upon.
Assuming the reaction to have been complete, the numbers that
were observed gave + 23*5 Gal. for the reaction

C6H1406 + 6HN03 = C6H2(HN03)e + 6H20,

or -f- 3 '92 Gal. per equivalent of fixed nitric acid.

2. The heat of formation of nitromannite from its elements is
calculated from the above figures, together with the heat of
formation of mannite, as deduced from its heat of combustion
(760 Gal.), which was obtained by M. Eechenberg. We thus
find—

C6 (diamond) -f H8 -f N6 -f 018 gives off + 156-5 Gal. for
452 grms.

Sarrau and Vieille deduced from the heat of combustion
of nitromannite itself its heat of formation, + 165*1 Gal. for
452 grms., a value sufficiently close to the above if we take
into account the heats of combustion given below; for the
difference between the heats of combustion calculated and those
found by experiment does not amount to one-hundredth.

The heat of combustion of nitromannite is the same as its heat
of decomposition, this substance containing, like nitroglycerin,
an excess of oxygen —

C6H2(HN03)6 = 6C02 + 4H20 +*Na + 02.

This reaction gives off, according to calculation, 564 -f- 276
- 1561 = + 683-9 Gal. for 452 grms.

Sarrau and Vieille found directly 678'5 Gal., or, for 1 grm.,
1501 cal.

4. Heat of Formation of Nitric Ethers in general

1. It is desirable to treat here of ethers formed from true
alcohols, which have simple functions (p. 268).

284 COMPOUNDS DERIVED FROM NITRIC ACID.

According to the preceding data, the formation of a nitric
ether, by means of alcohol and nitric acid, would give off, on an
average, -h 5 Cal. for each equivalent of fixed nitric acid. This
quantity may be used to calculate the heat of formation and
the heat of combustion of nitric ethers that have not as yet been
studied.

2. Let us suppose an ether to be formed from an alcohol,
represented by the letter K ; the ether being —

K + fiHN03 - 7&H20.

The heat of formation of the ether from its elements will be
deduced from the heat of formation, A, of the alcohol by the
following formula: —

A -f 41-671 -f 5n - 697i = A- 22'4n.

It is lower than the heat of formation of the original body ;
a fact which distinguishes ethers from nitro-compounds (p. 276),
the heat of formation of which, on the contrary, exceeds that of
the original substance by + $n Cal. The difference, which is
31 '4 Cal. for each equivalent of fixed nitric acid, denotes the
excess of energy of a nitric ether over that of an isomeric nitro-
derivative, formed from the same original substance ; benzyl
nitrate, for instance, as compared with nitrobenzyl alcohol.

3. The heat of decomposition of a nitric ether can thus be
calculated a priori, if its products be known ; as in the case in
which the substance contains an excess of oxygen.

4. The heat of total combustion of a nitric ether is deduced in
all cases from that of the original alcohol. This being equal to
Q, the formula of the ether deduced from n equivalents of nitric
acid will contain riH. less, and its heat of combustion will be —

Q - 34-5/1 -f 22-4^ = Q - 12'ln.

If, for example, we take nitroglycerin (n = 3), we shall get
Q = 392'5 Cal., according to M. Louguinine's data for glycerin.
The heat of total combustion of nitroglycerin, calculated by
the formula, will then be -f- 356-2. Messrs. Sarrau and Vieille
found by experiment, + 360*5. The discrepancy amounts to
one-hundredth, and includes both the error made in the heat
of combustion of glycerin and also that of nitroglycerin.

5. In order to make these points clear, let us calculate, accord-
ing to the above formula, the formation of methyl nitrate —

C + H3 + N + 03 = CH2(HN03).

The formation of methyl alcohol from its elements, A, = 62 ; we
shall therefore get + 39 '6 for the formation of methyl nitratefrom
the elements.

The heat of total combustion of this ether will be —

+ 157-9 for 77 grms., or 2050 cal. for 1 grm.

NITRIC DERIVATIVES FROM COMPLEX ALCOHOLS. 285

Assuming the following equation to represent the explosive
decomposition of this ether —

i[2CH2(HN03) = C02 + CO + N2 + 3H20],
the heat disengaged would be + 123'8 Cal. for 77 grms. But
if we prefer to assume that the decomposition answers to the
formula —

J(2C02 + N2 + H2 + 2H20),

we shall get + 1241 Cal., which is, to all intents, the same.
This gives for 1 grm., 1602 cal.
6. Let us also take the formation of ethylene nitrate —

C2H2(HN03)2,

A, = 11T7, derived from the heat of combustion of glycol, as
observed by M. Louguinine. The quantity 4- 66*9 Cal. for
1 equivalent = 152 grms. thus expresses the heat of formation
from the elements.

The heat of decomposition will in this case be identical with
the heat of total combustion —

C2H2(HN03)2 = 2C02 + 2H20 + N2
gives off -f 258-8 Cal. for 152 grms., or, for 1 grm., 1956 cal.

Since it does not contain any excess of oxygen, ethylene
nitrate must therefore be an explosive substance with maximum
effect.

5. Nitric Derivatives from Complex Alcohols.

1. We may now proceed to nitric derivatives produced from
alcohols of complex function. The only ones that have been
studied from a thermal point of view are cellulose and its
isomers, which are alcoholic ethers, themselves derived from
glucose, an aldehydic alcohol.1

2. These compounds, when treated with water or alkalis, do
not decompose in a simple manner, i.e. so as to reproduce the
original nitric acid and cellulose ; but give rise to complex
reactions, which are imperfectly known, and in which the
aldehydic function seems to play a part.

On the other hand, when treated with reducing agents, so
as to cause the destruction of the nitric acid, they reproduce
the cellulose, which still retains its original properties.

3. The greater stability possessed by this class of nitric de-
rivatives, when treated with agents of hydration, corresponds,
as we shall show, to the greater heat of nitration, i.e. to the
more considerable loss of energy in the act of preparation.2

Only two derivatives of this order have been studied from a
thermal point of view, viz. gun-cotton and xyloidin.

1 See the author's " Traits' e'le'mentaire de Chimie organique," torn. i. p. 371.
1881. Dunod.

2 See the theorem on p. 123.

286 COMPOUNDS DERIVED FROM NITBIC ACID.

6. Nitrostarch (Xyloidin).

1. This body answers to the formula in the following equa-
tion : —

C6H1006 + HNOS = C6H804(HN03) + H20,

or rather, to a multiple of this formula, if we admit that starch
is itself a condensed body, derived from several molecules of
glucose —

Since the value of n is not definitely known as yet, it is con-
venient, for the sake of simplicity, to reduce the data to a value
of n = 1.

2. Nitrostarch was prepared from a mixture of dry starch and
nitric acid, sp. gr. 1*5. It was found that the reaction —

C6H1005 + HN03 = C6H804(HN03) + H20,

gives off 12 *4 Cal, the nitrostarch separating out in a solid
form.

This is almost the same value for each equivalent of fixed
acid as we get for gun-cotton.

It will be noticed that this value is double that got for nitric
ether and nitroglycerin, while it is only a third of the heat
disengaged in the formation of nitrobenzene. Gun-cotton and
xyloidin behave as substances intermediate between nitro-com-
pounds and normal nitric ethers ; they also resist alkalis far
better than nitric ethers.

3. The heat of formation of nitrostarch from its elements may
be calculated, if we admit, with M. Eechenberg, that the heat
of total combustion of starch is equal to + 726 Cal. ; its heat of
formation will be equal then to 183 Cal We shall find, then,
that

C6 + H9 + N + 07 gives off + 183 + 41-6 + 12-4 - 69 =

+ 168 Cal. for 207 grms.,
or, for 1 grm., 812 cal.

4. The heat of decomposition could only be calculated if the
products of this decomposition were given ; but they have not
as yet been studied, and the quantity of oxygen contained in
the compound is far from being sufficient for its complete com-
bustion.

5. The heat of total combustion is equal to 706'5 Cal. for 207
grms., or, for 1 grm., 3413 cal.

7. Pernitro-cellulose, or Gun-cotton.

1. This substance results from the action of nitric acid upon
cellulose, the latter being taken under the particular form
of cotton. Nitric acid replaces the elements of water of the
cellulose, without altering in any way its physical appearance.

GUN-COTTON. 287

Several compounds may in this way be formed, distinguished
from each other by the amount of nitric acid which they contain.
For the sake of simplicity they are generally classified under
three heads —

Mononitrocellulose C6H804(HN03)

Dinitrocellulose C6H603(HN03)2

Trinitrocellulose C6H402(HN03)3

but these proportions are not always strictly observed. As
a matter of fact, the formula for cotton is higher than C6H1005 ;
it is a multiple of this quantity. Moreover, the quantity of
nitric acid indicated by the third formula is somewhat higher
than the maximum quantity that is ever united to the cotton ;
in fact, the latter falls appreciably below this value, according
to most exact analyses and syntheses. As no other thermal
experiments have as yet been made with gun-cotton, we pro-
pose to discuss this compound in detail.

Admitting the formula of cellulose to be €241140020, the for-
mulae of gun-cotton that best represent that formed in the
experiments are the following: —

C24H20010(HN03)10) or C21H1809(HN03)U.

The slight difference between the two formulae is owing to
the small quantity of carbon retained in the ashes under the
form of carbonate, which is disregarded in the second formula.
The latter, however, seems, on the whole, preferable.

2. Gun-cotton was prepared in a calorimeter by means of
nitrosulphuric acid, and under the same conditions as those in
the preparation of nitrobenzene (p. 270). 1*188 and 1-241 grm.
of dry cotton were used. The reaction being prolonged, the
experiment was each time stopped at the end of twenty minutes.
The gun-cotton was then washed, dried, and weighed, which
gave the proportion of acid fixed. This proportion was found to
be somewhat below that corresponding to complete nitration,
but the experiment had not lasted long enough for this. In
each case 9 equivalents of nitric acid, instead of 10 or 11, were
fixed on to C24H40020.

From the results obtained, we calculate the heat given off to
be 102 Cal. for 9HN03; or, + 11*4 Gal. for each equivalent of
fixed nitric acid. We may, therefore, admit that the fixing of
11HN03, according to the formula

CU^A,, + 11HN03 - 11H20,

would disengage 4-125*4 Cal. ; or + 114 Cal. for the formula
QJ^Oao 4- 10HN03 - 10H20,

which represents the conventional composition of gun-cotton.

3. The value + 11 '4 is very near that of + 124 found for
nitrostarch, which justifies us to a certain extent in assuming

288 COMPOUNDS DERIVED FROM NITRIC ACID.

that for each nitric equivalent fixed on to a carbohydrate, a
heat of about -f- 12 Cal., on an average, is liberated. This
value, it may be repeated, is double that of the heat of formation
of the nitric ethers properly so called.

4. In order to deduce from this the heat of formation of gun-
cotton from its elements, it would be necessary to determine the
heat of formation of cotton itself, which is at present unknown.

5. Messrs. Sarrau and Vieille have measured the heat given off
in the decomposition of gun-cotton. As this varies with the con-
ditions, they give results for the decomposition that furnishes
the foUowing products, 15CO + 9C02 -f HH + UN + 9H20.
From this we deduce, for the heat of total combustion of gun-
cotton —

QM + H48 + UN + °42 (= 1143 grms.), the value -f 633 Cal.

On oxidising the gun-cotton by means of ammonium nitrate,
they obtained a result leading to + 698 Cal. The discrepance
in the two values shows the difficulty of carrying out experi-
ments which are of this nature, and are based upon complicated
reactions. The above figures may, however, serve as approximate
data until the discovery of a more definite method.

According to the first value, the heat of total combustion of
gun-cotton in free oxygen would be 562*5 cal. for 1 grin.

The heat of formation from its elements would be 624 Cal. for
1143 grms.

6. We will now say a few words about the explosive decom-
position of gun-cotton conducted in a closed vessel and at
constant volume ; this formed the subject of a carefully studied
and very interesting paper by Messrs. Sarrau and Vieille.1
They found that the volume of the gases (reduced to 0° and
760 mms.), and also their relative proportion, vary with the
density of charge, i.e. with the pressure developed at the
moment of the explosion. These are some of the results —

Density of charge ...... 0-01 0*023 0'2 0-3

Volume of gases (reduced) per

grm. of material ...... 658-5 670-8 682-4 _

/CO 49-3 43-3 37-6 34-7

Composition of the gases <*>• *$ J4-6 27-7 30-6

per 100 volumes g 127 172 184 17;4

VCH4

0-0 trace 0-6 1-6

From this table it follows that the quantities of carbonic acid
and hydrogen increase with the density of charge ; whereas that
of carbon monnade diminishes. We notice, moreover, the pro-
duction of an appreciable and increasing quantity of marsh gas.

1 " Comptes rendus des stances de 1' Acad&nie des Sciences " torn. xc.
p. 1058.

GASES FROM GUN-COTTON.

289

If we disregard it, the following formulae express these facts : —

-

Density 0-01
„ 0-023
„ 0-21
0-2

33CO + 15C02 + 8H2 + 21H20 + 11N2
3000 + 18C02 + 11H2 + 18H20 + 11N2
2700 + 21C02 + 14H2 + 15H20 + 11N2
2600 + 22C02 + 15H2 + 14H20 + 11N2

Thus, with low densities of charge, the reaction produces
volumes of carbonic oxide, carbonic acid, and hydrogen, which
are represented, to all intents, by the simple ratio, 4, 2, 1,
whereas, under greater densities, the quantities produced
approximate more and more clearly to the limit —

24CO + 24C02 + 17H2 + 12H20 + UN,.

We may assume that the last formula fairly represents the
mode of decomposition realised under ordinary conditions of
practice in which gun-cotton, with great densities of charge, is
used.

It will be observed that neither nitric oxide nor any other
nitrous vapours are produced in the explosive decomposition of
gun-cotton in a closed vessel.

7. It is otherwise when the gun-cotton is ignited by means
of a red-hot wire, and the gases are allowed to escape freely,
under a pressure very nearly equal to that of the atmosphere,
so as to prevent their being heated. Under these conditions,
which are those of a miss-fire, the above-mentioned writers
obtained per 100 vols.

K) 24-7

See table, p. 33.

41-9

18-4

7-9

5-8

1-3

This again shows the multiplicity of decomposition that the
same explosive substance can undergo (see p. 7).

u

( 290 )

CHAPTER IX.

DIAZO-COMPOUNDS — DIAZOBENZENE NITRATE.

§ 1. GENERAL REMARKS.

1. NITROGENOUS organic compounds are derived from mineral
substances containing nitrogen by their combustion with non-
nitrogenous substances, this combustion being accompanied by
the separation of the elements of water.1

We thus obtain — either derivatives from the hydrogenated
compounds of nitrogen, such as those from ammonia, alkalis,
and amides, which were discussed in Chapter VI., and those
from hydroxylamine, with which we have nothing to do at this
point, or derivatives from oxygenated compounds of nitrogen,
such as the nitric derivatives, i.e. the nitric ethers and nitro-
compounds discussed in Chapter VIII. ; to these we may add, on
the same principle, nitrous derivatives, nitrous ethers, nitroso-
compounds, not as yet used as explosive substances, and
hyponitrous derivatives, hardly known.

2. The hydrogenated and oxygenated compounds of nitrogen
may also be associated two and two, three and three, etc., in
the formation of the same organic derivative ; they form bodies
of complex function, which are designated by the names diazo-,
triazo-, etc., derivatives.

Now, compounds of this order seem to be called upon to play
some part in the application of explosive substances. Let us
take the simplest of them, viz. those derived from ammonia
and nitrous acid, associated simultaneously with the same
organic compound. Such a one is diazobenzene, derived from
phenol and the two above-mentioned nitrogenous compounds —

C6H60 + HN02 + NH3 - 3H20
Such a body contains the nitrogenous residues both of ammonia

1 " Trait^ &£mentaire de Chimie Organique," by MM. Berthelot and
Jungfleisch, torn. ii. p. 313. 1881. Dunod.

DIAZOBENZENE NITRATE. 291

and of nitrous acid. Under certain conditions it takes up the
elements of water, reproducing phenol and free nitrogen —

C6H4N2 + H20 = C6H60 + N2.

In this case the nitrogen is produced by the reciprocal reaction
of the two nitrogenous components, precisely as in its produc-
tion from the direct reaction of ammonia and nitrous acid, the
original generators.

3. The heat disengaged in the formation of a diazo-compound
is far below thatfwniciiwould be produced in the formation of
nitrogen by the direct reaction of ammonia and nitrous acid.
In other words, the water eliminated in the original reaction
that engenders the diazo-compound, did not at the time of its
formation give off the same quantity of heat as if it had been
formed directly by the reaction of the two nitrogenous generators
in a free state. Thus, the diazo-compound contains an excess
of energy which renders it liable to sudden decomposition. It is
a highly explosive body. This theory leads us to foresee the
explosive properties of diazo-compounds. Only one of these
has, as yet, been studied from this point of view; namely,
diazobenzene ; and its properties fully bear out the forecasts
of this theory. For purposes of application diazobenzene
nitrate is especially worthy of study. It is a crystalline
compound, more easily handled than diazobenzene itself, and
containing, besides, a greater amount of energy, on account
of the additional presence of the nitric acid, which is calculated
to exercise an oxidising action upon the carbon. M. Vieille
and the author have studied its thermal and mechanical
properties.

§ 2. DIAZOBENZENE NITRATE.

1. Diazobenzene nitrate is an explosive substance which is
solid and crystalline. It answers to the formula —

C6H5N2N03,

its equivalent being equal to 167.

It has been proposed to use this body as a priming. In
virtue of its various modes of decomposition it is now employed
in industry in the manufacture of colouring matters.

M. Vieille and the author have studied its preparation,
stability, density, and also its detonation (both with respect to
the heat disengaged and also to the nature of the products), its
heat of combustion and of formation from the elements, and
lastly, the pressures developed by its detonation in a closed
vessel ; but the examination of this last branch of the subject
will be reserved for Book III.

2. Preparation. Aniline is the starting-point in the pre-

u 2

292 DIAZO-COMPOUNDS.

paration of diazobenzene ; that used in the experiments was
of excellent quality as regards purity.

3. Diazobenzene nitrate was prepared by the well-known
(Griess's) process of treating aniline nitrate with nitrous acid.
Five to 6 grms. of pure aniline nitrate were taken. This was
pounded and mixed with a little water, so as to form a paste,
which was placed in a tube surrounded with a refrigerating mix-
ture. A current of nitrous acid was then slowly introduced into
it, the mixture being continually stirred, so as to carefully avoid
any heating. The liquid at first turns a deep red, but afterwards
assumes a lighter tint. As soon as it begins to give off nitrogen
the operation is stopped. We then add to the liquid its own
volume of alcohol, and subsequently an excess of ether, which
precipitates diazobenzene nitrate. The latter is washed upon a
cloth with pure ether; it is then pressed and dried in vacuo.
In this way 67 per cent, of the theoretical yield was obtained.

4. Stability. Diazobenzene nitrate placed in a dry atmo-
sphere, and protected from the light, has been preserved for two
months and longer, without alteration, When exposed to day-
light, it becomes pink, and then changes more and more,
although slowly.

This alteration is much more marked under the influence of
moisture; the compound first emits an odour of phenol, and
assumes a peculiar tint ; then, after a time, it expands, becoming
black and giving off gases. Merely breathing upon this com-
pound will cause it to turn red.

On contact with water, it is immediately destroyed, giving off
nitrogen, phenol, and various other products —

C6H5N2N03 + H20 = C6H60 + N2 + HN03.

Diazobenzene nitrate is quite as sensitive to a shock as mercuric
fulminate; when struck by a hammer, or rubbed rather
vigorously, it detonates. It is much more susceptible to the
influence of moisture and light than the fulminate.

5. When heated beyond 90°, diazobenzene nitrate detonates
with extreme violence. Below this temperature it decomposes
gradually and without detonation ; at least, when it is heated
in small quantities. We see by this that diazobenzene nitrate
is much more sensitive to heat than mercuric fulminate — a
compound whose point of deflagration under the same conditions
is about 195°.

6. Density. The density of diazobenzene nitrate has been
found to be equal to 1/37, by means of the volumenometer,
or one-third that of the fulminate. A high pressure slowly
brought to bear on this body brings it to an apparent density
approximating to unity.

7. Composition. 0*5 grm. burnt by detonation in an atmo-
sphere of pure oxygen, gave the theoretical proportion of

EXPLOSION OF DIAZOBENZENE NITKATE. 293

carbonic acid to within about 3 J0th (less). There was neither
carbon monoxide nor any other combustible gas in the residue.
Experiments were made with 0'5 grm., suspended by means
of a metallic wire, capable of being made red-hot by an electric
current, in the centre of a platinum vessel filled with pure
oxygen. The average of two experiments gave 0*4296 carbonic
acid ; the quantity calculated being 0'43 grm.

8. Heat of formation from the elements. According to the
total heat of combustion, which will be given further on —

C6 (diamond) + H6 + N3 + 03 = C6H5N2N03, absorbs - 47'7 Cal.

The formation of nitric acid, H + N + 30 = HN03 (liquid),
gives off + 41*6 Cal. ; we therefore conclude that, taking into
account the nitric acid previously existing —

C6 + H4 + N2 + HN03 (liquid) = C6H^N2N03 absorbs - 89 Cal.

This value gives a more exact notion of the heat of formation
of diazobenzene itself. But we have to subtract from it the
heat disengaged by the combination of the diazobenzene with
the nitric acid. But free diazobenzene is itself a liquid body,
too imperfectly defined to have enabled one to study it.

However this may be, these negative values correspond very
well with the explosive properties so characteristic of this
compound.

The decomposition of diazobenzene nitrate by means of water,
with the reproduction of dissolved phenol and dilute nitric^
acid —

C6H5N2N03 + H20 = C6H60 + N2 + HN03 (diluted),

gives off + 1081 Cal.

9. Heat of detonation. This term is used to express the
heat given off by the simple explosion of diazobenzene nitrate,
an explosion that gives rise to complex products.

This explosion was effected in an atmosphere of nitrogen, in a
steel bomb lined with platinum ; it was ignited by means of
the galvanic heating of a fine platinum wire. The nitrate was
placed in a little tin cartridge, which was suspended in the
centre of the bomb, so as to avoid local actions arising out of
contact with the walls.

The results (in two experiments which were made upon
1-6 grms.) were : 688'9 and 686-6 Cal.; the mean being 687*7 Cal.
per kgm., or 6877 cal. per grm. This gives for an equi-
valent (= 167 grins.) 4- 114*8 Cal., at a constant volume.

10. The volumes (reduced) of the gases produced were 815*7
and 820 litres ; average = 817*8 litres per kgm., or 136*6
litres per equivalent (167 grms.).

11. Under the conditions of the experiments that were made.

294 DIAZO-COMPOUNDS.

i.e. with a low density of charge, the composition of these gases
was as follows : —

HCN
CO
CH4
H

N

3-2 or, for 136-6 litres . 4-4

48-65 „ . 66-4

2-15 „ . 2-9

27-7 „ . 37-9

18-3 „ . 25-0

100-0 136-6

It may be observed in this explosive decomposition—
(a) That a considerable quantity of hydrocyanic acid is
formed.

(&) That the whole of the oxygen, to within about one-hun-
dredth, is found as carbon monoxide ; i.e. the carbon takes up
all the oxygen, while water is not formed to any appreciable
extent in the detonation.

(c) That only three quarters of the nitrogen is disengaged in a
free state, one-fifteenth being given off as hydrocyanic acid. The
remainder is contained in the carbonaceous products of the ex-
plosion; a fraction, however — about one-fifth of the surplus
nitrogen — is found condensed as ammonia, as will be shown pre-
sently ; but, all allowance being made, the greater part (about
half an equivalent) remains united with the carbon, under the
form of a special fixed nitrogenous compound.

(d) That the free hydrogen amounts to almost three and a half
equivalents out of the five equivalents that the substance con-
tained ; one half equivalent goes to form marsh gas, another
half equivalent goes to form ammonia and hydrocyanic acid, and
the last half equivalent remains united with the carbon.

(e) That exactly half the carbon forms carbon monoxide. A
ninth of the remainder goes to form hydrocyanic acid and marsh
gas.

(/) That the solid residue contains nearly half (|) its weight
of carbon. A ninth of the remainder enters into the acid and
marsh gas. The gross composition of the residue approximates
pretty closely to the proportions represented by C5H2N2 ; it is
therefore a carbon rich in nitrogen and hydrogen, combined
under the form of condensed and polymerised bodies of the
paracyanogen type.

(g) That the gaseous products comprise, according to the
calculation of the preceding analyses, 75*9 per cent., by weight,
of the substance. A direct experiment effected by observing
the loss of weight of the apparatus when the gases are allowed
to escape freely after the explosion, gave 75*6.

(Ti) That, therefore, the solid residue comprises 241 per cent,
by weight. It exists as charcoal reduced to an impalpable
powder which is very voluminous and emits an ammoniacal
odour. The quantitative analysis of free ammonia in the

DECOMPOSITION OF DIAZOBENZENE NITRATE. 295

residue was effected without heat by means of the Schloesing
process, when it was found to represent 'Oil grm. per gramme
of the explosive compound. In the gases themselves, we found
0*00042 grm. of ammonia.

12. The following table sums up these results, the weights
being expressed in parts per thousand : —

Nitrogen ...-
Oxygen ...

Hydrogen

Carbon ...H

Gaseous proc
Residue ...

1 in the form of HCN

I combined in the charcoal ...
in the form of CO

16-7 [ 215-5 )
9-2) }• 251-2
35-6 j
287-6
20-5 \

fl 26*9)
2:o) 29'9
3-0 *
215-8)
14-3 \ 239-6 )
9-5J > 431-3
230-3 J

in the form of CH4

jj 5) -tlv^JM ...

„ NH3
combined in the charcoal ...
' in the form of CO
„ HCN
„ „ CH4
„ „ fixed matter

nets ... 769-7 1 1000
230-3/11

The result, 769*7, is higher than the weight of gas given above
(758-6), as it includes the ammonia.

13. Equation of decomposition. We see, from this table and
from the discussion that arose when these gases were being
studied, that, if we disregard the complications caused by
secondary formations (hydrocyanic acid, ammonia, marsh gas),
the principal reaction is reduced to the following : —

C6H5N2N03 = SCO + 3C + 5H + 3K

In reality, about one-tenth of the carbon that is not combined
with the oxygen, remains united with the hydrogen and nitrogen,
in a gaseous form, constituting marsh gas and hydrocyanic acid.
One-third of the hydrogen goes to form these same gases,
together with ammonia and fixed compounds. Lastly, one-
fourth of the nitrogen goes to form ammonia, hydrocyanic acid,
and nitrogenised charcoal.

14. The simple decomposition of diazobenzene nitrate so as
to give carbon monoxide and free elements —

SCO + 3C (diamond) + 5H + 3tf,

should disengage 201*6 Cal. at constant pressure ; i.e. 204*7 Cal.
at constant volume according to the heat of total combustion,
instead of -f 114'8, which was actually found. This proves that
the formation of secondary products has absorbed — 8 9 '9 Cal.

Such an absorption of heat results principally from the forma-
tion of the nitrogenised charcoal ; the exothermal formation of
ammonia and marsh gas almost counterbalancing the endo-
thermal formation of hydrocyanic acid.

296* DIAZO-COMPOUNDS.

This fact is in accordance with the general result ; according
to which the carburets, that are only slightly hydrogenated, and
the carbonaceous substances, retain a considerable portion of the
energy of their complex generators ; their energy exceeds more
or less that of the elements themselves.

This remark, which was at first made concerning acetylene,
has a very wide application in pyrogeneous decompositions ; it
explains the singular conditions under which certain endothermal
compounds are generated, at the very moment that organic com-
pounds are destroyed by heat.

15. Seat of total combustion. Combustion was started by
galvanic ignition of a fine platinum wire, in an atmosphere of
pure oxygen. It gave off, for 167grms. (1 equiv.), + 783 -9 Cal.
at constant volume (two experiments), which gives 782*9 Cal.
at constant pressure ; or, for 1 grm., 469 f4 cal. at constant
volume.

If the oxidation is complete, the reaction may be represented
by the following equation : —

i[2C6H5N2N03+ 230 = 12C02 + 5H20 + 3NJ

The Jieat of combustion lyy oxygen with reproduction of nitric
acid —

C6H5lSraN03 + 702 = 6C02 + 2H20 + 1ST2 + HlSTOg,

would give off, in addition, the heat of formation of nitric acid
combined with two equivalents of water —

HN03, 2H20,
or + 46-6 ; altogether + 829'5 Cal.

( 297 )

CHAPTEK X.

HEAT OF FORMATION OF MERCURIC FULMINATE.

1. WE know the part played by mercuric fulminate in the
manufacture of priming. This compound probably belongs to
the class of diazo-compounds. M. Vieille and the author have
studied its heat of decomposition, from which may be determined
the heat of formation.

2. The fulminate used in our experiments was taken from the
regulation detonators used by the Government. These deto-
nators contain 1/5 grm. of fulminate, and are manufactured at
Arras.

3. Its analysis gave —

Calculated.

C ... 8-35 C ... 8-45

0 ... 11-05 0 ... 11-30

N ... 9-60 N ... 9-85

Hg ... 71-30 Hg ... 70-40

100-34 100-0

The mercury was weighed as sulphide, the substance having
previously been oxidised by means of hydrochloric acid and
potassium chlorate. It is slightly in excess. This fact arises
from the presence of a small quantity of metallic mercury
mechanically mixed with the substance.

The nitrogen and hydrogen were determined volumetrically
after detonation of the substance. The hydrogen may be dis-
regarded, its presence being due to some accidental circumstance.
The carbon and oxygen were determined together, as carbon
monoxide after detonation, by which only slight traces of
carbonic acid are produced. In fact, five experiments gave for
one gramme of the substance, 234'2 cc., containing, per 100
volumes —

C02 ... 0-15

CO ... 65-70
N ... 32-26
H 1-80

Theory requires 235 '6 cc.

298 HEAT OF FORMATION OF MERCURIC FULMINATE.

The detonation should be effected in an atmosphere of
nitrogen in order to avoid the partial oxidation of the carbon
monoxide.

4. Heat of decomposition. Detonation, effected in the
calorimetric bomb, gave for one equivalent ( = 284 grms.) 4-
116 Gal. at constant volume, which corresponds to the following
decomposition : —

CHg(N02)CN = 200 + N2 + Hg,

or 114*5 Cal. at constant pressure, which for one grm. = 403
cal.

According to this equation, only carbon monoxide, nitrogen,
and mercury vapour are formed. One only of these bodies is a
compound ; it is stable and not susceptible of dissociation, which
accounts for the suddenness of the explosion. Moreover, the
heat is disengaged at first, and all the gases are produced without
the occurrence during cooling of any progressive recombination,
which would tend to moderate the expansion and diminish the
violence of the first shock.

The condensation of the mercury vapour, however, exercises
an influence of this kind ; but only after the principal cooling
has lowered the temperature below 360°.

5. Heat of formation from the elements. From the above
data we find that

C2 (diamond) + Hg + N2 + 02

absorbs + 51*6 - 114*5 = - 62*9 Cal. for 284 grms.

There is, therefore, absorption of heat in the formation of the
fulminate — a property in concordance with the explosive
character of the substance.

6. Heat of total combustion. Admitting the following re-
action—

CHg(N02)C£T + 02 = 2C02 + Hg + N2,

we shall get + 250 -9 Cal. for one equivalent ; or for one grm.,
883 cal.

This combustion may be effected, in the case of primings, by
mixing potassium chlorate with the fulminate, which causes
the heat disengaged to amount to + 262*9 Cal. per equivalent.
But in this instance we are heating 406*6 grms. of material
instead of 284 grms ; we get then for one grm., 647 cal.

We should also note the effects of expansion, due to the dis-
sociation of the carbonic acid, which renders the mixture less
sudden in its effects than pure fulminate.

( 299 )

CHAPTER XL

HEATS OF FORMATION OF THE CYANOGEN SERIES.

§ 1. HISTORICAL.

series of compounds in chemistry are of greater importance
than that of cyanogen, owing to the nature of the compound
radical that constitutes the characteristic pivot of the series. It
is the only electro-negative radical that has, up to the present,
been isolated. The exceptional properties of simple cyanides,
with their resemblance to the salts of the halogen elements, and
the still more singular properties of the double cyanides, add to
the interest of the cyanogen series. Several of the compounds
derived from it are employed in the manufacture of explosives.

In 1871, 1875, and finally in 1879 and 1881, the author
devoted long courses of experiments to the thermal study of
this series, which are published in the " Annals de Chimie et
de Physique," 5e serie, torn. v. p. 433; and torn, xxiii. pp.
178, 252.

These experiments, which were commenced during the spring
and summer of 1871, partly at Versailles and partly at Paris, in
the midst of the tumults of that year of trouble, presented
great difficulties and even serious dangers, for it was necessary
to use pure hydrocyanic acid and liquefied cyanogen chloride,
which are the most poisonous substances known.

The calculation of the fundamental quantities is based
principally on the measurements that were made of the heats
of formation of cyanogen, hydrocyanic acid, potassium cyanate,
and cyanogen chloride. The experiments on which the calcula-
tion of these quantities was based will be given in full. But
various alterations have been made in the values deduced from
them, principally on account of the complication caused in the
calculations by the intervention of the heat of formation of
ammonia. It has already been shown (pp. 237-242) to what
degree the former estimates of this quantity were inaccu-
rate, and the methods employed to rectify them. In 1882,

300 HEATS OF FORMATION OF THE CYANOGEN SERIES.

M. Joannis completed the author's results, in the laboratory of
the latter, by a prolonged study of various simple cyanides,
ferrocyanides, ferricyanides, and sulphocyanides. His paper
will be found in full, in the " Annales de Chimie et de Physique,"
5e se"rie, torn. xxvi. p. 482. The author's results concerning
explosive substances have been given in table x., p. 132.

§ 2. CYANOGEN.

1. The heat of formation of cyanogen has been measured in
two ways — by ordinary combustion and by detonation. The
following is the principle upon which the calculation is based.
The heat of formation required depends on the heat of formation
of carbonic acid, which is regarded as equal to 94 Cal. for

C (diamond) + 02 = C02.

On subtracting twice this quantity from the heat of com-
bustion of cyanogen, referred to the weight, which answers to the
equation —

C2N2 + 202 = 2C02 + N*

the difference represents the heat disengaged by the decomposi-
tion of the cyanogen. Consequently, this same difference taken
with the opposite sign, expresses the heat absorbed in the com-
bination of the carbon and nitrogen.

2. It is convenient to begin with ordinary combustion, by
means of which the following results were obtained.

The combustion of cyanogen by pure oxygen is easily effected
in the little glass combustion vessel shown on p. 241. With
a suitable excess of oxygen there is no formation of carbon
monoxide ; so that we are at once enabled to deduce the weight
of cyanogen consumed from the weight of carbonic acid which
is formed and collected in a bulbed tube (into which is subse-
quently introduced a lump of solid potash).

This combustion, however, presents a complication owing
to the formation of a little nitric peroxide. This body is
absorbed by the potash, together with the carbonic acid; its
weight should therefore be deducted. To do this, it is deter-
mined by consecutive operations. For example, assuming that
the original nitric peroxide has, on contact with the potash,
been converted into nitrous and nitric acids, we can then titrate
the nitrous acid by means of potassium permanganate. The
correction resulting from this is not of much importance ; in the
author's experiments it varied from one to three hundredths of
the total weight of the carbonic acid. This correction involves
another of still less importance, based on the fact that the
formation of nitric peroxide from its elements causes an
absorption of heat (— 2*6) which should be added to that

HYDROCYANIC ACID. 301

obtained by the calorimeter; but this new addition is insig-
nificant.

On calculation, the following values, referred to 26 grms. of
cyanogen, were obtained : —

Weight of cyanogen
consumed.

•419 grm 133-2 Cal.

•630 „ 130-0 „

•574 „ 131-3 „

•732 „ 129-6 „

Mean ... 131-6

i.e. for 52 grms. (= C2N2) -f 263'2 Cal.

3. The author had also recourse to detonation in the calori-
metric bomb (p. 148). In this way, the value + 261*8 was
obtained. This result was obtained at constant volume, but
it also applies to combustion at constant pressure ; the com-
bustion of cyanogen not giving rise to any change of volume.

It will be convenient to adopt the mean of the two results,
viz. -f 262 '5 Cal. Thomsen, who has repeated these experi-
ments quite recently, and after the publication of the above
results, obtained + 261'3, which comes as near as can be expected.

Dulong had obtained in 1843 -f 270 Cal. : the discrepancy
in this value will not seem excessive, when we take into con-
sideration to what perfection calorimetric methods have been
brought since that time.

From results made, it follows that

C2 (diamond) 4- N2 absorbs — 74*5.

If the carbon is supposed to be in the condition of charcoal, we
should only get — 68 -5. Thus cyanogen, CN, like acetylene,
CH,1 nitric oxide, NO, and all other substances acting as true
compound radicals, is a body formed with absorption of heat ;
a circumstance to which attention has already been called more
than once, as it seems calculated to account for the very
character of this real compound radical, which manifests in its
combinations an energy greater than that of its free elements.
The energy of these latter is increased by this absorption of heat,
instead of being weakened, as is the case in combinations that
give off heat, and this increase of energy renders the compound
system comparable to the more active elements.

§ 3. HYDROCYANIC ACID.

1. The heat of formation of hydrocyanic acid is deduced by
means of three methods, or series of independent measurements,
the results of which agree.

1 Acetylene and cyanogen are here considered under the same volume as
the simple radicals H and Cl.

302 HEATS OF FOKMATION OF THE CYANOGEN SEKIES.

The author first took, in 1871, as his starting-point —

(a) The conversion of hydrocyanic acid into formic acid and
ammonia.

(Z>) The conversion of mercuric cyanide by gaseous chlorine
and alkalis into carbonic acid, hydrochloric acid, mercuric
chloride, and ammonium chloride.

These two methods are based upon the employment of the
wet process. They require the knowledge of a great many
auxiliary data, and especially of the heat of formation of
ammonia. Now, the heat of formation of ammonia as adopted
in the first calculations, according to Thomson's measurements,
which were then universally accepted, was reputed to be equal
to + 35'15 (NH3 in solution). As this number should be
reduced to + 21, according to later conclusions (p. 242), the
correctness of which Thomsen has himself acknowledged, it
became necessary to deduct the difference between these two
values, i.e. 14*15, from the heat of formation (from the elements)
of hydrocyanic acid and also from that of cyanides. But it was
thought necessary to check this correction by measuring the
heat of formation of hydrocyanic acid by means of experiments
of another order, which are quite independent of the heat of
formation of ammonia, and in which the number of auxiliary
data was as limited as possible.

(c) This purpose was effected by burning a mixture of hydro-
cyanic acid gas and oxygen by detonation in the calorimetric
bomb —

2HCN + 50 = 2C02 + N2 + H20.

Three data only are required in this case, viz. the heats of
combustion of carbon, hydrogen, and hydrocyanic acid. The
experiments made according to this method will be described
first.

2. First Method. — Combustion of hydrocyanic acid. Pure liquid
hydrocyanic acid is introduced, by distillation, into little phials
of thin glass, care being taken to keep the weight of the acid
within suitable limits ('14 to *1 5 of a gramme in these experi-
ments). These limits are regulated by the capacity of the
calorimetric bomb, the tension of hydrocyanic acid vapour at
the temperature of the experiment, and the necessity of intro-
ducing into the bomb a sufficient amount of oxygen to obtain
total combustion. The tension of hydrocyanic gas being about
•59 of a metre at 18°, i.e. almost three-quarters that of the
atmosphere, it is easy to fulfil the conditions required.

The phial, sealed up and weighed, furnishes the exact weight
of hydrocyanic acid. This phial is carefully placed in the
bomb, which is then closed, and filled, by means of an orifice,
with pure dry oxygen at a suitable pressure. The orifice is
then carefully closed, and the phial containing the hydrocyanic

DETONATION OF HYDROCYANIC ACID AND OXYGEN. 303

acid broken to pieces by being violently shaken. The acid is
thus wholly converted into gas and constitutes with the oxygen
a detonating mixture.

This being done, the bomb is placed in the calorimeter, and
thermal equilibrium established ; we note the progress of the
thermometer and then proceed to detonate the mixture.
After the detonation, we again follow the progress of the
thermometer. The gas is then extracted by means of a mercury
pump, and caused to pass first through a drying apparatus, and
then through tubes containing potash. The bomb is then
purified by filling it several times with dry air, which is also
passed through the same tubes, so as to extract the whole of the
carbonic acid.

This can thus be weighed, which affords a valuable check on
the combustion.

Special trials showed that the quantity of nitrous compounds
formed in the combustion was negligable, but that a trace of
hydrocyanic acid always escaped. The latter was determined
each time in the potash, after the weighing; it amounted to
between half a hundredth and a hundredth of the original weight.
This was taken into account.

These points having been settled, the heat disengaged was
calculated in two ways ; either by considering it in relation to
the weight of hydrocyanic acid employed (minus the trace
which is not oxidised), or to the weight of carbonic acid
obtained (with the same deduction). The list of results observed
may be given. The heat absorbed, owing to the tension of the
aqueous vapour in the bomb at the temperature of the experi-
ment, was taken into account ; also + 0'4 was added to all the
results obtained, in order to allow for the fact that the detona-
tion was effected at constant volume. The heats of combustion
given below are supposed to be obtained at constant pressure.
We will now give the heat disengaged by the combustion of
27 grms. of gaseous hydrocyanic acid (HCN = 27 grins.), effected
by means of free oxygen at constant pressure.

According to the final weight According to the initial weight

of the hydrocyanic acid. of the carbonic acid.

158-9 163-4

160-0 161-3

154-2 155-6

159-0 160-4

160-1 160-3

Mean ... 158-4 Mean ... 160-2

The general mean, 159'3, of the two calculations will be
adopted. Thomsen, according to results which he published
after those just given, obtained by ordinary combustion
-j- 159*5, a value agreeing with that of the author as closely as
could be expected.

304 HEATS OF FORMATION OF THE CYANOGEN SERIES.

It is a fact worthy of mention that this number exceeds the
united heats of combustion of the carbon and hydrogen con-
tained in the hydrocyanic acid, whatever form the carbon may
be in.

C (diamond) + 02 = C02 + 94 (charcoal) + 97
£[H2 + 0 = H20 (liquid)] + 34-5 „ + 34-5

+ 128-5 + 131-5

According to these figures and this method, the formation of
hydrocyanic gas from its elements, H + C -f N = HCN, absorbs
+ 128-5 - 159-3 = - 30-2 when the carbon is in the form of
diamond, and - 27*2 when it is in the form of charcoal.

3. Second Method. — Conversion of the hydrocyanic acid into
formic acid and ammonia. This change is effected by means of
concentrated hydrochloric acid. In addition to the data con-
cerned in the direct experiment we must also have the heat of
formation of ammonia, the heat of combination of this base with
hydrochloric acid, the heat of dilution of hydrochloric acid,
and, lastly, the heat of combustion of formic acid, carbon, and
hydrogen, so that we have, in all, six auxiliary data.

A known weight of pure hydrocyanic acid was decomposed
in the calorimeter by means of very concentrated hydrochloric
acid. When the change was effected, it was proved to be com-
plete or practically so ; the mixture was diluted with a large
quantity of water, and the new quantity of heat evolved
measured. In a similar way was measured the heat disengaged
by the mixing of the same quantities of concentrated hydro-
chloric acid and water.

From this was deduced the quantity of heat that would be
disengaged by the following reaction : —

HCN (pure and liquid) + HC1 (diluted) +2H20 = H2C02 (in
solution) f NH4C1 (in solution); or -f 1115 Cal.

Experiments. — Some details of one of the experiments taken
as a type may now be given.

Preliminary operations. — The calorimeter contains 500 cms.
of water. It is placed in a double enclosure, in the centre of a
quantity of water, the temperature of which is exactly the same,
i.e. to within 0*1 of a degree, as that of the water in the calori-
meter, and that of the room in which the experiment is being
performed. This point is essential.

In the centre of the calorimeter is placed a little cylinder of
thin platinum, of a capacity of about 50 cms., with no opening
at the base, and closed at the top by means of a cork coated
with paraffin. This cylinder floats in the water of the calori-
meter, in which it is immersed nearly up to its top. We first
introduce into it 35 grms. of hydrochloric acid, which is concen-

EXPERIMENTAL DETAILS. 305

trated but not saturated ; then we place in the same cylinder a
glass phial containing 1*591 grm. of absolutely pure hydro-
cyanic acid — the phial itself weighs 1*568 grm. — it is very thin
and elongated into a point at each end, so as to be easily
broken when the cylinder is shaken.

These operations having been quickly performed, and the
phial being sealed, the cylinder is corked up, and the calori-
metric thermometer observed during an interval of ten minutes.
There was absolutely no variation during this interval in the
experiment performed. The temperature was about 20°.

First stage. — After the preliminary operations, we raise the
platinum cylinder a little by means of a pair of wooden pincers,
without, however, drawing it entirely out of the water, and
shake it violently so as to break the phial. It is then plunged
immediately into the calorimeter, and the course of the thermo-
meter again observed at the end of each minute. Eeaction
takes place, and the heat given off is gradually absorbed by
the water of the calorimeter. The variation of temperature is
most rapid at the commencement, but the maximum tempera-
ture is not produced until after a considerable length of time.
It exceeds the original by + 1'3°. It lasts for a quarter of
an hour and then the temperature slowly falls. We follow
this cooling for forty minutes, during which interval it only
amounts to 0*17 of a degree. This is the first stage of the
experiment.

Second stage. — We next incline the platinum cylinder and
open it under the water of the calorimeter, so as to fill it ; the
contents of the cylinder are mixed with those of the calorimeter
by stirring, until the thermometer, on being plunged alternately
into the calorimeter and cylinder, indicates exactly the same
temperature. This is the second stage of the operation. It
lasts about a minute, and gives rise to an excess of -f- 1*5° over
the temperature of the calorimeter at the beginning of this
stage, or -f 2-6° over the temperature at the beginning of the
entire experiment, i.e. at the beginning of the first stage. The
rate of cooling during an interval of five minutes is then
observed and the experiment is ended.

Verifications. — We then make sure, by means of suitable
reactions (the formation of Prussian blue), that the liquid
contains no appreciable quantity of hydrocyanic acid — the
latter having been entirely converted into formic acid and
ammonium chloride.

Moreover, in order to calculate the rate of cooling during
the first stage of the experiment, we proceed, by the addition
or subtraction of suitable quantities of water, to bring
the temperature of the liquid contained in the calorimeter
(the mass of water being kept constant during these fresh
mixtures) to + 1*3° above that of the enclosing vessel and

306 HEATS OF FORMATION OF THE CYANOGEN SERIES.

surrounding atmosphere, which should not vary to any appre-
ciable extent during the whole course of the experiment We
then follow for half an hour the rate of cooling, which corre-
sponds to this fresh increase of temperature, the conditions
observed being, as nearly as possible, the same as those of the
first stage.

The above experiment gave —

Initial temperature of the calorimeter ... +19*82

Initial temperature of the enclosing vessel ... + 19'98
Final temperature of the enclosing vessel ... + 20'06

Calculation from the experiment. — We now have to calculate
the actual quantity of heat disengaged during this experiment.
It is obtained, as we know, by multiplying the masses em-
ployed, reduced to units of water by the variation of tempera-
ture observed, this variation increasing with the lowering of
temperature produced during cooling.

Masses reduced to units of water. — Of the substance em-
ployed, the mass existing at the end of the experiment consists
of that in the water, which contains about ?-^o its weight of
hydrochloric acid, goW °^ ammonium chloride, and about as
much formic acid. Their weight being known from the original
data, the density is next taken, and then the volume calculated.
We assume that 1 cc. of this liquid absorbs 1 cal. for a rise in
temperature of 1°, an hypothesis sufficiently near the truth for
calculations of this kind.1

We reduce to units of water the various vessels of platinum
and glass that are used, and also the thermometer (that is, the
portion submerged), multiplying the weight of each vessel or
portion of vessel by its specific heat. The sum of all these
masses represents the total mass that has been subjected to the
variation of temperature observed.

The actual variation of temperature is the apparent variation
plus that corresponding to the heat lost during the first and
second stages of the experiment.

The calculation of these quantities will now be given, and
first of all, that during the second stage, as it is the easiest.

Heat lost during the second stage. — This is easily calculated, for
the duration of the second stage is only one minute, with a
final excess of temperature of 2 '5° above that of the enclosing
vessel. In fact, the loss of heat during the few following
minutes was measured and found to be almost uniform. We
calculate from this the mean loss during one minute ; then we
multiply the quantity thus got by the fraction f , for we assume
that the excess of temperature in the calorimeter, which
varied from 1*5° to 2*6° during one minute, has caused a loss

1 " Annales de Chimie et de Physique," 4* se"rie, torn. xxix. p. 163.

LOSS OF HEAT DUKING EXPERIMENT. 307

equal to that which would have resulted for a mean excess of
1*5 + 2-6 90

T = 2'

The resulting correction is very insignificant ; it only amounted
to 0-02 of a degree in 1'5°.

We then calculate the total heat, Qi, disengaged during the
second stage by multiplying the sum of the masses, reduced to
units of water, by the apparent variation of temperature, added
to that corresponding to the heat lost.

Heat lost during the first stage. — The loss of heat during the
first stage requires a more complicated mode of calculation.
This first stage is divided into periods of at least five minutes,
according to the rapidity of the heating, until the maximum
temperature is attained. The mean temperature of each period
is written down, and also the excess of this mean temperature
over the initial temperature.

The duration of the maximum temperature constitutes a
separate period, corresponding to the maximum excess of tem-
perature.

The time following this maximum is also divided into
periods of five minutes, and opposite each period are written the
mean temperature and the excess of temperature.

At the end of this time it was observed that the rapidity of
cooling was, for the same excess of temperature over the initial
temperature, exactly the same as in a check experiment made a
little later, and in which care was taken to reproduce this
excess under the same conditions.

This verification proves that the reaction was quite complete,
and that the data of the check experiment may be applied to
the calculation of losses of heat during the reaction itself.

In fact, this check experiment gives, without any hypothesis,
the losses of heat experienced by the calorimeter for a series of
excesses of temperature similar to those of the reaction, and
under conditions exactly parallel.

We then write down the loss of heat experienced by the
system in one minute for each mean excess of temperature
answering to each period ; we multiply this loss by the duration
of the period, generally five minutes (except in the case of the
maximum, which is longer). We then find the sum of all these
losses, and add them to the variation of temperature actually
observed.

To come to figures, it may be said that, the variation observed
being + 1/26°, the correction was -f 0'234°, which will not seem
too great in an experiment of such long duration.

We have now only to multiply the corrected variation of
temperature by the sum of the heated masses (reduced to units
of water), in order to obtain the quantity of heat, Q2, disengaged.
It is advisable to give these details, because they afford as

x 2

308 HEATS OF FORMATION OF THE CYANOGEN SERIES.

exact an idea as possible of experiments of this kind, and of the
difficulties which they present. Of course, we could not expect
the same degree of accuracy as in experiments of short duration,
but, nevertheless, the errors can hardly exceed '05 of the total
value.

4 Calculation from tfo theoretical reaction. It now remains
for us to deduce from the numbers observed the values which
are applicable to the reaction taken from a theoretical point
of view. For this purpose, the same weight is taken of the
same solution of concentrated hydrochloric acid, viz. 35 grms.
(or a weight very near to this, in which case the results are
afterwards referred to this weight by proportional calculation) ;
it is dissolved in the same quantity (500 cc.) of water at the
same temperature ; then the heat, Q3, disengaged is measured.

This quantity being known, the difference, Q! 4- Q2 — Q3,
represents exactly the heat disengaged by the conversion of the
weight employed of pure hydrocyanic acid by means of hydro-
chloric acid (diluted), into formic acid (diluted), and ammonium
chloride (diluted), as the initial state and the final state are
absolutely identical. Multiplying this quantity by the ratio of
the equivalent (HCN = 27 grms.) to the weight of hydrocyanic
acid actually employed, we obtain the heat disengaged in the
theoretical reaction —

HON" (pure and liquid) + HC1 (diluted) + 2H20
= H2C02 (diluted) + £TH4C1 (diluted).

The following numbers were found by experiment, + 11-54 and
•f 10'76, or, on an average, + 1115.

5. From this is deduced the heat of formation of hydrocyanic
acid from its elements, carbon (diamond), gaseous hydrogen, and
gaseous nitrogen —

H + C + N = HCN (pure and liquid), absorbs - 22'6.
In short, supposing the initial system to be

5H + C + N + 02 -f- HC1 (diluted),
and the final system

H2C02 (diluted) + NH4C1 (diluted),
we pass from one to the other by two different processes.

FIRST STEP.

H2 + C + 02 = H2C02 (pure) disengages + 93-00

H2C02 (pure) and water + Q'10

N + H3 = NH3 (in solution) + 21-00

NH3 (diluted) + HC1 (diluted) = NH4C1 (diluted) ... + 12-45

Sum + 126-55

HEAT OF VAPORISATION OP HYDROCYANIC ACID. 309

SECOND STEP.

2(H2 + 0) = 2H20

H + C + N = HCN (pure and liquid)

Reaction of HC1 (diluted)

Sum ... + 149-15
Thus,

x = - 149-15 + 126-55 = - 22'6.

This value applies to liquid hydrocyanic acid.

6. Vaporisation of hydrocyanic acid. In order to pass to the
gaseous state of the acid, we must determine the heat absorbed
in its vaporisation. To do this, the following method was
adopted, which may be applied to all liquids of similar volatility.
It consists in vaporising them in a current of dry gas and
measuring the heat absorbed. We pour into a glass phial a
known weight, say 1'396 grm., of pure hydrocyanic acid; we
then seal up the phial, which should be thin and provided with
two points easily broken. This is introduced into a little glass
receiver, fitted with a worm and arranged so that a regular
current of air may be made to circulate in it by means of an
aspirator, the gaseous current first passing through the recipient
and then through the worm.

This little system is plunged into the calorimeter, which con-
tains 500 grms. of water. It is immersed almost up to the
orifice of the receiver, which is closed by a cork through which
a tube is passed, by means of which the current of air may pass.

This air is perfectly dry, and its temperature during its
passage is shown by a thermometer indicating twentieths of a
degree; the volume of this air is determined sufficiently
accurately for the calculation into which it enters, by measuring
the volume of water that has flowed from the aspirator. It
may be added that a solution of alkali was placed between the
worm and the aspirator in order to absorb the hydrocyanic gas,
and thereby prevent its noxious fumes. These preparations
having been made, the phial being still closed, a certain volume
of air is allowed to circulate for twenty minutes through the
receiver and worm, in order to estimate the cooling. The
experiment for which the results are given gave a value of 0
for the initial cooling. This result is easily explained, as the
temperature of the water in the calorimeter was + 20'07, that
of the water of the enclosing vessel, + 20-22, and that of the
surrounding air, + 20'8.

The phial is then broken against the sides of the receiver by
the violent shaking of the latter. The gaseous current is allowed
to continue circulating, and the thermometer is read. The
experiment lasts twenty minutes, during which the liquid acid
has entirely disappeared, and the minimum temperature is
reached almost immediately. This minimum corresponds to a

310 HEATS OF FORMATION OF THE CYANOGEN SERIES.

fall of temperature of - 0'510°. The circulation of the gaseous
current is continued for twenty minutes longer, in order to
measure the re-heating, which, under these conditions, is very
slight.

We then possess all the data necessary for calculating the
heat of vaporisation of hydrocyanic acid under the conditions
described above.

It was found —

For HCN = 27 grms. (1st trial) 5'680

For HCN = 27 grms. (2nd trial) . . . 5-730

Mean ... 5-705

Thus the formation of gaseous hydrocyanic acid from its elements,
according to this method —

H + C (diamond) + N = HCN" (gas), absorbs - 28*3.

7. Solution of hydrocyanic acid. The solution of the liquid
acid in water may give rise to either a disengagement or to
an absorption of heat, according to the relative proportions,
and also to the temperature if the proportion of water be
small1

In this case only the heat disengaged in the presence of a
large quantity of water was measured. It was found that

HCN (liquid) + 60H2O, at 19°, disengages + 040.

8. Third Method. — Conversion of mercuric cyanide into mercuric
chloride, carbonic acid, and ammonia. This method consists in
dissolving gaseous chlorine in water contained in a closed calori-
meter, weighing it, and treating the solution obtained with an
exactly equivalent weight of mercuric cyanide ; the latter
becomes converted into mercuric chloride and cyanogen chloride
(in solution) —

i[2Cla (in solution) + Hg(CN)2 (in solution) = HgCl2 (in
solution) + 2CNC1 (in solution)].

The quantity of water to be employed must be calculated, so
as to be much greater than would be necessary to hold in
solution the whole of the carbonic acid finally formed. We
then add diluted potash, in proportions equivalent to the chlorine,
so as to obtain potassium chloride and potassium cyanate —

CNC1 (in solution) + K20 (diluted) = KCNO (in solution)
+ KC1 (diluted).

Without troubling ourselves whether the action is more or
less complete, we immediately pour an excess of diluted hydro-
chloric acid into the same calorimeter, so as to bring the whole

1 Bussy and Buignet, " Annales de Chimie et de Physique," 4e se*rie, torn,
iii. p. 235.

DECOMPOSITION OF MERCURIC CYANIDE 311

mass to the final state of carbonic acid (in solution), ammonium
chloride (in solution), potassium chloride, and mercuric chloride
(in solution). We thus get —

KCNO (diluted) + HC1 (diluted) + H2O = C02 (in solution)
+ NH3 (diluted) -f KC1 (diluted).

We measure the total heat disengaged in this series of re-
actions; the whole series occupying a period not exceeding
from twenty to twenty-five minutes. Then we make sure that
there is no cyanogen compound remaining in solution ; this is
confirmed by the quantitative estimation of the ammonia, made
in the cold by the Schloesing process.

The total heat disengaged by this series of reactions being
known, the following data are brought into the calculation — the
heats of combustion of carbon and hydrogen, the heat of oxida-
tion of mercury, the heat of chlorination of hydrogen, the
heat of formation of ammonia, the heats of combination of
mercuric oxide with hydrochloric and hydrocyanic acids, and
lastly, the heats of combination of diluted potash and dis-
solved ammonia with hydrochloric acid ; making in all, nine
auxiliary data.

In short, we proceed from the initial state, which is —

i[C2 + N2 + 4H2 + Hg + 2C12 + 202 + 2K20 (diluted)
-I- 4HC1 (diluted)],

to the final state —

J[2C02(in solution) + 2NH4C1 (diluted) + 2H20 + 4KC1
(diluted) + HgCl2 (diluted)].

By one mode of procedure, the compounds of the final state
may be formed directly ; the heat of formation of mercuric
chloride in particular being determined from the heats of forma-
tion of water, mercuric oxide, and hydrochloric acid, together
with the heat disengaged by the solution of the oxide in this
acid.

By a second method, the diluted hydrocyanic acid and mer-
curic oxide are formed first, and then combine.

H + C + N (liquid) disengages ... ......... x

HCN (liquid) and water ............... + 0-4

i[H
J[H

] ............... 4-15-5

di

HgO + 2HCN (diluted) = Hg (CN)2 (diluted) + H20] + 15.46

We then add to this sum the total amount of heat disengaged
in the calorimeter during the course of the operations, without
troubling about the chemical nature of the intermediate
reactions.

The details of the experiments will not be given here, as
they will be found further on, under cyanogen chloride. It
will merely be remarked that the quantity, x, calculated from

312 HEATS OF FORMATION OF THE CYANOGEN SERIES.

experiments of this order and from the value at present adopted
for the formation of ammonia (p. 242), was found to be equal
to — 24'3. This relates to liquid hydrocyanic acid. We get
then, according to this method —

H + C + N = HCN (gas) - 30.

9. In short, the following numbers have been obtained for
the formation of hydrocyanic gas —

By the first method (detonation) —30-2

By the second method (formic acid and ammonia) ... — 28*3

By the third method (mercuric cyanide and chlorine) ... — 30'0

Mean ... - 29'5

This mean value will be adopted to express the heat absorbed
by the combination of the elements —

H + C (diamond) + N = HCN (gas) - 29-5 Gal.

HCN (liquid), we should get —23-8 „

. HCN (in solution) — 23'4 „

10. From these figures it follows that cyanogen and hydro-
cyanic acid are both formed from their elements with absorption
of heat. This circumstance explains, as has already been said,
the character of cyanogen as a compound radical, and, in a more
general manner, the tendency of cyanogen and hydrocyanic
acid to form direct combinations and polymeric compounds,
and to give rise to complex reactions. The fresh determi-
nations which are here published confirm the views which
were expressed by the author on this subject twenty years
ago, with regard to cyanogen, acetylene, and endothermal com-
binations.1

11. It will be remembered that cyanogen, hydrocyanic acid,
acetylene, etc., may be regarded as following the general rule
applicable to chemical compounds, i.e. as being formed with
disengagement of heat; if we assume that the carbon, when
under the form of diamond or charcoal, does not correspond to
true elementary carbon, the latter would be comparable to
hydrogen, and would probably be in the gaseous state, charcoal
and diamond representing its polymeric forms. In passing
from the gaseous to the polymeric and condensed state, the
elementary carbon would give off a considerable quantity of
heat, which is greater than that absorbed in the formations of
acetylene (- 30'5 for C = 12) and cyanogen (-37'3).

The quantity of heat developed by the condensation of the
elementary carbon might even be estimated at 4- 42*6 for
diamond and + 39*6 for charcoal, if we assume that the successive
formation from the gaseous carbon of the products of the two

1 " Annales de Chimie et de Physique," 4e se*rie, torn. vi. pp. 351 et 433.

FORMATION OF HYDROCYANIC ACID. 313

degrees of oxidation of carbon, viz. carbon monoxide and
carbonic acid, gives off the same quantity of heat. These are
conjectures of some interest, and have been accepted by various
savants since they were first broached.1

12. However this may be, the figures actually obtained lead
us to conceive a very definite opinion, which is confirmed by
experiment. In fact, they show that the formation of hydro-
cyanic gas from cyanogen and hydrogen —

H + CN = HCN, gives off + 7-8 Gal.

This formation is therefore exothermal ; a circumstance which
led to the suspicion that it might be effected directly, notwith-
standing the negative experiments that had been made previously
by Gay-Lussac. In fact, the direct combination of the two
gases was effected directly by means of time and heat alone,
and under conditions comparable with those in the synthesis of
the hydracids of the halogen elements properly so called.2

13. The synthesis of gaseous hydrocyanic acid from acetylene
and free nitrogen, a synthesis very easy to effect through the
action of the electric spark, as was discovered in 1868 —

C2H2 + N2 = 2HCN, disengages + 2'1 Gal.,

a positive though very low quantity.

14. As to the formation of hydrocyanic acid from ammonium
formate and formamide, which is the simplest type of a general
reaction in organic chemistry, viz. that for the formation of
nitrils, it is worthy of special attention.

Let the reaction be as follows : —

NH4CHG2 = HCN + 2H20,

the water and the acid being supposed to be separate.

This reaction, if it could be effected with solid bodies at the
ordinary temperature, so as to produce water and liquid hydro-
cyanic acid, would absorb — 137 Gal. Effected with the dis-
solved salt, it would absorb - 104 Gal.

Let us again note the initial system —

H2C02 (pure), NH3 (diluted), HC1 (diluted) ;
and the final system —

HCN (pure), HC1 (diluted), 2H20.
We may pass from one to the other by two different methods.

1 " Annales de Chimie et de Physique," 4e se*rie, torn, xviii. pp. 161, 173,
and especially 175.

2 Ibid., 5e se'rie, torn, xviii. p. 378. The heat of formation of hydrocyanic
acid, admitted in the article quoted, was estimated, by means of the data then
known, at + 26'9 ; this is too high, but the sign of the phenomenon remains
the same, and, consequently, the preconceived idea of its beingsynthetic.

\( UNIVERSITY

314 HEATS OF FORMATION OF THE CYANOGEN SERIES.

FIRST STEP.

H2C02 (pure) + water ... +0-08

H2C02 (diluted) + NH3 (diluted) + 12-0

Separation of solid ammonium formate + 2'9

NH4CH02 (solid) = HCN (liquid) + 2H20 (liquid) ... + x
HC1 (diluted) no change.

Sum + 15-0 + x

SECOND STEP.

NH3 (diluted) + HC1 (diluted) + 12-4

H2C02 (pure) + water + O08

NH4C1 (diluted) + H2C02 (in solution) = HCN (liquid)

+ HC1 (diluted) + 2H20 -11-15

Sum ... + 1-3
x = - 15 + 1-3 = - 137 Cal.

In fact, the salt, when melted, is really destroyed, with the
production of water and hydrocyanic acid, both in a gaseous
form ; it thus absorbs, besides the — 137 Cal. mentioned above,
the heat necessary for the vaporization of these two substances,
i.e. a value approximately1 = — (57 4- 19*3) + F, where F is
the heat of fusion of the salt. This gives then - 387 4- F,
which amounts to about — 36 Cal.

A similar absorption would no doubt be produced if the
ammonium formate could exist in a gaseous form and be decom-
posed as such.

In short, the formation of formonitril from ammonium formate
absorbs, whatever hypothesis is adopted, a great quantity of
heat ; a result in accordance with what happens in most decom-
positions.

We may go further than this. In fact, the dehydration of
ammonium formate is effected by two stages, formamide and
water being first produced —

NH4CH02 = CNH30 + H20.

The liquid formamide was acted upon by means of con-
centrated hydrochloric acid, so as to give the opposite reaction.
From the numbers observed it was deduced that the theoretical
reaction, i.e. in the case of the use of diluted acid, would give
off + 1*4 Cal., a value which is probably too low, and is given
here with some diffidence, as the formamide, being in the liquid
state, cannot be guaranteed pure. It is practically applicable
to the change of formamide (in solution) into ammonium
formate (in solution), the value deduced being -f- 1.

We conclude, from these figures, that the decomposition of
the melted ammonium formate into gaseous formamide and

1 The term approximately is used, because the heats of vaporisation of the
bodies concerned in the experiment at the temperature of decomposition of
ammonium formate (180° to 200°) are not the same as at the boiling points.

POTASSIUM CYANIDE. 315

steam will absorb about + 18 Cal. (supposing the vaporisation
of the formamide to absorb from 6 to 8 Cal.).

The two stages of the dehydration of crystallised ammonium
formate, which is changed first into gaseous amide and then into
gaseous nitril, would correspond to thermal phenomena that
may be considered as having equal effects, the first stage absorb-
ing - 18 Gal, and the two stages together — 36 Cal. But this
equality would only exist when the products are considered in
the gaseous state.

Conversely, the fixing of the elements of water, either upon
the formamide or upon the formonitril (in solution), so as to
reproduce the ammonium salt (in solution), gives off a quantity
+ 1 Cal. for the amide, and + 104 for the nitril ; quantities
which are very unequal but both positive.

This is another proof of the disengagements of heat which
may result from simple hydration effected by the wet process,
and especially so from that of amides, which play a very im-
portant part in the study of the reactions of organic nitrogenous
principles, and that of animal heat.1

The formation of cyanides will now be explained.

§ 4. POTASSIUM CYANIDE.
1. It was found, by experiment, that at about 20° —

HCN (liquid), on being dissolved in 40 times its weight of water, Cal.

gives off ..................... +0-40

i[2HCN (diluted) + K20 (diluted)] ............ + 2'96

KCN (pure), on being dissolved in 50 times its weight of water,

absorbs ...... i.;"^ ......... ... -2-86

From this we deduce the heat disengaged by the formation of
solid potassium cyanide from the elements —

K -f C -f N = KCN (crystalline) disengages -f 30'3
The calculation is as follows : —

Initial system; £[&,+ C2 -f N2 -f H2 + 0].

Final system ; i[2KCN (crystal.) + H20 (liquid and separate)].

FIRST STEP.

H + C -h N = HCN (liquid) absorbs - 23-8

HCN + nAq ......... ... + 0-4

-"K2 + 0+riAq = K20 (diluted)] ............ +82-3

2HCN (in solution) + K,0 (in solution) = 2KCN (in

solution) + H20] ............... + 3*0

Separation of KCN (solid) ............... + 2'9

Sum ... + 64-8

1 "Annales de Chiruie et de Physique," 4e se'rie, torn. vi. p. 461.

316 EEATS OF FORMATION OF THE CYANOGEN SERIES.

SECOND STEP.

K + C + N = KCN (crystalline) x

i[H2 + 0 = H20 (liquid)] ,. ... + 34-5

Sum ... +34-5 + 03
x = + 64-8 - 34-5 = + 30-3.

2. The direct formation of potassium cyanide from the union
of its elements, as expressed by chemical equation, and the
corresponding disengagement of heat, is not really effected at
the ordinary temperature. But it is admitted that it does
actually take place at a very high temperature, when free
nitrogen is made to act upon charcoal impregnated with potas-
sium carbonate ; i.e. under conditions where potassium is
generated. At this temperature, the potassium cyanide is
melted or perhaps even gaseous, a change of state which causes
an absorption of heat ; but, on the other hand, the potassium is
gaseous, which compensates this absorption. If free nitrogen,
carbon, and potassium do really combine without any other
intermediate reaction, such as the formation of an acetylide
(this formation has not been proved to take place), we should
be led to admit that the total synthesis of potassium cyanide
disengages heat, under the actual conditions of the reaction.

Be the disengagement produced at once or by successive
reactions, it does not explain the total synthesis.

3. We come now to a clearer synthesis. The union of
cyanogen with potassium takes place, as we know, directly.
This union, calculated for the following states of the substances
concerned —

K (solid) -f CN (gas) = KCN (crystallised), disengages
+ 67-6 Cal.

These figures justify the direct synthesis of potassium cyanide
from cyanogen. But this quantity is lower than that dis-
engaged by the union of the same metal, in the solid state, with
halogen elements in the gaseous state.

Now,

Cl + K = KC1 (solid) gives off + 105*6

Br (gas) + K = KBr gives off +100*4

I (solid) + K = KI gives off + 85-4

This difference explains why chlorine, bromine, and iodine
decompose potassium cyanide in solution ; cyanogen is liberated
and combines, besides, with half the halogen, causing a slight
additional disengagement of heat —

[+ 1-6 for CNC1 (gas) ; + 4-2 for CNI (solid).]

4. It may also be mentioned, in order to complete this
parallel, that the formation of potassium cyanide from the
hydracid (diluted) and potash —

HCN (diluted) + KHO (diluted) = KCN (in solution) + H20,

AMMONIUM CYANIDE. 317

disengages much less heat (+ 3*0) than the formation of the
chloride, bromide, and iodide of potassium under the same
conditions (which disengages 13 "7).

This difference would be increased by 17 Cal., if the hydro-
chloric and hydrocyanic acids were considered in the gaseous state.

Hydrocyanic acid is therefore much less powerful than the
hydracids derived from halogen elements ; it is even displaced
in potassium cyanide in solution by most acids.1 This inert-
ness of hydrocyanic acid itself contrasts with the greater energy
of the complex acids which it forms when associated with
metallic cyanides ; hydroferrocyanic acid, for example ; this will
be referred to later on.

5. The conversion of potassium cyanide into formate —

KCN (in solution) + 2H20 = KCH02(in solution)
+ NH3 (in solution), gives off + 9*5 Cal.

This reaction does take place in solutions of the cyanide,
although slowly.

The same reaction, effected on the dry salt by means of
aqueous vapour, produces formate and ammonia gas ; it is much
more rapid, but gives off double the amount of heat : — 17'7.

If the temperature is raised, the reaction becomes complicated,
owing to the ultimate destruction of the formate by the heat or
excess of alkali ; this reaction, which takes place at about 300°
and finally transforms the potassium cyanide into potassium
carbonate —

KCN (solid) + KHO (solid) + 2H20 (gas) = K2C03 (solid)
+ NH3 (gas), gives off + 374.

This point is important, because it is one of the most
effective causes of the destruction of potassium cyanide during
its industrial preparation; in this case the melted salts are
operated upon, and this fact causes a slight modification of the
above values ; without, however, altering their general signifi-
cation. When exposed to the oxygen of the air, we know that
potassium cyanide is readily converted into potassium cyanate.
This reaction will be referred to at a later period.

§ 5. AMMONIUM CYANIDE.

1. It was found that the combination of hydrocyanic acid in
solution, with ammonia in solution, gives off -J- 1*3 Cal. The
solution of freshly prepared ammonium cyanide ( 1 part of salt
to 180 parts water) absorbs - 4'36 for NH4CN (= 44 grms.).

2. From these figures it follows that the union of hydrocyanic
gas and ammonia gas, with formation of solid cyanide —

HCN (gas) + NH3 (gas) = NH4CN, gives off + 20'5.
1 " Annales de Chimie et de Physique," 49 serie, torn. xxx. p. 492.

318 HEATS OF FOBMATION OF THE CYANOGEN SERIES.

This is only half the heat disengaged in the similar formations
of chloride ( + 42'5), bromide, and iodide of ammonium ; l the
acetate comes nearer (+ 28*2), and the hydrosulphide nearer
still (+23).

3. Starting from the elements, we should get —

N2 + C + 2H2 = NH4CN (solid) gives off + 40-5 Gal.

The analogous formation of ammonium chloride gives off
+ 767.

4. Lastly, the heat of formation of ammonium chloride from
the elements is less than that of potassium chloride by 28'3
Cal. ; whereas, between the formation of ammonium cyanide
and that of potassium cyanide, the thermal difference is 27'1.
The difference in the two cases is therefore almost the same;
i.e. this state does not depend on the halogen generator.

§ 6. METALLIC CYANIDES.

1. It was found that gaseous cyanogen combines directly, not
only with potassium, but also with certain true metals, such
as iron, zinc, cadmium, lead, and even copper ; but this tendency
towards direct combination does not extend to mercury and silver.

2. The reactions are effected by heating cyanogen and the
metals in sealed tubes, to 100° for the first-named metals, and
to about 300° for the two last.

3. Such combinations are always attended by a disengage-
ment of heat. In particular, according to M. Joannis —

J[Zn -f (CN)a = Zn(CN)2] gives off + 28'5.
J[Cd + (ON), = Cd(CN)2] gives off + 19-8.
From the elements, on the contrary —

J[Zn + C2 + N2 = Zn(CN)2] absorbs - 8-8.
£[Cd + C2 - N2 = Cd(CN)2] absorbs - 17-5.

§ 7. MERCURIC CYANIDE.

1. Formation from the acid and the oxide. It was found, by
experiment, that dilute hydrocyanic acid and mercuric oxide —

i[2HCN (1 equiv. = 2 litres) + HgO (precipitated and diluted
with 10 litres of water)], gives off -f 15-48.

An excess of hydrocyanic acid does not cause any alteration
in this value, which is considerable, exceeding even the heat
given off in the action of dissolved hydrochloric acid on potash.

It is owing to this difference in values that potash, combined
with hydrocyanic acid, with which, moreover, it gives off much

1 See p. 127.

MERCURIC CYANIDE. 319

less heat (3*0), is displaced by mercuric oxide. On the other
hand, the solution of this salt —

£[Hg(CN)2 (solid) + water (1 to 40)] -1-5 ;
consequently,

£[2HCN (in solution) + HgO = Hg(CN)2 (solid)] ... + 17'0
M2HCN (liquid and pure) + HgO = Hg(CN)2 (solid)] + 17-4
£[2HCN (gas) + HgO = Hg(CN)2 (solid) + H20 (gas)] + 18'3

We will compare this last result with the heat of formation of
mercuric chloride.

j[2HCl (gas) + HgO = HgCl2 (solid) + H20 (gas)] gives off

+ 23-5,

which value exceeds that of formation of mercuric cyanide by
4-8 only.

These figures, and the conclusions resulting from them, will
be referred to again.

2. Formation of mercuric cyanide from the elements.

l[Hg (liq.) + C2 (diamond) + N2 (gas) = Hg(CN)2 (solid)]
absorbs - 254 Gal.

Let the initial system be —

and the final system —

i [Hg(CTST)2 (solid) + H20 (liquid and separate)] ;
we pass from one to the other, by the two following steps : —

FIRST STEP.

*(Hg + 0 = HgO) gives off + 15-5

fl + c+ N= HCN (in solution) absorbs - 23-4

Union of these two bodies gives off ... ... ... + 15'5

Separation of Hg(CN)2 (solid) gives off + 1-5

Sum ... +9-1
SECOND STEP.

tH2+ 0 = H20 (liquid)] ... +34*5
Hg + C2 + N2 = Hg(CN)2 (solid)] x

Sum ... + 34-5 -f x
x = - 25-4.
The salt in solution, - 26-9.

There is, therefore, absorption of heat in the formation of
mercuric cyanide from the elements ; exactly as in the case of
hydrocyanic acid. The figures even are very much the
same (p. 312).

3. But, on the contrary, the union of gaseous cyanogen with
liquid mercury at the ordinary temperature —

i[Hg (liquid) + (CNa) (gas) = Hg(CN)2 (solid)], gives off
+ 37-3 - 25-4 = + H-9.

320 HEATS OF FORMATION OF THE CYANOGEN SERIES.

This quantity is 19 '5 less than the heat disengaged in the direct
formation of mercuric chloride —

i[Hg (liquid) + C12 (gas) = HgCl2 (solid)] + 314.

4. The same quantity of heat is, on the contrary, absorbed in
the ordinary preparation of cyanogen, by the decomposition of
mercuric cyanide. To this must be added the heat of vaporisa-
tion of the mercury, which brings the absorption of heat to
about — 194 for the reaction —

i[Hg(CN)2 (solid) = Hg (gas) + (CN)2 (gas)].
We may observe that this result is very near that ( - 224)
which accompanies the analogous decomposition of mercuric
iodide into gaseous iodine and gaseous mercury. But this last
reaction is accompanied by phenomena of dissociation due to the
opposite tendency of iodine and mercury to recombine — a
tendency which does not exist on the part of the components of
mercuric cyanide.1

5. Substitution of chlorine for the cyanogen and formation of
cyanogen chloride. The simple substitution of gaseous chlorine
for gaseous cyanogen —

i[Hg(CN)3 (solid) + C12 (gas) = HgCl2 (solid) + (ON* (gas)],

assuming the salts to be either in the solid state or in solution
(the heats of solution of both salts are the same), would give off
.+ 194,

In fact, this substitution is accompanied by a simultaneous
formation of cyanogen chloride —

^[Hg(CN)2 (solid) + 2C12 = HgCl2 (solid) + 2CNC1 (gas)] gives

off + 21-3,

or, if the cyanogen chloride be supposed liquid, + 2 9 '6.

All the bodies being in solution except the chlorine, we must
add to this quantity the heat of solution of cyanogen chloride.

In fact, the heat given off in this reaction, all the bodies
except the chlorine being in solution, was measured and found
to be = + 27*5 (the cyanogen chloride being also in solution).

This figure seems to indicate that the heat of solution of
gaseous cyanogen chloride is very near its heat of liquefaction,
as might be expected. Unfortunately, this action is not instan-
taneous, and this fact diminishes the certainty of accuracy of the
estimates, and leads us to fear some complication, attributable to
secondary reactions of the chlorine on the water.

6. Reciprocal displacements of the hydrochloric and hydro-
cyanic acids. According to the above remarks, the formation of
mercuric cyanide in solution, from the acid in solution and
precipitated mercuric oxide, gives off -f 1548, i.e. -f 6*02 more
than that for mercuric chloride (+ 946). The same difference

1 " Annales de Chimie et de Physique," 5e se*rie, torn, xviii. p. 382.

DECOMPOSITION OF CYANIDES AND CHLORIDES. 321

exists in the solid salts, if we always reckon from the diluted
hydracids. These latter being monobasic, the thermal inequality
that has been just mentioned indicates that dilute hydrocyanic
acid must entirely displace hydrochloric acid from its combina-
tion with mercuric oxide.

Here is an experiment that fully bears out this supposition :

/ £[Hg(CN)2 (1 eq. = 16 litres) + 2HC1 (1 eq. = 4 litres)] + 0\M - Mt = + 5-9
\|[HgCl2 (1 eq. = 16 litres) + 2HCN (1 eq. = 4 litres)] + 5'9 /calculated + 6'0

The reaction is all the more remarkable, as, according to
thermal observations made, the dilute hydrochloric acid
completely displaces the hydrocyanic acid in dissolved potassium
cyanide. It was, moreover, easy to foresee that this would be
the case in the last instance, for

i[2HCN (in solution) + K20 (diluted)] gives off + 2'96) M M . «.«,
J[2HC1 (in solution) + K20 (diluted)] „ „ + 13'59f Mi~

The decomposition of dissolved mercuric chloride by dilute
hydrocyanic acid is all the more remarkable from the fact
that solid mercuric cyanide is decomposed by concentrated
hydrochloric acid ; it is in this way that pure hydrocyanic
acid is prepared. But this decomposition — the opposite of
that which takes place in weak solutions — is easily explained
by thermal theories. In fact, it is due to the action of the
anhydrous hydrochloric acid contained in the liquors, when
we are operating without heat ; or formed under the influence
of heat, when we proceed by distillation. Now, this anhydrous
hydracid possesses, in relation to the hydrate of the same acid,
the energy which the latter has lost in forming a definite
hydrate ; the magnitude of which energy is sufficient to reverse
the reaction.1

Moreover, hydrochloric acid gas displaces, immediately and
without heat, the hydrocyanic acid gas of crystallised mercuric
cyanide. This process for preparing the latter gas has been
mentioned. According to calculation, the reaction disengages
+ 5*2 Cal. Attention2 has already been called to these two
reactions and their mechanism, which is frequently met with
under other circumstances, such as when we are comparing the
reactions of concentrated .acids or alkalis with those of the
same acids or alkalis diluted. It is the existence of a certain
proportion of acid (or alkali), either not combined with water
or combined in the state of a less advanced hydrate in the
concentrated liquids, and also the formation of such an acid,
dehydrated under the influence of heat, that causes the inverse
reaction ; and this is in proportion to the excess of energy that
the anhydrous acid possesses, in comparison with the hydrate of

1 " Annales de Chimie et de Physique," 5e se"rie, torn. iv. p. 465, and 4s
se*rie, torn. xxx. p. 494.

2 " Essai de Me'canique Chimique," torn. ii. p. 547.

Y

322 HEATS OF FOEMATION OF THE CYANOGEN SERIES.

the same acid, with which it co-exists in the liquors. This
excess of energy exactly measures the tendency to produce the
inverse reaction, which, however, ceases as soon as the anhy-
drous hydracid contained in the liquor is saturated.

But, on the contrary, the reaction could not be foreseen, as has
been supposed by various writers,1 from the knowledge of the
quantity of heat disengaged in the dilution of the concentrated
acid, the bulk of which becomes a dilute acid. Not only is this
mode of prediction not justified in principle, since it makes no
distinction between the anhydrous acid and its hydrates in
solution, but it leads to conclusions which are quite contra-
dicted by experiment. For example, mercuric cyanide is still
decomposed in the cold by hydrochloric acid of a density I'lO
(which nearly corresponds to HC1 -f 7H20) ; the dilution of
such a hydrochloric solution gives off only 1-7 Cal. Now, this
excess would have to be equal to + 6*0 for the reaction to be
reversed, according to the theory that we must reject ; i.e. that
the inversion is solely due to the heat of dilution taken in the
mass. This excess is so great that the dilution of even the most
concentrated hydrochloric acid could not make up for it.

The greater number of reciprocal displacements give rise to
the same observations, the heat disengaged by the dilution of
concentrated acids or alkalis being scarcely ever sufficient to
supply the whole of the body in solution with the energy
necessary to reverse the chemical reaction ; whereas this energy
is, on the contrary, supplied by the hydration of the portion of
acid (or alkali) which existed in the liquor in a dissociated
state.

7. But to return to mercuric cyanide. Theory indicates that
the displacement of hydrochloric acid by hydrocyanic acid in
mercuric chloride may be observed still more clearly if we
substitute an alkaline cyanide for the free hydrocyanic acid. In
fact, in this case, we shall get, besides, the difference of the
heats of neutralisation of both acids by the alkali. This is
confirmed by experiment.

i[2KCN(l equiv. = 8 litres) + HgCl2 (1 equiv. = 4 litres)]

disengages + 16*7.

J[2KC1(1 equiv. = 8 litres) -f- Hg(CN)a (1 equiv. = 4 litres)]
disengages -f 0.

Now, calculation gives —
(M-M1)-(M'-M'1) = (13'6 -3) -(9-5 -15-5) = + 16-6,

a result quite consistent with the above. Thus, the reality of a
double integral interchange between the bases and acids in
solution is fully established. This is one of the most glaring
cases in which the so-called saline thermo-neutrality which was

1 " Annales de Chimie et de Physique," 5s s6rie, torn. iv. p. 464.

SILVER CYANIDE. 323

formerly accepted is found to be at fault. The result observed
agrees perfectly with that calculated, when this calculation
is based upon the hypothesis of a total conversion into mercuric
cyanide and potassium chloride in solution. Moreover, this
does not affect the reciprocal reaction between the two last salts
when in solution ; i.e. the formation of a double cyanide, which
will be discussed presently.

8. A reciprocal action of this kind is easily shown between
potassium cyanide in solution, and solid mercuric iodide, which
enters into solution —

J[HgI2 (solid) + 4KCN(1 equiv. =16 litres)], total solution,

+ 97.

The solution of the solid body takes place, in this case, with
a considerable disengagement of heat, on account of the heats
of formation of the double salts that are generated and remain
in solution.

9. The formation of the mercuric oxycyanides from the
combination of the cyanide with the oxide may be mentioned
here. This combination is effected with disengagement of heat
(Joannis) —

i[Hg(CN)2 (solid) + HgO = Hg(CN)2,HgO (solid)] gives off

+ 1*2.

This oxycyanide, when heated, explodes, in consequence of
the combustion of part of its carbon by the oxygen which it
contains.

§ 8. SILVER CYANIDE.
1. Formation from the acid and base.

Some experiments were made on the heat of formation of silver
cyanide.
(6) AgN03(l equiv. = 16 litres) + HCN(1 equiv. = 4 litres)

+ 1572,
from which is got —

J[2HCN (in solution) + Ag?0 (precipitated) = 2AgCN
(precipitated)] gives off + 20'9.

(6) AgN03 (1 equiv. = 16 litres) + KCN (I equiv. = 4 litres)

+ 26-57,
from which is got —

£[HCN (in solution) -f Ag20 (precipitated) = 2AgCN
(precipitated)] gives off + 20*9,

a value identical with the above. It is, moreover, almost the
same as the heat disengaged in the formation of silver chloride.
Again, we get from the above results :

|[2HCN (liquid) 4- Ag20 (precipitated) = 2AgCN -i- H20 (liquid)] + 21'3
£[ 2HCN (gas) + Ag20 (precipitated) = 2AgCN + H20 (liquid)] + 27'0
i[2HCN (gas) + Ag20 (precipitated) = 2AgCN + H20 (gas)]... + 22-2

Y 2

324 HEATS OF FORMATION OF THE CYANOGEN SERIES.

this last value being only approximate, on account of the
physical changes experienced by the silver oxide and cyanide.
It is less by a third than the heat, 33*2, given off in the
analogous formation of silver chloride. These values explain
why hydrocyanic acid displaces nitric acid from its combina-
tion with silver oxide, and why silver cyanide resists the action
of dilute nitric acid.

2. Formation from the elements.
1. Ag -f C (diamond) + N + AgCN absorbs - 13'6.

The calculation of this value is as follows : —
Initial system :

£[Ag2 + C2 + N2 + H2 + 0],
Final system :

i[2AgCN (solid) + H20 (liquid)].

FIRST STEP.

J[Ag2 + 0 = Ag20] disengages ......... + 3-5

H + C + N = HCN (diluted) absorbs ...... - 23'4

|[Ag20 + 2HCN (diluted)] disengages ...... + 20'9

Sum ... + 0-8
SECOND STEP.

H20] ............... +34-5

Sum ... + 34-5 + x

whence, x = — 337.
2. Again, we have —

Ag + ON (gas) = AgGN gives off + 3*6.

Let us compare the differences observed between the heats
of formation of the chloride and cyanide, formed by the same
metal or system of elements. For potassium, the difference is —

105-67-6= +374;
for ammonium —

76-7 - 40-5 = + 36-2.

The value here is almost the same. But the figures become
very unequal for metallic salts, such as those from silver :

For silver and mercury we observe that the value is only
half that relating to alkaline salts.

DOUBLE CYANIDES. 325

The same differences exist for cyanides when compared with
bromides (Br gas) —

Difference.

K] 85-4 - 67-6 = 17-8

Hg] 22-4 - 11-9 = 10-5

Ag] 19-7 - 3-6 = 16-1

Also between cyanide and iodides (I gas) —

Difference.

K] 100-4 + 67-5 = 32-8

Hg] 30-4 + 11-9 = 18-1

27-7 + 3-6 = 24-1

[K]
[Hg]

[Ag]

These inequalities result from- the great quantity of heat
disengaged by the union of the hydrocyanic acid with metallic
oxides as compared with the small quantity disengaged in the
union of the same acid with alkalis.

§ 9. DOUBLE CYANIDES.

It is now necessary to find the heat of formation of double
cyanides, such as the cyanides of mercury and potassium, silver
and potassium, and also that of ferrocyanides, which are worthy
of particular attention.

1. Cyanide of mercury and potassium : Hg(CN2), 2KCN. This
compound offers a remarkable example of a double salt which
exists and is even undoubtedly generated in solutions. In fact,
it was found that its two components, when in solution, give off
a great quantity of heat if merely mixed together —

-J[Hg(CN)a (1 equiv. = 16 litres) + 2KCN (1 equiv. = 4 litres)]

+ 5-8.

This quantity represents nearly two-thirds of the heat dis-
engaged by the union of the two salts when in the solid state.
The latter value is calculated by means of the following data : —

KCN, on being dissolved in 40 times its weight of water — 2-96

i[Hg(CN)2], on being dissolved in 40 times its weight of water ... — 1-60
i[2KCN, Hg (ON) J, on being dissolved in 40 times its weight of water — 6'96

These data go to show that the combination

J[Hg(CN)2 (dry) + 2KCN (dry) = Hy(CN)2, 2KCN (dry)],
disengages -f 8 -3,

which is a considerable quantity of heat. It approaches, and
even exceeds, the heat disengaged in the formation of many
metallic salts, from the anhydrous acid and base. However,
the double cyanide, in solution, is immediately decomposed by
diluted hydrochloric acid, with separation of its components ;
the mercuric cyanide being regenerated unaltered in the liquor,
and the potassium cyanide being converted into potassium
chloride.

This was discovered by measuring the heat disengaged in the

326 HEATS OF FORMATION OF THE CYANOGEN SERIES.

reaction. This measurement proves, in fact, that dilute hydro-
chloric acid, acting on the solution of the double cyanide,
separates the components, with reproduction of potassium
chloride and hydrocyanic acid :

[iHg(CN)2, 2KCN (20 litres) + 2HC1 (1 equiv. = 2 litres], + 5*2

-4HC1 „ +0

The calculation, based upon these last data, indicates the
following value: -f 3'0 -f 5'8 + 5'2 = + 14;0 for the heat dis-
engaged in the union of hydrochloric acid with potash ; a value
which agrees to all intents with the actual value + 13 '6, that
is, if the liquors are as much diluted as in the above.

2. Cyanide of silver and potassium : AgCN, KCN. This salt,
so much used in electro-plating, acts in a manner similar to the
above. It is formed by the direct action of potassium cyanide
in solution on precipitated silver cyanide, the latter becoming
dissolved with disengagement of heat —
KCN (1 eq. = 4 litres) + AgCN" (precipitated) -f water (20 litres)

gives off -f 5*6.

The reaction disengages almost the same quantity of heat as
that in the case of mercuric cyanide, notwithstanding the solid
state of the silver cyanide.

This is a fresh instance of the solution of a precipitate being
effected with disengagement of heat, in consequence of the for-
mation of a double salt. The phenomena are dependent on this
formation, independently of the solubility or insolubility of the
original metallic cyanide (mercuric or silver), as the double salt
is formed with disengagement of heat, and is stable in the
presence of excess of the dissolving agent.

It was also found that the solution of the soluble salt —

AgCN, KCN (solid) (1 part = 40 parts of water), absorbs - 8'55.

"We conclude from these data, and also the heat of solution of
potassium cyanide, that the combination —

AgCN (precipitated) + KCN (dry) = AgCN, KCN" (dry),
disengages + 11 '2.

The double salt in solution is immediately decomposed by
dilute hydrochloric acid, with reproduction of potassium
chloride and hydrocyanic acid, as is proved by thermal
measurements. At the same time a precipitate of silver
chloride is produced, mixed with a considerable proportion of
cyanide, as might be expected ; for the formation of both salts
from the diluted hydracids and precipitated silver oxide causes
the liberation of about the same quantity of heat (-f 20'9).

The double cyanide of silver and potassium, however, con-
stitutes a firmer combination than is usually met with in
ordinary double salts. In fact, dilute acetic acid separates
silver cyanide from it only in a very incomplete manner, giving

POTASSIUM FERROCYAXIDE. 327

off only + 1'7 Cal., instead of -f 4*8, which would correspond
to a total decomposition. Tartaric acid gives similar results.
It would seem, then, that the liquors contain a hydro-argento-
cyanic acid, already mentioned by Meillet ; a complex acid,
which can only exist in the presence of water and another acid,
so as not to give rise to phenomena of equilibrium, and con-
sequently to a partial decomposition. The solutions of this
complex acid produce results in silver-plating that are almost
as well marked as those produced by alkaline cyanide solutions,
as there was occasion to prove.

This compound forms a very remarkable intermediate step in
the formation of those special molecular types that constitute
the complex cyanides.

3. Potassium ferrocyanide. A more decided stability charac-
terises the double cyanide of potassium and iron, known as
ferrocyanide. Although the thermal study of its formation
presents great difficulties, owing to the fact that we cannot start
with isolated iron cyanides, nevertheless it is undoubtedly worth
while giving the results of the experiments performed, with the
admission that they are, no doubt, imperfect.

4. The heat of solution of both dry and hydrated potassium
ferrocyanide was first measured, the former in fifty parts of
water, the latter in forty parts of water. It was found that
at 11°

^[K4Fe(CN)6, 3H20] (211-2 grins.), in dissolving, absorbs - 846.
i[K4Fe(CN)6 (dry)J „ „ - 5-98.

From these figures it follows that the union of the water with
the dry salt —

i[K4Fe(CN)6 + 3H20 (solid) = K4Fe(CN)6, 3H20 (crystal)], gives
off + 0-34, or + Oil for each

a quantity which is very small, but, according to certain experi-
ments,1 comparable to that which is disengaged in the formation
of the hydrated calcium and copper acetates.

5. The heat of neutralisation of hydroferrocyanic acid by bases
cannot be conveniently measured directly, on account of the
difficulty of obtaining the free acid in a perfect state of purity.
The latter object was attempted by indirect means, i.e. by dis-
placing the acid from its salts by more powerful acids.

On mixing a diluted solution of ferrocyanide —

i[K4Fe(CN)6] = 4 litres,

with diluted hydrochloric acid (1 eq. = 2 litres), we observe
that there is absolutely no change of temperature, either because
there is no reaction, or because the two acids disengage the
same quantity of heat in acting on the potash, in which case

1 " Annales de Chimie et de Physique," 5e se*rie, torn. iv. p. 127.

328 HEATS OF FORMATION OF THE CYANOGEN SERIES.

they would share between them the base in the liquor. This
last supposition seems the more probable.

In fact, on mixing the ferrocyanide with dilute sulphuric
acid we actually observe a progressive division of the base
between the acids and a displacement, which tends to become
total in the presence of a great excess of sulphuric acid. Among
the various experiments that were made with regard to this
question, the following only will be quoted : —

J[KJFe(CN)e (6 litres) + H2S04 (1 eq. = 2 litres)] disengages

+ 1107.

<L[K4Fe(CN)'6 (6 litres) + 2eH2S04 (1 eq. = 2 litres)] dis-
engages + 0'181.

Upon continuing the progressive additions of sulphuric acid,
an absorption of heat is produced owing to the formation of the
bisulphate.

With a large excess, added all at once —

i[K4Fe(CN)6 (4 litres) + 10H2S04 (1 eq. = 2 litres)] + 0-966.

These phenomena may be compared to those in the reaction of
sulphuric acid upon chlorides,1 although the values are some-
what different. They also show a progressive division of the
base between the two acids. If we admit that i[10H2S04] are
sufficient to abstract almost the whole of the potash from the
ferrocyanide, according to what happens in the case of chlorides,
nitrates, etc., we can calculate the heat, X, disengaged in the
action of dissolved hydroferrocyanic acid on diluted potash. In
short, + 15 '7 being the heat disengaged by the action of
sulphuric acid on potash, and — 1/75 the heat absorbed in the
action of J[4H2S04 (diluted)] on £[K2S04] in solution (so as to
form bisulphate), we shall get for the reaction we are seeking —

i[i(H4Fe(C]Sr)6 = 4 litres) + K20 (1 eq. = 2 litres)] gives off
X = 15-71 - 1-75 - J(0-97) = + 13-5.

This number is, to all intents, the same as that which repre-
sents the heat disengaged (13'6) by the combination of hydro-
chloric and nitric acids with potash, whence it follows that
hydroferrocyanic acid is a powerful one, and may be compared
with the mineral acids. We know, in fact, that it displaces
carbonic and acetic acids. The apparent absence of thermic
reaction between hydrochloric acid and ferrocyanide in solution
is consistent with these results.

6. Nothing is easier than to pass from this to the formation
of Prussian Uue. It was found, in fact, that -I12[3K4Fe(CN)6
= (4 litres) + 2Fe2(S04)3 (1 equiv. = 2 litres) = (Fe7(CN)18
(precip.) + 6K2S04 (diss.)] disengages + 2-54 to + 278, the

1 " Annales de Chimie et de Physique," 4" s6rie, torn. xxx. p. 524.

PRUSSIAN BLUE. 329

heat disengaged gradually increasing with the time; as fre-
quently happens in the formation of amorphous precipitates.1
Similarly :

1i2[3(K4Fe(C]Sr)6 = 4 litres) + 2Fe2(N03)6 (1 equiv. = 2 litres) =
Fe7(CN)18) (precipitated) 4- 2KN03 (in solution)] disengages
+ 0-725.

According to the result furnished by the ferric sulphate, the
substitution of potash for ferric oxide (K20 for Fe^) in Prussian
blue gives off -f 7'2 ; and according to the result furnished by
the nitrate, + 7'2 ; thus the two results agree. Admitting
that in the formation of potassium ferrocyanide,

£[H4Fe(CN)6 (diluted) + 2K20 (diluted)], disengages + 13'5
X 2 = + 27,

we conclude that the formation of Prussian blue, with the same
acid and precipitated ferric oxide —

£ [3H4Fe(CN)6 (diluted) + 2Fe2O3 (precipitated)], gives off
+ 6-3x2=-}- 12-6.

The value 6-3 is little different from the value 5*7, which
represents the heat of combination of nitric and hydrochloric
acids with ferric oxide ; this is a fresh proof of the analogy
existing between hydroferrocyanic acid and the mineral acids.
However + 6'3 is greater than 5'7, which explains why diluted
hydrochloric acid does not decompose Prussian blue, with
formation of ferric chloride.

7. Hydrocyanic acid, one of the weakest acids known, thus
constitutes, when combined with ferrous cyanide, a powerful
acid, comparable in every respect with nitric, acetic, and hydro-
chloric acids. This is a fresh proof, which helps to establish
the theory that the acid properties that are most marked, even
in compounds containing carbon and hydrogen, are not
necessarily connected in any way with the presence or propor-
tion of oxygen.

8. We now have to measure the heat disengaged in the
formation of ferrocyanide. The following results were found : —

i[FeS04 (1 equiv. = 2 litres) + 2Fe2(S04)3 (1 equiv. = 2 litres)
+ 6K20(1 equiv. = 2 litres)] gives off -f 23'2.

Adding to the above mixture ^[6HCN (1 equiv. = 4 litres)],
we observe a fresh disengagement of + 3 9 -3, which represents
the formation of ferrocyanide from hydrocyanic acid and the
two oxides —

£[6HCN(in solution) + 2K20 (in solution) + FeO (precipitated)
= K4Fe(CN)6 (in solution)] gives off + 39'3.

1 " Annales de Chimie et de Physique," 5e s&ie, torn. iv. pp. 174, 181.

330 HEATS OF FORMATION OF THE CYANOGEN SERIES.

As a proof,

J[6HC1] (1 equiv. = 2 litres)

was added to the liquid ; this gave off + 25'0 Cal., with the
production of an abundant precipitate of Prussian blue; the
heat disengaged varied, during this precipitation, from 23 to
25. To sum up, the hydrochloric acid should produce the
following reactions : —

£[2HC1 (diluted) + K20 (diluted) = 2KC1 (diluted)] ... +

£[6HC1 (diluted) + Fe203 (precipitated) = Fe2Cl6 (diluted)] + 11-4 , oc .
K2Fe2Cl6 (diluted) + 3K4Fe(CN)6 (in solution) = Fe7(CN)18 | + ^

+ 12KC1 (diluted)] ............... + 1-4J

The approximate consistence between the values 25 and 2 6 '4
is as close as can be hoped for in the study of such precipitates,
the state of which varies with the conditions.

9. We may conclude, further —

i[18HCN (diluted) + 3FeO (precipitated) + 2Fe203 (precipitated)

= Fe7 (CN)w (precipitated)] gives off + 24-9.

i[6HCN (diluted) + FeO (precipitated) = H4Fe (CN)6 (in solu-

tion)], + 12-3.

These values were verified by means of the direct formation
of Prussian blue from potassium cyanide and the two sulphates
of iron :

equiv. = 2 litres) + 3FeS04 (1 equiv. = 2 litres) +
2Fe2(S04)3 (1 equiv. = 2 litres) = Fe7(CN)18 (precipitated)
+ 9K2S04 (diluted)] gives off + 37*5.

The difference between the heat of formation of the alkaline
sulphate and that of the iron sulphates, reckoning from the
oxides, being 12'5 4- 111 - 471 = - 23'2, and the heat of
formation of 3KCN from potash being 8 '9, we can easily
determine from these data the heat disengaged in the formation
of Prussian blue from hydrocyanic acid :

i[18HCN (diluted) + 3FeO + 2Fe203 = Fe7(CN)18] gives off
37-5 + 8-9 -23-2 = +23-2;

a value sufficiently near to 24*9, which was obtained in another
way, but is a less accurate result.

10. It is now possible to draw a few general conclusions
from these results.

The first that occurs to us relates to the heat disengaged in
the formation of the ferrocyanide, starting from hydrocyanic
acid or from potassium cyanide.

J[6HCN" (in solution) + 3K20 (diluted)] gives off + 8'7.

J[6HCN (in solution) + 2K20 (diluted) + FeO (precipitated)]

gives off + 39-3.

We see that the substitution of ferrous oxide for potash, with

PRUSSIAN BLUE. 331

formation of ferrocyanide, gives off a considerable proportion
of heat: viz. 39'3 - 8*7 = + 30'6. Moreover, only a single
equivalent of ferrous oxide is required to constitute hydroferro-
cyanic acid.

These figures also explain the displacement observed, and
correspond to the constitution of a new molecular type, — that
of hydroferrocyanic acid.

In fact, we conclude from them that

i[6HCN (in solution) + FeO (precipitated)] gives off + 12*3,

which quantity exceeds the heat (-f 9'0) disengaged by the
combination of i[3K20 (diluted)] with i[6HCN].

There are here two simultaneous reactions, viz. the union
of six molecules of hydrocyanic acid into a type six times as
condensed, and the reaction of the ferrous oxide which enters
into the constitution of this new type : H4Fe(GN")6 .

In the same way, for Prussian blue, the fact has been
established elsewhere that

i[3H4Fe(CN)6 (diluted) + 2Fe203 (precipitated)] gives off + 12-6

= 6-3 x 2,

or almost the same quantity as in the union of the same oxide
with diluted hydrochloric and nitric acids.
From hydrocyanic acid itself we get —

i[18HON (diluted) + 3FeO + 2Fe203 = Fe7(CN)l8 (precipi-
tated)] + 24-9 = 8-3 x 3.

The magnitude of this last quantity, which is three times that
of the heat disengaged by potash in its combination with hydro-
cyanic acid, enables us, as before, to account for the formation
of the new molecular type of ferrocyanides, and, in a more
general manner, for the formation of double cyanides.

11. This consistency in effects also explains, by the greater
quantity of heat disengaged, the greater apparent affinity pre-
sented by ferrous oxide, as compared with potash, in its union
with hydrocyanic acid ; this is contrary to what happens in the
comparison of the formation of the ordinary oxysalts, sulphates,
nitrates, acetates, etc., from diluted acids and alkaline bases,
with that of those salts from the same salts with metallic
oxides.

12. Could we not explain by some similar circumstance why
silver and mercuric oxides, as well as ferrous oxide, give off
more heat than diluted potash in their combination with hydro-
cyanic acid? In a word, are mercuric and silver cyanides
really represented by the simple formulae

AgCN and Hg(CN)2,

those of potassium cyanide and hydrocyanic acid being KCN
and HCN ? or ought we rather not regard them as being them-

332 HEATS OF FORMATION OF THE CYANOGEN SERIES.

selves cyanides of a more condensed type, such as Hg2(CN)4
and Ag2(CN)2? The heat disengaged by their union with
potassium cyanide, so as to form double cyanides, such as
K2Hg(CN)4 ; KAg(CN)2, even in dilute solutions, would support
this opinion ; as it would result from the transition from the
simple type potassium cyanide to the complex type constituting
the double cyanides —

Hg(CN)2 + 2KCN = HgK2(ON)4

AgCN + KCN = AgK (CST)2.

Besides, hydrocyanic acid is not the only acid that gives
rise to the general inversion of ordinary affinities, as shown by
the corresponding thermal effects, between the alkaline oxides
and metallic oxides. It is exactly the same with hydrosul-
phuric acid.1

13. However it may be with regard to these last consider-
ations, it remains no less a fact that metallic oxides give off
more heat than alkaline bases in uniting with hydrocyanic
acid; which explains why they displace them. It may be
repeated that thermo-chemistry thus explains the constitution
of complex cyanides — new molecular types, very superior to the
primitive type, as regards the energy of their affinities towards
bases, and the stability of the resultant salts, very superior to
hydrocyanic acid, which aids in their formation by its conden-
sation.

Hydrocyanic acid, the common generator of these condensed
types, is moreover distinguished by the fact that it is formed
from the elements with an absorption of heat amounting to
29-5 ; in other words, its formation has stored up an excess of
energy that gives it a special tendency towards successive
combinations and molecular condensations.

14. We will give, in conclusion, the heat of formation of
potassium ferrocyanide from its elements, a quantity that
enters into the study of certain explosive substances.

From cyanogen, we should get —

J[2Ka + Fe + 3(CN)2 = K4Fe(CN)6 (solid)] gives off + 183*6,

or + 61-2 x 3.
From the elements —
£[2K2 + Fe + 60 + 31ST2 = K4Fe(C]Sr)6] + 717, or + 23'9 x 3.

These values are near those that correspond to the formation
of potassium cyanide ; viz. from cyanogen, + 67'6 ; and from
the elements, + 30'3.

The hydrated salt contains, in addition, three equivalents of
water, 3H20, the union of which in a liquid form with the
anhydrous salt gave off 2[ -j- 2 *48] ; which makes altogether for
the crystallised yellow prussiate from the elements and water,
+ 94-2 Cal.]

1 See " Annales de Chimie et de Physique," 5e se'rie, torn. iv. p. 186.

CYANOGEN CHLORIDE. 333

§ 10. CYANOGEN CHLORIDE.

1. Cyanogen chloride was prepared in the form of a colourless,
dry and pure liquid ; its purity was proved by the determina-
tion of the chlorine contained in a given weight of the com-
pound. This done, several samples of it were weighed into
sealed phials ; the weight of these samples was approximately
2 grms.; being 1/946 grin., 24675 grins., 2-137 grms., and
so on.

2. This cyanogen chloride was converted into carbonic acid
(in solution) and ammonium chloride —

CNC1 (liquid) + 2H20 + water = C02 (in solution) + NH4C1

(in solution).

The heat disengaged during this transformation was measured
by the following method, which consists in treating the cyanogen
chloride with potash and hydrochloric acid successively.

Preliminary operations. — A solution of potash containing
about 1 equiv. (471 grms.) in 10 litres of liquid was introduced
into the calorimeter, a proportion of it being taken that
would represent rather more than 3 equiv. for 1 equiv. of
chloride (CNC1 = 61*5 grms.) ; i.e. about 1 litre of the alkaline
solution.

The rate of cooling during an interval of ten minutes,
measured before the actual experiment, was found to be nil ;
which is explained by the fact that the temperature of the
liquid was 21*51°, that of the enclosing vessel being 21 '31°.

The phial containing the cyanogen chloride is surrounded
with a thick platinum wire, wound into a spiral, so as to add
weight to the phial and form a system that will remain at the
bottom of the water whether the phial be full, empty, or giving
off gases. This system is placed in a dry glass tube, which is
surrounded with ice, a little thermometer and a piece of potash
being put by the side of the phial; then we wait until the
thermometer records a temperature as near zero as possible ;
0*5° for example. It is necessary to take the precaution of
cooling the phial beforehand, if we wish to be able to open it
afterwards without loss or projection, after its introduction into
the calorimeter, seeing that cyanogen chloride boils at 12°, and
that it would be violently expelled if the point of the closed
phial were broken, in which it had been kept a liquid at a
temperature of 21°, in virtue of its own pressure.

First stage of the experiment. — The phial having been thus
prepared beforehand, and kept in a cold dry tube (dry, in order
to prevent the condensation of moisture on the surface of the
phial), the calorimeter is made ready. We then take the
platinum spiral surrounding the phial, and plunge the whole,

334 HEATS OF FORMATION OF THE CYANOGEN SERIES.

spiral and phial, into the calorimeter, the water in which should
completely cover the two points of the phial. We, however,
keep the lower point from touching the bottom of the calori-
meter, so as to prevent the possibility of its getting broken in
the course of the subsequent operations. When these pre-
parations are completed, we break the upper point of the phial,
by a sharp blow with a platinum hammer against a piece of
glass resting on this point, or in some similar way ; but the
lower point must be carefully kept intact. Under these con-
ditions, cyanogen chloride is at once given off, escaping in
gaseous bubbles from the broken point ; these are absorbed, as
they appear, by the alkaline solution. The operation proceeds
with extreme regularity, if all the prescribed precautions have
been observed.

Nevertheless, we follow at each minute the course of the
calorimetric thermometer. At the end of about 20 minutes,
the vaporisation is complete and the maximum temperature
attained. It exceeds the original temperature by about 2°.
The phial is then completely broken up with the hammer, in
order to destroy the last traces of cyanogen chloride, and
complete the mixture.

The first stage of the operation converts the cyanogen
chloride into potassium chloride and cyanate (or rather
isocyanate), mixed with a certain quantity of potassium and
ammonium carbonates.

The proportion of these products of a further reaction varies
according to the concentration of the potash and various other
circumstances ; it seems to increase little by little, by a slow
reaction. We cannot, therefore, stop at this point, as it does not
furnish a reliable basis for calorimetric computations.

Second stage of the experiment. — This is why, when the
maximum has been reached and the phial broken, we introduce
into the calorimeter a certain quantity of dilute hydrochloric
acid, rather more than would be required to exactly neutralise
the potash used in the experiment The temperature of this
acid is, moreover, obtained with exactness by special measure-
ment. A new reaction is immediately developed ; a reaction
which rather rapidly, but not instantaneously, converts the
potassium cyanate (iso-) into potassium chloride and carbonic
acid in solution. In order to avoid the liberation of this last
gas, in consequence of the liquid being mixed with air, this
liquid is agitated by means of a stirrer that moves horizontally

(P- W7).

The maximum is attained in six minutes. It exceeds the
original temperature by about 3°. It lasts three minutes, and
then the temperature begins to fall. The progress of this cool-
ing is followed during thirty minutes. The actual experiment
lasts about as long.

CALCULATIONS. 335

During all this time the temperature of the enclosing vessel
only varied from 21'21° to 21'37°.

Study of the cooling. — We then take a diluted solution of
potassium chloride, occupying the same volume as the above
liquid ; this is introduced into the same vessel and the
temperature raised so as to exceed by 3° that of the enclosing
vessel, just given. The progress of the cooling is again followed
for ten minutes ; then the excess of temperature is reduced to
2° by substituting a suitable volume of a cold solution of the
same composition for the heated solution contained in the
calorimeter. We then repeat the experiments on the rate of
cooling, corresponding to this last excess.1

Calculations. — We thus possess all the necessary data for
calculating the heat disengaged during the experiment.2 This
calculation is made without the aid of hypotheses, and according
to the rules already laid down for hydrocyanic acid (p. 306).

In this way we obtain the quantities of heat disengaged
during the two stages of the experiment ; or, ql -f- qz. The
correction due to the cooling is 8 per cent, for the first stage of
the experiment in question; it reaches 12 per cent, for the
second stage. The total quantity of heat disengaged represents
the conversion of the liquid cyanogen chloride into carbonic
acid in solution and ammonium chloride, plus the heat pro-
duced in the complete saturation of the potash employed by
dilute hydrochloric acid. P being the weight of this potash,
it would give off, if treated separately with hydrochloric acid,

p
— h 13 '6 Cal. Let $3 be the quantity of heat disengaged

by the conversion of the cyanogen chloride into carbonic acid
and ammonium chloride under the influence of pure water ; we
then get —

fc = 2i + ft - ~ 13-6 Cal.

We have now only to multiply the quantity, q3, by the inverse
ratio of the weight, p, employed to the equivalent of cyanogen

(\~\ •£*
chloride (61*5 grms.). It was found that q3 X = 61*68

Cal. (according to the average of the experiments). This value
represents the heat disengaged in the following reaction : —

CNC1 (liquid) + 2H20 + water = C02 (in solution) + NH4C1

(in solution).

1 For this method, see "Annales de Chimie et de Physique," 46 seYie,
torn. xxix. p. 158.

2 The only unknown quantity is the specific heat of liquid cyanogen
chloride. The approximate value, 0*4, was taken: it is near enough, on
account of the extreme smallness of the corresponding correction.

336 HEATS OF FORMATION OF THE CYANOGEN SERIES.

3. Heat of formation of cyanogen chloride from the elements.
This quantity is easily deduced from the above value. Let the
initial system be —

C + N + Cl + 2H2 +02 + water ;
and the final system —

C02 (gas) + NH4C1 (in solution).
We pass from one to the other, by two different methods :

FIEST METHOD.

C (diamond) +02 = C02 (gas) + 94-0

Solution of C02 + 5*6

HC1 = HC1 (in solution) + 39-3

NH3 = NH3 (in solution) + 21-0

Union of NH3 + HC1 = NH4C1 (in solution) ... + 12-45

Sum +172-35

SECOND METHOD.

+ 138

2(H2 + 0) = 2H20

C + N + Cl = CNC1 (liquid)

CNC1 (liquid) + 2H20 + water = C02 (in solu-
tion) + NH4C1 (in solution) + 61'7

Sum x + 199-7

whence,

x = - 27-35.

This is the quantity of heat absorbed during the formation
of cyanogen chloride from the elements —

C (diamond) + N + Cl = CNC1 (liquid) absorbed - 27'35.

We will pass on to the gaseous compound.

4. Vaporisation of cyanogen chloride. The heat absorbed in
this operation was measured in a direct manner, i.e. by sub-
merging in the water of a calorimeter (500 cc.) which was at
20° a phial containing a known weight, 2-069 grms. for
instance, of liquid cyanogen chloride. The phial had been
cooled beforehand to nearly zero and weighted with a platinum
wire, as already described. Only the upper point, instead of
being opened directly into the liquid of the calorimeter, was
fitted with a little worm through which the gaseous chloride
was given off. This gas was then conveyed outside the calori-
meter, and absorbed by a solution of potash. The operation
lasts about twenty-five minutes. All calculations being made,
the following number was found for the vaporisation of CNC1
( = 61-5 grms.) ; - 876 Cal. absorbed.

This value comprises —

(a) The heat of vaporisation of cyanogen chloride at + 12'7°.

(b) The heat absorbed by the liquid, which has been raised
from 1° to + 12-7°.

UNION OF CYANOGEN WITH CHLORINE. 337

(c) The heat absorbed by the gas, which has been raised from
127° to 197° (the average temperature during the vaporisation).

These two last quantities are relatively small. They could
only be estimated exactly if we knew the specific heats of
cyanogen chloride in the liquid and the gaseous states at
the temperatures indicated. In default of any direct data,
approximate values were used; the total value of these two
quantities, moreover, representing a very small quantity as
compared with the heat of vaporisation itself. Let us admit
for these specific heats the mean value + 0*4, deduced from
observation made on similar liquids. Consequently, the heat
absorbed in the accessory operations (b and c) will be estimated
at 0*46 ; a correction entailing probable error amounting to not
more than a fourth of its value. The heat of vaporisation of
cyanogen chloride will be, for CNC1 (= 61*5 grms.), — 8*3 Cal.

Thus the formation of gaseous cyanogen chloride from its
elements —

C (diamond) + N 4- Cl + CNC1 (gas), absorbs - 357.

These figures exceed in absolute value the heat absorbed in the
formation of hydrocyanic acid ; for —

H + C 4- N = HCN (gas), absorbs - 29'5.

The formation of cyanogen chloride from the elements is, there-
fore, endothermal, like that of hydrocyanic acid, but even in
a higher degree. This circumstance explains why cyanogen
chloride is so apt to undergo polymeric transformations and
other condensations.

5. Union of cyanogen with chlorine. From the above data
we deduce —

CN (gas) + Cl (gas) = CNC1 (gas) gives off + 37'3 - 357 =

+ 1-6.
CN + Cl = CNC1 (liquid) + 9-9.

These values were checked by another method.

We know that cyanogen chloride is easily formed by the
action of mercuric cyanide, in solution, on chlorine. The heat
disengaged in this operation was measured.

i[Hg(CN)2 (in solution) + 2C12 (gas) = HgCl2 (in solution)
+ 2CNC1 (in solution).

The value + 17'5 Cal. was found.

The calculation, supposing the cyanogen chloride to be
liquefied instead of being in solution, would give + 29'6 ; the
difference does not exceed the limits allowed in experiments of
this kind, or the differences which exist between solution and
liquefaction. This is, therefore, at least an approximate check
on the heat of formation of cyanogen chloride.

z

338 HEATS OF FORMATION OF THE CYANOGEN SERIES.

The heat of formation of liquid cyanogen chloride from
chlorine and cyanogen, or + 9*9, is quite comparable to the
heat of formation of iodine chloride and iodine bromide, under
the same form —

I (gas) + 01 = IC1 (liquid) + 9-8 ; IC1 (solid) + 12-1.
I (gas) + Br (gas) = IBr (solid) + 11-9.

This is another point of resemblance between cyanogen and
the halogens. In the formation of cyanogen chloride, as well
as in the combinations of the halogen elements with one another,
there is hardly any other heat given off than that which corre-
sponds to the change of state of the compound, i.e. to the con-
version of the gaseous body into liquid or solid.

6. Substitution of chlorine for the cyanogen. From the previous
results we conclude that the simple substitution of chlorine for
hydrogen in gaseous hydrocyanic acid —

Cl + HCN (gas) = CNC1 (gas) + H,

would absorb — 6'2. Such a reaction is therefore impossible,
unless accompanied by the formation of secondary products,
furnishing supplementary energy. On the contrary, the simple
substitution of chlorine for the cyanogen —

Cl + HCN (gas) = HC1 (gas) + CN (gas),

would give off + 14*2.

Lastly, the simultaneous formation of cyanogen chloride and
gaseous hydrochloric acid —

C12 + HCN (gas) = CNC1 (gas) + HC1 (gas), would give off

+ 15-8.

We see that this last formation answers to the maximum of
heat disengaged. This reaction, in fact, is produced in prefer-
ence to any other. It is all the more readily effected as the
combination between the cyanogen chloride and hydrochloric
acid gives off a fresh quantity of heat, at least if the substances
are in the liquid state, which again acts in the same way.
However, the effects of these direct reactions between chlorine
and free hydrocyanic aeid are complicated by various secondary
reactions which are imperfectly understood. The direct reactions
become more obvious, if we are operating on cyanides, the
corresponding bodies being always considered in comparable
forms :

KCN (solid) + da (gas) = KC1 (solid) + CNC1 (gas) would

give off + 39-0.
i[Hg(CN)2 (solid) + 2C12 (gas) = HgCl2 (solid) + 2CNC1 (gas)]

gives off + 21-3.

All these quantities of heat are positive. The formation of

CYANOGEN IODIDE. 339

liquid cyanogen chloride, together with mercuric chloride, gives
off -f- 7 '3 more ; or, altogether, -f- 29'6. If we suppose the
mercuric cyanide to be in solution, as well as the cyanogen
chloride, these figures do not vary to any appreciable extent. The
number -f 27*5, instead of + 29*6, was found by experiment.

7. We must also note the action of water on cyanogen chloride.
Water first dissolves it, and then slowly converts it into car-
bonic acid and ammonium chloride ; a reaction similar to that
in the case of amides. According to calculation —

CNC1 (liquid) + 2H20 (liquid) = C02 (gas) + NH4C1 (in solu-
tion) + 55*9.
CNCl(gas) + 2H20.(solid) = C02(gas) + NH4C1 (solid) + 65-4.

8. The heat of combustion of cyanogen chloride would be as
follows : —

CNC1 (gas) + 02 = C02 + N + Cl + 1297.

We know that this combustion does not take place directly.

§ 11. CYANOGEN IODIDE.

1. This substance was synthetically prepared from pure
potassium cyanide in aqueous solution (1 eq. = 6*5 litres)
and solid iodine. The reaction is easy and rapid. In a first
experiment, potassium cyanide was used, which had been pre-
pared beforehand and proved to be pure. It was found that

KCN (1 eq. = 6-5 litres) + I2 (solid) = CNI (in solution) + KI
(in solution) + 6'36 Cal.

In a second experiment, pure hydrocyanic acid was used, of
which a certain weight was dissolved in a diluted equivalent of
potash, so as to give a solution containing KCN (1 eq. = 2 litres).
On the addition of a corresponding quantity of solid iodine, the
value, + 6*21 Cal., was obtained. We will adopt the average,
+ 6-3.

2. Solution in water. The solution in water of crystallised
cyanogen iodide (6*3 grms. to 500 cc. of water) absorbs, at 20°,
- 2-78.

3. Formation from the elements.

Initial system : K 4- C (diamond) + N + I2 (solid) 4- water.
Final system : CNI (in solution) + KI (in solution).

FIRST STEP.

K 4 C (diamond) + N = KCN (solid) +30-3

Solution - 2-9

Keaction of I2 (solid) ... + 6- 3

Sum +33-7

z2

34U HEATS OF FORMATION OF THE CYANOGEN SERIES.

SECOND STEP.

K + I (solid) = KI (in solution) + 74-7

C (diamond) 4- N + 1 (solid) = CNI (in solution) ... x

Sum ... + 74-7 + x

X = - 41 ;

a value which applies to cyanogen iodide in solution. For solid
iodide, we get —

C (diamond) + N + I (solid) = CNI (solid), - 38-2.

4. Union of iodine with cyanogen.

CN + I (solid) = CNI (solid) + 0'9.
CN + I (gas) = CNI (solid) + 6-3.

These values are very little less than that of the heat of
formation of liquid cyanogen chloride, or + 9 -9. They are less
than the values relating to solid iodine chloride (+ 121) and to
solid iodine bromide (+ 1*9). The difference, however, is not
very great ; another confirmation of the general analogy of all
these components (see p. 338).

5. Substitution. The substitution of iodine for the hydrogen
of hydrocyanic acid, with the simultaneous formation of hydri-
odic acid and cyanogen iodide —

HCN (gas) + I2 (solid) = CNI (solid) + HI (gas), would absorb

- 131.

Thus this reaction does not take place in a direct manner. But,
on the contrary, we note the action of iodine on cyanides, which
takes place owing to the extra energy, due to the action of the
alkaline iodide. This is the calculation, all the bodies being
brought to a similar state —

KCN (solid) + I2 (solid) = KI (solid) + CNI (solid) ... + 13-3 Cal.

AgCN (solid) + I2 (solid) = Agl (solid) + CNI (solid) ... + 11-6 „
i[Hg(CN)2 (solid) + 21, (solid) = HgI2 (solid) + 2CNI (solid) ]+ 6-0 „

Here, it may be repeated, there is disengagement of heat due
to the formation of the metallic iodide : which formation is
necessary in order to ensure the combination between the
cyanogen and iodine.

6. It had already been found convenient to study the thermal
formation -of cyanogen bromide from bromide and dissolved
potassium cyanide, but this formation is almost immediately
followed by secondary reactions, which are indefinitely pro-
longed and cast a doubt on the numerical results observed at
first. Therefore it was thought advisable to suppress them.

§ 12. POTASSIUM CYAN ATE.

1. Pure potassium cyanate was decomposed by means of
dilute hydrochloric acid. If a quantity of water is used,

POTASSIUM CYAXATE. 341

sufficient to keep the whole of the carbonic acid formed in solu-
tion, the decomposition is complete at the end of a few minutes :

KCNO (in solution) + 2HC1 (in solution) + H2O
= C02 (in solution) + KC1 (in solution) + NH4C1 (in solution).
This reaction gives off, according to the experiment performed,
+ 28-8 Cal.

2. On the other hand, the solution of potassium cyanate,
KCNO (1 part of salt to 300 parts of water), absorbs - 5'2.

3. Formation of potassium cyanate from the elements. This is
deduced from the above data.

Initial system :

K + C (diamond) + N + 2H2 + 02 + Cl*
Final system :

C02 (in solution) + KC1 (in solution) + NH4C1 (in solution).

FIRST STEP.

= C02 ......... - 94-0

Solution ............ + 5-6

K + Cl = KC1 (in solution) ... + 100- 8

N + H3 = NH3 (in solution) ... + 21-0

H + Cl = HC1 (in solution) ... + 39-3

HC1 + NH3 = NH4C1 (solution) .. + 12-45

+ 273-15

SECOND STEP.

K+C+N+0= KCNO (solid) a

Solution ............ — 5-2

2(H + Cl) = 2HC1 (diluted) ... + 78-6

H2 + 0 = H20 ...... + 69-0

+ 142-4 + x
Reaction ... + 28-8

+ 171-2+a;
x = 273-15 - 171-2 = + 102.

Thus, the formation of solid potassium cyanate from the
elements, K +C (diamond) + N + O = KCNO, gives off + 102-0.
If the salt is in solution, •+ 96*8.

This same formation, from diluted potash —

i[C2 + N2 + 0 4- K20 (diluted) = 2KCNO (in solution)], gives

off + 15-5.

From gaseous cyanogen —

CN + K + 0 = KCNO (solid) ...... +139-3

M(CN)2 + 0 + K20 (diluted) = 2KCNO (in solution)] ... +51-8
CN2 + K20 (diluted) = KCNO (diluted) + KCN (diluted) + 34-2

4. All these values exceed the heats disengaged in the analo-
gous reactions of the halogen elements properly so called.

342 HEATS OF FORMATION OF THE CYANOGEN SEEIES.

For example —

C12 (gas) + K20 (diluted) = KC10 (diluted) + KC1 (in solution)
gives off only -j- 2 5 '4.

There is, besides, the difference that the complex nature of
cyanogen and its tendency either to form polymeric compounds
and other condensed bodies, or to reproduce ammonia and its
derivatives, gives rise to a number of secondary reactions,
which are not observed in the case of chlorine. These reactions
take place all the more readily in proportion as the heat
disengaged by the direct reaction is itself greater, and therefore
furnishes a greater reserve of energy for other transformations.

5. The union of dry potassium cyanide with gaseous oxygen,
to form solid cyanate —

KCN (solid) + 0 (gas) = KCNO (solid), would give off + 102
- 30-3 = + 717 ;

a very high value, being nearly three-quarters of the heat
(+ 94'0) disengaged in the combustion of the carbon contained
in the cyanide.

These figures relate to the bodies when considered in their
actual state ; under which condition, up to the present, the
absorption of oxygen by potassium cyanide has not been
observed, perhaps because it has not been looked for. On the
contrary, if the potassium cyanide be in the melted state, the
absorption takes place easily, as we know. Now, the values
which have just been calculated may be applied approximately,
at a high temperature, to the same bodies in the known
conditions of their actual reactions ; for the melting of the
cyanide and that of the cyanate must absorb quantities of heat
which are little different. As regards the heat disengaged by
the oxidation of its potassic compound, cyanogen resembles
iodine, but, on the contrary, differs from chlorine. In fact, we
get—

KC1 + 03 = KC103 (solid) absorbs -11

KBr + 03 = KBrO3 (solid) „ - 11-1

KI + 08 = KI03 (solid) gives off + 44-1

KCN + 0 = KCNO (solid) gives off + 71-7

a progression which is the reverse of that which characterises
the union of the same metal, such as potassium, with the same
series of halogen bodies, such as chlorine (+ 105*0), gaseous
bromine (+ 100*4), gaseous iodine (+ 8 5 '4), and cyanogen
(+ 67-6).

We can understand, from the above figures, why potassium
cyanide shows such a great tendency to become oxidised, either
under the influence of oxidising agents, or even under the
influence of the air.

The combustible nature of one of the elements contained in
cyanogen is, besides, opposed to its forming the peroxygenated

DECOMPOSITION OF CYAN ATE. 343

acids which are produced in the case of chlorine and the halogen
elements ; such compounds would have too great a tendency to
be converted into carbonic acid.

The total combustion of potassium cyanate in the solid state —

i[2KCNO + 03 = K2C03 + C02 + N2], would give off + 83*9.

6. The facility with which potassium cyanate is converted
into ammonia, even by the simple fact of its prolonged contact
with water, is easily explained ; for —

£[2KCNO (in solution) + 4H20 = K2C03 (in solution)
+ (NH4)2C03 (in solution)], gives off + 20 Gal.

This is again a reaction of amides.

The well-known conversion of melted potassium cyanate, by
means of aqueous vapour, into melted potassium carbonate,
carbonic acid and ammonia gas, gives off about + 9 Cal.

The change of potassium cyanide under the united influence
of oxygen and aqueous vapours at a high temperature into
carbonate and ammonia — a change which is so pernicious in the
industrial preparation of the prussiates — is as easily explained
by thermo-chemistry. In fact, we should get, at the ordinary
temperature —

J[2KCN (solid) + 02 + 3H20 (gaseous) = K2C03 (solid)
+ C02 (gas) + 2NH3 (gas)] + 79-3.

Towards a red heat this value should still keep the same, as
the cyanide and the carbonate are similarly fused.

A rapid resume has been made of the more immediate
deductions from the new values relating to the heat of forma-
tion of cyanogen, hydrocyanic acid, and cyanides. It would be
easy to develop and extend these conclusions to innumerable
other reactions, the subject being most fruitful. Any one can
do this as far as he deems expedient or interesting. The general
table of the thermal formation of the cyanogen compounds will
be found on p. 132.

( 344 )

CHAPTEK XII.

HEAT OF FORMATION OF THE SALTS PRODUCED BY THE OXYGEN-
ATED COMPOUNDS OF CHLORINE AND OTHER HALOGEN

ELEMENTS.

§ 1. GENERAL NOTIONS.

CHLORINE and the halogen elements form with oxygen a series
of compounds analogous with the oxygenated compounds of
nitrogen, and which comprise even an additional member, viz.
perchloric acid. Most of these compounds are energetic oxidis-
ing agents, either in the wet or in the dry way, and some of
their salts (especially the chlorates) play an important part in
the manufacture of explosive substances. This circumstance,
therefore, makes it desirable to measure their heat of formation.

§ 2. THERMAL FORMATION OF CHLORIC ACID AND CHLORATES.

1. The thermal formation of chloric acid and chlorates has
been already examined by Favre, Frankland, and Thomsen, but
with very different results.

Favre 1 tried to measure the heat liberated in the action of
gaseous chlorine on concentrated potash ; according to him, the
union of chlorine and oxygen to form chloric acid —

i[H20 + C12 + 0B + water = H20, C1205 (diluted)],

would absorb — 65 '2 Cal. The decomposition of solid potassium
chlorate into oxygen and potassium chloride —

KC103 = KC1 + 03, would then liberate + 64-9.

But the learned author's calculation is complicated, and is based
on uncertain data, such as the supposed insolubility of the salts
formed in the alkaline solution.

Frankland,2 having oxidised various organic substances, some

1 "Journal de Pharmacie et de Chimie," 36 s£rie, torn. xxiv. p. 316. 1853.

2 " Philos. Magazine," vol. xxxii. p. 184. 1866.

BARIUM CHLORATE. 345

with free oxygen, others with potassium chlorate, found in three
trials that the excess of heat developed by 9 -75 grms. of
potassium chlorate amounted to 378, 326, and 341, respectively ;
or, taking the average, to 348 Cal., which would give for

KC103( = 122-6 grms.) -h 4'37 Cal.,

a value subject to the doubts involved by such an indirect
determination.

Finally, Thomsen 1 reduced a diluted solution of chloric acid
by means of sulphurous acid — an operation which can be easily
carried out in a calorimeter ; on the other hand, he decomposed
potassium chlorate by means of the heat produced by the com-
bustion of hydrogen — a process which does not seem to admit
of such accuracy as the former one.

He deduced from his trials the following results : —

i[H20 + C12 + 05 + water = H20, C1205 (diluted)], - 10-2
KC103 (solid) = KC1 (solid) + O3, + 9-8.

2. The following are the results arrived at by the author.
Following the method he constantly adopted, he took, to start
with, a crystallised and definite salt, in preference to a titrated
solution, or a solution containing the acid prepared by precipita-
tion — solutions whose composition is always less exact.
Barium chlorate was used, which was in very fine crystals, and
corresponded to the formula Ba(C103)2 + H20. Analysis of
this salt gave —

BaS04 H20

Found 72-2 5'7

Calculated 72-3 5'6

This salt, when dehydrated and then heated in a tube, is im-
mediately decomposed with a very marked incandescence, and
a kind of explosion which throws off to some distance a white
powder, consisting of barium chloride ; these effects are observed
even when only a few grammes of the dry salt are operated
upon. It is well known that analogous phenomena may be
observed with potassium chlorate, but barium chlorate exhibits
them to a far greater degree. This proves that its decomposition
is exothermal.

3. A known weight, 2'5 grms., of this salt was dissolved, in
one case in 400 cc., in another in 900 cc. of water, and
reduced by means of 100 cc. of a moderately concentrated
solution of sulphurous acid. The barium chlorate is thus com-
pletely converted into barium sulphate and hydrochloric acid,
as was ascertained by various determinations. The reduction
is effected more rapidly according as the solution is less diluted,
all other things being equal. With 500 cc. of liquid, and 2'5

1 " Journal fur Praktische Chemie," Band xi. s. 138. 1875.

346 OXYGENATED COMPOUNDS OF CHLORINE.

grms. of salt, at 12*5°, it lasted six to seven minutes ; with
1000 cc., and the same weight of salt, it lasted fourteen to
fifteen minutes. At 22° the duration of the reactions was found
to differ but little ; which proves that the preceding difference
does not depend on the unequal heating produced by the re-
action (6-5° in the first trial, 3*2° in the second). The sulphurous
acid should be considerably in excess ; when operating with only
a very slight excess, the reaction effected at 23° lasted nearly
twenty minutes, instead of seven.

All these durations of the reactions can be clearly defined
according to the movement of the thermometer compared with
the rate of cooling of a similar liquid of the same weight, and
brought to the same heat, but in which no chemical reaction is
produced.

4. For the reaction

4[Ba(C103)a dissolved + 6 S02 (dissolved) = BaS04 (precipitated)
+ 2HC1 (dilute) + 5H3S02 (dilute)],

the following quantities of heat were obtained : —

Initial Heat liberated for

temperature. Weight of salt. equivalent weights.

23° .. 2-500 Ba(C108)a + HaO ... 213-8 Cal.

2-500 215-2 „

1-544 Ba(C103)a anhydrous ... 215-2 „ (2 trials)

2-500 Ba(C103)a + HaO ... 214-3

23°
22°
12°
12°

it «JW/ J-MVl V^l\_/o la T AAoV' ••• •** «* ),

2-500 Ba(C10s)a + HaO ... 212-3 „
At 19° we should have, on an average ... 214-3 „

On the other hand, direct trials gave for the heat liberated by
the action of dilute sulphuric acid on barium chlorate taken
with the same degree of dilution as in the preceding experi-
ments—

i[H2S04 dilute + Ba(C103)a (dissolved)], at 19°, + 4-6 ;

whence we obtain for the union of dilute chloric acid with
baryta —

i[2HC103 (dilute) + BaO (dissolved)] + 13-8 ;
and for the reduction of free chloric acid —

^[2HC1O3 (dilute) + 6S02 (dissolved) = 6H2S04 (dilute)
~ + 2HC1 (dilute)] liberates + 214*3 - 4-6 = + 209-7.

5, Let us note now the following data : —

J[SOa (dissolve^ + 2H90 + Clf (gas) = HaS04 (dilute)

+ 2Iiri ^dissolved)] liberates + 36*95 l

H + Cl = lirl Dilute) + 39-3

tfH, + 0 = HaO (dilute)] + 34-5

1 Thomson.

CHLORIC ACID. 347

We deduce from these numbers —

J[S02 (dissolved) 4- 0 (gas) + H20 = H2S04 (dilute)] 4- 3215;
and consequently —

i[H20 4- C12 + 06 -f water = H20, C1205 (dilute)] - 12'0.
This number depends on the heats of formation of water,
hydrochloric and sulphuric acids.

6. Hence it is that the formation of dilute chloric acid from
its elements —

£[C12 4- 06 4- H2 4- water = H20, C1205 (dilute)], liberates

+ 22-5.

The conversion of dilute chloric acid into dilute hydrochloric
acid and gaseous oxygen —

HC103 (diluted) = HC1 (diluted) + 03, liberates 4- 16'8.

This value plays an important part in oxidations.

7. It is the same for dissolved chlorates, decomposed into
chlorides and free oxygen ; for the heat liberated in the action
of various bases on hydrochloric and chloric acids is essentially
the same. In fact, the following numbers were found at
19°:—

J[2HC1 (diluted)] J[2HC10, (diluted)]

+ *[K20 (diluted)] +13-7 +13-7

+ |[Na20 (diluted)] + 13-7 + 13-7

+ |[BaO (diluted)] + 13-85 + 13-8

8. Returning to the heats of solution of chlorides and chlorates,
the values of which have been given elsewhere,1 we obtain the
heat of decomposition of chlorates into chlorides and oxygen
(referred to the ordinary temperature) —

KC103 (solid) = KC1 (solid) 4- 03 4- ll'O ;
instead of 4- 9 '8, given by Thomsen. These values are very near
each other. It was also found that

NaC103 (solid) = NaCl (solid) 4- 03 4- 12'3.
J[Ba (C103)2 (solid) = BaCl2 (solid) 4- 30 J 4- 12-6.

Even at the temperature of the reactions, that is to say, at
500° or 600°, the quantities of heat, for solid salts, are nearly
the same, as may be established by calculation. For instance,
the specific heat of potassium chlorate for the equivalent weight
KC103 is equal to 23'8 ; for KC1 4- 03, the sum of the specific
heats amounts to 23 '3. The term U — V, which expresses the
variation of the heat of combination, is therefore equal to

4- 0-5 cal. (T - t) ;

or, for an interval between zero and 500°, 0'25 Cal., which is
an insignificant increase in comparison with 4-11-0.

1 " Annales de Chimie et de Physique," 5' se*rie, torn. iv. pp. 103. 104.

348 OXYGENATED COMPOUNDS OF CHLOKINE.

It follows from these numbers that combustion effected ly
solid potassium chlorate liberates more heat than the same com-
bustion effected by means of free oxygen; viz. by + T83 Cal.
for each equivalent of oxygen (0 = 8) consumed (p. 134).

9. The formation of chlorates from the elements —

K + Cl + 03 = KC103 (solid) liberates + 94'6 Cal.
Na + Cl + 03 = NaC103 (solid) liberates + 85'4 Cal.

These quantities scarcely vary with the temperature ; at least
when the metals are solid. In fact, the specific heat of the
system of elements, K + Cl + 03, is 21*3 ; heat of the compound
KC103 is 23-8. We have then U - V = - 2-5 cal. (T - t)t
or, what is the same thing, - 0'0025 Cal. (T - t)y if we adopt
the same unit as for the formation of chlorates. An interval of
100°, then, only produces an increase of - 0'25 Cal. in the heat
liberated.

10. Various reactions. The action of gaseous chlorine on
diluted potash may be considered as forming either hypo-
chlorite, or chlorate, or free oxygen.

(a) With hypochlorite —

6C1 + 3K20 (diluted) = 3KC10 (dissolved) + 3KC1 (diluted).

According to the experiments performed,1 the reaction liberates

+ 25-4 x 3 = + 76-2.

With soda, we have, + 75'9 ; with baryta, + 75'8.

(b) This reaction may also form chlorate —

6C1 + 3K20 (diluted) = KC103 (dissolved) + 5KC1 (dilute),

which liberates, with potash, + 94'2 ; with soda, + 94'2 ; with
baryta, + 95'0.

(c) The formation of potassium perchlorate and chloride,
referred to the same weight of chlorine as the preceding —

|[8C1 + 4K20 (diluted) = 7KC1 (dissolved) + KC104
(dissolved)], liberates + 11 TO.

With soda, + 111-0 ; with baryta, -f 111-8.

(d) Finally, the same reaction may develop chloride and free
oxygen — p

6C1 + 3K20 (diluted) = 6KC1 (dissolved) + 03,

which liberates, with potash, + 111-0; with soda, + 110*0; with
baryta, + 111'8.

It follows from these numbers that the formation of the
hypochlorite corresponds to the least liberation of heat ; then
comes the chlorate, and finally the perchlorate and free oxygen,
which liberate the most heat, the two quantities being, moreover,

1 " Annales de Chimie et de Physique," 5e seVie, torn. v. pp. 335, 337, 338.

6C1 + 3H20 + water = 3HC10 (diluted) + 3HC1 (diluted) liberates

SUCCESSIVE DEGREES OF OXIDATION. 349

sensibly the same. When hypochlorite is changed into
chlorate —

3KC10 (dissolved) = KC103 (dissolved) + 2KC1,
heat is liberated to the amount of

-f 18'0, for potassium salts ; 4- 18'3, for sodium salts ; + 19'9,
for barium salts.

The second decomposition, that of the dissolved chlorate into
chloride, liberates -f- 16*8 for the three salts as before stated.

The conversion of dissolved chlorate into perchlorate liberates
sensibly the same quantity of heat.

It is seen that the relative stability of the solutions keeps
increasing from the hypochlorite to the chlorate, then to the
perchlorate and to free oxygen, which is consistent with what
we know from chemistry.

Finally, if we refer the actions to the formation of the acids
themselves, which gives at least the heat liberated by their
union with the bases, we have —

+ 5-7

_____ + 12-0
liberates +28-9
28-8

The thermal relations, therefore, remain the same.

11. Successive degrees of oxidation. Let us now examine the
heats of formation of the different acids of chlorine.

From the experiments performed —

A2 + C12 + 02 + water = H20, C1202 (dilute)] absorbs ... - 2'9
H20 + C12 + 05 + water = H20, C1206 (dilute)] absorbs - 12-0

H20 + C12 + 07 + water = H20, C1207 (dilute)] liberates + 4-9

There is then, in the first place, an absorption of heat which
increases according as the proportion of oxygen united to equal
weights of chlorine increases in the compound ; this is the case
for the first two compounds.

But, on the contrary, heat is liberated in the case of the
third, to the extent of -f- 16*9 for 0 ; formation of perchloric
acid.

The same relations subsist if we take oxygen as the unit and
vary the chlorine. For a given weight of oxygen, such as
£[05] = 40 grms., united to 01 = 35 '5 grms. in dissolved
chloric acid, there is an absorption of — 12'0. The same weight
of oxygen, when united to C16 = 177*5 in dissolved hypo-
chlorous acid, gives rise to an absorption of

-2-9x5=- 14-5,

which is more considerable. But this increase in the heat
absorbed does not extend to perchloric acid ; it, on the contrary,

350 OXYGENATED COMPOUNDS OF CHLOKINE.

gives rise to a liberation of heat of -f 3 '5, for the same weight
of oxygen.

These relations are the more remarkable in the acids of
chlorine, since the formation of successive combinations of one
and the same element with oxygen generally liberates heat, as
is shown by the history of the oxygenated combinations of
sulphur, selenium, phosphorus, arsenic, etc.

12. Nevertheless, similar anomalies are found in the study of
the combinations of iodine and nitrogen with oxygen. In fact, if
we compare hypoiodous, iodic, and periodic acids, J [I2 + 0 -f-
water = I20 (in solution)] absorbs, according to the author's
experiments, a quantity of heat notably superior, in absolute
value, to — 5*2.

*[L + Ofi + water = L05 (dissolved)] liberates ... + 21-5 (Thomsen)

or ... + 22-2 (Berthelot)
i[I2 + 07 + water = I207 (dissolved)] liberates ... + 13-5 (Thomsen)

Hence it is seen that the heat liberated presents a minimum
and a maximum, neither of which corresponds to the highest
degree of oxidation. The combinations of nitrogen and oxygen
present an analogous minimum for nitric oxide (p. 84). These
numbers are given here in order to show how difficult it is to
generalise the relations between the quantities of heat disengaged
or absorbed, and the multiple proportions of the successive
combinations of two elements.

If the minimum of heat liberated or the maximum of heat
absorbed corresponded in all cases to the first term formed by the
successive union of two elements, and if the heat liberated then
increased regularly with the proportion of the variable element,
it might be supposed that one of the elements — the one con-
sidered as constant — undergoes a special isomeric modification
preceding the combination, and from which, as a starting-point,
the quantities of heat should be reckoned. But it seems
difficult to admit this hypothesis in the oxygenated series of
nitrogen, iodine, and even chlorine, series in which the thermal
minimum and maximum correspond neither to the first nor to
the second degree of oxidation.

13. It was thought desirable to pursue the comparison
further, and to extend it to chlorous acid. To this end it was
attempted to prepare a definite salt, barium chlorite, which,
according to Millon, should be a crystallised salt. By closely
following the author's instructions, a crystallised salt was
obtained, presenting a scaly appearance — as he says — and
giving by analysis numbers which essentially correspond to the
formula Ba(C102)2; but a closer examination showed that the
salt was nothing but a mixture of barium chloride and per-
chlorate in equivalent proportions (with a small percentage of
chlorite) —

BaCl2 + Ba(C104)2.

PERCHLORIC ACID. 351

This substance presents the same percentage composition as the
chlorite. It appears that its formation in the action of chlorous
acid on the alkalis must account for the discoloration of the
solution which takes place, as is known, after some time.

§ 3. THERMAL FORMATION OF PERCHLORIC ACID AND ITS SALTS.

1. The result of the researches made on the oxy acids of
chlorine and other halogen elements led to the study of the
heat of formation of perchloric acid; the results, which are
with great difficulty obtained, demonstrate a certain number of
new chemical facts. They show at the same time how thermo-
chemistry explains the differences of stability and activity which
exist between pure perchloric acid and the same acid when
combined with a more considerable quantity of water.

2. In fact, it is known, principally from the researches of
Eoscoe,1 that there exist several hydrated perchloric acids,
namely, monohydrated acid, properly so called, or HC104, a
crystallised hydrate, HCIOJIA and a hydrate, HC1042H20,
volatile at 200°, and partly dissociable, even under the conditions
of its distillation.

These experiments were reproduced; it was even succeeded
in obtaining the first acid in the crystallised form. It suffices
to take the liquid acid, which contains a slight percentage of
water in excess, and to place it in a cooling mixture. The acid
becomes crystallised, and the mother-liquid is decanted. It is
allowed to liquefy, and then recrystallised, which finally gives
an acid fusible at 15°, a point of fusion which is still not high
enough. Its composition was proved by analysis. It is a body
which eagerly absorbs water and emits dense fumes on contact
with the air.

3. The solution of monohydrated liquid HC104, at 19°, in one
hundred times its weight of water, liberates + 20*3 Cal. This
is rather a delicate experiment, owing to the rapidity with
which the acid attracts moisture while being weighed, and also
to the violence with which it reacts on the water at the time of
the calorimetric part of the experiment. The preceding figure
is enormous; it exceeds the heat of solution of all ordinary
monohydrated acids — being, for instance, more than double that
of hydrated sulphuric acid, H2S04. It is even nearly equal to
the heat of solution of anhydrous sulphuric acid (+ 187) and
anhydrous phosphoric acid (+ 20'8), which are the highest
hitherto known; but they refer to anhydrous bodies. The
figure + 20*3 also exceeds the heats of solution of hydracids,
although the latter are increased from 6 Cal. to 8 Gal, owing to
their gaseous state.

This enormous heat of hydration of perchloric acid explains
3 " Annalen der Chemie und Pharmacie," Band cxxxi. s. 376. 1861.

352 OXYGENATED COMPOUNDS OF CHLOKINE.

the great difference which exists between the reactions of this
acid when diluted with water — a condition under which it is
almost as stable as dilute sulphuric acid — and those of the
monohydrated acid, which ignites hydriodic acid gas, and acts
with explosive violence on oxidisable bodies. This subject will
be referred to again later on.

4 Monohydrated perchloric acid decomposes spontaneously,
as Eoscoe observed. It is at first colourless, but it assumes a
yellow, then a red and brownish-red colour, and eventually
liberates gases which render the containing vessels liable to
burst ; this is the more to be feared since the necks of emery
flasks soon become stopped up through the formation of crystals
of the second hydrated perchloric acid.

The acid which has suffered a partial decomposition is not
suitable for measuring the heat of hydration on account of its
becoming less and less considerable, owing to the formation of
water which accompanies this decomposition. Notwithstanding
this formation of water, the acidimetric value of the acid —
referred to the equivalent weight of perchloric acid — does not
fall, and it may even apparently increase a little, since the
lower oxygenated acids of chlorine have an equivalent less than
that of perchloric acid. This is a cause of error which should
be noted.

5. A similar decomposition is produced under the influence
of heat, and prevents the re-distillation of perchloric acid. It
takes place in the very conditions of its preparation from
potassium perchlorate and sulphuric acid, as is shown by the
constant liberation of chlorine which accompanies the dis-
tillation. It appears that the monohydrated acid cannot be
separated unless carried away in a current of gas ; it is only
obtained in small quantities. This depends on the fact that the
decomposition of perchloric acid liberates heat. Even in its
preparation from potassium perchlorate and concentrated sul-
phuric acid, the reaction, when once started by the action of an
external source of heat, continues of itself, after the removal of
this source, and sometimes with a violence sufficient to give
rise to an explosion. This fact proves that the reaction is
exothermal. At the same time, chlorine and oxygen are
liberated, which carry away the perchloric vapour and render
its condensation difficult.

6. Some details may now be given respecting the oxidising
characters exhibited by perchloric acid.

In dilute solution, this acid is not reduced by any known
body. No action is caused by sulphurous, hydrosulphuric,
hydrosulphurous,1 or hydriodic acid, by free hydrogen, zinc

1 It was especially proved by accurate weighing, that this acid, which was
recently declared capable of reducing perchlorates, does not act in reality
except on the small quantities of chlorates often contained in perchlorates.

MODES OF DECOMPOSITION OF PERCHLORIC ACID. 353

in the presence of acids, sodium amalgam in the presence
of pure acidulated or alkaline water, or by electrolysis. Per-
chloric acid and perchlorates in solution are even as stable as
sulphates.

The hydrates HC1O4, 2H2O (liquid) and even HC104, H20
(crystallised), hydrates whose heat of solution only rises to
-f 5-3 Cal. for the former, and + 7*7 Cal. for the latter,1 seem
to be scarcely more active than the diluted acid itself, from
determinations made with hydriodic gas, sulphurous acid, and
arsenious acid.

Monohydrated perchloric acid acts quite differently ; which
can be explained by the fact that it also liberates -f- 20*3 Cal.
in its solution. When brought into contact with oxidisable
bodies it sometimes remains almost inactive just as in the case
of nitric acid and iron (the passive state) ; on the other hand, it
sometimes attacks them suddenly and with explosive violence.
It causes the ignition of hydriodic acid and sodium iodide ; it
attacks arsenious acid, etc,, very energetically. With hydro-
genised bodies the formation of water modifies the action — by
converting a portion of the acid into a higher hydrate.
Arsenious acid does not present this disadvantage ; it produces
an oxychloride, intermediate between this body and arsenic
acid, and to which reference has already been made while
treating of the reciprocal displacements of oxygen and halogen
bodies.2 Nevertheless, it has been impossible to utilise this
reaction for calorimetric measurements, even by dissolving these
products in soda, owing to the uncertain constitution of the
arsenic acid formed, which presents differences similar to those
of the several phosphoric acids. Hence it is that the saturation
of arsenic acid by soda liberates much less heat than that of
normal arsenic acid. It is this fact that interferes with all the
calculations.

7. It is only necessary to cite the following figures, which
show the multiplicity of the simultaneous modes of decom-
position of perchloric acid.3 1*175 grm. of this acid, in
presence of a great excess of arsenious acid, was distributed in
the following manner: 0*264 grm. yielded all its oxygen (04)
to the arsenious acid ; 0*139 grm. was destroyed in HC1
-f 04, which fixed itself on the same acid ; 0*145 grm. in C12
+ 07 (fixed on the arsenious acid) -f H20 ; 0*645 grm. left
unchanged.

According to a special determination made, only a few
milligrammes formed chlorous acid.

The heat liberated in the combination of perchloric acid with
various bases, at 18°, was measured.

1 About + 11*7 in the liquid state.

2 " Annales de Chimie et de Physique," 5" serie, torn. xv. p. 211.

3 See pp. 6, 7.

2 A

354 OXYGENATED COMPOUNDS OF CHLORINE.

i[2HC104 (1 equivalent = 6 litres) + Na20 (1 equivalent = 6

litres)] disengages + 14'25

£[2HC104 (1 equivalent = 6 litres) + 2 equivalent Na20 (1 equi-
valent = 6 litres)] disengages + 0'07

£[2HC104 (1 equivalent = 6 litres) 4- BaO (1 equivalent = 6

litres)] disengages +14-47

i[2HC104 (1 equivalent = 6 litres) + 2 equivalent BaO (1 equi-
valent = 6 litres)] disengages 4- 0'08

HC104 (1 equivalent = 6 litres) + NH3 (1 equivalent = 4 litres)

disengages + 12'90

HC104 (1 equivalent = 6 litres) + 2 equivalent NH3 liberates ... + O'OO

Potash liberated the same quantity of heat as soda, but the
potash solutions were taken twice as weak in order to avoid
precipitation of the perchlorate.

The heat of solution of the perchlorate may now be added :

KC104 + water absorbs —12*1.

NaC104 (at 100°) - 3-5

i[Ba(C104)2] at 100° - 0-9

NH4C104at20° ... ... ... - 6-36

8. Let us moreover examine the heat of formation of perchloric
acid and perchlorates from their elements.

The author and M. Vieille determined the heat of formation
of potassium perchlorate by mixing it in precisely equivalent
proportions with a combustible substance, such as potassium or
ammonium picrate, explosive in itself and therefore capable of
giving rise to an instantaneous reaction. This same substance
being burnt, on the other hand, by means of free oxygen, the
difference between the two quantities of heat measured
represents the excess of heat developed by the reaction by
means of free oxygen over the heat developed by the reaction
by means of combined oxygen : that is, the heat absorbed (or
liberated) by the decomposition of potassium perchlorate into
free oxygen and potassium chloride —

KC104 (solid) = KC1 (solid) + 202.

This quantity is derived from two experimental data only ; it is
independent of the heats of combustion of potassium, carbon,
and hydrogen, as also of that of ehlorination of potassium.

It was found each time that the weight of the potassium
chloride formed (determined as silver chloride) was within ^J^
of that which corresponded to the complete decomposition of the
perchlorate. On the contrary, the combustion of the picrate
was not found to be complete when the operations were con-
ducted in an atmosphere of nitrogen; a certain deficit being
noticeable in the carbonic acid, which is accounted for by the
free carbon and carbon monoxide.1 For this reason it was

1 A corresponding fraction of the ox}'gen of the perchlorate is liberated
owing to the simultaneous decomposition of this salt, but this in no way affects
the calculation. Moreover, let us bear in mind that the combustion of
potassium picrate converts the potash into bicarbonate, as was proved.

HEAT OP FORMATION OF PEKCHLORIO ACID. 355

deemed advisable to operate in an atmosphere of oxygen, which
completes the combustion, as was proved by the determination
of the carbonic acid.

Three series of experiments were made; namely, with
potassium picrate, ammonium picrate, and picric acid ; but
only the first two gave satisfactory results, as the combustion
of the picric acid was never complete, probably owing to its
being, to a certain extent, volatile.

The numbers obtained with potassium picrate burnt, in the
one case by pure oxygen, and in the other by perchlorate,
differ by — 8*6 Cal. ; the numbers obtained with ammonium
picrate, by — 6*5 Cal. ; results which agree as closely as could
be expected for values which represent the difference of
numbers that are far too high. We shall adopt the mean
— 7'5 Cal. as corresponding to the reaction —

KC104 (solid) = KC1 (solid) + 202 gas.

This decomposition, when effected at the ordinary temperature,
would then absorb heat, contrary to what takes place in the
decomposition of potassium chlorate, which liberates heat to
the amount of + ll'O Cal.

It is easy to calculate the heat of formation of potassium
perchlorate from the elements. For, if we admit that

K + Cl = KC1 (solid) liberates + 105-0 Cal.

K + Cl + 04 = KC104 (solid) liberates + 112-5 „

From this figure and the preceding data it follows that

H + Cl + 202 = HC104 (liquid, pure) disengages +19-1

H + Cl + 202 + water = HC104 (diluted)

K + Cl + 202 = KC104 (solid) + 112-5 dissolved

Na + Cl + 202 = NaC104 (solid) + 100-2 dissolved

N + H4 + Cl + 202 = NH4C104 (solid) + 79-7 dissolved

+ 39-35
+ 100-4
+ 96-7
+ 73-3
KC103 + 0 = KC104 (solid) + 17-9

9. We derive from these figures —

HC104 (pure, liquid) = HC1 (gas) + 04 liberates ... ... + 2'9

i[2HC104 (pure, liquid) = C12 + 07 + H20 (gas)], + 9'9 ; H20

(Kquid) + 14-9

HC104 (diluted) = HC1 + 04 nil

J[2HC104 (diluted) = Cl, + 07 + H20 liquid)] - 4-9

numbers which account for the difference between the stability
of the concentrated and diluted acids, and also for the easy decom-
position of the concentrated acid.

10. The perchlorates in solution are converted into chlorides
with scarcely any thermal phenomenon, but it is different with
the solid salts. In fact —

KC104 (solid) = KC1 (solid) + 202 - 7-5

NaC104 (solid) = NaCl (solid) +202 - 3-0

£[Ba(C104)2 solid = BaCl2 (solid) + 402] ... - 1-1

The conversion of a solid perchlorate into chloride, at the
Ordinary temperature, therefore, absorbs heat ; that is to say, it

2 A2

356 OXYGENATED COMPOUNDS OF CHLORINE.

could not become explosive; whereas, according to the
measurements made, the contrary is the case with chlorates.

The sign of the phenomenon, however, does not seem neces-
sarily to change with elevation of temperature ; the molecular
heat of potassium perchlorate (2 6 -3), for instance, being smaller
than the sum of those of the chloride and oxygen (33'9) ; that is
to say that, towards 400°, the divergence in absolute value
would be increased by about 3 Cal.

11. The conversion of potassium chlorate into perchlorate is
therefore exothermal, as might be foreseen.

At the ordinary temperature,

4KC103 = 3KC104 + KC1 would liberate + 63.
This is, moreover, in conformity with the thermal relation
already observed between the hypochlorites and chlorates, the
latter being more stable than the former, but also being formed
with a smaller absorption of heat.

12* The thermal relations also show that the decomposition
of ammonium perchlorate must be explosive, for —

NH4C104 (solid) = Cl + 02 + N + 2H20 (liquid) liberates

+ 58-3 Cal.

NH4C104 (solid) = Cl + 02 + N + 2H20 (water as vapour)
liberates + 38'8 Cal.

With the salt in a melted state we should have, in addition, the
heat of fusion. This is verified by experiment. In fact,
ammonium perchlorate^ when heated, first melts ; then the
liquid becomes incandescent, assuming a spheroidal form; the
brilliant bead thus produced is decomposed with great rapidity
into free chlorine, oxygen and water, with the production of a
yellowish flame. The salt does not, however, detonate; at
least when only a small quantity is operated on. These
phenomena resemble those of the decomposition of ammonium
nitrate (nitrum flammans), but possess rather more intensity.

13. Reference has been already made to the great heat of
solution (+ 20*3) of hydrated perchloric acid HC1O4, which is
more than double that of all the other monohydrated acids, and
is comparable to that of the most powerful anhydrous acids.
The greatness of the heat liberated is followed up to the
secondary hydrates.1 That of the second hydrate,

HC104 (liquid) + H20 (liquid) = HC104, H20 ;

liberates (the hydrate being solid) -f- 12*6 Cal. ; and, if con-
sidered as a liquid, about + 8*6. The formation of the third
hydrate —

HC104, H20 + H20 = HC104, 2H20 (liquid),

1 With regard to the heat of dilution of perchloric acid, which presents
some remarkable peculiarities, see researches made, " Annales de Chimie et de
Physique," 5e se'rie, torn, xxvii. p. 222.

OXYGENATED COMPOUNDS OF BROMINE. 357

liberates also -f 7'4 ; a value comparable to the heat of for-
mation of the secondary hydrated sulphuric acid. These
numbers tend to support the opinion that considered the
hydrated perchloric acids as the last indication of the penta-
basic character recognised in perchloric acid. Such a character
would only declare itself by the formation of hydrates with an
abundant liberation of heat, perchloric acid only forming
monobasic salts. In another series — HRQ3 — it has been shown
already how to pass from monobasic chloric and nitric acids to
tribasic phosphoric acid by means of iodic acid, which presents
certain intermediate characters.1

We see by these details how thermo-ehemistry accounts for
the characteristic properties of perchlorates, and especially for
the singular opposition which exists in the energetic oxidising
reactions of the concentrated acid, and the great stability of the
diluted acid.

§ 4. OXYGENATED COMPOUNDS OF BROMINE.

1. The thermal formation of bromic acid, potassium bromate,
and potassium hypobromite was studied.

2. Very pure potassium bromate was used, which had been
prepared and analysed by the author himself.

It was dissolved in water.

KBr03 -f water (1 part of salt -f 50 parts of water), at 11°,
absorbs - 9'85 Cal.

This solution was reduced by means of an aqueous solution
of sulphurous acid ; also the heat liberated by the union of the
diluted bromic acid with the potash was measured ; this is
essentially the same as the heat of neutralisation of chloric,
hydrobomic, and hydrochloric acids (4- 13*7) by the same
base.

3. All calculations being made, it was found for bromic acid
that— -

J [Br2 (liq.) + 05 -f H20 + water = H20, Br205 (diluted)]

absorbs - 24'8.

Thomsen, having reduced the same acid by means of stan-
nous chloride, found — 21'8, but on substituting in the
calculation of his experiments the number 3 8 '5, which seems to
be the more accurate,2 for the number 38'0, which he adopted
for the chlorination of stannous chloride, he also arrived at
- 24-8.

From this we get —

i[Br2 (gas) + 05 -f H20 + water = H20,Br2O5 (diluted)], - 2O8 ;

1 " Annales de Chimie et de Physique," torn. xii. pp. 313, 314.

2 Ibid., 5e se>ie, torn. v. p. 330

358 OXYGENATED COMPOUNDS OF CHLORINE.

a number almost double that of the heat absorbed by the
formation of chloric acid (— 12*0).

4. Again, for bromic acid (and bromates in solution) —

HBr03 (diluted) = HBr (diluted) + 03, + 15-5 ;
and for solid potassium bromate —

KBr03 (solid) = KBr + 03, +121 ;

values which are essentially the same as for chloric acid in
solution (+ 16*8) and for solid potassium chlorate ( + ll'O).
Finally, from the elements —

K -f Br(gas) + 03 = KBr03 (solid) liberates + 89'3.

& Hypobromous add. — The hypobromites are easily formed
by the action of bromine on alkaline solutions.

It was found, in presence of an excess of alkali, the bromine
being liquid, that —

Na20 (1 equiv. = 3 litres) H- Br (14-318 grms. and 3-365 grms.) at 9° + 6-0
K20 (1 equiv. = 4 litres) + Br (15-801 grms. and 5'734 grms.) at 11° + 5-95
BaO (1 equiv. = 6 litres) + Br (12-096 grms. and 12-339 grms.) at 13° + 5-7

Admitting that diluted hypobromous acid, when combining
with bases, liberates the same quantity of heat as hypochlorous
acid ; that is to say, + 9*5 ; it may be deduced from the pre-
ceding figures that —

J[Br2 (liquid) + O + water = Br20 (diluted)], - 67;
i[Br2 (gas) + 0 + water = Br20 (diluted)], - 31.

The latter number is essentially the same as what was observed
for the formation of hypochlorous acid ( — 2'9).

The alkalies, moreover, dissolve a greater quantity of bromine
than that which corresponds to the formation of hypobromous
acid. Thus baryta water dissolves in the cold nearly 2 eq. of
bromine : or Br4 for BaO. These facts are explained by the
simultaneous formation of alkaline bromides1 and of hypo-
bromites.

Before pursuing these comparisons further, it is advisable to
study the thermal formation of the oxygenated compounds of
iodine.

§ 5. lODIC ACID AND lODATES.

1. The results will be given which were obtained by the
action of iodine on potash, by which are formed hypoiodous and
iodic acids. It will then be convenient to examine the reaction
of iodic acid on water and alkalies, and also finally to compare
the thermal formation of the oxygenated salts derived from
chlorine, bromine, and iodine, endeavouring at the same time to
deduce therefrom some new data for molecular mechanics.

1 " Annales de Chimie et de Physique," 58 se*rie, torn. xxi. pp. 375, 378.

OXYGENATED COMPOUNDS OF IODINE. 359

2. If iodine be dissolved in diluted potash, at the ordinary
temperature, with the aid of the crusher described in p. 247,
two thermal effects succeed each other very rapidly. During
the first minute a lowering of the temperature is observed, which
reaches — 0'30° when we dissolve, for instance, 31 grms. of iodine
in 500 cc. of a solution containing one quarter of an equivalent
of potash per litre.

This initial phenomenon corresponds to the solution of the
greater portion of the iodine employed. Effects of the same
sign take place with solutions twice and four times as diluted.
As soon as these effects are produced the thermometer begins to
rise again in consequence of a new reaction, which lasts four to
five minutes, while the whole of the iodine becomes dissolved.
All the reaction can be effected with equivalent proportions
(excepting a trace of free iodine or some other compound which
turns the liquor slightly yellow). At this moment the solution
contains potassium iodate and iodide, according to the well-
known reaction —

3I2 + 3K20 (diluted) = 6KI (dissolved) + KIO3 (dissolved).

3. It may be that the initial phenomenon is due to the
formation of a hypoiodite —

I2 + K20 (diluted) = KIO (diluted) + KI (diluted) ;

but this body has only a momentary existence, and is changed
forthwith into iodate at the ordinary temperature.

4. It is well known that the same reaction with the hypo-
chlorites is only produced very rapidly at 100°.

The hypobromite, with an excess of alkali, resists much
longer, as has been proved.

5. This unequal stability of the three salts is explained by
the inverse progression of the stability of the chlorates, bromates,
and iodates, as will be seen by-and-by. Free hypochlorous acid
is, on the contrary, the most stable of all ; for it can be dis-
placed unchanged when cold by carbonic acid, and even by
acetic acid, whereas either of the latter acids, when in presence
of the hypobromites, separate the bromine at once, as Ballard has
observed from the beginning. This bromine is probably mixed
with some other compound, as was ascertained from the
measurement of the heat liberated.

6. Let us, however, return to the formation of the hypoiodite.
When iodine is added to diluted potash in successive fractions —
for instance, in twice or three times — each addition gives rise to
the same succession of phenomena, namely, to a lowering of
temperature, immediately followed by an increase of heat;
which shows that the effect is very characteristic of the reaction
itself, and independent of the fractions of iodine and potash
already combined. These singular effects, which only the

360 OXYGENATED COMPOUNDS OF CHLOKINE.

thermometer can reveal to us, require to be determined by
figures :

i[I2 + K20 (1 equiv. = 2 litres)] at 14°—

First effect : absorption ... . . ^ ... — 0*58

Second effect : liberation + 0'65

Total effect ... +0-07

I2 + K20 (1 equiv. = 4 litres)] at 15°—

1 'ing half the iodine : first effect - 0'38

second effect +0-30

Total effect ,,. -0-08

Adding the surplus of iodine : first effect ... - 0-19

„ „ second effect .,. +0-17

Total effect ... - 0-02

The total effect of the two effects united ... - 0-10

I2 + K20 (1 equiv. = 8 litres)] at 15°—

"irst effect - 1-27

Second effect +1-18

Total effect ,,. - 0-09

Let us note here that the first thermal effect, namely, the
cooling, does not afford an exact measure of the heat absorbed
in the corresponding reaction (formation of hypoiodite), but only
a superior limit, since the fresh rise in the temperature succeeds
too rapidly.

7. It being admitted that the final product of the preceding
reaction is potassium iodate in solution —

61 (solid) -f 3K20 (diluted) = KIO3 (dissolved) + 5KI (dilute);
and also that the formation of diluted potassium iodide —
K -f I -f water = KI (diluted) liberates + 747 Cal.;

we pass from this to anhydrous iodic acid, monohydrated acid,
and solid potassium iodate, by means of the following data : —

(a) Potassium iodate in solution.

i[2HI03 (1 equiv. = 1 litre) + K20 (1 equiv, = 1 litre) = 2KI03

(dissolved)], at 13° + 14-30

i[2HI03 (1 equiv. = 4 litres) + K20 (1 equiv. = 4 litres) = 2KI03

(dissolved)] ... ., + 14-25

These numbers slightly exceed the heat of neutralisation of
nitric acid by potash. This excess was ascertained by the
method of reciprocal double decompositions ; that is, by treating
alternately dissolved potassium iodate with diluted nitric acid,
and potassium nitrate with iodic acid, in presence of the same
quantities of water.

(b) Solution of hydrated iodic acid.

HI03 (crystallised) (1 part to 45 parts of water) 4- water, at 12°,

- 2-67.

FORMATION OF IODIC ACID. 361

Ditte found - 2'24, and Thomsen - 217, at a slightly different
temperature.

(c) Dilution of iodic acid.

HI03 (1 equiv. = 1 litre) + its volume of water, at 13° ... - 0-30
HIO (1 equiv. = 2 litres) + „ „ ... - 0-08

HI03 (1 equiv. = 4 litres) + „ „ ... - O'O

(d) Solution of anhydrous iodic acid. — This body was pre-
pared pure, and its composition ascertained by analysis.

^[I205 (1 part to 45 parts of water, at 12°) 4- water, - 0'81.

Ditte found — 0*95, and Thomsen — 0*89, at a slightly different
temperature.

(e) Solution of semihydrated iodic acid. — This body is
crystallised and well defined. The composition was ascertained.

J[2HI03I205 (1 part 4- 45 part of water, at 12°) 4- water, - 2-86.

(/) It was thought necessary to ascertain whether the three
solutions formed by anhydrous, monohydrated, and semi-hydrated
acid contain the acid in the same molecular state.

To this end, these solutions were treated, as soon as they
were made, with potash (1 equiv. == 2 litres). They all liberated
the same quantity of heat —

For£[I006] + 14-28

ForHIO, +14-31

i[2HI03,I2OJ ... +14.35

(g) Solution of the potassium iodates, — Neutral iodate (crystal-
lised)—

KI03 (crystallised) (1 part 4- 40 parts of water) 4- water, at 12°,

- 6-05.
Dilution.

KI03 (1 equiv. = 2 litres) 4- its volume of water, at 13°, - 0-36
KI03 (1 equiv. = 4 litres) -f „ „ „ - 0-0

Acid iodate (crystallised) —

KI03,HIO3 (crystallised) (1 part 4- 40 parts of water)
4- water -11-8

(h) Formation of iodic acid from the elements. — From the pre-
ceding data we deduce —

J[I2 solid 4- 05 + water = H20, I205 diluted], + 22-6.
This number, obtained by the synthetical method, is con-
sistent with the value 4- 21'5 found by Thomsen, by means of
analytical processes.
We have, moreover —

_ + 06 = I205 (anhydrous)] ... + 18-0 Cal.

I2 (gas) + 06 = I205 (solid) +23-4 „

+ I + 03 + water = HI03 (dissolved) ... + 57-1 „

H + I + 03 = HI03 (crystallised) + 59-8 „

J[H20 (solid) + I205 (solid) = 2HI03 (crystallised)] + 1-13 „

362 OXYGENATED COMPOUNDS OP CHLORINE.

According to the last number the hydration of the iodic acid
does not liberate more heat than the hydrated salts do, and
about the same quantity as anhydrous nitric acid.
We have also —

J[IA (solid) + 2HIO (solid) = 2HI03, I205], + 0'62 ;
HIO (dissolved) = HI (dissolved) + 03 (gaseous), - 43-9.

(i) Salts.

|[I205 + K20 = 2KI03 (solid)] + 55'5

i[I205 + BaO = Ba(I03)2 (solid)] + 34'9

HI03 (cryst.) + KHO (solid) = KI03 (cryst.) + H20 (solid) ... + 31-5

i[2HI03 (cryst.) + Ba(HO)2 (solid) = Ba(I03)2 (solid) + 2H20 (solid)] + 25'6

The formation of solid potassium iodate, shown by the above
figure, liberates far less heat than the sulphate (4- 71'1,
anhydrous substance ; 407, hydrated substance) and potassium
nitrate ( + 64*2, anhydrous; +42*6, hydrated). On the contrary,
it exceeds to a notable degree that of the monobasic organic
salts, such as the acetate (+ 55*1, anhydrous ; -f 21'9, hydrated).
It is, however, comparable to that of the salts of the most
powerful organic acids, such as potassium oxalate (+ 29 '4, see
table, p. 127), or again, acid iodate —

KI03 (crystallised) + HI03 (solid) = KI03, HI03 (solid), + 31,

the value of the class of ordinary double salts. We have
finally from the elements —

K + I (solid) + 03 = KI03 (solid) + 123-9

With I (gas) + 129-3

KI03 (solid) = KI (solid) + 03 - 44-1

KI03 (in solution) = KI (in solution) + 03 ... - 43'4

8. The heat liberated by the formation of solid potassium
iodate from the elements (+ 129'3) exceeds that of the solid
bromate and chlorate. In fact, it was found that —

K + Cl + 03 = KClOg disengages + 94-6

K + Br (gas) + 03 = KBr03 disengages ... + 87 "6
K + I (gas) + 03 = KI03 disengages ... + 129'3

It is well known that the relative stability of the three salts
goes on increasing from the bromate to the chlorate and then to
the iodate. This becomes still more evident by the comparison
of the heat brought into play when the three solid salts are
decomposed, with the liberation of oxygen.

KClOg = KC1 + 03 disengages + 11-0

KBr03 = KBr + 03 „ + 11-1

KI03 = KI + 03 absorbs - 44-1

Not only is the decomposition of the iodate more difficult
owing to its endothermal character, but it is accompanied by
phenomena of dissociation, the dry potassium iodide absorbing
the free oxygen.1 Chloric (- 12'0), bromic (- 24'8), and iodic

1 " Annales de Physique et de Chimie," 5" se'rie, torn. xii. p. 313.

REACTIONS OF HALOGENS AND ALKALIS. 363

(+ 22*6), acids diverge still more, one from the other, and
present differences which are not the same as for their salts.

9. Let us now compare the three principal reactions to which
the systems formed by halogens and alkali are susceptible.

(?)

3C12 gas + 3K20 (diluted) = 3KC10 (dissolved) + 3KC1 (dissolved) + 76-2

KC103 (dissolved) + 5KC1 (dissolved) + 94-2

6KC1 (dissolved) + 03 +111-0

The liberation of heat and the stability continue to increase
from the hypochlorite to the chlorate and free oxygen.

w

3Br2 (gas) + 3K20 (dissolved) = 3KBrO (dissolved) + 3KBr (dissolved) + 57-6

KBr03 (dissolved) + 5KBr (dissolved) + 54'0

6KBr (dissolved) + 03 + 74'4

The formation of hyprobromite liberates a rather greater
quantity of heat than that of the bromate, which explains the
relative stability of the former compound. However, the forma-
tion of bromide and oxygen is still the reaction which liberates
most heat. It is, moreover, well known that concentrated
potash can yield oxygen by its action on free bromine.

" w

3I2 + 3K20 = 3KIO (dissolved) + 3KI (dissolved) ... + 24-9 - 3a

KI03 (dissolved) + 5KI (dissolved) l + 31-8

6K1 (dissolved) + 03 ' - 12'3

Here the formation of the iodate exceeds all the others. The
liberation of oxygen would even involve an absorption of heat,
contrary to what takes place with the chlorate and bromate.
Moreover, this liberation does not take place at the ordinary
temperature ; it is only effected with the aid of a foreign energy
which is got in the act of heating.

We see that the principal chemical circumstances attending
the formation of the combinations between oxygen and the
halogens are in harmony with thermal data.

1 Calculated from the figures on p. 360, admitting that they represent a
maximum value for the formation of hypoiodite.

( 364 )

CHAPTEE XIII.

METALLIC OXALATES.

1. THERE exists a certain number of non-nitrogenised com-
pounds, formed in a regular manner, i.e. from the elements, in
consequence of a succession of exothermal reactions, which,
nevertheless, through heating or a shock capable of determining
decomposition, give rise to explosive phenomena. They are
compounds of such a kind that their elements have not reached
the most stable state of combination, i.e. the state to attain
which they have liberated the greatest possible amount of
heat.

We have, for instance, silver and mercuric oxalates— bodies
which detonate when suddenly heated or submitted to a violent
shock. Such a decomposition converts them into carbonic acid
and metal, in consequence of a real internal combustion by
which the oxygen of the metallic oxide attacks the oxalic acid
and completely oxidises it. This combustion, however, is only
possible when the heat it liberates surpasses that of the
oxidation of the metal plus the heat of neutralisation of the
metal. In other words, in order that an oxalate may possess
such properties, the reaction

M2C204 = 2C02 + M2

must be exothermal. Such is the fundamental condition which
distinguishes explosive oxalates from such as are not.

2. Let us elucidate these notions by calculating the heat
brought into play by the decomposition of the principal metallic
oxalates.

For this purpose the heat of formation of dissolved oxalic
acid from its elements * was first measured —

H2 + C2 (diamond) + O4 water = H2C2O4 (dissolved) (90 grms.)
liberates + 1947 Gal.

1 " Annales de Chimie et de Physique," 5e s£rie, torn. vi. p. 304.

HEAT OF FORMATION OF OXALATES.

365

We find, moreover, the heat of formation of metallic oxides in
the tables (p. 130).

These are the values relative to the more common metallic
oxalates : —

Zn + 0 = ZnO
Pb + 0 = PbO
Cu + 0 = CuO
Hg + 0 = HgO

Ag2 + 0 = Ag20

+ 86-4
+ 53-4
+ 38-0
+ 31-0
+ 7-0

By the method of double decomposition the heat liberated
by the union of the metallic oxides with oxalic acid was
measured ; or l

H2C204 (diluted) + ZnO (precipitated) = ZnC204 + H20
H2C204 „ + PbO „ = PbC204 + H20

H2C204 „ + CuO „ = CuC204 + H20

H2C204 „ +HgO „ = HgC204 + H20

H2C204 „ +Ag20 „ = Ag2C204 + H20

These data having been obtained, it is only necessary to add
together the heats of formation of the oxalic acid, the metallic
oxide and that of their reciprocal combination, and then deduct
the heat of formation of water, H20 (69 CaL), in order to
find the heat of formation of the metallic oxalate from its
elements.

+ 25-0
+ 25-6
+ 18-4
+ 14-0
+ 25-8

Acid (solid)
Zinc salt
Lead salt
Copper salt
Mercuric salt
Silver salt

H2 + C2 + 04 = H2C204

Zn + C2 + 04 = ZnC204

Pb + C2 + 04 = PbC204

Cu + CL + 04 = CuC204

.. ,, . Hg + d; + 04 = HgC204

, Ag2 + C2 + 04 = Ag2C204

+ 197-0
+ 237-1
+ 204-7
+ 182-1
+ 170-7
+ 158-5

3. If we note the heat of formation of 2 eq. of carbonic acid
from carbon (diamond) and oxygen, or

2(0 + Oa) = 2C02 liberates + 188-0,

it is easy to calculate the heat brought into play when an
oxalate is decomposed into gaseous carbonic acid and free
metal, the reaction being referred to the ordinary temperature —

H2C204 (solid) = H2 + 2C02
ZnC2O4 = Zn (solid) + 2C02
PbC204 = Pb (solid) + 2C02
CuC204 = Cu (solid) + 2C02
HgC204 = Hg (liquid) + 2C02
Ag2C204 = Ag2 (solid) + 2C02

- 9-0

- 49-1

- 16-7
+ 5-9
+ 17-3
+ 29-5

4. We see from this that zinc and lead oxalates cannot be
decomposed into carbonic acid and metal with a liberation of

1 The calculation is made here on the supposition that the precipitated
oxalates are anhydrous, or rather, that the heat liberated is essentially the
same for the anhydrous and precipitated salts ; which, in fact, has been
proved to be the case for the salts of mercury and silver.

366 METALLIC OXALATES.

heat. In fact, this reaction does not take place ; at least, not
without a strange complication.

It would seem at first sight that oxalic acid is in the same
position ; but this is only true when we start from the acid in
a solid state. In fact, the acid partly assumes the gaseous
state, at the moment of decomposition ; for observation proves
that a portion is always volatilised under these conditions.
However, this volatilisation of the solid acid must from analogy
absorb about 8 to 12 Cal. Taking this quantity into considera-
tion, we see that oxalic acid, when gaseous, is on the confines of
an exothermal decomposition, which explains its instability.
When the acid is in solution, the decomposition is in reality
exothermal, for

H2 + C2 + 04 + water = H2C204 liberates + 1947,
while

2C02 (gas) + water = 2C02 (dissolved) + 199-2.

The difference," + 4'5 Cal., represents the heat liberated in the
reaction. Copper oxalate is also on the confines, and even
beyond, its decomposition being exothermal. Finally, that of
mercuric and silver oxalates is positively exothermal.

5. Nevertheless, as regards mercuric oxalates the heat
liberated is limited, from a certain temperature, by the vola-
tilisation of the mercury, which absorbs — 15 '4; but this
restriction does not exist in the case of silver oxalate ; and, in
fact, this compound is very explosive. It explodes very
energetically when subjected to a shock or when heated to
about 130°. At 100° and lower it decomposes slowly and
progressively.

We see from these facts how thermo-chemistry explains the
explosive properties of certain metallic oxalates, and also the
difference which exists between the conditions of decomposition
of these and other oxalates.

BOOK III.

FORCE OF EXPLOSIVE SUBSTANCES IN
PARTICULAR.

CHAPTER I.

CLASSIFICATION OF EXPLOSIVES.

§ 1. DEFINITION OF EXPLOSIVES.

1. ANY system of bodies capable of developing permanent gases
or substances which assume the gaseous state in the conditions
of reaction, such as water above 100°, mercury above 360°, etc.,
may constitute an explosive agent. Even gaseous bodies assume
the same character if compressed beforehand, or if their volume
increases in consequence of some transformation. For this
purpose it is not necessary that the temperature of the system
should rise, although this condition is generally fulfilled and
tends to increase the effects.

2. Nevertheless, this definition of explosive agents, although
exact from an abstract point of view, is too wide for practice,
which only utilises such systems as are susceptible of a rapid
transformation and accompanied by the liberation of great
heat.

3. Moreover, the initial system should be able to subsist of
itself, at least for some time ; its transformation only taking
place if provoked by some external circumstance, such as fire,
shock, friction, or again by the intervention of small quantities
of a chemical agent, acting either in consequence of its own
reactions, which propagate themselves chemically (sulphuric
acid in presence of potassium chlorate mixed with organic sub-
stances), or because it produces a sudden shock, determining by
its mechanical effects the production of the explosive wave
(p. 88) and general explosion.

368 CLASSIFICATION OF EXPLOSIVES.

§ 2. GENERAL LIST OF EXPLOSIVES.

1. Let us enumerate the explosive bodies which fulfil these
conditions. They belong to eight distinct groups of substances.
These are —

First group. — Explosive gases, such as —

(1) Ozone, hypochlorous acid, the gaseous oxides of chlorine,
etc., which detonate under very slight influences — for instance,
slight heating or sudden compression.

(2) Various gases also formed with absorption of heat, but more
stable, gases which do not explode under the influence of pro-
gressive heating or moderate compression. Nevertheless, they
may explode through the detonation of mercury fulminate.
Such are acetylene, nitric oxide, cyanogen, arseniuretted
hydrogen, etc. (p. 66).

2. Second group. — Detonating gaseous mixtures formed by
the association of oxygen or chlorine, oxides of nitrogen with
hydrogen, hydrogenated gases, and carburetted and hydro-
carburetted gases or vapours.

3. Third group. — Explosive inorganic compounds, definite
bodies, liquids or solids, capable of exploding by shock, friction,
or heating, such as —

(1) Nitrogen sulphide, nitrogen chloride, and nitrogen iodide.
Mercury nitride and some other metallic nitrides. Fulminating
gold and mercury oxides, which are also nitrated derivatives.

(2) The liquid oxacids of chlorine and concentrated per-
manganic acid.

(3) Solid ammoniacal salts formed by the oxacids of chlorine,
nitrogen, chromium, manganese, and similar substances.

4. Fourth group. — Explosive organic compounds, definite bodies,
solid or liquid, capable of exploding by shock, friction, or heat-
ing, such as —

(1) Nitric ethers properly so called; nitric ether, nitro-
glycerin, nitromannite, etc.

(2) The nitric derivatives of the carbohydrates : cotton, paper,
wood, various kinds of cellulose, dextrine, sugar, etc.

(3) Nitro-derivatives, especially aromatic derivatives — for
instance, trinitro-phenol and its salts (picric acid and picrates),
nitro-oxyphenol (oxypicric acid and oxypicrates), tetranitro-
methane, chloropicrine (chloronitro-methane). Nitromethane
and its homologues, as well as their derivatives, are also classed
here.

(4) The diazo derivatives, such as diazobenzene nitrate and
similar bodies, nitrolic acids and other polynitro-derivatives,
nitro ethane, to which the fulminates of mercury and silver,
etc., seem to belong.

(5) The derivatives of highly oxygenated mineral acids, such
as, on the one hand, nitrites, nitrates, chlorates, perchlorates,

LIST OF EXPLOSIVES. 369

chromates, permanganates of organic alkalis ; on the other hand,
nitrous ethers, perchloric ethers, etc.

(6) Here we may also add the explosive derivatives of
hydrogen peroxide ; ethyl, acetyl, etc., peroxides.

(7) The hydrocarbon derivatives of mineral oxides which can
be easily reduced, especially the salts of silver and mercury
oxides, such as silver oxalate, mercury oxycyanide, etc.

(8) The derivatives of the hydrocarbons and other bodies
characterised by an excess of energy with relation to their
elements, such as metallic acetylides, etc.

5. Fifth group. — Mixtures of definite explosive compounds with
inert bodies. Each of the preceding compounds, whether solid
or liquid, can be mixed with inert bodies, destined to attenuate
the effects. Dynamite, properly so called, with a silica or
alumina base, wet gun-cotton, or soaked with paraffin, nitro-
glycerin dissolved in methylic alcohol, camphorated gun-cotton
and dynamite, etc., constitute such mixtures.

6. Sixth group. — Mixtures formed ~by an explosive oxidisaUe
compound and a non-explosive oxidising body destined to complete
the combustion of the former. Such are —

(1) Gun-cotton mixed with potassium or ammonium nitrate
potassium picrate mixed with potassium chlorate or nitrate, etc

(2) Also the mixtures of nitric acid with nitro compounds,
such as dinitrobenzene, the nitro toluenes, picric acid (trinitro-
phenol), etc., generally mixed in the form of paste.

(3) The mixtures of nitric peroxide and nitro compounds are
also classed here.

7. Seventh group. — Mixtures with an explosive oxidising base.

(1) The mixtures formed by an explosive body containing
an excess of oxygen (nitroglycerin, nitromannite) and an
oxidisable body such as carbon dynamite.

(2) Analogous bodies, in which the oxidising and oxidisable
bodies are both explosive, such as blasting gelatin formed by
the association of nitrocellulose and nitroglycerin, etc.

8. Eighth group. — Mixtures formed by oxidisable and oxidising
bodies, solid or liquid, neither of these being explosive separately.
This group comprises —

(1) Black powder formed by the association of sulphur and
carbon with potassium nitrate and constituting the varieties
designated as service, sporting, and blasting powder, etc.

(2) The various powders formed by the association of hydro-
carbon compounds, charcoal, coal, wood, sawdust, various kinds
of cellulose, starch, sugar, ferrocyanide, or by the association of
sulphur and metals with potassium, sodium, barium, strontium,
lead, etc., nitrates.

(3) The liquid or pasty mixtures formed by the association of
liquid nitric acid either with a combustible liquid or with a solid
substance on which it does not exercise an instantaneous action.

2B

370 CLASSIFICATION OF EXPLOSIVES.

(4) Here we may class the mixture of liquid nitric peroxide
with various oxidisable substances, such as carbon disulphide
or petroleum spirit.

(5) The powders formed by the association of combustible
bodies with chlorates and perchlorates.

(6) The powders formed by the association of combustible
bodies with various combustive bodies, such as potassium
bichromate, chromic acid, the oxides of copper, lead, antimony,
bismuth, etc.

(7) To the mixtures of this group may be assimilated the
mixtures formed by the association of a sulphide, a metallic
phosphide or an analogous binary compounds with another metal
capable of displacing the former under the gaseous form
(mercury, for instance) with the liberation of heat.

§ 3. DIVISION otf THE THIRD BOOK.

The variety of explosive mixtures thus practically created
with a view to their being applied is indefinite. Nevertheless
the number of the Usual Compounds is limited, and we will
designate the principal ones we intend to examine specially;
but first of allj in Chapter II. we shall present the general data
which it is necessary or useful to know in order to define the
manufacture and employment of a given explosive.

Chapter III. will comjttise the study of explosive gases,
detonating gaseous mixtures, and analogous substances (groups
1 and 2).

Chapter IV. is devoted to nOn-caYbonated explosive com-
pounds (3rd group).

In Chapter V. we shall treat of nitric ethers properly so
called (4th group).

The sequence of the substances belonging to this group is
studied in the following four chapters; which also comprise the
mixtures of the 5th, 6th, and 7th groups. The dynamites will
be examined in Chapter VI.

Gun-cotton and allied bodies in Chapter VII.

Picrates in Chapter VIII.

Dinitro compounds in Chapter IX.

Lastly the eighth group will be examined, viz. : Powders with
a nitrate base in Chapter X. ;

Powders with a chlorate base in Chapter XI.

And we shall conclude with some general considerations.

( 371 )

CHAPTER II.

GENERAL DATA RESPECTING THE EMPLOYMENT OF A GIVEN
EXPLOSIVE.

§ 1. THEORETICAL DATA.

1. EXPLOSIVE bodies cannot be employed profitably and securely
unless they are characterised by a certain number of data,
theoretical as well as practical, which will now be enumerated.

2. First as regards theoretical data. They have been given in
principle in Book I. ; but it seems desirable to summarise them
here from a more special point of view. These data refer to
eight orders of measurements, namely :

(1) The chemical equation of transformation.

(2) The heats of formation of the components and products.

(3) Their specific heats.

(4) Their densities.

(5) The pressures developed.

(6) The initial work which determines the reaction (tempera-
ture of inflammation, nature of shock, etc.)

(7) The law which determines the rapidity of the transforma-
tion with reference to temperature and pressure.

(8) The total work which an explosive substance can effect
(potential energy).

Each of these orders of measurements embraces several dis-
tinct determinations.

3. The chemical equation of the explosive transformation
comprises :

(1) A knowledge of the original bodies and of the products as
regards their nature and relative weight.

(2) The knowledge of the volume of the permanent gases, re-
duced to 0° and 0*760 metres, which the transformation develops
(p. 18). This volume may be calculated a priori, or measured
directly and as an essential element of chemical analysis.

(3) A knowledge of the gaseous volume (reduced by calculation
to 0° and 0*760 metres) of the products actually liquid or solid,

2B2

372 GENEKAL DATA.

but capable of assuming the gaseous state at the temperature
of explosion. Much discussion often arises on this head.

(4) The knowledge of the state of dissociation of the products
at the moment of explosion and during the period of cooling
(p. 8).

In fact, up to the present this datum is known with precision
for scarcely any compound body, and our ignorance in this
respect is one of the principal causes of the divergence observed
between the practical results and the data of theoretical calcula-
tion.

(5) The knowledge of the weight of oxygen actually employed
in the explosive reaction.

(6) The knowledge of the weight of oxygen required for total
combustion is deduced from the preceding.

4. The heats of formation of the components and products
comprise :

(1) The knowledge of the heats of formation of these various
bodies from their elements ; quantities given in the thermo-
chemical tables (p. 125 and following).

(2) Their heat of total combustion by free oxygen, or ly the
oxidising compounds (nitrates, chlorates, oxides, etc.).

(3) The knowledge 'of the heat of vaporisation of bodies
actually liquid or solid, but capable of assuming the gaseous
state in the conditions of the explosion (p. 140).

(4) The heat liberated by the explosive transformation is also
deduced from the foregoing data, which are supposed to be
known. On the other hand, it may be measured directly and
employed in the inverse calculation of these same data.

5. The specific heats of the components and products are
generally known by the tables for the ordinary temperature
(pp. 141-143). For high temperatures, such as are developed
during the explosion, our knowledge on this point is very im-
perfect.

From the mean specific heat of the products is deduced the
temperature developed during the explosion. The calculation is
made according to the knowledge of the quantities of heat
(pp. 11 and 19) ; but the accuracy of the result is subordinated
to the knowledge of the dissociation and that of the specific
heats (see p. 18).

Processes of direct measurement for the temperatures would
be preferable ; but hitherto it has not been possible to try them
with any probability, except in one single case, namely with
black powder.

6. The densities of the components and products may be
measured at the ordinary temperature (p. 144).

(1) The molecular volumes are obtained from them. A
knowledge should be added of the co-efficients of expansion of
the various solid, liquid, or gaseous bodies, so as to deduce

GENERAL DATA. 373

therefrom the exact volume of the products at the tempera-
ture of explosion. Unfortunately these are data which are bu't
little known, and we generally content ourselves with the
densities in the cold for solids and liquids, and the densities
calculated according to Mariotte's and Gay-Lussac's laws for

(2) These data are necessary to calculate a priori, according
to the same laws, the theoretical pressure which the explosive
would develop when detonating in its own volume (p. 30).

(3) They would be equally useful for calculating the theoretical
pressure under any density of charge (p. 30), that is to say, the
real volume occupied by the gases at the moment of explosion ;
but for this purpose the real density of solid, liquid, and
gaseous products should be known exactly.

7. The pressures developed must be measured directly (p. 20).

(1) Under various densities of charge.

(2) A curve is deduced therefrom which permits us to esti-
mate according to the experiments themselves, the real pressure
developed under a density equal to the unit, viz. the specific
pressure (p. 30) as well as,

(3) The maximum pressure developed by the explosive. It
is that of a body detonating in its own volume (p. 30).

If we admit that there exists a proportion between the
pressures and high densities of charge (p. 30), the specific
pressure, namely the pressure developed under a density, equal
to the unit, will characterise the force of the explosive.

The effective measurements thus obtained for the real
pressures should be compared with the theoretical pressures
calculated, as has been said, with the aid of Mariotte's and
Gay-Lussac's laws. In this calculation the volume occupied
by the solid or liquid products must be taken into account.

(4) A more certain datum, and one that is more easily calcu-
lated a priori and verified experimentally, is the permanent
pressure exercised by the gases of explosion reduced to 0° in
a determinate and sufficiently resisting capacity (p. 32). It is
often limited by the liquefaction of the products, such as car-
bonic acid.

(5) As a term of comparison, the characteristic product, if not
absolute at least relative, can be given, namely, the product of
the heat liberated multiplied by the reduced volume of the
gases and divided by the specific heat of the bodies formed
(p. 32). This product gives essentially in theory the same
relations between the various explosive substances as the theo-
retical pressure.

8. The initial work which determines the reaction seems
to be summed up in a knowledge of the following data : —

(1) The temperature of incipient reaction, a temperature which
must be measured directly.

374 GENERAL DATA.

(2) The smallest shock which will cause decomposition, also
the effects due to the shock, or the application of fire would be,
no doubt, derived therefrom in a complete theory.

In the absence of this theoretical datum, we measure the
minimum fall of a given weight which is required to cause the
substance to explode when placed in definite conditions.

More generally, but in a vaguer manner, we ascertain whether
it explodes by the shock of iron on iron, bronze on bronze,
stone on stone, wood on wood, iron on bronze, stone, wood,
bronze on stone or wood, stone on wood, or by friction exercised
in various conditions, etc.

9. The law of the rapidity of decomposition, in cases of
simple ignition, and the rapidity of propagation of the explosive
wave in other cases (p. 88), is of primary importance, but this
law is generally not known.

10. The total work performed by an explosive substance
in given conditions corresponds to the difference between the
heat liberated by the chemical transformation effected without
external work and the heat really liberated in the conditions of
the experiment, a difference which might, if necessary, be
measured experimentally.

In principle the maximum work would be measured by the
liberated heat itself (potential energy), but we have only to con-
sider the work which may be performed by the gases developed
by the explosion in the case of indefinite expansion. The theory
of these effects has only been broached for service powder
(p. 17).

11. In practice this deficiency is made up by empirical
notions drawn from the study of the effects of each explosive on
various kinds of vessels and materials. These effects are more-
over complex, for they often result at the same time from the
total work, the pressure exercised, the law of rapidity, and the
nature of the materials.

Without entering into circumstantial details, may be cited
as an instance the trial of the force of an explosive substance
according to the size of the capacity produced by its explosion
in a block of lead (Abel's process). For instance, a block of
lead is taken, 250 mm. square, 280 mm. high, and weighing
175 kgnL Following the axis, a cylindrical channel is bored
with a diameter comparable to that of a miner's boring tool
(28*5 mm.), and 178 mm. deep. A determinate weight of the
explosive substance (10, 20, or 30 grms.) is placed at the bottom,
and if necessary it can be arranged under an impermeable
covering. A detonator is introduced at the end of a fuse of
suitable length, and the hole is then filled up with water, which
serves as tamping. The explosion is then effected, and the
capacity of the pear-shaped chamber produced is afterwards
measured. The proportion between the increase of the capacities

EXPLOSIONS IN BLOCK OF LEAD. 375

produced under the influence of equal weights of the various
explosives may be taken as comparative measurements of their
power. When the substance is too active, a system of rents is
produced, following almost a diagonal direction in any vertical
section passing through the axis of the block, and tending to
detach a kind of truncated cone in the total mass. This accident
can, however, be avoided by diminishing the weight of the
substance.

It has been found that the relations of the increases of capacity
obtained with variable weights of different materials remain the
same, the weight being moreover very small in comparison with
that of the block. Here are some of these relations which
express the increase of capacity produced by I grin, of explosive
according to the experiments of the Commission des substances
explosives : —

cc.

Nitromannite 43

Nitroglycerin .. ... ... ... 35

Dynamite 75% ... .„ .., ... 29

Dry gun-cotton ... ... ... ... ... ... ... 34

Ditto (0-40 grm.) + ammonium nitrate (0-60 grm.) ... ... 32

Ditto (0-50 grm.) + potassium nitrate (0'50 grm.) 21

Mercury fulminate ..13-5

Ditto, eliminating the weight of the mercury by calculation ... 45
Panclastites : 1 vol. carbon disulphide + 1 vol. nitric peroxide 25

2 vols. CS2 + 1 vol. N02 ' 18

3 vols. CS2 + 5 vols. N02 (complete oxidation).,.. ... ... 28

1 vol. essence of petroleum l + 1 vol. N02 28

2 vols. essence of petroleum l + 1 vol. N02 ... 18

1 vol. nitrotoluene + 1 vol. N02 ... 29

This process furnishes very interesting comparative data, but
it does not apply to slow powders, such as black powder, as the
tamping is then driven forward before the chamber has been
enlarged.

In the case of rapid powders the relations are not the same as
those resulting from the quantities of heat and of the gaseous
volumes. Thus these two quantities are nearly the same for
nitroglycerin and nitromannite, whereas the capacities are greater
by a fourth in the case of the latter substance, doubtless because
its explosion is effected in a shorter time.

The classification of the relative force of explosives according
to their effects changes very much according as the operation
is carried out with or without tamping. Generally speaking,
studies of this kind are only fully valid for works, effects, and
materials comparable to those which formed the object of the
preliminary experiments.

12. Such is the ensemble of the scientific data we must
endeavour to obtain before laying claim to the complete theory
of a given explosive substance.

1 Containing one-tenth of its volume of carbon disulphide.

376 GENEKAL DATA.

In fact and in practice these data are less numerous than
might be inferred from the preceding statements. In the present
state of our knowledge they are reduced practically to the
following : —

(1) Chemical equation of the transformation.

(2) Heat developed by this transformation,

(3) Volume reduced to 0° and 760 mm. of the gases and
bodies capable of being rendered gaseous in the conditions of
the transformation.

(4) Pressures developed.

(5) More or less crude empirical indications referring to the
work effected.

These five orders of data regulate our knowledge of the force
of explosive substances.

Let us remark here that the first three measurements are
deduced simply from the chemical equation of the phenomenon,
and the thermo-chemical tables ; the fourth and fifth would be
calculated by the preceding if the laws respecting the thermo-
dynamics of gases and those of the resistance of substances were
sufficiently well known,

§ 2. PRACTICAL QUESTIONS RESPECTING THE EMPLOYMENT OF
EXPLOSIVE SUBSTANCES.

1. In practice an explosive substance must satisfy a certain
number of conditions which we will now summarise. These
conditions refer to the employment, manufacture, preservation,
and stability of the explosive substance. Let us commence with
the employment.

2. The explosive substance placed in a small volume and under
a moderate weight should develop a considerable quantity of gas
and a great amount of heat, circumstances which exclude ex-
plosive gases and detonating gaseous mixtures, at least in most
applications.

3. The chemical transformation which the substance under-
goes should be produced in a very short space of time, so that the
heat may not be gradually dissipated, which would greatly
reduce the pressure.

Let us remark, moreover, that the effort of a sudden pressure
produces very different effects of rupture on a given substance
to what would have been the case if the same pressure had been
exercised slowly.

In mining works, or with firearms, a slow reaction would tend
to let the gases escape little by little through the interstices of
the earth or the charge.

4. The empirical measurement of the force of an explosive
substance will be effected by means of a system of tests approach-
ing as far as possible the conditions of its practical employment.

CONDITIONS OF EXPLOSION. 377

In the absence of these conditions, which are not very
suitable for precise comparisons, trials are made on a small
scale, such as —

The use of the testing mortar on ballistic pendulum for powders
intended to throw projectiles from firearms ;

The use of bombs of different thicknesses from which the
bursting charge (p. 58) and the mode of fragmentation are
studied ;

The rupture of freestone, rails, T-iron, iron girders, masses of
rolled, cast, or wrought iron, beams of different kinds of wood,
and different scantlings, by charges laid on their surface ;

The curve imparted to thick iron plates in comparative
conditions ;

The crushing of a small block of lead by a charge placed on
its surface, with or without tamping ;

The crushing of a copper cylinder (p. 20) ;

The form and size of the chambers produced in a mass of
clay or lead by the explosion of an internal charge (see p.
374), etc.

We shall refer to the technical treatises and memoirs for the
description of these various tests, as it would be almost impos-
sible to give the exact theory of them at present.

5. The explosive substance should be capable of being handled
and transported by road or railway with at least relative safety,
and it must not be too sensitive to shocks or friction. This is
the reason why pure nitroglyeerin and chlorate powders are
almost excluded in practice.

The same circumstance forbids the employment of dynamite
and pure gun-cotton in warfare, since these substances explode
from the shock of a ball.

6. The substance should only explode in conditions which are
precisely known, and capable of being produced or avoided at
pleasure ; for instance — special ignition, the use of certain caps
and fuses ; the employment of electricity to heat a wire or produce
a spark; the shock of two metal pieces arranged beforehand;
definite chemical reaction — for instance, that of sulphuric acid
on potassium chlorate mixed with a combustible body, etc.

The conditions under which the explosive substance is
brought to explode should be realisable without too much
difficulty ;' thus the explosion of paraffined gun-cotton becomes
almost impossible above a certain quantity of paraffin. In the
same way a mixture of petroleum spirit and nitric peroxide in
equal volumes does not explode under the influence of an
ordinary fulminate cap, while it does so by the addition of a
tenth part of carbon disulphide, etc.

7. The explosion should produce effects foreseen beforehand,
at least in a certain limit, such as direction, general characters,
and intensity.

378 GENERAL DATA.

Thus too sudden a reaction brought about in a firearm causes
its rupture before the projectile has time to be displaced. Any
substance capable of producing such effects must be excluded,
and this prevents the employment of pure nitroglycerin or
potassium picrate in firearms.

A shell should be broken into large pieces and not pulverised
by the explosion of the internal substance, and this circumstance
opposes the use of pure mercury fulminate. The reaction of
the powder in the weapon should be sufficiently progressive for
the projectile to acquire a determinate initial velocity.

8. From a more particular point of view, the explosive sub-
stance should not injure the weapons ; either by chemical
reaction, sulphurising, oxidation, etc,, or by fouling (ash and
fixed substances, leading, etc.), or by mechanical wear and tear.

9. In subterranean works the explosive substance must not
produce any deleterious gases capable of suffocating the workmen
(carbonic oxide, sulphuretted hydrogen, nitrous vapours,
hydrocyanic vapours, etc,).

In general it should not produce too much smoke in warfare.

10. On the contrary, in certain military operations it may be
useful to produce a great deal of smoke, in order, for instance, to
mask a movement or some works.

It may also be useful to produce deleterious gases in order to
render the gallery of a mine, etc., impracticable for some time.

11. The pyrotechnical effects, such as signals, lighting, bon-
fires, etc., represent quite a different order of special conditions
to be fulfilled, but on which we shall not dwell, as this subject
is foreign to the present work.

12. The necessity of dividing the explosive substances, or of
making them into a determinate form, enters into consideration
sometimes.

Thus dynamite and the powders properly so called are more
easily divided than gun-cotton into small pulverulent masses,
destined to be introduced into some cavity whose cracks and
fissures they fill up, such as a blast^hole.

On the other hand, compressed gun-cotton may be easily
divided and worked with tools so as to give it a special form
independent of any covering ; special care is taken to impregnate
it beforehand with paraffin, a substance which moreover has the
advantage of diminishing the explosive sensitiveness of gun-
cotton.

13. In various cases the explosive substances are compressed
or agglomerated under an hydraulic press in order to increase the
density and modify the law of propagation of the ignition. Black
powder and gun-cotton are very suitable for this operation, which
it would be perilous to attempt with fulminate or chlorate powders.

14. Let us cite again the employment of fulminating sub-
stances under the form of caps, ordinary or strong detonators,

KEEPING OF EXPLOSIVES. 379

which are destined to provoke the explosion of a considerable
mass of another substance (p. 54).

They are treated in small quantities, and precautions are
taken against the dangers presented by their preparation and
manipulation, dangers which would not be accepted in industries
for a substance manufactured or employed in large masses.

We shall restrict ourselves to the indications which have just
been enumerated and which correspond to the principal uses of
explosive substances in war and industry, As regards the
effects themselves which it is proposed to accomplish, it can
easily be understood that the diversity of these special effects
required from explosive substances is unlimited.

§ 3. PRACTICAL QUESTIONS DEFERRING TO THE MANUFACTURE.

1. The manufacture of explosives ought to be effected under
conditions of cost proportioned to their industrial uses, one and
the same effect being produced in mines or industries in general
at the lowest possible price. In military matters this condition
also intervenes, but in a minor degree, since facility and safety
of employment outweigh all other considerations.

2. The manufacture must be carried on regularly and without
danger, or at least with as little danger as possible to the work-
people and neighbourhood.

3. The inconveniences resulting through noxious gases, noise,
and damage arising from accidental explosions must also be
taken into consideration.

§ 4. PRACTICAL QUESTIONS RESPECTING PRESERVATION,

1. It should be possible to keep explosives without any
spontaneous decomposition in the ordinary state of the atmo-
sphere, in various climates, under moderate conditions of tempera-
ture and light, in an average hygrometric state, etc.

2. Direct sunlight is bad for nitro compounds, as it often
leads to their chemical decomposition.

3. Extensive variations of temperature also exercise an im-
portant influence, particularly if they determine the freezing
of certain ingredients, such as nitroglycerin in the dynamites,
or if they increase the fluidity of certain bodies, such as nitro-
glycerin itself, and consequently their tendency to exudation.
The separation between nitroglycerin and its absorbent can thus
take place by the fact of repeated variations of temperature or
even of repeated freezing and thawing. Under the influence of
a somewhat high temperature, such as occurs in practice,
especially in hot countries, certain compounds may gradually
evaporate slowly and modify the primitive composition of the
mixtures. This occurs, for instance, to ordinary dynamite heated

380 GENERAL DATA.

for a long time on a sand-bath, as the nitroglycerin gradually
evaporates and the substance consequently loses part of its power.
The elevation of the temperature might also give rise to the
rapid vaporisation of certain components and consequently to
their elimination, for instance in the case of compounds con-
taining nitric peroxide, which boils at 26°.

4. The state of preservation should remain satisfactory even
in very varied hygrometric conditions of the surrounding
atmosphere.

It is this condition which has led to deliquescent bodies such
as sodium nitrate being excluded from the manufacture of
service powder. This salt should also be avoided in the manu-
facture of dynamite, seeing that the accidental formation of a
concentrated solution of sodium nitrate due to the deliquescence
of the solid salt determines the separation of the existing
nitroglycerin and transforms this substance into a non-homo-
geneous and very dangerous mixture.

Diazobenzene nitrate becomes completely decomposed under
the influence of moisture.

5. The salts with which sea air is impregnated constitute a
special cause of change which must be borne in mind, especially
as regards explosives which are to be employed on board ships,
or even conveyed by them, since the air eventually penetrates
into the best closed vessel, owing to the variations of tempera-
ture and pressure.

6. From this point of view it is useful to know whether an
explosive substance resists the action of liquid water, which may
accidentally moisten explosive substances, especially at sea.
It is well known that water destroys service powder by dis-
solving the saltpetre : by a kind of liquefaction it gradually
displaces the nitroglycerin in silicious dynamite.

Dynamites which contain nitrates are also decomposed by
water.

Silicious dynamite deposited in running water gradually
loses its nitroglycerin by way of solution, since nitroglycerin is
slightly soluble in water.

On the other hand, pure water does not affect gun-cotton
whether the latter be simply moistened or plunged into running
water. The inflammability of the substance, which is checked
by the presence of water, reappears completely after drying.

Moistened gun-cotton can moreover be kept and even
employed in that state with less danger of accidental ignition
than in the dry state.

However, gun-cotton which is kept moistened for a long time
may become the seat of mould and other microscopic plants
which alter the properties in the long run.

7. The slow exudation of the nitroglycerin in dynamites
made with bad materials forms an obstacle to their preservation,

TESTS OF STABILITY. 381

and also a serious danger, for it tends to substitute pure nitro-
glycerin for a substance which is but little sensitive to shocks
or friction, while the former is, on the contrary, extremely
sensitive.

It has been stated how freezing followed by thawing, and
even the action of water, might also give rise to exudation.

8. The possible separation of the various ingredients of a
mixture under the influence of jolting arising from conveyance
by sea or land is also to be considered.

9. The slow action which the metals, constituting metallic
cartridges, exercise on the saltpetre and the sulphur contained
in cartridges, especially if these are even slightly hygrometric,
may determine the oxidation and sulphurising of these metals at
the expense of the saltpetre and sulphur. Hence there arises
at length a certain weakening of the effects obtained with recent
powders, according to the experiments made by Colonel Pothier.

We then see how the preservation of explosives gives rise to
very varied special problems. It suffices at present to have
pointed out the preceding.

§ 5. TESTS OF STABILITY.

1. The tests of stability to which a given explosive is
subjected in practice, comprise the most essential conditions
among those which have been just enumerated. These are —

2. Stability on exposure to air. The substance must maintain
itself, when in contact with air, without evaporation, lique-
faction, or apparent alteration, even after having been kept
several days. It must not attract atmospheric moisture.

3. Neutrality. It should in general be neutral and preserve
this neutrality; above all, it must not liberate acid vapours
even when heated for some minutes in a bath kept about 60°.

4. Exudation. It must not allow the liquid substances it
contains, such as nitroglycerin, to exude, either spontaneously
or by a slight pressure such as is applied when pushing back
the substance gently with a wooden piston in a brass tube
pierced with lateral holes. In this trial the piston should not
be pressed by hand but by a weight, which is gradually
increased until exudation takes place.

When heated to about 55° to 60° in a bath, the substance
should not give rise to the separation of small drops even under
a slight pressure.

When subjected to a temperature below zero, arid then
brought back to the ordinary temperature, and that several
times, it ought also not to produce exudation.

Nor should exudation take place under the influence of air
saturated with moisture ; for instance, should the substance be
left for a fortnight in a chamber containing damp tow.

382 GENERAL DATA.

It should also be ascertained whether the substance, when
subjected for several days to a series of shocks in conditions
similar to such as would arise during conveyance by sea or land,
occasions the separation of some of its components.

These exudation tests are, above all, essential as regards
dynamites, as the separation of the nitroglycerin tends to make
them very dangerous.

5. Shock. It should be tried whether the substance explodes
by the shock of a hammer on an anvil, or better still by the
fall of a given weight falling from various heights on a portion
of the substance placed on an anvil.

An explosive should not explode through the shock or friction
of wood on wood or of wood on metal (bronze or iron). Some
substances do not explode by the shock of bronze on bronze,
but do so by iron on iron.

The accidental introduction of some grain or fragment of
sand or other hard rock facilitates the explosion, especially by
friction.

The action of the shock of bullets at different distances
should be studied, especially in the case of substances intended
for military operations.

6. Immersion. The explosive substance is placed under
water without any covering for fifteen to twenty minutes. It
ought neither to, dissolve nor split up, nor give rise to the
separation of small liquid drops. This test is only applicable to
substances which are liable to be in contact with water when used.

7. Heat. It is first ascertained whether the substance
becomes inflamed when in contact with an ignited body, and
how it burns in this condition.

The influence of very slow progressive heating is also studied
in order to see whether it gives rise to the partial evaporation
of any of its components.

We then proceed to more rapid heating, placing, for instance,
a small quantity of the substance in a thin metallic capsule,
which is laid on the surface of an oil or a mercury bath l main-
tained beforehand at a fixed temperature. It is ascertained at
what temperature the explosion takes place, and whether simple
burning or even progressive decomposition can take place at a
lower temperature.

These general questions being defined, we proceed to the
study of the various groups and kinds of explosive substances.
Let it, however, be remembered that it is not intended to give
an individual and a practical history of each of them in all its
details, which would lead us too far ; but we especially wish to
point out the scientific data which characterise them by study-
ing the principal explosive bodies hitherto known, these bodies
being considered as typical of all similar substances.

1 The capsule must then be made of platinum.

( 383 )

CflAPTEK III.

EXPLOSIVE OASES AND DETONATING GASEOUS MtXTtJRES.

§ 1. DIVISION OF THE CHAPTER.

THIS chapter comprises the study of definite explosive gases ;
of detonating gaseoUs mixtures formed, for instance, by the
association of oxygen with a combustible gas ; of liquefied
mixtures of gas ; and, finally, of the mixtures of gas with com-
bustible dust. The Study of all these systems is connected
with that of the gases themselves.

§ 2. EXPLOSIVE GASES.

1. There exists a certain number of definite gases, capable of
transforming themselves with explosion under the influence of
a shock, sudden compression, heating, the electric spark, etc.
Such are ofcone and the oxygenated compounds of chlorine,
which explode through sudden compression or heating. These
bodies are characterised by the fact that their formation, either
from ordinary oxygen, as in the case of ozone, or from their
elements, as in the compound gases, takes place with absorption
of heat.

This last characteristic belongs also to other gases, whose
explosive decomposition could not be determined for a long
time, such as the oxygenated compounds of nitrogen, acetylene,
and some other hydrocarbon gases, arseniuretted hydrogen,
cyanogen, the vapour of hydrocyanic acid, cyanogen chloride,
the vapour of carbon disulphide. Latterly, however, the author
has succeeded in making gases of this kind explode under the
influence of mercury fulminate (p. 66).

2. The heat liberated by the decomposition of explosive
gases is known. It is precisely equal to the heat absorbed in
formation (p. 115). Starting from this datum, we can then
calculate the pressure and the temperature developed by the
explosion according to Mariotte's and Gay-Lussac's laws, and

384 EXPLOSIVE GASES AND DETONATING GASEOUS MIXTUKE8.

by employing the specific heats of the gaseous elements
measured at the ordinary temperature. Let us note that here
there pan be no question of dissociation, since the products of
the explosion are elementary gases.

Belying on these principles, the heat liberated, the tempera-
ture produced, and the pressure developed for ozone and
hypochlorous gas, will first be given. As regards chlorous and
hypochlorous gas, no measurement has been taken up to now.
A summary of the results referring to nitric oxide, cyanogen,
and acetylene will be added.

3. Ozone is changed into ordinary oxygen at the ordinary
temperature. This transformation is all the more rapid accord-
ing as we operate on a mixture of oxygen and ozone richer in
ozone, for the latter has never been isolated in a state of purity.1

It is accelerated with the temperature and becomes explosive
under the influence of sudden compression.2 The heat liberated
is equal to 14-8 Cal. for 24 grms. of ozone, occupying 11 '16 lit.
or 29'6 Cal. for the molecular weight, Oz = 03 (48 grms.),
according to the author's experiments,8 that is, '616 Cal. per
kgm. of substance.

The specific molecular heat of oxygen being equal to 6*95 for
32 grms. (or 02) at constant pressure, if we suppose this specific
heat to be invariable, the temperature attained by pure ozone
when being transformed into oxygen would then be 2840° at
constant pressure. .At constant volume the specific molecular
heat is 5'0 for 02, and the heat liberated reaches 29-9 Cal.
Consequently, the specific heat being supposed constant, the
temperature produced would be 3987°. .

The pressure developed at constant volume, calculated
according to this datum, would be equal to 2 3 '4 atm.

Such are the characteristic data of ozone, supposing it to be
pure and taken under the normal pressure. If this be dwelt
upon, it is because this transformation represents a typical case
in the theory of explosive bodies, since it is only a question of a
simple gas changing as regards condensation.

In practice, since pure ozone has never yet been obtained, the
transformation is effected in a mixture of ozone and ordinary
oxygen. Let us give, moreover, the calculation of the pressure
developed for a mixture capable of supplying after transforma-
tion a weight of oxygen proceeding from the ozone equal to a
sixteenth of the total weight (6 '2 hundredths), a mixture which
can be easily prepared under ordinary circumstances with the
author's apparatus (p. 220).

1 Upon the rapidity of the transformation, see " Annales de Chimie et de
Physique," 5" s^rie, torn. xiv. p. 361, and torn. xxi. p. 162.

* Chappuis et Hautefeuille, " Comptes rendus des stances de 1'Academie
des Sciences," torn. xci. p. 522.

3 "Annales de Chimie et de Physique," torn. x. p. 152

HYPOCHLOEOUS ACID, NITRIC OXIDE, ACETYLENE. 385

The heat liberated is always the same for a given weight of
ozone, but it is distributed between the oxygen derived from it,
and the excess of the same gas which pre-existed. Consequently,
the temperature produced at constant volume will be 245°, and
the pressure developed about 1*9 atm.

4. Hypochlorous acid explodes under the influence of a tem-
perature above 60°, or under the influence of a spark, shock, etc.

Thus it liberates 7*6 Gal. by C120 = 43*5 grms., occupying
1T6 lit. or 15'2 Cal. for the molecular weight (87 grms.).
C12O = C12 + O liberates 15 -2 Cal. at constant pressure, or
175 cal. per gramme of substance.

The specific heat of 0 being 3'5 and that of C12 8 '6, the sum
is 12il at constant pressure, and the temperature developed in
the final mixture of the elements in consequence of their sepa-
ration will be then ' = 1256°.
\.2t' J.

At constant volume the sum of the specific heats of the
elements is reduced to lO'l, and the heat developed rises to
15'5 Cal. The temperature produced rises then to 1530°, and
the pressure calculated to 9*9 atm.

5. It has been deemed useful to give these results, since they
are typical, owing to the gaseous character of the components
and products and the elementary nature of the latter. From
the same point of view, it is also interesting to mention the
explosions of nitric oxide, acetylene, and cyanogen, although
they only take place under the influence of mercury fulminate.

6. The decomposition of nitric oxide into elements, as it is
brought about by fulminate (p. 72), becomes complicated, owing
to the combustion of carbonic oxide produced by the detonation.
If it could be produced isolated, it would develop less pressure
than pure ozone. In fact, we arrive at the following figures : —
Heat liberated,

Q = -f 21-6 Cal. for NO (30 grms.) ;
temperature developed at constant volume,

t = 4204° ;
pressure produced,

p = 16*4 atm.

7. The detonation of acetylene, also induced by fulminate
(p. 69), gives rise to the following effects : —

Heat liberated,

Q = 61 Cal. for C2H (26 grms.) ;
temperature developed at constant volume,

t = 6220°;
pressure produced,

p = 23-8 atm.

2c

386 EXPLOSIVE GASES A.ND DETONATING GASEOUS MIXTURES

8. The detonation of cyanogen caused by fulminate (p. 71)
corresponds to the following effects : —
Heat liberated,

Q = 74-5 CaL for C2N2 (52 grms.) ;
temperature developed at constant volume,

t = 7600°;
pressure produced,

p = 28-8 atm.

In these calculations it is supposed that the molecular heat
of solid carbon is equal to that of gaseous oxygen at constant
volume.

We see from these figures that the temperature developed
and the pressure produced by acetylene and cyanogen would
exceed the effects produced by all other explosive gases, even if
we take into consideration the solid state of the carbon.

§ 2. DETONATING GASEOUS MIXTURES.

1. Chlorine and oxygen are the only simple gases which can
supply explosive gaseous mixtures by their association with
combustible gases, hydrogenated or carburetted. Among the
compound gases, the chlorine and nitrogen oxides share this
property.

2. In the following table the characteristic data have been
given for the principal detonating gaseous mixtures constituted
by these various gases, whether combustive or combustible.

Here the heat liberated results from the formation of certain
compound bodies ; consequently the maximum pressure, calcu-
lated theoretically, might be considerably diminished in practice,
owing to dissociation. It might also be diminished owing to
the variation of the specific heats. We shall revert to this
subject later on, but first give the theoretical values.

3. According to this table, the maximum work which can be
accomplished by one kgm. of the various explosive gases, work
which is in proportion to the heat liberated, that is, the
potential energy of these mixtures, varies only from single to
double for gases containing carbon and hydrogen mixed with
pure oxygen (the water being supposed to be gaseous).

Moreover, this work is nearly the same for the various
hydrocarbon gases.

Such work exceeds, moreover, that of all the solid or liquid
explosive compounds taken under the same weight. With
hydrogen and oxygen, for instance, it is four times as great as
that of ordinary powder, and twice as great as that of nitro-
glycerin.

EXPLOSIVE GASEOUS MIXTURES.

387

li

A

9

!!

J!i

ri

I 0 00 W5 0 T* 00 THW-*<03i
tl II £f il <M <N<N CO^^^

»0»0 ^ »O
'^'* 0 T*

'OOOOOOOOO

Cil>QOlO^C<ICO-^OO5CC
1-1 i-H C^ <N I-H ri (M <N CO <M <M

t>
(N

(M <N (M IM

.OOOCpCpCpCpO^QplpO^THf^lfyH

cp co h- t> N <?* »ft gi

S SS

COi-irH

t +

W W

~ ^ o

00 0 0 0 0 000 0 00

I!?

Us

as..

fit

al-

HI
ill

•I!

iill

f> fe S^

|-2 £g

&2H.2

S5S!

388 EXPLOSIVE GASES AND DETONATING GASEOUS MIXTURES.

With hydrocarbon gases it is three times that of powder, and
one and a half times as much as that of nitroglycerin. How-
ever, the advantages which might result from the potential
energy of explosive gaseous mixtures compared to that of solids
and liquids are counterbalanced in practice by the difficulties
arising from the greater volume of the gaseous mixtures and the
necessity of keeping them in resisting envelopes. From this
point of view of the potential energy of gaseous mixtures,
referred to the unit of weight, no coinbustive, generally speak-
ing, rivals pure oxygen, seeing that every other oxidising com-
pound contains inactive elements (useless weight), which share
the heat without supplying sufficient compensating energy at
the moment of the destruction of the oxidising compound.

4. We must remark that the theoretical pressures calculated
for the various explosive mixtures scarcely vary, except from
single to double, these being limits which we shall find by-and-
by, between the pressure really observed, notwithstanding the
diversity of composition and the condensation of the gases
taken into consideration.

5. Moreover, the pressures calculated are purely theoretical,
and only intended to serve as terms of comparison.

In fact, the figures measured by observers are much lower,
which is explained either by the short duration of the state of
integral combination which seems to correspond to the explosive
wave, or by the inaccurate estimation of the specific heats
employed in the calculations, or, finally, by dissociation.

Let us follow up this question.

It suffices to admit the existence of a certain dissociation in
order to reduce the pressures by one-half, or even one-third, of
the calculated values.

Nevertheless, the rapidity of propagation of the explosive
wave as it has been measured (p. 101) seems to indicate that at
the moment of its production the explosive system contains all
the heat liberated by an integral combination. The propagation
of the wave is, however, so rapid that the pressure observed
probably corresponds in every kind of apparatus to a system
which is already partially cooled, and it is this reduced pressure
which seems to correspond to the case of ordinary combustion.
We might .also explain the results observed by accepting the
variation of specific heats, especially if we double the mean
specific heat of water vapour or of carbonic acid.1

Experience has not yet expressed a definite opinion respecting
these different manners of conceiving the phenomenon. It
tends, however, to show that the part played by dissociation had
been exaggerated at first.

6. Let us now cite the figures really observed for pressures
subject to the reservations just named.

1 See " Essai de Me'canique Chimique," torn. i. pp. 344 et 346.

PRESSURES PRODUCED BY GASEOUS MIXTURES. 389

According to Bunsen's experiments,1 made by raising a valve
loaded with a weight, a mixture of carbonic oxide and oxygen
burnt at constant volume only develops 10*3 atm., instead of
24 as calculated. The number observed would correspond to the
combination of only one-third of the mixture on the hypotheses
of dissociation. Such a calculation is, however, based on the
employment of far too low a specific2 heat for the carbonic acid.

A mixture of hydrogen and oxygen, burnt at constant volume,
develops also, according to Bunsen, 9*6 atm. instead of 20 atm.
as calculated. The number observed would correspond again
to the combination of a third of the mixture on the hypotheses
of dissociation, but it is subject to the same objection for the
specific heat.

Mallard and Le Chatelier arrived at approximate experi-
mental values by their measurements, based on the employment
of a metallic manometer; say 8*6 atm. for the mixture of
carbonic oxide and oxygen, 9*2 atm. for the mixture of hydrogen
and oxygen, 14 atm. for methane and oxygen, 8 atm. for chlorine
and hydrogen, etc.

The following are the numbers observed by the author and
M. Vieille with the principal detonating mixtures, by another
method based on the registration of the pressures by means of a
movable piston : —

Hydrogen and oxygen : H2 + 0 ......

Hydrogen and nitrogen monoxide : H2 + N20
„ nitrogen and oxygen : H2 + N2 + 0

„ : H2 + 2N3 + 0
Carbonic oxide and oxygen : CO + 0 ......

„ „ and nitrogen monoxide : CO + N20
„ nitrogen and oxygen : CO + N2 + 0
„ :CO + N + 0
„ „ hydrogen and oxygen : CO + H2 + 02

atm.

7'7 atm. to 9 63
11*1
8'2
7-4
9-4
9 -7
7'7
8-0
7-8

„ 6 5 8-3

Methane and oxygen : CH4 + 04 ............ 13-6

Acetylene and oxygen : C2H2 + 06 ............ 13'7

Ethylene and oxygen : C2H4 + 06 ............ 13-8

Ethane and oxygen : C2H6 + 07 ............ 11-9

Ethylene, hydrogen, and oxygen : C2H4 + H2 + 07 ...... 13-3

Cyanogen and oxygen : 2CN + 04 ............ 19*5

Cyanogen, nitrogen, and oxygen: 2CN + N2 + 04 ...... 15'6

Cyanogen gives the maximum pressure according to theory.
However, the values observed are only two-fifths of the
theoretical values for hydrogen, carbonic oxide, and methane.
They are reduced to about a third for the other hydrocarbons
and for cyanogen.

It results from these indications that the real relations of the

1 " Annales de Chimie et de Physique," 4e se*rie, torn. xiv. p. 446. 1868.

2 See the author's remarks on this point (" Annales de Chimie et de Phy-
sique," 5' s^rie, torn. xii. p. 306).

3 According as the experiment was made in a chamber of 300 cc. or 4 litres.

390 EXPLOSIVE GASES AND DETONATING GASEOUS MIXTUKES.

pressures observed do not differ very much from the theoretical
pressures, so that, if necessary, the latter may be employed in
the comparisons, at least for a first approximation.

7. By replacing pure oxygen by its mixture with nitrogen,
that is, by atmospheric air, in order to effect the combustion of
the gases and vapours, we obtain systems which are very
interesting in their applications. In fact, it is a similar mixture
of air and methane which constitutes the fire-damp so much
dreaded in mines.

A similar mixture, composed of air and coal gas, has often
given rise to serious accidents in houses and sewers.

The vapour of ether, carbon disulphide, and petroleum
spirit, associated with air, have more than once produced fires
and explosions in manufactories and laboratories. Let us now
examine more closely the effects of this substitution of air for
oxygen.

8. It does not change the heat liberated, and consequently it
does not affect the maximum work which can be developed by
a given weight of the combustive body.

9. On the contrary, it modifies the pressures, and that in two
ways. In fact, at first sight it may be conceived that the
theoretical pressures should decrease by one-half, or even more,
owing to the necessity of heating the nitrogen, and even the
excess of oxygen, which lowers the temperature. For instance,
hydrogen mixed with five times its volume of air would not
develop, according to theory, more than 8'5 atm., instead of
20 atm., and only 5'1 atm. with ten times its volume of
air.

10. These figures are still above the real values, for the same
reasons that lower the pressures with pure oxygen, that is to
say, on account of dissociation, or rather, the increase of the
specific heats (p. 388).

However, the influence of these causes is limited by the
lowering of the temperature. Thus, according to Bunsen, one-
half of the mixture of carbonic oxide and oxygen would burn,
instead of one-third, as soon as the temperature falls below
2560°. Below 1146° the quantity burnt would again increase,
and continue to do so until total combustion took place.
Nevertheless, the last figures must be looked upon as doubtful.
In fact, they have been derived from observed pressures, assum-
ing the specific heats to be constant, which is not admissible ; l
now the effects observed can be explained equally by the varia-
tion of the specific heats, a variation which cannot be disputed
for compound gases.

For instance, since the specific heat of carbonic acid increases
with the temperature, the gaseous mixture which contains it is
brought to a lower temperature by a given quantity of heat,
1 See " Annales de Chimie et de Physique," 5" se*rie, torn. xii. p. 305.

PRESSUKE DEVELOPED. 391

and the pressure developed is diminished to the same extent.
The difference is, however, diminished by the introduction of a
certain quantity of inert gas, which tends of itself to lower the
temperature. The pressure will even be reduced proportionately
still more for such mixtures than for explosive mixtures con-
taining no inert gases.

11. This is confirmed by experience. As far back as 1861,
Him measured the pressure developed by the combustion of air
mixed with one-tenth of its volume of hydrogen, and he found
3*25 atm., instead of 5*14 atm.. The reduction would be about
one-third, instead of being greater than the half, as with pure
oxygen.

Mallard made similar observations 1 on various mixtures of
air and combustible gases.

Finally may be cited the recent experiments of Mallard
and Le Chatelier on the pressures developed by mixtures of
air and methane, and also on mixtures of air and coal gas.2
The measurements of these authors were effected by means
of a hollow spring, which served as a registering manometer
and communicated with a combustion chamber of 4 litres
capacity.

atm.

0-94 (CO + 0) mixed with 0-06 of inert gas (nitrogen and water vapour) 8-6

0-31 (CO + 6) „ 0-66 of C02, 0-02, 0-01 water vapour 6-0

0-955 (H2 + 0) „ 0-03 N + 0-015 water vapour 9-2

0-67 „ „ 0-32 0 + 0-01 „ 8-3

0-65 „ „ 0-34 H + 0-01 „ 8-1

0-49 „ „ 0-490 + 0-02 „ 7-2

0-32 „ „ 0-67 H + 0-01 6-3

0-33 „ „ 0-65 N + 0-02 6-3

0-19 „ „ 0-54 H + 0-25 N + 0-02 water vapour 5-15

0-17 „ „ 0-14 H + 0-69 N + 0-02 „ 5-0

0-95 (H + Cl) „ 0-03 H + 0-02 water vapour 8-1

0-74 „ „ 0-25 C1 + 0-01 „ 7-1

0-51 „ „ 0-47 H + 0-02 „ 7-0

0-41 „ „ 0-59 H + 0-01 „ 6-0

The detonating mixture with a methane base (CH4 + 04),
mixed with three times its volume of air, gave pressures ap-
proaching 7 atm.

With the same mixture, when pure, the figure rose to 14 atm.

The following is a table of some observations which M.
Vieille and the author made by means of a movable piston : —

I. Mixture of two combustible gases.

atm.

CO + H2+02 7-8

2CO + H6 + 06 8-3

C2H4+H2 + 07 13-3

1 " Annales des Mines," torn. vii. 1871.

8 " Journal de Physique," 2e se'rie, torn. i. p. 182.

392 EXPLOSIVE GASES AND DETONATING GASEOUS MIXTURES.

atm.

8-2

7-4

9-5

7.7

8-0
15-6

11. Mixture of detonating gases with an inert gas.

H2 + 0 + N2
H2 + 0 + N4
H2 + N2 + N20
CO + 0 + N2
CO + 0 + N
CN + 02 + N

From these various measurements very important results
may be deduced for the theoretical study of the temperatures
of combustion, specific heat and dissociation ; but this discussion
would lead us too far, and it suffices to cite the above-mentioned
figures as terms of comparison.

12. The temperature may be lowered to a limit at which the
inflammation ceases to propagate itself, and this limit is interest-
ing, since it is the same as that which commences to produce
the inflammation of the mixture in an adverse sense.

We have here two distinct notions to define : the composition
limit,1 and the temperature limit.

13. Composition limit of inflammability. An explosive gaseous
mixture ceases to burn when the relative proportion of one
of its components falls below a certain proportion. For instance,
3 vols. of electrolytic gas, formed by 1 vol. of oxygen and 2 vols.
of hydrogen, cease to ignite when mixed with 27 vols. of oxygen
or with 24 vols. of hydrogen.

A similar volume of water vapour, above 100°, also prevents
ignition. It is the same at the ordinary temperature with
18 vols. of nitrogen, 12 vols. of carbonic oxide, 9 vols. of
carbonic acid, 6 vols. of ammonia gas, hydrochloric acid or
sulphurous acid, etc.

Three vols. of gas, formed by 1 vol. of oxygen and 2 vols. of
carbonic oxide, ceases to ignite when mixed with 10 vols. of
carbonic oxide or 29 vols. of oxygen.

The mixture of methane with air only gives rise to an exact
combustion when it is formed by 9*5 vols. of air for 1 vol. of
methane. It ceases to burn where the proportion of air exceeds
17 vols. to 20 vols. These are very important data, owing to
the presence of fire-damp in mines.

The combustion is incomplete near the limits of inflamma-
bility.

These limits, however, vary considerably according to the
process of inflammation, and, above all, with the temperature
and mass of the body in ignition, which serves to produce the
combustion.

They also vary according to the nature of the electric spark,
when the latter is employed to produce ignition, the spark pro-
duced with the aid of a condenser being much more efficacious

1 See " Essai de M<?canique Chimique," torn. ii. pp. 73 et 343.

LIMITS OF EXPLOSIONS. 393

than ordinary sparks. All this can be easily understood, since
the igniting agent propagates the combustion around itself in a
sphere which is more or less extended, according to the quantity
of heat it supplies itself.1

Hence variations and strange phenomena result in a mixture
limit, the mixture becoming filled with small disseminated
flames, which are propagated hither and thither, and whose
production precedes the state of general combustion. These
curious effects have been the object of special study by Schlce-
sing and Demondesir.

The singular phenomena witnessed in the Solfatara at Pozzuoli
could also be cited. Towards certain points, especially in a
depression, vaporous wreaths are liberated, irregular jets of
water vapour mixed with a trace of sulphuretted ^hydrogen.
It suffices to bring near them some ignited body, such as
tinder, when the sulphuretted hydrogen burns in contact with
the air in which it is disseminated, with the production of a
cloud which gradually extends and propagates itself all round
to a considerable distance.2

The easy ignition of sulphur and its compounds is a great
factor in this circumstance, but has nothing to do with explo-
sive phenomena.

From the point of view of mechanical effects produced by
a detonating mixture, the rapidity with which the ignition is
propagated is very essential, the latter taking place sometimes
by ordinary combustion, and sometimes owing to a real ex-
plosive wave, which proceeds with incomparably greater rapidity
(pp. 49, 55, 88, 90). Now, the limits of composition at which
the explosive wave ceases to be produced are far higher than
those which correspond to simple ignition. This is a very
important result as regards applications (see p. 110). The limit
of inflammability, and especially the more or less easy propaga-
tion of the inflammation, is influenced by the pressure, which
increases the mass of heated matter in a given time and extent,
and consequently checks the influence of cooling.

The limit is also influenced by the initial temperature of the
mixture. That is, the excess of temperature of the body which
produces ignition above that of the inflammable mixture ought
to be less according as the latter mixture is raised beforehand to
a higher temperature (see p. 64).

Generally speaking, in order that the propagation of the com-
bustion may take place, it is necessary that the heat liberated
by the ignition of the first parts should be sufficient to repro-
duce in the adjoining portions the initial temperature at which
the combustion commenced.

1 " Essai de M^canique Chimique," torn. ii. pp. 338, 343 et 346.

2 Melloni et Piria, " Annales de Chimie et de Physique," 2" se'rie, torn.
Ixxiv. p. 331.

394 EXPLOSIVE GASES AND DETONATING GASEOUS MIXTUKES.

This is also a question in which there intervene, at the same
time, the quantities of heat liberated, the specific heats of the
products of combustion, and those of the gases in excess with
which these products are mixed. The variation of the specific
heats of the compound gases with the temperature enters
then into consideration here. If this were not the case, it would
always be easy to calculate d priori the temperature limit.
We will cite various facts respecting the latter temperature.

14. Temperature of inflammation. This temperature, which
corresponds to the minimum of work required to produce the
reaction, presents a certain amount of interest as regards appli-
cations. It has been frequently studied since the time of
H. Davy. On this point, the following are the most recent
data, due to Mallard and Le Chatelier 1 : —

2 vols. H + 1 vol. 0
1 „ H + 2 „ 0

1 „ air + 2 „ H

2 „ air + 1

M , , cm. ~r JL jj o-j. .. ... *ju\j

1 ,,0+2 „ H + 3vols.C02 ... 560° „ 590°

1 ,,0+5 „ CO

... 550° to 570°

... 530°

... 530° „ 570°

550°

630° „ 650°
650°

650° „ 660°

.0

1 ,,0+2 ,,CO

2 ,,0+1 „ CO

1 ,,0+2 „ CO + 3 vols. C02 ... 700° „ 715<

2 „ air + 1 „ CO + 3 „ C02 ... 715° „ 725°

0 nil PIT (650° explosion

58 »l »bil* \600° slow combustion

1 „ 0 +2 ,,CH4 650° to 660°

1 ,,CH4+9 „ air ... ... lower than 750°

It is remarkable how the ignition temperature of detonating
mixtures formed by the association of oxygen either with
hydrogen, carbonic oxide, or with methane, is but slightly
modified by the introduction of even a considerable volume of
foreign gases. The same happens at least as long as the limits
at which the mixture ceases to burn are not approached.

Nevertheless, the addition of an equal volume of carbonic
acid has a greater influence on carbonic oxide than on hydrogen,
as if the very products of the combustion of the mixture
exercised a special influence on the ignition temperature.

The authors also observed that there exist very notable
differences between the different intervals of time required to
ignite a gaseous mixture brought to a given temperature. Thus
the mixtures containing hydrogen or carbonic oxide ignite im-
mediately, whilst a certain time is required for the mixtures of
methane with air or oxygen. Hence it is that a bar of iron
when brought to a red heat does not ignite these mixtures,
since the gases escape before having been subjected to the
influence of this temperature for a sufficient time. These

1 " Comptes rendus des stances de TAcade'mie des Sciences," torn. xci.
p. 825.

TEMPERATUKE OF IGNITION. 395

observations are very important as regards the study of fire
damp.

15. The oxidation of gases and of organic substances heated
to 300° or 400° may be slowly effected, with a phosphorescent
glow, only visible in the dark, such as is seen when ether or
absolute alcohol is poured on a red-hot brick. The very products
of the oxidation are thus changed, as aldehyde is formed by
means of ether. If, however, these reactions are prolonged,
especially in the presence of a porous body of small mass, the
oxidation is rendered more active by the very heat it liberates,
and it may raise the temperature of the system up to the sudden
and explosive degree of ignition. This happens sometimes with
cotton impregnated with oil, with slowly burning tinder, with
brown coal, etc. It has been observed that in manufactories
and powder magazines serious accidents have been caused by
this cause of ignition, that is to say, due to the elevation
of temperature, owing to slow oxidation, which gradually
accelerates.

16. Gases containing sulphur ignite at much lower tempera-
tures than hydrocarbon gases, from 250° for example. With
reference to this matter may be instanced the following experi-
ment. Taking two flat dishes, ether is poured into one and
carbon disulphide into the other. If, then, a piece of red-hot coal,
but emitting no flame, be introduced into the ether it is extin-
guished ; but if it be only rolled in it so as to make the super-
ficial incandescence disappear, and we introduce it at once into
the carbon disulphide, the latter becomes ignited and can then
ignite the ether placed beside it.

Certain compounds, such as chlorinated and brommated
acetylene, ignite spontaneously in contact with air in consequence
of analogous phenomena. The same is the case with several
phosphoretted compounds.

17. Let us now return to the question of pressures. Instead
of burning a combustible gas by pure oxygen, we should be
inclined to expect some advantage arising from nitrogen mon-
oxide or nitric oxide, seeing that by their own decomposition
these gases supply an additional volume of nitrogen and a
supplementary quantity of heat. Nevertheless these advantages
are nearly compensated by the necessity of heating the nitrogen
(Table on p. 387).

18. It would be quite another thing if we only considered
the total work, for since this is proportional to the heat
liberated it is increased with nitrogen monoxide and nitric
oxide.

There also exist certain oxidising solids, such as potassium
chlorate, which supply more heat than free oxygen. On the
other hand, pure oxygen produces more than poljassium nitrate,
and more than most of its liquid or solid compounds.

396 EXPLOSIVE GASES AND DETONATING GASEOUS MIXTURES.

The heating of the elements, excepting oxygen, consumes,
moreover, a portion of this work, and this limits the elevation of
the temperature and pressure, as has just been said. Moreover,
let us note that the storage of oxygen in its compounds is always
very costly.

Hence it can scarcely be expected that economical engines
can be invented which will derive their motive power from solid
explosive substances, such as ordnance powder, as Papin once
imagined. Nevertheless, such machines, if they could be con-
trolled, would perhaps be applicable to special conditions, where
the reduction of the volume of the apparatus would be of
paramount consideration.

19. From these facts and considerations it results that the
employment of gaseous mixtures appears more economical in
machines than that of other explosive mixtures, solid or liquid.
In fact, gas engines are based on the combustion of coal gas by
air. However, in this case, the combustible and the combustive
are introduced from without and the products are gradually
discharged, and this limits the volume of the apparatus.

20. The gaseous mixtures we have under consideration have
been supposed to be produced under atmospheric pressure ; the
theoretical pressures which they then develop, being comprised
between 18 atm. and 51 atm., are far removed from the pressures
developed by most explosives, whether solid or liquid. The
effective pressures are even far less, since they do not exceed
20 atm. (p. 389), a result which differs from the opinions enter-
tained by most persons during the siege of Paris.

21. It would be advantageous to compress gaseous explosive
mixtures beforehand, but the pressures developed would become
comparable to those of solid or liquid mixtures only by employ-
ing enormous compressions capable of reducing the initial
volume of the mixture to one-hundredth, or a still smaller
fraction, that is by bringing it to a density equal to that of solids
or liquids. Apart from the practical difficulties attending such
a compression, it would result in liquefying most of the hydro-
carbon gases without liquefying the oxygen at the same time,
which would destroy the homogeneity of the explosive mixture
and the possibility of its immediate ignition.

§ 4. MIXTURE OF LIQUEFIED GASES AND ANALOGOUS LIQUIDS.

In this case certain advantages could be obtained by the
employment of nitrogen monoxide in the liquid state or of liquid
nitric peroxide, a compound which may be likened to a liquefied
gas owing to its great volatility.

The oxides of chlorine, whose combustive properties would be
extremely valuable if they were not too dangerous to manipulate,

MIXTURE OF LIQUEFIED GASES. 397

in consequence of their liability to explode spontaneously, need
not be considered. The oxides of nitrogen, on the contrary, are
stable in the cold.

Now, the liquid oxides of nitrogen can be associated with
liquefied hydrocarbons in hermetically closed vessels. Thus we
obtain mixtures whose theoretical explosive force is comparable
to that of the most energetic compounds, such as nitroglycerin
or the mixtures of potassium chlorate with gun-cotton or
potassium picrate.

Such mixtures of liquefied gases formed by the oxides ot
nitrogen do not detonate directly, but may do so under the
influence of primings of mercury fulminate, and this makes up
the resemblance between such mixtures and dynamite.

During the siege of Paris the author made some trials of this
kind with liquid nitrogen monoxide.

Eecently M. Turpin thought of having recourse to nitric per-
oxide, which is more easily handled, since it remains liquid up
to about 26° and may then be easily mixed with various com-
bustible compounds, such as carbon disulphide, ether, petroleum
spirit, etc. This is the base of the panclastites patented by this
ingenious inventor.

It is not yet known how far such a volatile body as nitric
peroxide, the vapour of which it is so dangerous to breathe, and
which is so corrosive, could be applied. It may, however, be
noted that this body nearly represents liquid oxygen, the loss of
energy being almost nil in its formation (p. 128). Its explosive
decomposition presents the disadvantage of heating the nitrogen
which does not intervene in the combustion.

The study of mixtures of this kind presents very great variety,
but the reactions they develop are but imperfectly known, except
as regards the systems which correspond to total combustion.
We shall therefore limit ourselves to these.

The following figures will serve to show the theoretical
energy of the mixtures formed by liquid nitrogen monoxide and
nitric peroxide.

398 EXPLOSIVE GASES AND DETONATING GASEOUS MIXTURES.

Explosive material.

Heat disen-
gaged by 1 grm.
water gaseous.

Reduced
volume of
the gases
formed.1

Permanent pres-
sure at 0° under
density of charge
1 3
n*

Theoretical
pressure at
moment of ex-
plosion.

Nitrogen monoxide and)
liquefied ethane . >
C2H6 + 7N20 . .
Ethylene8 or analogous)
liquid hydrocarbons . >
C2H4+6N20 . . .)
Liquefied acetylene * . \
C2H2 + 5N20 . . ./
Liquid benzene . . .\
C6H6+15N20 . . ./
Liquefied cyanogen 8 .)
C2N2 + 4N20 . . ./
Liquidcarbon disulphide\
CS2 + 6N20. . . ./

Liquid nitrobenzene . \
2C6H5N02 + 25N20 ./

cal.
1356

1418

1564
1339
1416
1012
1346

m.c.

0-79

0-76

0-73
0-73
0-69
0-59
0-71

atm.
590

21,700

«- 016
610

n

22,200
n
24.300

n - 0-12
640

n - 0-07
640

n

19,600

n - 0-07
690

n
26,900
n
15,850

n
590
n
630

n
22,100

n - 0-04

n

Liquid nitric peroxide j
and liquid * . >
2C2H6 + 7NO . . .)
Liquefied cvanogen . . \
C3N2 + 2N02 . . ./

Liquid nitrobenzene .\
4C6N5N02 + 25N02 ./
Liquidcarbon disulphide\
CS2 + 3N025 . . ./

Nitroglycerin . .

1794

1800
1568
1129
1460

0-79

0-62
0-60
0-47
0-72

644

23,800
n
25,400

n - 0-28
620

n

480

n - 0-12
470

n
20,000

n
15,040

n
480

n
19,000

n - 0-20

n

1 This volume ought really to be multiplied by (1 + a 0 to render the
water actually gaseous at t.

2 One grm. in n cubic centimetres ; water liquid. These figures are only valid
when «is sufficiently large for the carbonic acid not to be liquefied, or in the case
of carbon dlsulphide, the sulphurous acid.

3 Favre found the heat of liquefaction of N20 = 44 grms. to be = 4'4 Cal.
This figure has been taken for the other liquefied gases.

4 Nitric peroxide is incompatible either with ethylene or benzene.

5 Turpin employed barely half the proportion of nitric peroxide indicated
here; this gave rise to incomplete combustion with deposition of sulphur.

MIXTURES OF AIR AND CHARCOAL. 399

§ 5. GAS AND COMBUSTIBLE DUSTS.

1. A gas may form explosive mixtures, not only by its asso-
ciation with another gas, but also with a solid or liquid dust.
Hence we obtain systems of a very special order. Their explo-
sive nature may easily be conceived, seeing that these systems,
when once ignited, give rise to sudden expansion, accompanied
by an increase of pressure. However, the explosion of such a
system is necessarily slower than that of a purely gaseous
mixture, since the propagation of the reaction only takes place
as each solid particle is reached by the incandescent gases
arising from the combustion of the neighbouring particles.
Hence we may conceive the influence exercised .by the slightest
trace of combustible vapour or gas already mixed with air in
facilitating ignition.

2. Explosions of this kind have been observed in coal mines,
in flour mills and warehouses, and in places containing sulphur
in the form of an impalpable powder.

The clouds formed by petroleum vapours and other volatile
hydrocarbons have also given rise to similar explosions in
cellars or magazines, or even in the open air, but in this case
the effects are of a mixed character, owing to the peculiar
vapour tension of these hydrocarbons, a portion of which should
be considered as gaseous in these mixtures.

3. Eeference will only be made to mixtures formed by air
associated with a combustible dust. Let us first define the
limits which correspond to the maximum effect with mixtures
of air and combustible dust, supposed to be effected in suitable
proportions at the moment of explosion.

(1) Mixtures of air and charcoal. One cubic metre of air
may by its oxygen generate 208 litres of carbonic acid reduced
to 0° and 760 mm. The same volume of air would burn
112 grms. of pure carbon. Now such a system, namely, an
intimate and as uniform a mixture as possible of air and
carbon in the form of powder would develop a theoretical
pressure of 15*5 atm. if it were burnt at constant volume. If
the quantity of charcoal were doubled (224 grms.) and the
whole could be changed into carbonic oxide, we should obtain
416 litres of the latter gas, and the pressure developed would
be 6 '7 atm.

If necessary, carbonaceous dusts may be assimilated to carbon
for similar effects.

At any rate we see that the maximum limit of theoretical
pressures which can be developed by the combustion of a
carbonaceous dust is similar to the pressures developed by fire-
damp itself.

(2) Mixtures of air and starch. Let us take it to be starch
dust which, to facilitate calculation, we may substitute for

400 EXPLOSIVE GASES AND DETONATING GASEOUS MIXTURES.

flour : 1 cubic metre of air would burn 255 grms. of starch
(C6H10O5), developing a theoretical pressure rather above that
which carbon would produce (owing to the aqueous vapour).

(3) Mixtures of air and sulphur. Finally 1 cubic metre of air
would burn about 300 grms. of powdered sulphur, developing a
pressure of 11 atm.

4. The limits we have just defined presuppose a uniform dis-
tribution of the dust in the air, which, however, can only be
realised under very special conditions of movement and division
of the dust.

It is, moreover, difficult to reproduce them by experiment.

Such systems, moreover, supposing them to be produced
instantaneously, cannot exist in the same state, without violent
agitation, since the action of gravity tends to separate the com-
ponents, contrary to what occurs with systems formed by the
mixture of two gases.

In a system consisting of gas and dusts the relative propor-
tions are therefore continually modified by time, as are also the
combustible properties of the system which can only maintain
their maximum for a very short period.

5. On the contrary, however, combustible dusts mixed with
air remain inflammable far beyond the combustible limits of
purely gaseous mixtures, and one single grain in a state of
ignition suffices to propagate the flame, either to the neighbour-
ing strata, or to the surface of the surrounding solid bodies.

Such seem to be the most ordinary conditions of the accidents
produced by inflammable dusts at the bottom of mines. They
are due to a propagated inflammation rather than to a real ex-
plosion. Nevertheless, the expansion of the gases is sufficiently
sudden to produce violent mechanical effects, which are very
dangerous.

6. The propagation of fire in a mixture of air and combustible
dust is intensified by the movements of expansions and the
projection of gaseous masses, inflamed at the very outset.
Hence it is as regards coal-mines that experience has led to
attributing a very dangerous part to carbonaceous dust, raised
like a whirlwind when a blast is fired, and which propagates
fire and asphyxia even to a great distance in the galleries.
Thus it has happened that a blast, the flame of which did not
extend beyond 4 metres, has propagated combustion through
the dust that was raised, to a distance of more than 14 metres,
and reached workmen who thought they were out of danger.

Blastings which blow out are especially dangerous in this
respect.

At the outset a real amplification of the flame is produced ;
afterwards it is a simple propagation of the ignition of the dust.

The finer the dust is the more the volume of the initial flame
provoking the phenomenon can be limited.

MIXTURES OF AIR AND CHARCOAL. 401

7. The proportion of volatile substances which coal dust can
supply also plays an essential part, for these substances,
reduced to vapour by combustion, in their turn promote the
propagation of the ignition. This dust, however, only burns in
an incomplete manner and by means of a kind of distillation
which deprives it of its hydrogen, leaving as a residuum portions
of coke adhering to the walls and wood-work. Owing to this
fact it is not the mixture of air and dust effected in theoretical
proportions which is the most combustible, but a mixture which
is richer in carbon, seeing that only the superficial layers of the
grains take part in the combustion.

8. Finally, the propagation of the inflammation is effected all
the better if the air in the mine already contains a small
quantity of some combustible gas, such as methane, the propor-
tion of which is often too feeble for it to form by itself a
detonating mixture with the air in the mine.

In mixtures of this kind, even an inert dust, such as magnesia,
lowers the limits of combustibility ; a mixture containing only
2'75 of fire-damp may thus burn, but in this case the combus-
tion does not propagate itself. This circumstance seems to be
due to the storage of heat by the magnesia, which then heats
the neighbouring gaseous particles, and consequently lowers
their limit of combustibility (p. 393).

Combustible dusts are evidently more efficacious. They
increase, moreover, the violence of the explosion produced by
fire-damp, owing to the volume of the gases and the supple-
mentary heat they supply. Besides this, they tend to increase
the quantity of carbonic oxide which is so dangerous to the
mines.

All these circumstances, observed by engineers and managers
of mines, have been made the object of methodical experiments
by Galloway and Abel in England, and also by Mallard and
Le Chatelier in France,1 in an inquiry recently instituted by the
fire-damp commission. For further details the reader is referred
to the publications issued by that commission.

1 " Annales des Mines," Janvier et FeVrier, 1882.

2 D

( 402 )

CHAPTER IV.

DEFINITE NON-CARBURETTED EXPLOSIVE COMPOUNDS.
§ I-

THE general list of these compounds has been given on
p. 368. The only ones which have been the object of sufficiently
accurate experiments to speak of them here are — nitrogen
sulphide, nitrogen chloride, potassium chlorate, and certain
ammoniacal salts of the higher oxygenated acids, such as
ammonium nitrate, perchlorate, and bichromate.

§ 2. NITROGEN SULPHIDE : NS.

1. Nitrogen sulphide contains, for 1 equiv. = 46 grms.,
32 grms. of sulphur and 14 grms. of nitrogen. Or for 1 kgm.,
sulphur 696 grms., nitrogen 304 grms. Its density is equal to
2*22. It is solid and crystallised. Heated to 207° it is decom-
posed, suddenly and explosively, into sulphur and nitrogen.

2. According to the thermal study which we have made of
this body (p. 262), its explosive decomposition at constant
pressure,

NS = SN,

liberates + 32*2 Cal. for 46 grms. ; at constant volume 4-
31-9 Cal.

3. It develops 1116 litres of nitrogen.

This gives for 1 kgm. 694 Cal. and 242-6 litres of nitrogen
reduced to 0° and 760 mm.

At the temperature of explosion, sulphur should be regarded
as gaseous and even as possessing its theoretical density, which
it acquires beyond 800°, according to Troost and Deville. The
total volume of the gases for 1 kgm. would then be at the
temperature t : 48 5 -2 litres (1 + at).

To calculate the theoretical pressure at constant volume it is
necessary to know the specific heats of sulphur under its
various states, and the heats of transformation of this body in
passing from the solid to the liquid state, and from the liquid

PRESSURE OF EXPLODED NITROGEN SULPHIDE. 403

to the gaseous state, lastly from the gaseous state developed
towards 448°, in which sulphur has a density treble its theo-
retical density, to the state in which it resumes its normal
density. This calculation cannot be performed solely upon the
basis of experimental data, which are partly wanting. We
have shown how they can be compensated for, up to a certain
point (p. 27). The reader will there find the data for the
calculation, of which the results will simply be given here.

4. The theoretical temperature developed by the explosion
of nitrogen sulphide may be estimated at 4375°.

5. Let us now estimate the pressures.

Take first the permanent pressure, that is, the pressure after
cooling, the explosion having taken place in a constant capacity.
For a density of charge equal to unity, the pressure at 0° would
be 242'6 atm., if the volume occupied by the sulphur were
nil. But one litre in reality contains 340 c.c. of solid sulphur ;
the permanent pressure will therefore become 367'6 atm. ; or
390 kgm. per square centimetre, admitting Mariotte's law.

If the nitrogen sulphide had exploded in an entirely filled
capacity, that is to say in its own volume, one kgm. would
occupy only 450 c.c. After explosion, the volume of the solid
sulphur being deducted, there would remain 110 c.c. for the
nitrogen; which would bring the theoretical pressure to
2205'6 atm., or 2340 kgm. per square centimetre.

In general, one kgm. of this substance enclosed in a capacity

of n litres, that is to say, supposing the density of charge to be -,

n
the permanent pressure per square centimetre will be —

250*4 kgm.
n - 0-340 *

a theoretical value which will be the nearer the real one the
greater n is.

6. The calculation of the pressures developed at the moment
of explosion is more hypothetical ; we shall, however, refer to it
as a term of comparison (see p> 28). This calculation should
be performed on the supposition of the sulphur being gaseous at
the time of the explosion. The pressure developed will then
be, for a density of charge equal to unity —

485-2 atm. (l -f^).
Supposing t = 4375°, as has been said above,

17>0'

and the above product becomes 8246 atm. ; or 8555 kgm. per
square centimetre.

2 D 2

404 DEFINITE NON-CABBURETTED EXPLOSIVE COMPOUNDS.

If the nitrogen sulphide exploded in its own volume, we
should have 18702 kgm.

More generally for the density of charge - we shall have —

8555

n

These are the theoretical figures.

7. The following are the real figures which we have obtained
with the apparatus described (p. 21) : —

Density of
charge. Pressures.

0-1 ............... 815kgm.

0-2 ............... 1703 „

0-3 ............... 2441 „

which gives for a density equal to unity, 8150, 8515, and 8137 ;
mean, 8270 ; a value only slightly lower than the figure 8555,
deduced from theory.

These pressures are nearly the same as those of mercury
fulminate. However, nitrogen sulphide is much less sudden in
its effects, owing, doubtless, to a certain expansion produced by
the successive transformations which the sulphur undergoes in
cooling — change of gaseous density, liquefaction, and solidi-
fication. Hence it follows that the effects produced by the two
substances, regarded as detonators and playing the part of caps,
must be very dissimilar.

§ 3. NITROGEN CHLORIDE.

1. Nitrogen chloride is considered to be one of the most
dangerous bodies to handle, owing to the facility with which it
explodes, by shock, friction, or contact with various bodies.

2. Its equivalent = 120*5 grms.

3. Composition —

Nitrogen ............ 116

Chlorine ... ......... 884

1000

4. It is liquid, but may, however, be evaporated in a current
of air at ttie ordinary temperature.

5. Its density is equal to T65.

6. Nitrogen chloride is decomposed when heaterl even below
100°, and is slowly destroyed at the ordinary temperature. It
explodes on contact with a great number of bodies.

7. Nitrogen chloride explodes, resolving itself into its ele-
ments —

NC13 = N + C13.

It develops in this way 44*64 litres of permanent gases, or
370 litres per kilogramme.

EXPLOSION OF NITEOGEN CHLORIDE. 405

The quantity of heat liberated in this reaction is considerable,
but not well known. Indeed, the experiments of Sainte-Claire,
Deville, and Hautefeuille on this point1 have given two
numbers which, calculated with the values actually adopted for
the heats of formation of ammonia and its chloride (pp. 237 and
243), vary almost from the single to the double.

8. The permanent pressure may, however, be calculated. For

' . 1 . , , . 370-4 atm. 3827 kgm.

a density of charge — it would be , or s —

n n n

per square centimetre, supposing n large enough, in order that
the chlorine may not assume the liquid state.

On the contrary, if the chlorine be liquefied, the density of
liquid chlorine being 1*33, 1065 grms. of this body will occupy
807 c.c., and hence the pressure developed by the nitrogen,
which formed only the fourth of the gaseous volume at the

normal pressure, will be *> a much lower figure than

n — 0*80

that yielded by nitrogen sulphide.

9. The maximum work which can be developed by nitrogen
chloride is considerable, but the actual data tend to show that
this work is greatly inferior to that of black powder, when an
equal weight of both these substances are exploded in any equal
capacity. These are results which seem at first sight to contra-
dict what is known of the terrible phenomena produced by
nitrogen chloride. Nitrogen chloride is, in fact, regarded as the
type of these shattering substances, which cannot be employed
in firearms to effect the same work of projection wnich powder
realises by its progressive expansion.

10. We shall now try to account for these differences.

The principal one must doubtless be attributed to the nature
of the products of explosion and the complete absence of every
compound capable of dissociation. The pressure and the work
result from the heat liberated by the decomposition of the
nitrogen chloride. Now, the latter gives rise to elementary
bodies which have no tendency to recombine, whatever be the

1 "Comptes rendus des stances de TAcad^mie des Sciences," torn. Jxix.
p. 152. The authors employed two reactions — that of chlorine on ammonium
chloride in presence of water, and that of hypochlorous acid on the same
salt, and they believed the results which follow from their measurements
were concordant. But the values deduced from the data they adopted,
putting aside certain errors in calculation, would be 51'7 and 39'3. By
reckoning, still with the aid of their measurements but by means of the heats
of formation actually received for ammonia, hydrochloric acid, and ammonium
chloride, we find : 57'8 and 37-8. The discordance in these results is prob-
ably owing to the reactions not taking place entirely according to the formulae
indicated. It would be well to resume these measurements, operating upon
pure nitrogen chloride and by the decomposition method, synthesis being
here very uncertain.

406 DEFINITE NON-CAEBUKETTED EXPLOSIVE COMPOUNDS.

temperature and pressure. The initial pressure will therefore
at once attain its maximum, and nitrogen chloride at once yield
the whole work of which it is capable, whether in dislocating
the materials on which it acts, or by crushing them, if they are
not sufficiently compact, or, lastly, by communicating to them its
energy under the form of movements of projection and rotation.

Moreover, the pressure will decrease very suddenly, as much
by the fact of these transformations as by that of the cooling
and of the expansion of the gases ; and it will decrease without
any fresh quantity of heat, gradually reproduced, intervening to
moderate the rapid fall in pressure. An enormous initial
pressure, becoming almost suddenly lowered, are conditions
eminently favourable to the rupture of vessels containing
nitrogen chloride.

These conditions contrast with those which accompany the
combustion of powder, as in the latter the final state of combina-
tion of the elements is not produced at the very first in a com-
plete manner, but becomes more advanced according as the
temperature falls. The initial pressure could therefore be less
with powder than with nitrogen chloride. But, to compensate
this, it decreases less quickly, owing to the intervention of the
fresh quantities of heat produced during the period of cooling.
These considerations have already been insisted upon (p. 12).

In order to fully explain the differences observed between the
properties of nitrogen chloride and those of ordinary powder,
the duration of the molecular reactions must also be taken into
account.

The almost instantaneous transformation of nitrogen chloride
develops pressure of which the sudden increase does not give
the surrounding bodies time to put themselves into motion, and
thus gradually yield to these pressures. It is well known that
a film of water on the surface of nitrogen chloride is sufficient
to produce such effects.

11. This would be the proper place to speak of nitrogen
iodide, a compound so sensitive to shock and to friction that it
is hardly possible to isolate it. Everybody has seen the experi-
ments of which this body is the subject in public lectures.
But it is so unstable that up to the present it has not been
possible to determine its composition with certainty. No
attempt has been made to measure its heat of formation.

§ 4. POTASSIUM CHLORATE: C103K

1. Potassium chlorate is not explosive by simple shock or
friction at the ordinary temperature. However, the powdered
salt, wrapped in a thin piece of platinum foil and strongly struck
with a hammer on an anvil, yields some chloride; that is to
say, it undergoes partial decomposition.

POTASSIUM CHLORATE. 407

When melted and heated too suddenly, it is decomposed with
incandescence, and sometimes causes dangerous explosions.

2. The equivalent of potassium chlorate is 122*6.

3. Composition —

Oxygen 392

Potassium 319 \

Chlorine 289 /

608

1000

4. Density, 2'33.

5. Heat of formation —

Cl + O3 + K = C103K liberates + 94 Cal.

6. The salt melts at 334°, without undergoing decomposition,
at least if the operation is carried on at constant temperature.
It decomposes slowly at 352°, but more rapidly if the tempe-
rature be suddenly raised.

This decomposition is effected by two distinct processes.
The salt heated with precaution yields a large quantity of
potassium perchlorate —

4C103K = KC1 + 3C104K,

a reaction which liberates + 51*5 Cal., but which would give
rise to no gas if it were developed alone.

As a matter of fact, it is always accompanied by another
transformation, effected on a considerable portion of matter,
viz. the direct decomposition of potassium chlorate into potassium
chloride and oxygen —

C103K = KC1 + 03.

The latter reaction becomes more and more predominant,
according as the operation takes place at a higher temperature,
or as the substances are superheated. It even seems to be the
only one that takes place in presence of copper oxide or of
manganese dioxide.

7. This decomposition, referred to the ordinary temperature,
liberates +11 Cal. at constant pressure, or + 11*8 at constant
volume.

This makes per kilogramme, 81 '6 Cal. at constant pressure
and 8 7 '4 Cal. at constant volume.

At 350° and upwards, this reaction liberates more heat, the
potassium chlorate being melted, but the exact figure cannot be
given, the melting heat of the salt not having been measured.

8. We thus obtain 33'48 litres of gas (reduced volume), or,
per kilogramme, 2731 litres at the normal pressure and at 0°.

9. The molecular specific heat of potassium chloride being
12*9 and the special molecular heat of oxygen, O3, at constant
volume, 7'4, this makes in all 20 -3. From this we conclude

408 DEFINITE NON-CABBUBETTED EXPLOSIVE COMPOUNDS.

that if these data remained constant, the theoretical temperature
of the products would be 581° at constant volume.

The initial body being taken at t, the theoretical temperature
would be 581° + t. Take, for instance, t = 400°, the tempe-
rature developed by the decomposition would reach 982°. It
would even then be increased by some hundred degrees, on
account of the heat of fusion of potassium chlorate.

None of these theoretical data are too much at variance with
observable results, if the incandescence developed at the moment
of the explosive decomposition of potassium chlorate be taken
into account.

10. The permanent pressure, after cooling, is calculated, de-
ducting the volume of the potassium chloride, or 304 c.c. per
kilogramme of the fixed capacity, in which decomposition took

place. For a density of charge -, we have —

2731 atm.

n - 0-304
or, which is the same thing —

282-2 kgm.
n - 0-304

which makes for n = 1 : 405 kgm. per square centimetre.
If the chlorate be supposed to explode in its own volume,

n = (r^-= 0*429, that is, the permanent pressure would be

'

2306 kgm.

11. At the temperature of decomposition, the latter being
supposed produced without the aid of an external heating, the
theoretical pressure is nearly trebled. It becomes, in fact,
neglecting the dilation of the potassium chloride —

_ 855 atm. __ 869 kgm.

n - 0-304 n - 0'304 n - 0-304

per square centimetre.

This makes for n = 1 : 1248 kgm.

§ 5. AMMONIUM NITRITE : NH3HNOa.

1. The equivalent is equal to 64 grms.

2. The following is the composition : —

Nitrogen 437-5

Hydrogen 62-5 \ w . nro ^

Oxygen 500-0 / Water 562'5

1000-0

AMMONIUM NITRITE 409

The density is not known.

3. The dry salt may explode when suddenly heated, even
below 80°.

4. It is decomposed principally into water and nitrogen,
N02HNH3 = N2 + 2H2O, which yields 22'32 litres of permanent
gases, or, for 1 kgm., 349 litres.

5. The same reaction liberates -f 73'2 Gal. at constant
pressure and -f- 734 Cal. at constant volume, or, for 1 kgm.,
1144 Cal. at constant pressure, 1153 Cal. at constant volume.

6. At the temperature of the explosion the water is gaseous,
which trebles the volume of the gases. The latter therefore
occupy

66-96 litres (l + —

On the other hand, the heat developed must be referred to
the formation of gaseous water, which reduces it to 4- 5 3 '8 Cal.

7. The theoretical temperature of the products is obtained by
dividing 53,800 by 19'2, which gives 2800.

8. The permanent pressure is obtained by subtracting from
the fixed capacity the volume of the water, or 562 c.c. for
1 kgm.

For the density of charge, — , it will therefore be —

349 atm. 361 kgm.
n - 0-5625 °r n - 0'56

for n = 1 : 820 kgm. per square centimetre.

9. At the theoretical temperature of decomposition the
pressure becomes, the water being gaseous —

1147 \l + 273J 12961 atm. 13393 kgm.
- = _ _ or - -,

n n n

which makes for n = 1 : 13,393 kgm. per square centimetre.

§ 6. AMMONIUM NITRATE: N03HNH3.

1. Equivalent = 80 grms.

2. Composition —

Nitrogen ......... 350

Hydrogen ......... 50 Water ... 450

Oxygen ......... 600 Excess of oxygen 200

1000

3. Density, 1'707.

4. Heat of formation from the elements —

N2 + H4 + O3 = N03HNH3 + 87'9 Cal.

410 DEFINITE NON-CARBURETTED EXPLOSIVE COMPOUNDS.

5. This salt commences to decompose a little above 100°, not
without being partly sublimed (p. 243). Towards 200°, it
separates in a sufficiently definite manner into nitrogen mon-
oxide and water, without, however, there being a fixed tempera-
ture at which this destruction takes place.

If the salt be superheated, and especially from 230° upwards,
the decomposition grows more and more rapid (nitrum flammans)
and ends by becoming explosive at the same time as the salt
becomes incandescent (p. 6).

6. A sudden decomposition yields at the same time as
nitrogen monoxide various products corresponding to simul-
taneous decompositions, so that ammonium nitrate can undergo
eight distinct transformations, several of which are simultaneous
in certain explosive decompositions. We proceed to enumerate
them, calculating for each of them the heat developed, the
permanent pressure, the theoretical temperature and pressure.

7. — (1st) The integral volatilisation absorbs an unknown
quantity of heat, and therefore affords no opportunity for
calculation.

8. — (2nd) The integral dissociation into acid and base,

N03HNH3 (solid) = N03H (gas) -f NH3 (gas), would absorb - 41 '3 ;
The fused salt, - 37'3.

This makes, for one kgm. of solid salt, 516 Cal. Hence this
reaction is not explosive and cannot be produced without foreign
energy.

9. — (3rd) The formation of nitrogen and free oxygen —

N03HNH3 (solid) * N? + O + 2H20,
would, on the contrary, liberate heat, viz. at constant pressure.

The water being liquid, 4- 50*1 Cal. ; the water being gaseous,
4- 307 ; at constant volume these figures become -1- 50*9 and
4- 337. There is, therefore, produced, at the temperature t, a
gaseous volume equal to

33*5 litres (1 4- ^IT ), the water being liquid ;

or 781 litres ( 1 4- 0=5), tne water being gaseous.
Or for 1 kgm., 4187 litres (l 4- <Tzr ),the water being liquid;

> Zfo/

or 976 litres (l 4- ^), the water being gaseous.
V 2*1 &'

The theoretical temperature developed at constant volume would

»«••

DECOMPOSITION OF AMMONIUM NITRATE. 411

The permanent pressure at 0°, taking into account the volume
of the liquid water (450 c.c. for 1 kgm.), will be, for a density

of charge — —
n

4187 atm. Qr 432 kgm.

n - 0-450 n - 0450

For n = 1, we should have in theory 787 kgm. per square
centimetre. The salt being decomposed in its own volume,
that is, 1 grm. occupying 0'585 c.c., the permanent pressure
becomes 3200 kgm.

At the temperature developed by decomposition, the water
being gaseous, the theoretical pressure would be —

976 V + 273~/ 6344 atm. 6555 kgm.

— =? - - or

n n n

The salt being decomposed in the same volume which it
occupies in the solid state, 11,200 kgm,

These values represent the maximum of the effects which can
be produced by the decomposition of ammonium nitrate, all the
following reactions producing less heat and a less volume of gas.

10. — (4th) The formation of nitrogen monoxide is the pre-
ponderating reaction when we proceed by progressive heating.
This reaction,

N03HNH3 (solid) N20 + 2H20,

would liberate —

Liquid water, -f 29*5 Cal. at constant pressure, 4- 301 at con-
stant volume.

Gaseous water, 4- 10'2 Cal. at constant pressure, -f 121 at con-
stant volume.

The volume of the gases produced at the temperature t will
be—

22-3 litres (l 4- ^), the water being liquid,

66-9 litres (l + ^r ), the water being gaseous ;
\ Zio/

or for 1 kgm. —

278-7 litres ( 1 + 7=r\ the water being liquid,
> 2 to/

836'2 litres (l + •7—- \ the water being gaseous.
\ 27o/

The theoretical temperature at constant volume is —

412 DEFINITE NON-CARBURETTED EXPLOSIVE COMPOUNDS.

The permanent pressure, at 0° —

2787 atm. 288 kgm. m
n _ 0-45 °rw - 0-450*

but this value ^ only applicable when n is large enough for the
nitrogen monoxide not to be liquefied. For high densities of
charge it becomes imaginary.

At the theoretical temperature, the water being gaseous, the
pressure would be —

+ 2737 2559 atm. 2642 kgm.

^^ or •

n n n

The salt being decomposed in the volume which it occupies
in the solid state, 4500 kgm. All these values hardly amount
to more than the third of the figures corresponding to the forma-
tion of free nitrogen.

11. — (5th) The formation of nitric oxide,

N03HNH3 (solid) = N +NO + 2H20,
would liberate —

Liquid water, -f 28*2 Cal. at constant pressure, + 29*3 at constant

volume.
Gaseous water, -|- 9*2 Cal. at constant pressure, + 11 '2 at constant

volume.
The volume of the gases produced at the temperature t —

3 3 *5 litres (l + ^7;), the water being liquid ;

78*1 litres (l + ^/> the water being gaseous.

This volume is the same as in the case of the formation of
free nitrogen.

The theoretical temperature at constant volume, =

21'6

518°.

The permanent pressure at 0° is the same as for the formation

of free nitrogen, viz. — ' ; it is, moreover, imaginary for

n — 0*450

high densities of charge, nitric oxide becoming liquefied.

At the theoretical temperature, the water being gaseous, the
pressure would be —

976 V1 + 273) 2753 atm. 2840 kgm.

= QJ. .

n n n

The salt decomposed in the volume which it occupies in the

DECOMPOSITION OP AMMONIUM NITRATE. 413

solid state, 4860 kgm. ; values nearly the same as those corre-
sponding to nitrogen monoxide.

12. — (6th) The formation of nitrogen trioxide —

3N03H, NH3 (solid) = 4N + N203 + 6H20.

Liquid water, + 42*5 Cal. at constant pressure, 4- 43*1 at constant

volume.
Gaseous water, 4- 23'3 Cal. at constant pressure, 4- 251 at constant

volume.

The volume of the gases at the temperature t is the same as
for nitrogen monoxide, viz. —

4- 22*3 litres (1 4- T^A the water supposed liquid ;
\ 273/

4- 66'9 litres (\ -J -- \ the water supposed gaseous.
\ 273/

In all cases this reaction can only be developed upon a
fraction of matter ; nitrogen trioxide existing only in the dis-
sociated state in presence of nitric oxide and nitric peroxide,
which are in excess. Hence it appears useless to give the
calculations relative to the pressures and temperatures, a remark
which is equally applicable to the following reactions.

(7th) The formation of nitric peroxide,

2N03HNH3 (solid) = N3 4- NOa 4- 4H2O,
would liberate — •

Liquid water, 4- 48 Cal. at constant pressure, 4- 49*5 at constant

volume.
Gaseous water, 4- 29*5 Cal. at constant pressure, 4- 31*4 at constant

volume.

The volume of the gases, at the temperature t, is the same as
for nitrogen monoxide and for nitrogen trioxide.
(8th) The formation of gaseous nitric acid,

5N03HNH3 (solid) = 2HN03 4- 8N 4- 9H20,

would liberate, the acid and the water being gaseous, and not
combined, + 33'4 Cal. at constant pressure, 4- 351 Cal. at con-
stant volume.

The volume of the gases at the temperature ty the water and

the acid assuming the gaseous state, would be 67 litres (1 4- ^«\

( t
That of the permanent gases, 17 '8 litres (l -f ;~A being the

least of all. On the contrary, the heat liberated is the greatest.
But this mode of decomposition is accessory.

13. We have deemed it useful to develop the study of the
manifold and simultaneous modes of decomposition of am-

414 DEFINITE NON-CARBURETTED EXPLOSIVE COMPOUNDS.

monium nitrate, as typical in the study of explosive substances ;
this multiplicity of simple reactions not being known, generally
speaking, with precision, for the other bodies. It will be noticed
with regard to this point that in the explosive decompositions of
this salt, the heat at constant volume may vary from + 351 Cal.

to 11*2 Cal. ; the volume of the gases, from 62*5 litres ( 1 -f- ~^o

/ t \ ^ *1

to 78-1 litres

§ 7. AMMONIUM PERCHLOHATE : C104H, NH3.

1. Equivalent, 117-5.

2. Composition —

Cl ... .. .......... = 302

N ............... = 119

H ...... ;., ^ ... = 34

0 ,,. .<* ......... = 545

1000

3. Heat of formation from the elements —

Cl + N + H4 + 04 = C104H, NH3 + 797.

4. The decomposition of this salt by heat has already been
studied (p. 356). The principal reaction is C104H, NH3 (solid)
= Cl + O2 + N + 2Hj,0 (gaseous) + 38*3 Cal. ; the water being
liquid, + 58 ; or, for 1 kgm., 4963 Cal.

At constant volume we should have, -h 59 '5 Cal., the water
being liquid, and -f- 40 '7, the water being gaseous.

5. This reaction produces at the temperature t} the water

being supposed liquid, 44'6 litres M -f — - ) ; the water

\ 273/

gaseous, 89'3 litres ^1 + — J; of, for one kgm., 379'6 litres

(l + ^|)i the water being liquid, and 759*2 litres (l + -~^t

the water gaseous.

6. Theoretical temperature at constant volume,

40700 - 1563°
~2W'

7. The permanent pressure at 0°, for a density of charge — ,

taking into account the volume of the liquid water (307 c.c. for
1 kgm.), would be —

379-6 atm> 392 kgm.

n - 0-307 °1% n - 0'307

AMMONIUM BICHROMATE. 415

But this figure is only applicable to low densities of charge.
For high densities the chlorine is liquefied and occupies 227 c.c.
The volume of the permanent gases is in this way diminished
by one-fourth.

The permanent pressure then becomes - -^ — - , which makes,

n — Q'534

for n = 1, 631 kgm. per square centimetre.

At the theoretical temperature of decomposition, the water
and the chlorine being gaseous, the pressure becomes —

893 atm.l

. 6Q()4

- = _ - or - §_,
n n n

figures which are not very remote from the maximum effects of
which ammonium nitrate is capable.

The decomposition which served as base for the foregoing
calculations is not exclusive, a small quantity of perchlorate
being decomposed at the same time with formation of hydro-
chloric acid ; now

2C104NH3 = 2HC1 + 3H20 + N + 05 liberates + 30-8 Gal,
producing 1004 litres (l -f ^zr) of gas. But this reaction is

\ Z7o/

accessory.

§ 8. AMMONIUM BICHROMATE: C&£)6 (NHJaO.

1. We shall take this salt as a type of the ammoniacal salts
formed by the metallic oxacides.

Its equivalent is 126*4.

2. Composition —

N ............... =1 111

Cr ............... = 415

H ............... = 32

0 ............... = 442

1000

3. The heat of formation from the elements cannot be
calculated, the heat of oxidation of chromium being unknown.
But the decomposition of the salt not producing any oxide
lower than the chromium sesquioxide, it is sufficient to
calculate its formation from this oxide and the pre-existing
water contained in salts of ammonia.

The author has thus found : l

Cr203 (precip.)-h 04 + H8 + N2 = Cr2O3 + 2NH3 + H20 (solid)

+ 79-0 Cal.

1 " Comptes rendus des stances de 1'Acade'mie des Sciences," torn. xcvi.
pp. 399 and 536.

416 DEFINITE NON-CARBURETTED EXPLOSIVE COMPOUNDS.

Cr203 (precip.) + H20 (liquid) + O3 + H6 + N2 = Cr206,

(NH4)20 (solid) liberates + 44'5.

Cr203 (precip.) + 03 + 2H3N (dissolved) + H20 (liquid) =
Cr206, (NH4)2O = solid + 23'5 ; dissolved + 17'3.

Some remarks are here necessary.

The above figures are relative to a special state of chromium
oxide, namely when precipitated cold from dilute chrome alum,
by dilute potash, used in strictly equivalent quantity. But
they vary according to the manifold states of this oxide ; l the
variation by the precipitated oxide may amount to as much as
4- 6*9 Gal. according to the author's observations. With the
anhydrous oxide, and especially with the state produced by
ignition,2 the difference would be even greater, a circumstance
which explains the greater resistance to the acids of calcined
chromium oxide. In the case of the formation of the chromates
the heat liberated must be diminished by the heat of transfor-
mation of the ordinary chromium oxide into calcined oxide, say
for example — q. This quantity is, on the contrary, added to
the heat liberated by the explosive decompositions in which
the chromates intervene.

4. Ammonium bichromate, briskly heated, becomes incan-
descent and is tumultuously decomposed with formation of water
and chromium oxide in virtue of a real internal combustion —

Cr206(NH4)20 = Cr203 + 4H20.+ N2.
This reaction liberates —

The water liquid, + 59 Cal. at constant pressure, +60*4 Cal.

at constant volume.

The water gaseous, 39'0 Cal. + q at constant pressure, +
40*4 Cal. at constant volume.

The direct reaction of chromic acid on ammonia gas, in the
absence of water, would liberate nearly the double

C206 (solid) + 2NH3 (gas) = Cr203 + N2 + 3H20 (gas) +
73 '3 Cal. + 2 at constant pressure.

5. The explosive decomposition of ammonium bichromate
produces the following gaseous volumes : —

The water liquid, 11 -2 litres (l + -^~) »

\ 2il6/

The water gaseous, 557 litres ( 1 -f- — \

\ ZiioJ

1 " Comptes rendus des stances de FAcad^mie des Sciences," torn. xcvi. p. 87.
1 Beraeims.

AMMONIUM BICHROMATE. 417

Or for 1 kgra. —

88-6 litres (l + — \ the water liquid;

44*3 litres (1 4- 970)' ^e wa^er gaseous.

6. Theoretical temperature, at constant volume (the chromium

40400
oxide solid), -f = 1300°, or, more accurately, 1300°

ol'l

JL
311:

7. Permanent pressure at 0°, taking into account the volume
of the liquid water and of the chromium oxide (density -f 5 "2) —

88-6 atrn. ^ 91*6 kgm.
n - 0401 ( r n - 0401*

For n = 1, we have 153 kgm. per square centimetre, a much
lower pressure than that of the foregoing bodies.

At a high temperature, the vaporisation of the water tends to
increase fivefold the pressure which would be attributable to
nitrogen alone.

Hence, at the theoretical temperature of the decomposition,
we should have the pressure1 —

273 / 2570 atm. 2656 kgm.
or

n - 0116 " n - 0116 ' n - 0116

For n = 1, this figure is increased to about 2990 kgm.
All these values are much lower than those relative to the
foregoing bodies ; that is, ammonium nitrate and perchlorate.

1 Neglecting the quantity q>

2 E

( 418 )

CHAPTEK V.

NITRIC ETHERS PROPERLY SO CALLED.

§ i. ,;••••' ~—

WE shall describe the following ethers regarded as types of
the nitric derivatives of monatomic and polyatomic alcohols: —

Nitric ether of ordinary alcohol, of which the author has
measured the heat of formation.

Nitro-methylic ether, employed lately in dyeing.

Dinitric ether of glycol, remarkable because its decomposition
corresponds to a total combustion without an excess of any
element.

Trinitric ether of glycerin or nitroglycerin.

Lastly, hexanitric ether of mannite, or nitromannite.

It would be advisable to add to these, for the sake of com-
pleteness, the explosive mixtures formed by the association of
an organic compound with fuming nitric acid, or with nitric
peroxide, but these mixtures are elsewhere studied.

§2. NITRO-ETHYLIC ETHER: C2H4(HN03).

1. Equivalent, 91.

2. Composition —

C ......... ... rr 264

H ............... = 55

N ............... = 154

O ............... = 527

1000

3. The body is liquid, boiling at 86°.

4. Density : 1132 at 0°.

5. This ether can be inflamed when a small quantity of the
liquid is operated upon. In this way nitrous vapour is formed
in abundance, but if the vapour of the ether be superheated
beforehand, it explodes violently.

NITRIC ETHER. 419

6. The heat of formation from the elements has been found
(p. 279)-

C2 (diamond) + H5 + N + 03 = C2H4(N03H) (liquid) +49'3 Cal.

In the gaseous state it should be nearly + 42 Cal.

The heat of total combustion of the liquid body by an excess
of oxygen, + 311'2 Cal.

7. The following decomposition,

C2H4(N03H) = 2CO + H20 + 3H + N,

would liberate, the ether and water being liquid, + 71'3 Cal. at
constant pressure, -f 73 '5 Cal. at constant volume. All the
bodies being supposed gaseous, the heat liberated must have
nearly the same value.

Lastly, the ether being liquid and the water gaseous, we
should have, -f 61'3 Cal. at constant pressure, -f 64'6 Cal. at
constant volume.

For 1 kgin. we should have at constant pressure, the ether
and the water being liquid, 783'5 Cal. ; at constant volume,
791-6 Cal.

We shall not here examine the other possible modes of
decomposition.

8. The volume of the permanent gases, at the temperature t,
will be, for 1 equiv. = 91 grms. —

89 -3 litres (l + 7^)> the water being gaseous, 111 '6 litres

\ 2ii o/

(l + 7^3) ; °r> for 1 kgm, 981-3 litres (l + j~^ for the per-
manent gases, 1226 litres (l + — f— J if gaseous water be added.

9. The theoretical temperature, at constant volume —

64000

- = 2424°.
26-4

10. The permanent pressure, at 0° (the liquid water occupying
198 c.c.)—

981 atm. 1016 kgm.

n = 0198 °r n - 0198*

But this formula is only applicable to low densities of charge,
owing to the liquefaction of carbonic acid produced at high
densities.

11. The pressure developed at the theoretical temperature and
calculated according to the laws of gases —

12137 atm. nr 12541 kgm.
n n

2 E 2

• W

420 NITRIC ETHERS PROPERLY SO CALLED.

§ 3. NlTRO-METHYLIC ETHER :

1. Equivalent, 77.

2. Composition —

C = 156

H = 39

N = 182

0 = 623

1000

3. This body is liquid, boiling at 66°.

4. Density at 20° : 1182.

5. This ether can be inflamed in small quantities at the
ordinary temperature, but its vapour, superheated to about 150°,
explodes violently. It may even explode cold, on contact with
a flame,1 and communicate the explosion to liquid ether.

6. Heat of formation from the elements (p. 284) —

C (diamond) + H3 + N + 03 = CH2(N03H) (liquid) + 39'6 Cal.

7. The ether being gaseous, this figure must be nearly + 32.
The heat of total combustion of the liquid body, + 157'9.

8. Admitting the following decomposition —

2CH2(N03H) = C02 + CO + 3H20 + N2,

the heat liberated at constant pressure would be, the ether being
liquid, the water gaseous, +113 Cal. at constant pressure,
+ 114 Cal. at constant volume; the ether and water being
both liquid, + 123*8 Cal., or for 1 kgm. 1605.

All the other bodies being gaseous the heat liberated remains
nearly the same.

9. The volume of the permanent gases for 1 equiv., 33*5 litres
/. t \ / t \

\ * 273/' the Water Saseous 66'9 lltres V1 + 273J;or'for

1 kgm., for the permanent gases 435 litres (I + 070)' wnen

the water is gaseous 869 litres (l + ^=^

10. Theoretical temperature at constant volume —

4S5""*-

11. Permanent pressure, at 0° (the liquid water occupying
351 c.c.) —

435 atm. 448 kgm.
or

n - 0-351 n - 0:351
This value is applicable only to low densities of charge, that
1 Explosion at Saint-Denis, November 19, 1874.

DINITROGLYCOLIC ETHER. 421

is to say, within the limits in which carbonic acid retains the
gaseous state.

12. Pressure ab the theoretical temperature, calculated
according to the laws of gases —

669 atm. (l +

15052 atm.

n n

or 15,534 kgm. per square centimetre.

The permanent pressures are much lower than for nitro-
e thy lie ether, but the heat liberated is more than double, which
gives an advantage to the theoretical pressure.

§ 4. DINITRO-GI/TCOLIC ETHER : C2H2(N03H)2.

1. Equivalent, 152.

2. Composition^

C ......... ... = 158

H ............... = 26

N ... .. .......... = 184

0 ............. ,. = 632

1000

This composition is very nearly the same as that of nitro-
methylic ether.

3. The body is liquid.

4. The heat of formation from the elements calculated from
the formula on p. 285 —

C2 (diamond) + H4 + N2 + 06 = C2H2(HN03)2 (liquid) -f 66*9 CaL

5. The heat of total combustion and the heat of explosive
decomposition coincide —

C2H2(N03H)2 = 2C02 4- 2H20 + Na + 2591 Gal.

at constant pressure, the water being liquid; -f- 2607 Cal. at
constant volume.

For 1 kgm. we shall have 1705 Cal. at constant pressure,
1715 Cal. at constant volume.

6. Volume of the permanent gases for 1 equiv., 86*9 litres

\ ~*~ 273/' ^e water being gaseous 111-6 litres (l + 7^); or>

for 1 kgm., 440 litres ( 1 -f- — - ) for the permanent gases,
\ 2if of

734 litres (l + oijo) when the water is gaseous.

7. Theoretical temperature —

241-900 = 7982°
33-6

422 NITKIC ETHERS.

8. Permanent pressure, at 0° (the liquid water occupies
237 c.c.)—

440 atm. 455 kgm.

n - 0-237 ' C n - 0*237'

a value only applicable to low densities of charge.

9. Pressure at the theoretical temperature calculated accord-
ing to the laws of gases —

734

273 / 22170 atm.

n . n

or 22,910 kgm. per square centimetre.

§ 5. NITROGLYCERIN: C3H2(N03H)3.

1. Nitroglycerin is considered the most powerful of explosive
substances. In spite of terrible accidents, its extraordinary
properties have been taken advantage of for industrial purposes.
The manufacture of nitroglycerin in France commenced on a
large scale during the siege of Paris, at the instance and under
the direction of the Scientific Committee of Defence. Since
then it has assumed an ever-increasing importance, dynamite
tending to replace blasting powder in the greater number of its
uses.

It is not our intention to make here a complete study of
nitroglycerin and dynamites, nor of their industrial or military
applications ; but it enters into the scope of the present work
to present the figures which express the heat and pressure
developed by the explosive decomposition of nitroglycerin.

We shall, therefore, devote a paragraph to pure nitroglycerin,
reserving the study of dynamites for the following chapter.

2. Formula: C3H2(N03H)3.
Equivalent, 227.

3. Composition —

C = 159

H = 22

N = 185

0 = 634

1000

4. Nitroglycerin is liquid, but solidifies at -f- 12°. These
circumstances play an important part in the properties of
dynamite.

The density of liquid nitroglycerin is T60.

This body is very soluble in alcohol or ether, but only very
slightly soluble in water. Nevertheless, in presence of a
sufficient quantity of water, it is entirely dissolved, which does

HEAT OF FORMATION OF NITROGLYCERIN. 423

not permit of allowing nitroglycerin, either free or associated
with a pulverulent substance, to remain long in a current of
water.

It is poisonous.1

5. Nitroglycerin is very sensitive to shock, and explodes
easily by the shock of iron on iron or silicious stone. The fall
of a flask or of a stone jar has occasionally sufficed to cause
explosion. The shock of copper on copper, and especially of
wood on wood, is considered less dangerous ; however, there are
instances of explosion provoked by a shock of this kind.

6. Pure nitroglycerin keeps for an indefinite time. The
author has kept a bottle of it in his collections for ten years
without it showing any signs of alteration. But a little
moisture, or a trace of free acid, is sufficient to excite a
decomposition, which, when once commenced, is sometimes
accelerated up to the point of inflammation, and even of ex-
plosion of the substance.

The action of s)lar light also causes the decomposition of
nitroglycerin, as well as that of the nitric compounds in general.

Electric sparks inflame it, though with difficulty. They may
even cause it to explode under certain conditions ; for instance,
under the influence of a series of strong sparks, nitroglycerin
changes, turns brown, then explodes.

Submitted to the action of heat, it is volatilised to an ap-
preciable extent, especially towards 100°; it may even be
completely distilled, if this temperature be long maintained.
But if the temperature be suddenly raised to about 200°, nitro-
glycerin ignites; and a little above it, explodes with terrible
violence.

Its inflammation, caused by contact with an ignited body,
gives rise to nitrous vapour and a complex reaction, with the
production of a yellow flame, without explosion, properly so
called, at least as long as small quantities of matter are operated
upon. But if the mass be too great it ends by exploding.

The explosion of nitroglycerin corresponds to a very simple
decomposition —

2C3H2(N03H)3 = 6C02 + 5H20 + 6£T + 0.
It will be seen that nitroglycerin possesses the exceptional ^
property of containing more oxygen than is necessary for com-
pletely burning its elements.

7. Heat of formation from the elements (p. 282) —

C3 (diamond) + H3 + N3 + O9 = C3H2(N03H)3 liberates
+ 98 Cal.

8. The heat of total combustion and the heat of decomposi-

1 For the preparation with the aid of two binary mixtures made beforehand,
see Boutmy et Faucher, "Comptes rendus des stances de 1'Acade'rme des
Sciences," torn. Ixxxiii. p. 786.

424 NITRIC ETHERS PROPERLY SO CALLED.

tion are identical, according to what has just been said. The
reaction will, therefore, be represented by the following
formula : —

= 6CO2 + 5H30 + 6N + O.
The heat developed will be—

The water being liquid, at constant pressure + 356*5 Cal., at

constant volume + 358*5 Cal.
The water being gaseous, at constant pressure + 331*1 Cal,, at

constant volume + 335'6 Cal.

For 1 kgm., at constant pressure, the water being liquid,
1570 Cal. ; at constant volume, 1579 Cal.

Sarrau and Vieille found 1600, a figure the difference between
which and the above does not exceed that which might be due
to experimental errors.

In the miss-fires, treated of on page 283, the heat liberated is
necessarily less, the combustion being incomplete.

9. Volume of the permanent gases for 1 equiv. —

t
106 litres / 1 4- -rzr Y the water being liquid ;

161*8 litres fl + "^^)* the water being gaseous.

Or, for 1 kgm., 467 litres ( 1 -f 5=5), the water being liquid.
\ 273/

Sarrau and Vieille found 465 litres, at 0°, by experiment.

We should have 713 litres (l -f y^>\ the water being gaseous.
For a litre of liquid nitroglycerin we should have, lastly —
747 litres (I -f ^ J the water being liquid ;

1141 litres (l + A tne water being gaseous.

10. Theoretical temperature, at constant volume —

335600

-IT " 6980°'

11. Permanent pressure (the liquid water occupies 198 c.c.) —

467 atm. 482 kgm.
- or - — —
n - 0-198 n- 0198*

This figure is only applicable to low densities of charge and
with the usual exception as to the liquefaction of carbonic
acid.

PRESSURE OF EXPLODED NITROGLYCERIN. 425

12. Pressure at the theoretical temperature —

713 V1 + 27s) 18966 atm.

— — = - — , or 19,t)80 kgm. per square

n n

centimetre.

13. Let us compare this result with the pressures observed
by Sarrau and Vieille by means of the crusher and with dyna-
mite at 75 per cent.

They found under the density of charge —

- = 0-2 1420 kgm.

n 0-3 2890 „

0-4 4265 (3984 and 4546) kgm.

0-5 6724 (6902 and 6546) „

0-6 9004 kgm.

The volume occupied by the silica, after undergoing the
temperature of the explosion, may be estimated at Ol c.c. for
1 grm. of dynamite. Consequently, the volume occupied by
the gas yielded by 1 grm. of pure nitroglycerin would be equal

4.
to - (n — 0*1), neglecting the expansion of silica by heat.

o

We find in this way, referring the corrected densities of
charge to nitroglycerin, and calculating the pressures from the
results of the " crushers " —

n' = 6-5 c.c. ?r= 9230 kgm.
n

4-3 c.c. .. „. 12430 kgm.

3'2 „ 13640 „

2-5 „ 16800 „

2-1 „ 18900 „

F

It will be noticed that the values of -— are not constant.

n

But here intervenes the new theory of crushing manometers by
Sarrau and Vieille (p. 23), which accounts for these variations
by the duration of decomposition of dynamite and tends to
reduce by one-half the figure obtained with very high densities
of charge.

According to their new trials, made with very heavy piston,

for the density ~ = 0'3, we obtain a pressure of 2413 kgm.,

7l>

which corresponds to n' = 4*3 c.c.

— , = 10,376 kgm.

71

If we wish to compare with strict accuracy these figures with
the theoretical pressures, the heat yielded to the silica must be
taken into account in calculating the latter. Let the specific

426 NITRIC ETHERS PROPERLY SO CALLED.

heat of silica be supposed constant and equal to 0*19, which
makes for 73*7 grms. of silica 14'4, the theoretical temperature
becomes —

335600 _

62-4

The corresponding pressure will be —
/ 8378\
V + ~273~) _ 14759 atm. 15281 kgm.

n n n

a value higher by a third than the actual figure found for high
densities.

14 To sum up — weight for weight nitroglycerin produces
three and a half times as much permanent gases reduced to 0°
as nitrate powder, and twice as much as chlorate powder.

At equal volumes it produces nearly six times as much
permanent gases as ordinary powder. As, moreover, it produces
weight for weight more than double the heat, the difference
between the effects of the two substances taken in equal
weights is easy to foresee.

At equal volumes this difference is still greater. Thus
one litre of nitroglycerin weighs 1'60 kgm., whilst one litre of
ordinary powder weighs about 0*906 kgm. At the same volume
as powder, nitroglycerin will develop a pressure ten or twelve
times greater ; which may be actually realised in a completely
filled capacity, as in the case of a blast-hole, or when operating
under water. Under these conditions the maximum work
developed by one litre of nitroglycerin may amount to a value
treble that of the maximum work of ordinary powder at the
same volume. These colossal figures, no doubt, are never
attained in practice, especially owing to phenomena of dis-
sociation, but the fact that they are approached is sufficient to
explain why the work, and especially the pressures developed
by nitroglycerin, exceed the effects produced by all the other
explosive substances industrially employed. The relations
which these figures show between nitroglycerin and ordinary
powder, for example, agree pretty closely with the empirical
results observed in the working of mines.1

15. The rupture into fragments and the explosion of wrought
iron,2 effects which cannot be produced by ordinary powder,

1 See the experiments cited in the small treatise " La Dynamite," by Trauzl,
extracted by P. Barbe, pp. 91 and 92 (1870). The usual effect of nitro-
glycerin in quarries has been found to be five or six times greater than that,
of blasting powder, weight for weight. For an equal volume in blast-holes,
there is obtained with dynamite about eight times the effect produced by
powder ; that is to say, eleven times the same effect for a given weight of pure
nitroglycerin employed under this form. This refers to effects of dislocation,
which depend especially upon the initial pressures.

2 Same work (pp. 98 and 99).

EFFECTS OF EXPLODED NITROGLYCERIN. 427

are fresh proofs of the enormous initial pressures developed by
nitroglycerin. The question of the rapidity of decomposition,
moreover, intervenes here (p. 35).

Although nitroglycerin is shattering, it nevertheless fractures
rocks without crushing them into small fragments. The facts
observed during the study of the pressures exerted by the
crushers, at various densities of charge, would lead us to foresee
this property. It may also be accounted for by the phenomena
of dissociation. The elements of water and carbonic acid will
be partly separated in the first instance, which diminishes the
initial pressures ; but the formations of water and carbonic acid
being completed during expansion, successively reproduce fresh
quantities of heat, which regulate the fall of the pressures.
Nitroglycerin will therefore act during expansion in a similar
manner to ordinary powder. However, the dissociation will
be less with nitroglycerin, because the compounds formed are
simpler and the initial pressures higher.

In short, nitroglycerin combines the apparently contradictory
properties of various explosive substances : it is shattering, like
nitrogen chloride ; dislocates and fractures rocks without crush-
ing them, like ordinary powder, though with more intensity ;
lastly, it produces excessively great effects of projection. All
these properties, recognised by observers, can be foreseen and
explained by theory.

16. It could further be shown that inflammation induced at
a point of the mass is less dangerous with nitroglycerin than
with chlorate and even nitrate powder, seeing that the com-
bustion of the same weight of matter raises the temperature of
the neighbouring parts to a less extent, either owing to the
cooling produced by contact with the surrounding liquid parts,
or, especially, owing to the specific heat of nitroglycerin, which
appears to be much greater than that of potassium chlorate and
nitrate powders.

With regard to the theory of the effects of shock on nitro-
glycerin, the reader is referred to p. 52.

17. Lastly, let us compare nitroglycerin with ordinary powder
from the point of view of the best use of a given weight of
potassium nitrate. According to the equivalent, 303 parts of
nitre produce either 404 parts of ordinary powder, or 227 parts
of nitroglycerin, that is to say, a weight less by half. But as
a set-off the latter can develop, under the most favourable
circumstances, a pressure from eight to ten times greater than
the same volume of powder.

It follows from these numbers that a given weight of
potassium nitrate, if it could be changed atomically, and with-
out loss, into nitroglycerin, would develop in a blast-hole a
pressure treble that yielded by ordinary powder, made with the
same weight of nitrate.

428 NITRIC ETHERS PROPERLY SO CALLED.

§ 6. — NITROMANNITE : C6H2(N03H)6.

1. Equivalent, 452.

2. Composition —

C ... = 159

H = 18

N = 186

0 = 637

1000

The body crystallises in fine white needles. It must be
carefully purified by being re-crystallised in alcohol to free it
from -the products of incomplete nitrification.

3. Its apparent density is 1'60, but by melting it under
pressure as much as 1/80 may be observed at 20°.

4. It melts between 112° and 113°, and solidifies at 93°.
The temperature of the melting point given by various authors

falls to 70°, but this is for an impure product

5. Mtromannite commences to give off acid vapours from the
melting temperature. But this emission is very slow ; it is
accelerated with the rise in temperature. When suddenly
heated to about 190° it takes fire ; towards 225° it deflagrates,
towards 310° it explodes.

When the heating has been progressive, and accompanied by
a commencement of decomposition, which alters the composition
of the residuum, inflammation and explosion can no longer take
place.

6. Mtromannite purified by crystallisation in alcohol and
kept protected from sunlight can be kept for several years
without alteration,

But if care be not taken to re-crystallise it, it contains much
more changeable products, which cause its progressive decom-
position. These products also lower its melting point to
about 70°.

7. Nitromannite explodes by the shock of iron on iron more
readily than nitroglycerin, but with rather more difficulty than
mercury fulminate. It is intermediate in its shattering pro-
perties. It explodes by the shock of copper on iron or copper,
and even of porcelain on porcelain, provided the latter shock
be violent.

8. The heat of formation of nitromannite from the elements
has been found (p. 283), + 1561 Cal., according to a calculation
founded on the heats of formation of mannite, nitric acid, and
nitromannite, or -f 161*4 according to the heat of combustion
observed by Sarrau and Vieille.

9. The heat of total combustion coincides with the heat of
decomposition (see p. 283). It is equal to + 683'9 Cal. at con-

NITROMANN1TE. 429

stant pressure, the water being liquid, or 689 -6 at constant
volume. Or, for 1 kgm., at constant pressure, 2513 Cal. ; at
constant volume, 1529 Cal. —

C6H2(N03H)6 = 6C02 + 4H20 + 3N2 + 02.

Sarrau and Vieille found 1512 at constant volume, and
proved, further, that the decomposition really takes place
according to the above equation.

The heat of combustion is inferior to that of nitroglycerin
and of nitroglycol, an inferiority due to the formation of a
larger amount of free oxygen.

10. Volume of the permanent gases for 1 equiv. —

223 litres ( 1 + -^r J; the water being gaseous, ( 1 + — - J 312 litres.

The water being gaseous, we should have for 1 equiv. at
constant pressure + 603*9 Cal., at constant volume + 612 Cal.

Or, for 1 kgm., 494 litres (l + — \ for the permanent gases,

692

litres (l + TZJJ) the water being gaseous.

612000 Cf7ino

11. Theoretical temperature, = b/iu .

oL a

12. Permanent pressure at 0° (the liquid water occupies

159 c.c.)—

494 atm»_ 510 kgm.,

n - 0159 °r n - 0159

subject to the usual proviso as to the lowness of the densities
of charge and of the limit of liquefaction of carbonic acid.

13. Pressure at the theoretical temperature, calculated
according to the laws of gases —

/ 6710\
692 (14- — - — )

V 273 / 17220 atm, 17760 kgm.
-- = - or -
n n n

or 23,510 kgm. per square centimetre, a value very close to
those which belong to nitroglycol (22,910) and nitroglycerin
(19,580), as might be expected.

14. The pressures actually exerted in the explosion of nitro-
mannite have been measured by Sarrau and Vieille. These
authorities found —

At the density of charge 01, 2273 kgm,
At the density of 0'2, 4634 kgm.

22950

Or as mean, - — , a value very near the theoretical figure.
n

430 NITRIC ETHERS PROPERLY SO CALLED.

But the new theory of the authors would tend to reduce it to
the half (p. 23).

But this pressure is so quickly developed that the piston of
the crusher is often broken, which shows the shattering charac-
ter of nitromannite. The same property intervenes in the tests
founded on the capacity of chambers hollowed in leaden blocks
by various explosives (p. 374). Now, the capacity hollowed by
a given weight of nitromannite is greater by a fourth (43 c.c.
for 1 grm.) than that hollowed by nitroglycerin (35 c.c. for
1 grm.) Nitromannite, moreover, manifests a much more
marked tendency to tear the leaden blocks in diagonal direc-
tions. These facts contrast with the theoretical calculation of
the pressures or of the maximum work, which give nearly the
same value for nitromannite and nitroglycerin. They show
that the empirical method of chambers hollowed by an explo-
sive does not really measure either the pressure or the work,
but certain more complicated effects.

( 431 )

CHAPTER VI.

DYNAMITES.1

§ 1. DYNAMITES IN GENEKAL.

1. IN 1866, iii consequence of terrible accidents caused by
explosions,2 the use of this substance was going to be forbidden
everywhere, when a Swede, Mr. Nobel, conceived the idea of
rendering it less sensitive to shocks by mixing it with an inert
substance, a well-known artifice for attenuating the effects of
the ordinary powder, but which leads to unexpected results in
the present case. Nobel added to it first a little methylic
alcohol ; then, this expedient being insufficient, he mixed it
with amorphous silica. He designated this mixture by the
name of dynamite.

He soon recognized, and this was a very important discovery,
that the explosion requires the use of special mercury fulminate
detonators, and that it acquires in this way an exceptional
violence ; it can then be produced even under water. By using
these detonators tamping may be dispensed with, when
absolutely necessary, in blasting with dynamite.

This name has since been extended to very diversified
mixtures, with nitroglycerin as base, and at the present day a
score of different dynamites are distinguished. Mixtures con-
taining liquid explosives other than nitroglycerin have even
been designated by the same word. Dynamites have the
common property of not exploding either by simple inflamma-
tion, slight shock, or moderate friction. But they explode, on
the contrary, by the use of strong caps, called detonators, gene-
rally composed of mercury fulminate.

Dynamites are divided into several classes.

2. In some, containing an inert base, the nitroglycerin is

1 See " La Nitroglycerine et les Dynamites," par Fritsch, 1872 (" Memorial
de Tofficier du Genie ") ; " Manuel de pyrotechnic a 1'usage de I'ArtilJerie de
Marine," torn. ii. ; " Traits' de la poudre," etc., revu par Desortiaux, p. 798.
1878.

2 Stockholm, Hamburg, Aspinwall, San Francisco, Quenast in Belgium.

432 DYNAMITES.

associated with silica, alumina, magnesium carbonate, calcined
alum, brick-dust, tripoli, sand, boghead ashes, etc., all these
being substances intervening only to a slight extent, or not at
all, by their chemical composition, but only by their physical
constitution and their relative proportion. They check the
propagation of the molecular shocks, the harmonious succession
of which gives rise to the explosive wave (p. 78). After
deflagration, they are more or less modified.

3. Others, containing an active base, may themselves be
separated into three groups.

4. Some dynamites (those with ammonium nitrate or potas-
sium chlorate base) are formed by the association of nitroglycerin
with an explosive substance, which explodes simultaneously
without the elements of the one intervening chemically in the
decomposition of the other. They might be termed dynamites
with simultaneous active base.

5. Other dynamites with simple combustible base are manu-
factured by taking advantage of the fact that the explosion of
nitroglycerin sets free a certain quantity of oxygen (3*5 per
cent.) in excess of that which is necessary to convert the whole
of the carbon into carbonic acid and the whole of the hydrogen
into water.

There is then added to the nitroglycerin, whether pure or
already mixed with an inert substance, a certain quantity of a
combustible body (coal, wood sawdust, starch, straw, bran,
sulphur, spermaceti, etc.) for the purpose of utilising this excess
of oxygen.

6. But the quantity of oxygen is generally too small for the
corresponding proportion of combustible matter, such as 1 per
cent, of coal or spermaceti, or 2 per cent, of wood sawdust, or
3 '5 per cent, of powdered sulphur, to be sufficient to absorb the
whole of the corresponding nitroglycerin. Hence in practice
the complementary substance must be employed in great
excess, which constitutes the mixed base dynamites. We will
only mention black dynamite, a mixture of charcoal and sand,
capable of absorbing 45 per cent, of nitroglycerin. Such an
excess of combustible matter changes the character of the
chemical reaction, which may cease to be a total combustion.

7. Dynamites with a combustible explosive base may also be
prepared by employing as combustible complement a compound
explosive in itself, but which does not contain enough oxygen
to undergo total combustion.

Such are gun-cotton, the several varieties of nitro-cellulose
and nitro-starch, picric acid, etc.

They belong to two principal groups.

8. — (1st) Dynamites with nitrate base; such as dynamite
with black powder as base (100 parts of black powder associated
with from 10 to 50 parts of nitroglycerin).

VARIETIES OF DYNAMITES. 433

Dynamite with Hasting powder as base.

Dynamite with saltpetre and charcoal as base.

Dynamite having as base barium nitrate and resin, or charcoal,
with or without the addition of sulphur.

Dynamites having as base sodium nitrate, charcoal, and sulphur,
etc.

Dynamites formed by nitroglycerin, saltpetre and wood saw-
dust, or starch, or cellulose.

9. — (2nd) Dynamites having as base pyroxyl, such as Trauzl
dynamite, formed of nitroglycerin and gun-cotton in a paste.

Abel's glyoxylin, formed of the same substances, with the
addition of saltpetre.

Dynamites having as base a nitrified ligneous substance (paper
pulp, or wood pulp), and analogous ones.

Blasting gelatin, formed by the association of 93 to 95 parts
of nitroglycerin, and 5 to 7 parts of collodion cotton.

10. We should here note that the relative proportions of
nitroglycerin and of the combustible or explosive base, which
are the most useful in practice, are not always those which
correspond to a total combustion ; either because an incomplete
combustion gives rise to a greater volume of gas, or because the
rapidity of decomposition and the law of expansion vary accord-
ing to the relative proportions and the conditions of application.

11. It can further be seen that the inert, the simple com-
bustible, and the explosive combustible substances may be
associated in various proportions, and this constitutes fresh
dynamites with mixed base, extremely varied.

The requirements of practice and the imagination of inventors
are daily multiplying these varieties, designated by the most
diversified and sometimes the most pompous names : Hercules
powder, giant powder) petralites, etc. ; but they all belong to the
five foregoing types.

12. Among these practical requirements we shall point out
some of those which play the most important part, inde-
pendently of the question of the first cost. The most important
point lies in the strength of the mixture. Indeed, the additions
have generally the effect of lowering the strength, by reducing
the amount of nitroglycerin. It is sought in this way to retard
decomposition, so as to change the shattering agent into a
propulsive agent. But if the retardation be too great, we enter
into the category of the slow powders (p. 2), and lose the
advantages due to the presence of nitroglycerin. There is,
therefore, a practical limit to these additions, if it be desired to
obtain the greatest useful effect. The use of mica, on the
contrary, increases the rapidity of explosion.

The homogeneousness and stability of the mixture are of the
highest importance ; it is, in fact, requisite that the nitroglycerin
should be entirely absorbed by the substance which serves as

2F

434 DYNAMITES.

base, and that this mixture should remain uniform without
chemical change and without exudations due to shocks in
transport or to variations in temperature, otherwise we should
be brought back to the drawbacks and dangers of pure nitro-
glycerin. The absorbent substance must, therefore, have a
special structure opposing itself to the spontaneous separation
of the nitroglycerin. Dynamites having as base ordinary sand,
brick-dust, and powdered coke have thus been set aside owing
to their instability.

The presence of an excess of nitroglycerin beyond the satura-
tion point may even diminish the strength of a dynamite instead
of increasing it, owing to the difference of the mode of propaga-
tion of the explosive wave in the liquid and in the porous
mixture. It is in this way that the crushing effects upon a
leaden block are more marked with 75 per cent, dynamite than
with a richer dynamite, and even with pure nitroglycerin.

13. This tendency to separation is increased by a special
property of nitroglycerin, which plays an important part in the
application of all dynamites formed by this agent, viz. the
solidification of nitroglycerin at about 12°. In fact, in becoming
solidified, the explosive more or less completely separates itself
from its absorbent, and thenceforth constitutes a new system,
endowed with special properties.

On the one hand, solid nitroglycerin seems less sensitive to
shocks, and especially to their transmission step by step. It
requires more powerful fuses to explode it, which generally
renders it necessary to reheat the cartridges in order to liquefy
it, and to reconstitute the original dynamite, an operation which
has occasioned numberless accidents in mines.

On the other hand, nitroglycerin thus liquefied, after having
been partly separated from its absorbent by crystallisation, may
not mix with it again in so intimate a manner as before,
especially if the absorbent be not of good quality, and if it be
submitted to pressure.

14. The degree of sensitiveness to shock of dynamites is
a circumstance of fundamental importance, particularly for
military applications. Thus it is necessary to put into the
hands of soldiers a substance which does not explode during
transport, nor under the shock of a ball. Ordinary dynamite
with silica base does not satisfy this condition, which has often
caused compressed gun-cotton to be preferred, though the latter
is not entirely free from danger in this respect.

15. It has been attempted to gain the end in view by adding
certain foreign substances to dynamites — camphor, for instance,
to the amount of a few hundredth parts ; but this mixture is
only moderately efficacious.

The condition sought after is especially realised by blasting
gelatin, formed of nitroglycerin and collodion cotton. But here

DYNAMITE PROPER. 435

we meet another stumbling-block ; the substance requires special
capsules and too great a quantity of fulminate to explode it.
This must be compensated for by employing a small inter-
mediate cartridge of compressed gun-cotton, primed itself with
fulminate, which complicates the question. It appears that
even in this way it is sometimes difficult to effect the explosion
of blasting gelatin.

16. This technical discussion will not be further entered
upon here except to observe that the absence of explosion by
simple ignition, and the necessity for special detonators, are
among the number of essential characteristics which distinguish
dynamite from service powder and all analogous kinds.

Hence arise fresh complications in the use of these substances.
Thus, owing to this circumstance and the risk of explosions by
influence, the detonators should be carefully kept apart from
the stores of dynamite, in magazines, and during transport.
Many accidents are due to the neglect of this precaution.

17. These general notions being set forth it would require
a whole volume to enter into the study and the discussion of
the properties of all the dynamites proposed, or even only of
those actually employed. This is why we shall confine our-
selves to treating with more detail three interesting varieties of
dynamite in order to show how our theories are applied to their
study. They are —

1st. Dynamite proper with silica as base.

2nd. Dynamite with ammonium nitrate as base.

3rd. Blasting gelatin with collodion cotton as base.

§ 2. DYNAMITE PEOPER*

1. We have said above how Nobel had invented this substance
to obviate the terrible effects which result from the propagation
of shocks in liquid nitroglycerin. Now, dynamite proper, being
less sensitive to shocks than nitroglycerin, can be transported
and handled almost without danger, provided certain rules be
observed.

2. For many years dynamite has been employed in mines
and in tunnel boring to rupture and reduce very hard or
fissured rocks, as well as in harbour and other works. It has
been applied to break up blocks of stone, masses of cast or
wrought iron, blocks of pyrites, beds of flint, accumulated ice,
to break up and lighten soils intended for vine growing, etc.,
and its applications are daily being developed.

Dynamite also plays a most important part in warfare
(torpedoes, mines, the destruction of palisades, the levelling of
trees, buildings and bridges, the destruction of rails and rail-
ways, the bursting of cannons, etc.).

2F2

436 DYNAMITES.

3. Dynamite proper results, as we have said, from the
association of nitroglycerin with amorphous silica. At the
outset Nobel employed for this purpose Kieselguhr, that is,
the silicious earth of Oberlohe (Hanover) ; but there have since
been found in various places natural silicas, such as randanite
(Auvergne), which answer the same purpose.

The special structure and the organic origin of these varieties
of silica, formed for the greater part of shells and infusoria
(Diatoms), were at first regarded as indispensable for the
fabrication of dynamite. But amorphous silica, prepared by
a chemical process — for instance, that resulting from the action
of water on silicon fluoride — is no less suited for this preparation ;
it even stores up at least as large quantities of nitroglycerin
(more than nine times its weight) as natural silica.

4. Dynamites are also distinguished according to their origin —
as Nobel and Iboz dynamites, Vonges dynamites, etc. ; and
according to their strength — No, 1 dynamite, with 75 per cent.
of nitroglycerin ; No. 2 dynamite, with 50 per cent. ; No. 3
dynamite, with 30 per cent.

5. Preparation. The silica is first dried in ovens, without
however heating it to too high a temperature, and sifted to
eliminate the large grains 5 then the nitroglycerin is incorporated
with it. A few hundredth parts of lime or magnesia carbonates
or of sodium bicarbonate are added in order to prevent the
mixture from becoming acid, a transformation which is the
prelude to its spontaneous decomposition.

6. Properties. The substance thus obtained is grey, brown,
or reddish (according to the foreign ingredients), rather greasy
to the touch, forming a pasty mass. It should not give rise to
considerable exudations of nitroglycerin; The absolute density
of dynamite is a little more than 1-60. The relative density,
obtained by the gravimetric method, is 1'50 for dynamite at 75
per cent.

In preparing dynamite an apparent contraction of the
materials is observed; that is to say, that the nitroglycerin
occupies a volume less than the air interposed in the silica.
Nitroglycerin freezing at 12°, dynamite is transformed at about
this temperature, or slightly below, into a hard mass, expanding
at the same time. The properties of dynamite are then
extremely modified, and it requires much stronger detonators
to explode it ; say 1'5 grm. of fulminate, instead of 0'5. How-
ever, the explosive force remains the same. This circumstance
forms one of the most serious drawbacks to the keeping and
use of dynamite. Indeed the necessity for thawing it frequently
occasions serious accidents, especially if this operation be
effected at an open fire and without precautions. It was in
this way that at Parma, in 1878, a lieutenant of cavalry having
placed on a brazier a can containing one kgm. of dynamite, an

SPONTANEOUS DECOMPOSITION. 437

explosion immediately occurred, eighty persons being killed
or wounded.

Moreover, thawing may occasion exudations of pure nitro-
glycerin, the latter expanding by the fact of solidification. It
is thus exposed and may explode by subsequent shock of friction.
It is sometimes enough to bring about an accident, to cut a
frozen cartridge with an iron tool. Ramming is even dangerous
with it. Moreover, frozen dynamite has not lost the property
of exploding by influence.

7. Action of heat. Dynamite, submitted to the action of a
gentle heat, undergoes no change, even under the prolonged
influence (an hour) of a temperature of 100°. Heated rapidly,
it takes fire near 220°, like nitroglycerin. If ignited, it burns
slowly and without exploding ; but if it be enclosed in a
hermetically sealed vessel with resisting walls, it explodes under
the influence of heating. The same accident is sometimes pro-
duced in the inflammation of a large mass of dynamite, owing to
the progressive heating of the interior parts, which brings the
whole mass to the temperature of explosive decomposition.

Dynamite, moreover, becomes more sensitive to shock, as do
also explosive substances in general, according as it is raised
nearer to temperature of decomposition.

Direct solar light can cause a slow decomposition, as with all
the nitro and nitric compounds. Electric sparks, generally
speaking, ignite dynamite without exploding it, at least when
operating in the open air.

8. Spontaneous decomposition. Dynamite prepared with
neutral nitroglycerin appears to keep indefinitely if care be
taken to add to it a small quantity of calcium carbonate, or
alkaline bicarbonate, thoroughly mixed. Contact with iron and
moisture changes it in course of time. Dynamite which has
commenced to undergo change becomes acid and sometimes
explodes spontaneously, especially if contained in resisting
envelopes. Nevertheless, neutral and well-prepared dynamite
has been kept for ten years in a magazine without loss of its
explosive force.

9. Action of water. Water brought into contact with
dynamite gradually displaces the nitroglycerin from the silica.
This action is slow but inevitable. It tends to render all wet
dynamite dangerous. However, ordinary dynamite hardly
attracts the atmospheric moisture.

It has been observed that a dynamite made with wood saw-
dust can be moistened, then dried without marked alteration,
provided the action of the water has not been too prolonged.
Fifteen to twenty per cent, of water may be added to cellulose
dynamite, rendering it insensible to the shock of a ball without
depriving it of the property of exploding by a strong fuse. But
nitroglycerin is then separated under a slight pressure.

438 DYNAMITES.

10. Action of shock. Dynamite requires a much more violent
shock than nitroglycerin to explode it. It explodes by the
shock of iron on iron, or of iron on stone, but not by the shock
of wood on wood.

Dynamite is the more sensitive the more nitroglycerin it
contains.

When dynamite is struck with a hammer, the part directly
affected by the shock alone explodes, the surrounding portions
being simply dispersed.

Owing to this circumstance the effects may vary greatly, unless
the dynamite be contained in a resisting and completely filled
envelope, or placed at the bottom of a receptacle. It explodes
by the direct shock of a ball at a distance of 50 m., and even
more, a very important matter in military applications.

11. The detonation of dynamite in tubes entirely filled with
this substance propagates itself with a speed of about 5000 mm.
per second.

12. Its explosion, when complete, does not produce noxious
gases, like gunpowder ; but if it burn by simple inflammation
(miss-fires), it produces nitric oxide, carbonic oxide, and nitrous
vapour, which are deleterious (p. 283).

13. The heat liberated by the sudden decomposition of dyna-
mite is the same as its heat of total combustion, and pro-
portionate to the weight of nitroglycerin contained in the
dynamite.

It can therefore be easily calculated from the data on page
424.

14. The volume of gases liberated by any dynamite, and the
theoretical pressure which it can develop, are also calculated in
this way, taking into account the volume occupied by the
silica (see p. 425), and the heat absorbed in raising its tem-
perature.

The experiments of Sarrau and Vieille on this question have
been described above.

15. It will be shown in a general way that thermal theories
favour the employment of dynamite. In the fii st place, dyna-
mite is less shattering than nitroglycerin, because the heat
liberated is shared between the products of explosion and
the inert- substance. In consequence there is a less rise in
temperature, which diminishes the initial pressures propor-
tionately.

For instance, the silica and anhydrous alumina, which may
be mixed with nitroglycerin, have nearly the same specific heat
(019) as the gaseous products of explosion of the latter at
constant volume. Weight for weight, and in a completely filled
space, they will lower the temperature, and consequently the
initial pressure by half.

For an equal weight of nitroglycerin the shattering properties

DYNAMITE WITH AMMONIUM NITRATE BASE. 439

will therefore be diminished proportionately to the weight of
the inert matter in the mixture ; while the maximum work will
retain the same value, being always proportional to the weight
of nitroglycerin.

The same circumstances will render the propagation of simple
ignition of a small portion of the mass into the neighbouring
parts more difficult, since the latter explode only when raised
suddenly to a temperature approaching 200°. Hence the ex-
plosion produced by a detonator requires a greater initial
disturbance in order to take place.

If deflagration be produced by the shock of a hard body, or
of a fulminating fuse, the solid particles interspersed in the
liquid divide the energy of the shock between the inert and
the explosive substance, in a proportion depending on the
structure of the inert substance. The latter thus changes the
law of explosion ; it opposes itself to some extent to the propa-
gation of the explosive wave, except in the case of extremely
violent shocks, and introduces an extreme diversity into the
phenomena, as follows from the experiments of Nobel, and those
of Grirard, Millot, and Vogt, on nitroglycerin mixed with silica,
alumina, ethal, or sugar.

It is, moreover, evident that the useful effects of the inert
substance could only be completely produced when the mixture
is homogeneous, and without any separation of liquid and nitro-
glycerin, for the liquid which has exuded retains all its pro-
perties, hence the necessity of the special structure in a solid
substance.

§ 3. DYNAMITE WITH AMMONIUM NITRATE BASE.

1. This substance is very interesting on account of the great
energy which is derived both from nitroglycerin and ammonium
nitrate, whether associated or not with a complementary com- .
bustible substance.

It has been proposed on various occasions by inventors, with
certain variations due to the introduction of the complementary
bodies (charcoal, cellulose, etc.), the latter being for the double
purpose of utilising the excess of oxygen supplied both by nitro-
glycerin and ammonium nitrate, and for completing the absorbent
properties of the substance.

But this dynamite presents a certain drawback, because
ammonium nitrate is hygroscopic, especially in an atmosphere
saturated with moisture. Moreover, water immediately separates
nitroglycerin from it.

2. The relative proportions of nitroglycerin, ammonium nitrate,
and combustible substances may vary extremely, even when it
is subjected to the condition of a total combustion. We shall
consider only the mixtures in which charcoal constitutes the

440 DYNAMITES.

combustible substance, and for the sake of simplicity the char-
coal will be considered as pure carbon.

All systems which satisfy the condition of total combustion
reduce themselves to the following formula : —

4C3H2(N03H)3 + £0] + </(N03NH4 + C).
They produce

The corresponding weight is

(230z + 86y) grms.
The heat liberated (gaseous water) is equal to

325-7z + 79% ;
Or, for 1 kgm.,

The volume of the gases is

(7iz + 4Jy) 22-32 litres (1 + at)=

Jti

The permanent and the theoretical pressure at a density of
charge -— are immediately deduced from the above data.

3. For instance, let

jc«=l;y=16

be the corresponding weight : 230 + 1376 = 1606 grms.
The percentage of the immediate composition will be —

Nitroglycerin ............ 14-1

Ammonium nitrate ............ 79-6

Charcoal ............... 6-3

4. The heat liberated = 1593 Cal. (gaseous water), or 1938
Cal. (liquid water) ; or, for 1 kgm., 992 Cal. gaseous water, or
1207 Cal. liquid water.

^ 5. The reduced volume of the gases (gaseous water) = 1769
litres, or 642 litres (liquid water) ; or, for 1 kgm., 1101 litres
gaseous water, or 400 litres liquid water.

6. The permanent pressure (liquid water) = 4QO atm-> with

fli — U'O«7

the usual reservation as to the liquefaction of the carbonic acid
when n falls below a certain limit.

7. The theoretical pressure = -- 5*5t a vaiue higher

than the theoretical figure for ordinary 75 per cent dynamite
(p. 425).

DYNAMITE WITH NITROCELLULOSE BASE. 441

This is in conformity with the practical tests which point to
the approximate equal power of 60 per cent, dynamite and the
mixture formed of 75 parts of ammonium nitrate, 3 parts of
charcoal, 4 parts of paraffin, and 18 parts of nitroglycerin.

§ 4. DYNAMITE WITH NITROCELLULOSE BASE.

1. The association of nitroglycerin with gun-cotton was first
proposed in 1868 by Trauzl, in Austria ; but the product thus
obtained was dangerous and difficult to manufacture, and was
not adopted in practice. However, at the present day there is
a tendency to return to active base dynamites of a similar
formula (dualines). They are sometimes associated with
potassium nitrate (lithofracteur), etc. Mixtures containing
40 parts of nitroglycerin and 60 parts of gun-cotton or nitro-
lignite, with the addition of 2 per cent, of ammonium carbonate,
are those which are more especially manufactured. These
mixtures do not correspond to a perfect combustion, but they
will produce effects very closely approaching the mean of their
components. Dynamite with ligneous nitrocellulose base is
somewhat less sensitive to shock and freezing than that con-
taining gun-cotton. If potassium nitrate be superadded it
allows of the combustion being completed, but it increases the
sensitiveness.

2. Some years since Nobel conceived the idea of forming a
compound of quite a different order by dissolving collodion
cotton in nitro-glycerin in the proportion of 93 parts of the
latter and 7 parts of the former, and in this way obtained the
substance called blasting gelatin, explosive gelatin, or gum
dynamite, a clear, yellow, gelatinous, elastic, transparent com-
pound, more stable than ordinary dynamite, especially from a
physical point of view, for it gives rise to no exudation, even by
pressure. It is unchangeable by water (see further on). Lastly,
it is much more powerful than Kieselguhr dynamite and com-
parable in this respect to pure nitroglycerin.

By adding to blasting gelatin a small quantity of benzene, or,
better still, of camphor (from 1 to 4 per cent.), it is rendered
insensible to mechanical actions which cause the explosion of
ordinary dynamite, such as friction, the shock of a bullet at a
short range, etc. Its strength is appreciably diminished by this
mixture, but it is no longer developed except under the influence
of very strong charges of fulminate or of a special primer formed
of nitrohydrocellulose (4 parts), nitrocellulose and nitroglycerin
(6 parts), which itself may be ignited by a small charge of
fulminate.

The work of the initial shock necessary to explode blasting
gelatin has been calculated at six times that which would be
required for ordinary dynamite, coeteris paribus, a difference

442 DYNAMITES.

which is doubtless attributable to the cohesion of matter ; that
is to say, to the greater mass of particles in which the energy
of the shock transformed into heat causes the first explosion
which is the origin of the explosive wave (p. 54).

Owing to these circumstances blasting gelatin is far less
sensitive to explosions by influence. All these conditions are
very favourable to its use as an explosive for military purposes.

3. The properties of this substance will now be more par-
ticularly considered. Blasting gelatin does not absorb water ;
it merely turns white on the surface under this influence, owing
to the solution of the nitroglycerin contained in the superficial
stratum, but the action does not go any further. The collodion
couon, separated by the action of the water on the first stratum
of substance, being insoluble in this agent, envelops the whole
of the rest of the mass in a protecting film. Blasting gelatin
therefore remains unaltered, even after having been kept for
forty-eight hours under running water. The explosive force
has been found to be the same after this test.

Neither does freezing change its shattering force, but it causes
it partly to lose its insensibility to shock.

4. The density of blasting gelatin is 1*6, i.e. equal to that of
nitroglycerin, as might have been expected, from its com-
position and its homogeneous structure without pores. This
density is higher than the apparent density of dry gun-cotton
(TO) or damp gun-cotton (1*16), which constitutes a real and
important advantage.

5. Blasting gelatin burns in the open air without exploding,
at least when small quantities are operated upon and a previous
heating is avoided. It has been kept for eight days at 70°
without being decomposed.

After having been kept for two months between 40° and 45°
it lost the half of the camphor and a small quantity of nitro-
glycerin without further alteration.

Slowly heated it explodes towards 204°.

If it contains 10 per cent, of camphor, it no longer explodes,
but it fuses.

6. Let us now estimate the strength of blasting gelatin by
our ordinary calculations.

As an example, a blasting gelatin formed of 91*6 parts of
nitroglycerin and 8 '4 parts of collodion cotton, which are the
proportions corresponding to a total combustion.

The collodion cotton is here taken as corresponding to the
formula —

Such a dynamite is formed in the proportions —

51C3H2(N03H)3 +
Its equivalent weight is 12,360 grms.

PRESSURE OF EXPLODED BLASTING GELATIN. 443

The explosion produces

177CO2 + 143H20 + 81N2.

7. The heat liberated by its explosion is equal to 19381
Cal. (gaseous water) ]or 2241 Gal. (liquid water) ; or, for 1 kgm.,
1535 Cal. (gaseous water), or 1761 Cal. (liquid water).

8. Eeduced volume of the gases = 8950 litres (gaseous water)
or 5759 litres (liquid water) ; or, for 1 kgm., 709 litres (gaseous
water), or 456 litres (liquid water).

9. The permanent pressure (liquid water) = - :, with

n — 0*41
the usual reservations.

10. The theoretical pressure = 1, value nearly

Tit

identical with that of nitroglycerin (p. 425).

It might have been supposed that the pressure and the heat
developed would have been greater owing to the complete
utilisation of the oxygen, but there is a compensation on account
of the greater loss of energy which takes place at the outset in
the union of the elements, and afterwards in the combination of
nitric acid with the cellulose, which liberates 11*4 Cal. per
equivalent of fixed acid instead of 4'9 Cal. liberated in the case
of nitroglycerin (see p. 282).

Hence it will be seen that blasting gelatin considerably sur-
passes ordinary dynamite in the ratio of 19 : 14 according to
theory. The ratio of the actual effects of the two substances has
been estimated by Hess, by the aid of practical tests based
on the rupture of strong pieces of wood. It has been found to
approach the numbers 78 : 56, which notably are in accord.

( 444 )

CHAPTEE VII.

GUN-COTTON AND NITROCELLULOSES.

§ 1. HISTORICAL.

1. IN 1846 Schonbein proposed to replace service powder by a
new substance, the composition of which he kept a secret. This
was gun-cotton, the discovery of which is the starting-point of
the works since accomplished with the new explosive substances.
In 1832 Braconnot and Pelouze had already made known some
similar nitric compounds.

Numerous experiments carried out up till 1854 led to gun-
cotton being regarded as more powerful for equal weights than
black gunpowder, that it possessed shattering properties which
hardly admitted of its continued use in firearms. Soon, terrible
explosions and accidents in powder factories 1 gave evidence of
the existence of spontaneous decompositions, which put a stop
to its manufacture almost everywhere; nevertheless, experi-
ments were still carried out in Austria, under the direction of
Lenck, until the occurrence of a fresh explosion in a magazine
at Simmering in 1862. Another explosion occurred in 1865 at
Wiener-Neustadt.

2. In England, however, Abel succeeded in almost entirely
removing risks by a very careful process of manufacture, namely,
by reducing the cotton to pulp, which enabled it to be more
completely washed, and finally, by the compression of the cotton
(1865) by hydraulic presses.

Compressed gun-cotton thus came into use. Brown discovered
in 1868 that it could be detonated by means of mercury
fulminate.

The explosion which happened in 1871 in the Stowmarket
factory, and in which twenty-four persons perished, was at-
tributed, rightly or wrongly, to imperfect supervision, and the
manufacture of compressed gun-cotton is still carried out in
England. It has been carried out also in France for some time
at the " Moulin Blanc " factory.

1 Bouchet and Vincennes, 1847.

GUN-COTTON AND DYNAMITE. 445

3. Gun-cotton is practically only used for military purposes,
since its high price prevents it becoming a rival of dynamite,
which, besides, is more easily adapted to the requirements of
miners.

In Austria, Eussia, and France, even up till recently, dynamite
has been preferred to it as a war explosive, whereas in England
and Germany gun-cotton has the preference. The Marine
Artillery in France l also uses it, and the French army autho-
rities evince a tendency to go back to its use on account of its
safer preservation.

4. However, gun-cotton being, like dynamite, susceptible of
detonation from the shock of a ball at a short distance, en-
deavours have been made to reduce this sensitiveness. In order
to effect this it suffices to incorporate with it from ten to fifteen
per cent, of water or paraffin. Damp gun-cotton is much better
able to resist mechanical agents.

In this state it cannot be inflamed by contact with a body in
ignition, or by spontaneous decomposition. Gun-cotton, when
mixed with paraffin, is also less sensible to shock, but it is not
safe from the risk of inflammation.

On the other hand, the detonation of moistened or paraffined
gun-cotton is more difficult ; it requires the employment of a
very strong dose of fulminate, or a small hand-made cartridge
of dry gun-cotton primed with fulminate.

The presence of water, as also of paraffin, further lessens the
force of the explosion.

The application of water is subject to variations owing to
spontaneous evaporation, which is a serious difficulty.

In the German army paraffin is employed. The application
of this is simpler, and it is not subject to variations on account
of the weather. Nevertheless, sensibility to detonators does not
appear to be the same in paraffined gun-cotton which has been
recently or for some time prepared, probably on account of the
change in structure, which is the result of the slow crystallisa-
tion of the paraffin.

5. Gun-cotton does not, like nitroglycerin, contain a suffi-
cient quantity of oxygen for the combustion of its elements ;
hence the proposal to associate it with potassium, barium, or
ammonium nitrate, or with potassium chlorate ; bodies which
would supply it with oxygen.

Abel's glyoxyline contains potassium nitrate and nitro-
glycerin.

The most varied compounds have from this point of view
been proposed, and continue to be proposed daily. We shall
particularly mention Schultze powders, formed by nitrified

1 See " Memorial des Poudres et Salpetres " (Rapport sur 1'emploi du coton-
poudre aux operations de guerre), par H. Sebert, Commissions des Explosive
Substances, p. 109. 1882.

4.46 GUN-COTTON AND NITRO-CELLULOSES.

wood-pulp associated with various nitrates, an explosive which
has assumed some importance recently.

6. In the following paragraphs we shall merely treat of
ordinary gun-cotton alone and when with water or nitrates,
these three substances being regarded as types. We shall, as
usual, regard them chiefly with reference to the degree of heat
liberated, the volume of gases, and the pressures developed.

§ 2. NlTRO-CELLULOSES : THEIR COMPOSITION.

1. The nitrification of cellulose under its various forms (cotton,
paper, straw, wood-pulp, etc.) is accomplished by means of
nitric acid of various degrees of concentration, with or without
the addition of sulphuric acid, and working at different tempe-
ratures. The products are numerous, and they have been the
object of many researches. Here we shall content ourselves by
reproducing the results of the most recent experiments, namely,
those by Vieille l carried out at 11°, in the presence of an excess
of acid sufficient to prevent the water formed by the reaction
modifying the composition to any appreciable extent.

The highest nitrification is obtained with nitro-sulphuric
mixtures ; it corresponds sensibly to the formula of an ende-
canitric cellulose —

C21H18(N03H)1109.

This is gun-cotton intended for military purposes.

With nitric acid alone, corresponding to the composition

(N03H + |H20),

and when experimenting at 11°, we obtain a decanitric cellulose ;
that is to say, less rich in acid —

C24H20(N03H)10010,

a body which is completely soluble in acetic ether, but almost
insoluble in a mixture of alcohol and ether. This is still gun-
cotton.

When the acid is rather more diluted —

(HN03 + -34H20),

it yields collodion cotton, the composition of which is very
similar to that of the enneanitric and octonitric celluloses —
C24H22(N03H)9011, and 024H24(N03H)8011, /y

bodies which are soluble in acetic ether and in a mixture of
alcohol and ether.

With the acid N03H+ iH20, a cellulose is obtained which
answers to the characteristics of a heptanitric compound —

1 " Comptes rendus des stances de 1'Academie des Sciences," torn. xcv.
p. 132. 1882.

GUN-COTTON. 447

yet still preserving the aspect of the cotton, but which becomes
gelatinous without actually dissolving in a mixture of alcohol
and ether and in acetic ether.

If the acid is more diluted, such as

(N03H + -75H20),

the cotton becomes dissolved in such an acid, producing a viscous
liquor which can be precipitated by water. The product ob-
tained is similar in its characteristic features to hexanitric
cellulose —

CMH28(N03H)6014.

It swells in acetic ether without dissolving. A mixture of
alcohol and ether does not act on the substance.

With the acid N03H mixed with + 1/375 to 1-5 H20, we
obtain friable products, without any action on acetic ether or
on the mixture of alcohol and ether, and which vary between
the following formulae : —

C34H32(N03H)4017.1

With a more diluted acid, the nitrification is incomplete, the
products still being darkened by iodine ; that is to say, it is no
longer possible to distinguish the nitro compounds properly so
called from their mixture with the unaltered cellulose.

§ 3. GUN-COTTON PROPERLY so CALLED.

1. Gun-cotton2 preserves the appearance of cotton, although
it is slightly rougher to the touch. It is not hygroscopic, and
it also possesses the property of becoming electrified by friction.
Plates for electric machines have even been constructed with
nitrified paper.

Gun-cotton is soluble in acetic ether, but insoluble in most
other solvents (water, alcohol, ether, acetic acid, and ammoniacal
copper oxide).

It may be moistened, and when dried resumes its properties.

When in lumps, its apparent density is only 01 ; if it be

1 Table of the volumes of nitric oxide obtained by Schloassing's process from
various celluloses by Vieille. One grm. gives —

I Collodion cotton 1178

Heptanitric „ 162

Hexanitric „ ... 146

Pentanitric „

Tetranitric „ 108

2 For its preparation see " Traite" sur la poudre," par Upman et Meyer,
traduit et augmente" par Desortiaux, etc., p. 350.

448 GUN-COTTON AND NITRO-CELLULOSES.

twisted into thread, it increases to 0'25; when subjected, in
the form of pulp, to hydraulic pressure, it becomes TO; but
these densities are apparent, the absolute density of gun-cotton
being 1'5.

Nitrohydrocellulose prepared with cellulose disintegrated by
hydrochloric or sulphuric acid (A. Girard's process) has a
pulverulent form, which is very convenient for practical use.
Its composition and the force are the same as for gun-cotton.

2. Gun-cotton is an extremely explosive compound, which is
ignited by contact with a heated body or by shock, or, again,
when it is raised to a temperature of 172°. It burns suddenly,
with a large yellowish-red flame, but almost without smoke or
residue, and liberates a large volume of gas (carbonic acid,
carbonic oxide, nitrogen, steam, etc.).

Compressed gun-cotton previously heated to 100° may explode
when ignited. It is, therefore, more liable than dynamite to
explode on simple inflammation.

Gun-cotton kept at 80° to 100° decomposes slowly, and may
end by inflaming.

It has been shown that a thin disc of compressed gun-cotton
may be pierced by a ball without explosion ; but if the thick-
ness of the disc be increased, or if resisting envelopes be used,
an explosion occurs.

3. Sunlight causes it to undergo slow decomposition.

4. Gun-cotton should be neutral to litmus, when it has been
carefully freed from all acid products by washing with alkali.
Nor should it emit acid fumes even, after keeping for some
time. A little sodium or ammonium carbonate is incorporated
with it to increase its stability.

In the French navy, gun-cotton is submitted to a heat test,
which consists in heating it to 65°, until it gives off sufficient
nitrous vapour to turn the iodised starch paper blue, or more
simply to redden litmus. It should stand this test for eleven
minutes. The heat test may be carried out either on the raw
material or on the washed product (the washing frees it from
alkaline carbonates), compressed between blotting paper, dried
at a low temperature, then left some time in the open air.

5. The indefinite stability of gun-cotton has always been
regarded as doubtful, both by reason of its chemical constitution
and by the presence of the accessory products arising from the
original reaction or formed by accidental causes, which it is
hardly possible to avoid indefinitely. A slow decomposition
produced in this way sometimes becomes considerably accele-
rated by the heat which it liberates and by the reaction of the
products originally formed on the rest. It may become violent,
and end by exploding (see p. 45).

Nevertheless, gun-cotton has been preserved for ten years
and more without any alteration. It has also been kept dry on

THEORETICAL CONSIDERATIONS. 449

board vessels during long voyages, even in high temperatures in
the tropics.

6. Gun-cotton is very susceptible to explosions by influence.
According to experiments made in England, a torpedo, even
placed at a long distance, may explode a line of torpedoes
charged with gun-cotton.

7. The velocity of the propagation of the explosion in metallic
tubes filled with pulverised gun-cotton has been found to be
from 5000 to 6000 mms. per second in tin tubes, and 4000 in
leaden tubes (Sebert).

Gun-cotton loosely exposed in the open air burns eight times
as quickly as powder (Piobert).

8. It is admitted that the effect of gun-cotton in mines is
very nearly the same as that of dynamite for equal weights. It
requires stronger detonation, and it gives rise to a large quantity
of carbonic oxide, which is sometimes difficult to disperse,
because the earth remains impregnated with the gas. Carbonic
oxide being very deleterious, the use of gun-cotton is dangerous
to workmen in mines, But the form of compressed gun-cotton
is more convenient, because it does not require resisting
envelopes, and because it preserves the form which is given to
it. Besides, it is less sensitive to shock by reason of its special
structure. Its use for firearms has been abandoned.

9. Let us now examine gun-cotton a little closer from a
theoretical point of view. Its force depends upon its composi-
tion, and upon the nature of the products, which vary with the
density of the charge; that is to say, with the pressure
developed.

10. We have, at p. 288, given a summary of the very in-
teresting researches of Sarrau and Vieille on this question.

Let us simply remember that the substance studied by these
authors contained —

C 24-4

H 2-4

N 12-8

0 56-5

Water 1-4

Ash 2-5

that is to say, abstracting the water and the ash —

C ... ... 25-4

H 2-5

N 13-3

0 58-8

the formula C24Hi8(N03H)1109 requires —

C ... 25-2

H 2-6

N 13-5

0 58-7

2 G

450 GUN-COTTON AND NITRO-CELLULOSES.

11. The equivalent of this substance is 1143.

12. The heat liberated by the formation of gun-cotton from
the elements under constant pressure —

€24 (diamond) + H^ + N^O*,

amounts to 624 Cal. for 1143 grins., or 546 Cal. for 1 kgm.
The heat of formation of collodion cotton —

C* + H31 + N9 + 038 = CatHa(N03H9)Ou,

is 696 Cal. for 1053 grms., or 661 Cal. for 1 kgm.

Soluble gun-cotton made in Norway is very near this com-
position.

13. The heat liberated in the total combustion of gun-cotton
by free oxygen —

2[C24H18(N203H)1109] + 0*! = 48C02 + 29H20 + 11N2,

at constant pressure, is 2633 Cal. for 1143 grms. (water liquid),
or 2488 Cal. (water gaseous). Say for 1 kgm. of gun-cotton,
2302 Cal. (water liquid), or 2177 Cal. (water gaseous). ^~*, /
The total heat of combustion of collodion cotton —

2[C24H22(N03H)9011]051 = 48C02 + 31H«02 + 9N2,

at constant pressure, the water being liquid, -f 2627*5 CaL ; the
water being gaseous, + 2474'5 Cal.

It will be seen that it is nearly the same at equal equivalents 0
as for gun-cotton.

For 1 kgm. of collodion we should have 2428 Cal. (water
liquid), 2351 Cal. (water gaseous).

14. The heat of decomposition of gun-cotton in a closed
vessel, found by experiment at a low density of charge (0'023),
amounts to 1071 Cal. for 1 kgm. of the substances, dry and free
from ash, or 1225 Cal. for 1143 grms. (water liquid).

We proceed to compare this result with the heat calculated
from the equation for the decomposition.

15. Equation for the decomposition. From the analysis of the
products, the decomposition of the gun-cotton which yielded
this quantity of heat practically agreed with the following
equation (low densities of charge) : —

(1) 2[CaiH18(N08H)1109] = 30CO = 18C02 + 11H2 + 18H2O

+ 11N2.

But the quantity of heat changes with the equation of decom-
position, the latter approximating to

(2) 24CO + 24C02 + 12H20 + 17H2 + 11N2

for high densities of charge, according to Sarrau and Vieille (p.
289). There is, moreover, no nitric oxide under these conditions.1

1 Karolyi, Sarrau and Vieille.

HEAT OF DECOMPOSITION. 451

On the contrary, in a miss-fire (progressive combustion) the
carbonic oxide increases and nitric oxide appears (p. 289).
We shall treat here only of the explosive combustion.

16. Let us now calculate the heat liberated at constant
pressure.1 According to equation (1), which corresponds to low
densities of charge, the reaction liberates 1230 Gal. (water
liquid), or 1140 Cal. (water gaseous).

That is to say, for 1 kgm.,2 1076 Cal. (water liquid), or 9977
Cal. (water gaseous).

According to equation (2), which represents the limit of re-
action for high densities of charge, we should have 1228 Cal.
(water liquid), or 1168 Cal. (water gaseous). That is to
say, for 1 kgm., 1074 Cal. (water liquid), or 1022 Cal. (water
gaseous).

It will be remarked that the heat liberated is practically the
same according to equations (1) and (2). It therefore varies
but little with the density of charge, an observation which
appears applicable to explosive substances in general. Thus the
numbers 1074 CaL, and 1076 Cal., which correspond to the two
equations, are very close to each other, and also to the figure
1071 Cal. found by experiment.

17. The volume of the reduced gases, calculated from equation
(1), will be 781 litres (water liquid), or 982 litres (water
gaseous) ; that is to say, for 1 kgm.,2 684 litres, or 849 litres.
Sarrau and Vieille found 671 litres, with a substance leaving
2 '4 per cent, of ash, which agrees. According to equation (2),
the volume of the gases will be the same, the water being sup-
posed liquid ; it would be raised to 743 litres per kilogramme,
the water being gaseous. Hence the volume of the gases does
not change much with the density of charge.

18. The permanent pressure according to equation (1)

(low densities) = — -f- This formula is only applicable for
n — 0*14

densities - low enough for the carbonic acid not to be liquefied.
n

16400 atm

19. The theoretical pressure, from equation (1), = -•

fir

From equation (2) = 1675° atm"

Sarrau and Vieille actually found, by means of the crusher
and for densities of charge -, the following pressures, P' ex-

7i

pressed in kilogrammes : —

1 At constant volume these figures must be increased by one per cent.

2 The substance supposed dry and free from ash.

2 G2

452

GUN-COTTON AND NITBO-CELLULOSES.

P.

I".

0-10
0-15
0-20
0-30
0-35
0-45
0-55

1185
2205
3120
5575
7730
9760
11480

11850
14700
15600
18600
22100
21700
21500

But these results should be interpreted in accordance with
their new researches on the calibration of " crushers " (p. 23).

The latter gave for - = 0'20, a maximum pressure of 1985

n

F F

kgm. ; which would make — = 9825 kgm. The limit — , that

n n

is, the specific pressure, relating to gun-cotton would therefore
seem to need to be reduced to about 10,000 kgm., in round
numbers, for high densities of charge. The theoretical pressure
calculated from our formula would, on the contrary, be
applicable to low densities. To obtain the maximum effect of
gun-cotton, theory, in accordance with the latest experiments,
shows that this powder must be compressed and reduced to the
smallest possible volume. For the initial pressures are thereby
increased.

20. Let us now compare gun-cotton with other explosive
substances. It is especially distinguished by the magnitude
of the initial pressures. Thus, according to theory, the initial
pressure will be more than treble that of ordinary powder,
which is, in fact, the empirical ratio given by Piobert.1 This
theoretical pressure, calculated from the reactions of the final
state, will, moreover, be diminished in practice, as in the case of
ordinary powder, owing to the incomplete state of combination
of the elements and the complexity of the compounds which
tend to be formed. Hence results a less sudden and more
regular expansion, following upon a combination which has
become more complete during cooling.

On the contrary, pure nitroglycerin, weight for weight,
realises a work greater by half than gun-cotton, the initial
pressure being nearly the same. It is not surprising, there-
fore, that, nitroglycerin should have been found preferable for
industrial purposes, at least in the form of dynamite ; the more
so as the latter needs no previous compression, is easier to divide,
and, above all, more economical. But it is easier to distribute
non-compressed gun-cotton in a uniform manner over a con-
siderable space, which offers certain advantages in practice.

1 " Trait^ de I'Artillerie," 2' Edition, p. 496.

WET GUN-COTTON. 453

§ 4. WET GUN-COTTON.

1. We have explained how it has been found advisable to
employ gun-cotton saturated with water, in order to lessen its
sensitiveness to shock and to render its direct inflammation
impossible, which limits the risks due to a fire. Three per
cent, of water is sufficient to diminish the sensitiveness, but
more than 11 per cent, is required to prevent direct inflamma-
tion. The standard quantity is 15 per cent, of water; but it is
difficult to maintain constant and uniform in the whole mass.
Thus regular saturation, followed by compression, leaves about
25 per cent, of water in the mass, which renders a partial
drying necessary. Besides, the moist substance, if it be not kept
in hermetic receptacles, tends to lose the water by spontaneous
evaporation.

2. Damp gun-cotton retains the property of exploding under
the influence of a powerful fulminate detonator, or of a small
intermediate cartridge of gun-cotton dry, or mixed with nitrate,
with fulminate cap. Thus a torpedo containing 100 kgm. of
gun-cotton requires a priming cylinder containing 0'560 kgm.
of dry gun-cotton. It will be useful to examine the influence
of the water thus introduced on the pressures developed.

3. Granted that the chemical reaction is the same as with
high densities of charge (which, however, has not been verified),
the heat liberated remains the same. The volume of the gases
produced by gun-cotton also remains the same, whether it be
calculated from that of the permanent gases alone, or the water
derived from the gun-cotton be supposed to retain the gaseous
state at the first instant of the explosion ; an hypothesis which
the experiments made on the explosive wave (p. 99) justify us
in regarding as possible. Condensation will, moreover, take
place almost immediately ; the water vapour thus ceasing to be
active beyond the first instant.

Nevertheless, the water imprisoned in gun-cotton also absorbs
heat, and may even be regarded as assuming, either wholly or
partly, the gaseous state, simultaneously with the water pro-
duced by the reaction.

We will calculate the pressure developed at the moment of
explosion according to the various hypotheses.

4. Take, for example, gun-cotton with the addition of 20 per
cent, of water —

CaiH18(N03H)u09 + 26H20,

and gun-cotton with 10 per cent, of water —
C24H18(N03H)U09 + 13H20.

The heat liberated by decomposition with a high density of
charge will be 1168 Cal. (water gaseous), or 1022 Gal., for 1
kgm. of the dry substance. This heat falls to 908 Cal. for the

454 GUN-COTTON AND NITRO-CELLULOSES.

same weight of dry gun-cotton with 20 per cent, of added
water. It is 1038 Cal. with only 10 per cent, of added water.
This makes, in other terms, for 1 kgm. of the damp substance
containing 16*7 per cent, of water, 662 Cal., and for 1 kgm. of
the substance containing 91 per cent, of water, 882 Cal., the
whole of the water being supposed gaseous. The heat is there-
fore reduced by a fifth in the latter case, and by a third in the
former, owing to the vaporisation of the added water.

5. The volume of the reduced gases will be for 1 kgm. of the
dry substance, with 20 per cent, of added water, 1563 litres ;
or 1139 litres for 1 kgm. damp.

We shall have further, for 1 kgm. of dry matter with 10 per
cent, of added water, 1272 litres ; or 1133 litres for 1 kgm. of
the damp substance.

The gaseous volume is therefore increased by the addition
of water, as might be expected, supposing vaporisation to take

place.

391 kgm. f

6. The permanent pressure = - for the substance

7t — Uol

with 20 per cent., and — -f— - with 10 per cent, of water
n — 0'2o

added, with the usual reservations regarding the limits of
liquefaction of carbonic acid.

7. The theoretical pressure = — - for the substance

n

with 20 per cent., and — ' with 10 per cent, of added

n

water. It will be seen that it is diminished by a third in the
latter case, and that it is reduced almost to the half in the case
of the most hydrated substance.

8. Paraffined gun-cotton. Instead of adding water to gun-
cotton it has also been proposed to paraffin it, which yields
mixtures which are more stable and even capable of being cut
arid wrought by tools working at high speeds. But it is
difficult to render them uniform, unless by adding so great a
quantity of paraffin that the mixture only explodes with great
difficulty; 100 parts of gun-cotton absorb as much as 33 of
paraffin.

Hence the operation is often confined to paraffining the
cartridges superficially. To explode paraffined gun-cotton an
auxiliary cartridge of ordinary gun-cotton is employed, ignited
by a fulminate detonator.

9. The use of camphor and plastic substances diminishes still
further the liability of gun-cotton to explode. We may also
mention here celluloid, a variety of nitro-cellulose, nearly
corresponding to C^H^NOaH)^, to which camphor and

GUN-COTTON AND AMMONIUM NITRATE. 455

various inert substances are added so as to render it non-
sensitive to shock. This product may be worked with tools, in
the manner of ivory, and is very plastic when heated towards
150°. But it must not be forgotten that it then tends to
become sensitive to shock, and that large quantities of such
substances might become explosive during a fire, owing to the
general heating of the mass and the evaporation of the camphor.
Heated celluloid may even explode, when greatly compressed,
and press accidents have occurred in factories. When main-
tained at 135° in an oven celluloid decomposes quickly. This
is not all, for in an experiment made in a closed vessel at 135°,
and the density of charge 0'4, it ended by exploding, developing
a pressure of 3000 kgm.

It is therefore a substance the working of which calls for
certain precautions, though it is not explosive under ordinary
circumstances, even with very powerful detonators.

§ 5. "NITRATED" GUN-COTTON.

Mixture formed with ammonium nitrate.

1. We will examine gun-cotton mixed with ammonium
nitrate, and also with potassium nitrate, these two products
having been studied in a special manner by Sarrau and Vieille.

We will first observe that gun-cotton —

C24H18(N03H)U09 = 1143 grms.,

requires 41 equivalents of oxygen (328 grms.) for complete com-
bustion, and that it then develops at constant pressure
2633 Cal., the water being liquid'; ors 2488 Gal., the water
being supposed gaseous. The volume of oxygen employed is
equal to 229 litres; the carbonic acid produced occupying
536 litres, the nitrogen 123 litres, and the water vapour
(reduced volume) 324 litres.

2. This being granted, the total combustion by ammonium
nitrate corresponds to the formula —

2[C24H18(N03H)1109] + 41N03NH4 = 48C02 + 111H20 + 52N2

Or 1640 grms. of nitrate for 1143 of gun-cotton ; in all, 2783 grms.
The substance, then, contains in 1 kgm. 589 grms. of nitrate
and 411 grms. of gun-cotton, all the products being supposed
dry and free from fixed ash. Sarrau and Vieille used 60 parts
of nitrate to 40 parts of gun-cotton. The substances were
triturated together, 24 parts of water having been previously
added to the gun-cotton, then the whole dried at 60°. It was
ascertained that the combustion of the mixture yielded only
carbonic acid and nitrogen, these two gases being in the ratio
of 54 : 46 volumes ; the difference between which and the
theoretical figures, or 52 : 48, corresponding to the slight

456 GUN-COTTON AND NITRO- CELLULOSES.

deviation of the composition employed from the composition in
equivalents.

In a miss-fire, on the contrary, combustion ceases to be total.
The authors, for instance, have observed, out of 100 volumes of

NO

29-5

CO

15-8

C02

24-8

H

2-9

N

2-7

3. The heat liberated by the total and regular reaction
amounts, according to the calculation, to 3678 Gal. (water
liquid), 3117*5 Cal. (water gaseous) ; or, for 1 kgm., 1321 Cal.
(water liquid), or 1120 Cal. (water gaseous). Sarrau and Vieille
actually found 1273 Cal. (water liquid) for a composition con-
taining only 40 per cent, of gun-cotton instead of 41. The
difference between the figure observed (1273) and the calculated
figure (about 1288) does not exceed the limits of experimental
error.

4. The reduced volume of the gases = 1116 litres (water liquid),
2399 litres (water gaseous) ; or, for 1 kgm., 401 litres (water
liquid), and 862 litres (water gaseous). Sarrau and Vieille
found 387 litres, with the composition containing 40 per cent,
of gun-cotton instead of 41 per cent.

401 atm.

5. The permanent pressure = - under the ordinary

71 — U ot)

reservations.

a TU ^ .• i 14900 atm.

6. The theoretical pressure = .

n

It is somewhat less than for gun-cotton.
Sarrau and Vieille actually found, with the composition con-
taining 40 per cent, of gun-cotton, and by the crusher method —

Density of charge 0'2, P = 3270 kgm.
„ 0-3, P = 5320 kgm.

which would make for the density 1, 16358 and 17730 ; mean,
17000 kgm. approximately, a figure which is rather higher than
14900. But it is possible that it ought to be reduced to the
half by a more exact estimation of the force of calibration
(p. 23).

§ 6. GUN-COTTON AND POTASSIUM NITRATE.

1. The total combustion of gun-cotton by potassium nitrate
corresponds to the formula —

10[C24H18(N03H)11OJ + 82KN03 = 199C02 + 41K2C03 +
145H20 + 96N2.

GUN-COTTON AND POTASSIUM NITRATE. 457

Note further that during cooling the potassium carbonate is
charged into bicarbonate, which gives finally —

158C02 + 82KHC03 + 104H20 + 96N2.

Or 828 grms. of nitrate for 1143 grms. of gun-cotton ; in all, 1971
grms. The substance contains, therefore, for 1 kgm., 420 grms.
of nitrate and 580 grms. of gun-cotton.

2. Sarrau and Vieille operated with equal weights, to assure
total combustion. These proportions correspond practically
with —

6[C24H18(N03H)1109] + 68KN03 = 110C02 + 34K2C03 +

87H20 + 470 + 67N2
or after cooling —

76C02 + 68KHC03 + 53H20 + 67N2 + 470.
The authors have found, with high densities of charge (0'3 and
0'5), that a mixture of carbonic acid, nitrogen, and oxygen, in
the following ratios of volume, is obtained —

52-3 ; 371 ; 10'7.
The formula gives — 54'9 ; 33*4 ; 11'7.

The difference shows that there probably exists a certain
quantity of nitrite. With low densities of charge (0'023) the
relative proportion by volume of carbonic acid increases (59*5),
nitrogen diminishes (33*8), oxygen disappears, and carbonic
oxide (5*0) and hydrogen (1*8) are obtained; the nitrite is
necessarily here present in a considerable quantity. Lastly,
in a combustion experiment at the atmospheric pressure, a
condition comparable to a miss- fire, the authors obtained for
100 vol.—

NO 363

CO 29-5

C03 29-0

H 1-6

N 3-4

3. We shall give the calculations for the proportions (1) and
(2), which correspond to total combustion. Equation (1)
represents an exact combustion, without excess of oxygen;
giving rise to a liberation of 1606 Cal. (initial formation of
neutral carbonate and gaseous water), or of 1766 Cal. (bicar-
bonate, liquid water) ; l or for 1 kgm. of the substance, 815 Cal.
or 891 Cal. Note that each molecule of liquefied water H20
increases the heat by + 10 Cal. Each equivalent of carbonate
changed into bicarbonate —

K2C03 + C02H20 liquid = 2KHC03,

further increases the heat by 4- 12 '4 Cal. If the water were
gaseous to commence with, the increase would be + 17'4 Cal.
Equation (2) represents a combustion with excess of nitrate and

1 Neglecting the dissolving action of this water on the salt.

458 GUN-COTTON AND NITRO-CELLULOSES.

consequently of oxygen. It corresponds to a liberation of
2240 Cal. (carbonate, gaseous water), or 2560 Cal. (bicarbonate,
liquid water). Or for 1 kgm. of the substance, 980 Cal. or
1120 Cal. Sarrau and Vieille found 954 Cal. for a substance
of this order ; but, in reality, the combustion in their experiment
did not give rise either to the total destruction of the nitrate, or,
probably, to the integral and immediate change of the carbonate
into bicarbonate.

4. The volume of the reduced gases will be, according to
equation (1), 1062*5 litres (carbonate and gaseous water) or 728
litres (bicarbonate and liquid water) ; or, for 1 kgm., 475 litres
or 271 litres.

Sarrau and Vieille found only 196 litres ; a figure which is
too low, for the reasons given above.

5. Owing to these considerable differences between the
theoretical and the real equation, it appears useless to calculate
the theoretical pressure of this powder. We will only mention
that Sarrau and Vieille found, by the crusher process, operating
on the product containing an excess of nitre —

Density of charge. Pressure.

0-20 1315 kgm.

0-30 3100 „

0-40 4900 „

0-50 5520 „

values which are nearly the half of those given by pure gun-
cotton or the same mixed with ammonium nitrate.

6. This is, moreover, what theory would enable us to fore-
see in a general way.

In fact, 1 kgm. of gun-cotton, decomposed under a high
pressure, develops 859 litres and produces 1020 Cal. (water
gaseous). On the other hand, 1 kgm. of gun-cotton mixed with
ammonium nitrate develops 862 litres and produces 1120 Cal.
(water gaseous). Whilst 1 kgm. of gun-cotton mixed with potas-
sium nitrate can only develop 475 litres and produce 980 Cal.

The volume of the gases with the latter mixture is therefore
nearly half that produced by the two other substances, the heat
being slightly less. Consequently the pressures will fall to
about the half for the same density of charge. Dissociation,
moreover, will intervene to lower the initial pressure and
moderate the fall of the successive pressures.

7. On the whole, theory does not show that the addition of
potassium nitrate to gun-cotton, which is rather inconvenient
to realise in practice, offers any very great advantages, except
in the way of economising the gun-cotton, rendering expansion
less abrupt and suppressing the carbonic oxide. The experi-
ments which have been made with similar mixtures formed of
various nitrocelluloses impregnated with potassium nitrate
seem to point to this conclusion.

SCHULTZE'S POWDER. 459

8. The Faversham " Cotton Powder " consists of a mixture of
equal weights of gun-cotton and barium nitrate. Gun-cotton
may also be mixed with sodium nitrate, the hygroscopic pro-
perties of which lessen the risk of inflammation. But special
detonators must then be employed. The relations by weight of
total combustion would be 51'6 of gun-cotton to 484 of barium
nitrate. The heat liberated is practically the same (pp. 4
and 134) as for an equivalent weight of potassium nitrate ; but
the barium nitrate mixture weighs 2223 grms. instead of 1971
grms., or one-eighth more.

The volume of the gases gives rise to the same relations, this
volume being identical for equivalent weights (provided only
the carbonate be neutral), but less at equal weights.

9. Schultze's powder, made from nitrated wood meal, will now
be considered.

It is prepared from wood reduced to small grains, which are
freed from resinous, nitrogenous, and incrusting matters by the
following treatments. It is boiled for six to eight hours with
sodium carbonate; washed, dried, and treated successively by
steam, cold water, bleaching powder ; then 16 parts of a mixture
of nitric acid of 1*50 specific gravity, with twice its volume of
concentrated sulphuric acid, is allowed to react for from two to
three hours. In this way is obtained a substance nearly related
by its composition to heptanitrocellulose —

In reality it is a mixture of several unequally nitrified products.
It is washed in cold water, then in a weak solution of sodium
carbonate. This done, the substance is steeped in a concentrated
solution of potassium or barium nitrate, pure or mixed, and
dried at 45°. The nitrate and the ligneous grains, which are
impregnated beforehand with 20 to 25 per cent, of water, can
again be incorporated under light edge runners. The com-
position of the final substance varies with the amount of the
nitrates. The following is the result of some of the analyses : —
Nitrocellulose soluble in alcohol ether 13-1 \ KQ ^
„ „ insoluble „ „ 44-9 /

Foreign matters soluble in alcohol 2-3

Potassium nitrate 6-2

Barium nitrate 30-0

Water 3-5

100-0
Another sample —

Nitrocellulose .. ... ... ... ... 66*5

Barium nitrate 15-0

Potassium nitrate 15-0

Water 3-5

100-0
1 One grm. produces 166 cc. of nitric oxide by Schlcessing's process.

460 GUN-COTTON AND NITRO-CELLULOSES.

The latter had as gravimetric density, 0'416.
Density taken with mercury, 0'944.

The substance gave 7300 granules to the gramme. It is a3
sensitive to shock as black powder, keeps well, decomposes
towards 174°. It gives a light smoke which dissipates rapidly.

This powder has been much used for sporting purposes.

§ 7. GUN-COTTON AND CHLORATE.

1. Some data may now be given regarding this mixture, the use
of which, however, has been abandoned owing to the dangerous
character of the chlorate powders.

2. The complete combustion corresponds with the following
equations : —

e] + 41KC103 = 144C02 + 87H20 + 33N2

+ 41KC1.

It corresponds to the proportions, 1143 grms. of gun-cotton and
838 grms. of chlorate ; in all, 1981 grms. ; or, for 1 kgm. of the
mixture, 577 grms. of gun-cotton and 423 grms. of chlorate.

3. It liberates 2708 Cal., the water being liquid,1 and 2563
Cal., the water being gaseous ; or, for 1 kgm., 1367 Cal. (water
liquid), or 1294 Cal. (water gaseous), figures which are some-
what higher than those for pure gun-cotton, but the volume of
the gases is far less.

4. The volume of the reduced gases is 978*6 litres (water
gaseous), or 653 litres (water liquid) ; or, for 1 kgm., 484-5 litres
(gaseous water), or 323*5 litres (liquid water), figures lower by
half than those for pure gun-cotton. They are likewise less
than those for gun-cotton mixed with ammonium nitrate.

* m, . 323-5 atm. ...

5. The permanent pressure is , provided n be large

enough for the carbonic acid not to be liquefied.

a rru 4.U .• i - 13175 atm'

6. The theoretical pressure is .

n — O'Oo

This number is less by a third than that for pure gun-cotton,
and by an eighth than that for gun-cotton mixed with
ammonium nitrate. The smallness of the gaseous volume
would enable this inferiority to be anticipated. Gun-cotton
mixed with potassium nitrate would alone yield volumes of
nearly equal magnitude. Hence it will be seen that chlorated
gun-cotton does not present, from the point of view of strength,
the same advantages over the other varieties of gun-cotton,
which have often been attributed to the chlorate powders. When
we add that it is much more sensitive to shocks and friction,
and therefore much more dangerous, it will be easy to understand
the reasons which have led to the use of it being given up.
1 Neglecting the action of this water on the potassium chloride.

( 461 )

CHAPTER VIII.
PICRIC ACID AND PICRATES.

§ 1. HISTORICAL.

TRINITROPHENOL, otherwise termed picric acid, heated towards
300°, decomposes with a sudden explosion, and its salts behave
in a similar manner. But the decomposition is complex, and
only takes place at a temperature higher than that of nitro-
glycerin, when oxidising bodies, such as potassium nitrate or
chlorate, are added. It occurs at a lower temperature than
with the pure acids and salts, and yields simpler products.
Powders of various natures are obtained in this way, some
having as base picric acid and sodium nitrate (Borlinetto
powders), others having as base potassium picrate associated
either with potassium nitrate (Designolle powders) or chlorate
(Fontaine powder); other powders again having as base
ammonium picrate with potassium nitrate (Brugere powder and
Abel powder). The chlorate powder has been proposed for
torpedoes only, it being very dangerous. On the other hand,
the powders formed with the nitrates can be employed in fire-
arms, especially ammonium picrate powder, which has of late
been greatly studied in France. We shall successively examine
the picric acid, potassium picrate, and ammonium picrate
powders.

§ 2. PICRIC ACID.

1. Picric acid is a yellow body, in laminated and friable
crystals, having a bitter taste, very stable in itself, not easily
soluble in water, but soluble in all other solvents.

When heated it melts, and can even be sublimed when
very small quantities are operated upon. But if the quantity
be at all considerable, or the acid be suddenly heated, it explodes
very violently. This property has occasioned serious accidents.
For instance, it has happened that experimenters have been
injured by throwing powdered picric acid into a furnace from a

462 PICRIC ACID AND PICRATES.

flask to show its explosion, the latter having been propagated
backwards along the trail of dust up to the principal mass.

2. The formula for picric acid is C6H3(N02)30, its equivalent
229.

3. Its heat of formation from the elements (p. 277) —

C6 (diamond) + H3 + N3 + 07 = C6H3N307 + 491.

This body hardly contains more than half the oxygen necessary
for its complete combustion.

4. Its heat of total combustion by free oxygen —

2[C6H3(N02)30] + 013 = 12C02 + 3H20 + 3N2,

is equal to -f 61 8 '4 Cal. (water liquid), according to the results
of the experiments of Sarrau and Vieille.

5. The equation representing its explosive decomposition has
not been studied. Admitting provisionally the following —

2[C6H3(N02)30] = 3C02 -f SCO + C -f 6H + 6N,

the heat liberated would be + 13O6 Cal., or 570 CaL per
kilogramme.

6. The reduced volume of the gases would be 190 litres per
equivalent, or 829 litres per kilogramme.

7. This figure divided by n, or -, practically represents

the permanent pressure, owing to the small volume occupied by
the carbon, with the usual exception of the liquefaction of the
carbonic acid.

10942 atm.

8. Lastly, the theoretical pressure = . These

values are only given with all due reserve.

9. To obtain a total combustion of picric acid, recourse must
be had to a complementary oxidising agent — nitrate, chlorate,
etc. It has been proposed, for instance, to mix picric acid
(10 parts) with sodium nitrate (10 parts) and potassium
bichromate (8 '3 parts). These proportions would furnish a
third of oxygen in excess of the necessary proportion.

But it is doubtful whether this powder has ever been pre-
pared on a large scale or kept. In fact, the mechanical mixture
of bodies of this nature can only be executed without danger
on the condition of wetting the pulverised substances before
incorporating them under the millstone or otherwise. Now,
as soon as water intervenes, the picric acid displaces the nitric
acid of the nitrates, even in the cold, and this volatile acid dis-
appears wholly or partly during the drying in the stove. This
circumstance hardly permits of employing free picric acid in
the manufacture of powders.

An analogous reaction renders its mixture with potassium
chlorate particularly dangerous.

EXPLOSIVE DECOMPOSITION OF POTASSIUM PICBATE. 463

§ 3. PICRATE POTASSIUM.

1. Potassium picrate,

C6H2K(N02)30,

crystallises in long orange-yellow needles, very slightly voluble
in water.

2. It explodes, when heated above 300°, much more violently
than picric acid. It also explodes by contact with an ignited
body, which renders it still more dangerous than black powder.
In the dry state its fine and light dust takes fire at a distance,
and may cause the whole mass from which it emanates to
explode. Operators have been wounded in public lectures by
throwing upon lighted coals potassium picrate contained in a
flask. This kind of accident is even more to be feared with
potassium picrate than with picric acid. The catastrophe in
the Place de la Sorbonne (1869) appears due to this property.
Potassium picrate is sensitive to shock, and even much more so
than picric acid. The addition of 15 per cent, of water deprives
it of this sensitiveness. Potassium picrate does not contain
enough oxygen to produce complete combustion. Hence the
necessity for mixing it with potassium nitrate or chlorate.

3. Its equivalent is 267.

4. Its heat of formation from the elements —

C6 (diamond) + H2 + K + N3 + 07 = C6H2K(N02)30,

is equal to -f 117*5 CaL, according to the data of Sarrau and
Vieille.

5. The heat of total combustion by free oxygen —
2[C6H2K(N02)30] + 0* = 2KHC0310C02 + H20 + 6N,

amounts to 619'7 Cal. (potassium bicarbonate and liquid
water). The explosive decomposition of potassium picrate
yields products which vary with the conditions, as is generally
speaking the case with bodies which do not contain a sufficient
quantity of oxygen to produce complete combustion (p. 7).

Sarrau and Vieille have studied this decomposition minutely.
The following are the results obtained by them,1 with various
densities of charge, per 100 vols. —

Densities of charge.

HCy

C02

CO

0-023
lit.

1-98
10-66
62-10
0-17
10-31
16-88

0-3
lit.

0-32
13-37
59-42
2-38
6-77
17-74

0-5

lit.

0-31
20-48
50-88
5-39
2-68
18-26

N

Volume of gases disengaged per 1 kgm., 5741, 557'9.

1 " Comptes rendus des stances de 1' Academic des Sciences,11 torn, xciii. p. 6.

464 PICKIC ACID AND PICRATES.

6. The solid residuum is formed of potassium carbonate and
cyanide, with a trace of carbon. The proportion of potassium
changed to cyanide, in 100 parts, amounted respectively to 29 '8,
347, 24'3. At the density 0'5, the reaction approximates to
the following formula : —

16C6H2K(N02)30 = 4KCN + 6K2C03 4- 21C02 + 52CO +

44N + 6CH4 + 8H + 70.

It tends towards 4CH4 + 50 ; that is to say, the methane is
formed in an increasing quantity, according as the density
augments. On the contrary, the methane tends to disappears
for low densities.

7. Heat of decomposition. The formula given above would
correspond to + 208*4 Cal. for 1 equiv. of picrate decomposed,
or 781-2 Cal. for 1 kgm.

8. Volume of the gases. It would yield 146*5 litres (re-
duced volume) of gases per equiv., or 549 litres for 1 kgm.

n ™ .* . . T 5600 atm.

9. The theoretical pressure =± TTTT-

n — 014

Sarrau and Vieille found 6700 kgm. at low densities of
charge, such as 0*023.

It has been seen that for high densities the gaseous volume

found tends to approach the theoretical figure. Now, at these

•p
high densities the ratio — has been found by the same authors

n

at nearly 12,000 kgm., a figure which should be corrected
according to their recent experiments (p. 23). These point kto
about the half, or 6600 kgm., a value near the theoretical
figure, which would correspond, for n = 1, to 6700 kgm. It
will further be seen that it is greatly lower than the pressures
developed by nitroglycerin, or by gun-cotton, for the same
density of charge (p. 425). This is, in fact, as it should be,
according to theory, the heat liberated being less, weight- for
weight, as well as the volume of the gases.

Potassium picrate, therefore, does not offer the advantages
which had been anticipated from it at first, from the abruptness
of its explosive effects.

-§ 4 POTASSIUM PICRATE WITH NITRATE.

1. The total combustion of potassium picrate by potassium
nitrate corresponds to the following formula : —

5C6H2K(N02)30 + 13KN03 = 9K2003 + 21C02 + 5H20

+ 28N.1

2. The total weight of the substance in equivalents is 267
grins, of picrate and 263 grms. of nitrate ; in all, 530 grms. For

1 The slow formation of 2 equiv. of bicarbonate is here neglected.

POTASSIUM PICKATE AND CHLORATE. 465

1 kgm. both bodies are nearly in equal weights, 504 grms. of
picrate to 496 grms. of nitrate. This composition is that of
torpedo powders.

3. The heat liberated amounts to + 538-0 Gal. (water liquid),
or + 528-2 Cal. (water gaseous) ; or, for 1 kgm., 1015 Cal., or
997 Cal.

4. The reduced volume of the gases = 170 litres (water
gaseous), or 116 litres (liquid water and bicarbonate) ; or, for
1 kgm., 326 litres, or 246 litres.

. 6320 atm.

5. The theoretical pressure is — — ; it does not greatly

fb ~— \J*£i\.

differ from the value for pure potassium picrate.

6. The potassium picrate powders proposed for cannons and
guns have a different composition. The amount of picrate has
been diminished in order to reduce its shattering properties, and
it has been replaced by carbon ; for cannons, 9 parts by weight
of picrate, 80 of nitre ; for guns, 23 parts of picrate, 69 of nitre,
8 of carbon, etc.

§ 5. POTASSIUM PICRATE WITH CHLORATE.

1. The total combustion of potassium picrate by potassium
chlorate corresponds to the formula —

6C6H2K(N02)30 + 13KC103 = 3K2C03 + 33C02 + 6H20 +

18N + 13KC1,
or rather

6KHC03 + 30C02 + 3H20 + 18N + 13KC1.

2. The equivalent weight is 267 grms. of picrate to 265*7 of
chlorate ; in all, 532'7 grms.

For 1 kgm., 502 grms. of picrate and 498 grms. of chlorate ;
that is to say, nearly equal weights. The composition is, more-
over, very nearly the same by weight for the nitrated and the
chlorated powders, owing to a numerical coincidence in the
equivalents.

3. The heat liberated will be 622-2 Cal. (gaseous water and
carbonate), or 647'6 (liquid water and bicarbonate) ; or, for
1 kgm., 1168 Cal., or 1214 Cal.

4. The reduced volume of the gases, 178 -6 litres (gaseous
water), or 145 litres (liquid water, bicarbonate) ; or, for 1 kgm.,
335 litres, or 272 litres.

5. The permanent pressure = - ', with the usual

exception.

8200 atm

6. The theoretical pressure, n 01 ', is about a third

n — U'AL

greater than that of nitrated picrate and that of pure picrate.
But it hardly reaches half of that of dynamite or gun-cotton.

2H

466 PICRIC ACIDS AND PICRATES.

Hence it will be seen that " chlorated " picrate does not bear
out, by an exceptional strength, the hopes which the vivacity
of its explosion had given rise to at the outset. It therefore
does not compensate in this direction for the considerable
dangers which result from its great sensitiveness to shock,
friction, and inflammation, as well as the easy propagation of
the latter by dust trails. Its use, therefore, seems to be almost
abandoned.

§ 6. AMMONIUM PICRATE.

1. This is an orange-yellow salt, in needles, less hard than
potassium picrate. It is far less sensitive to shock. Ignited in
the open air it burns like a resin, with a smoky flame. It has
been used in pyrotechny as a fusing substance. However,
when burnt at a high density of charge, or in a confined space,
from which the gases only escape by a small orifice, its com-
bustion may change into detonation.

2. Its formula is —

C,H2(NH4)(N02)30;
its equivalent, 246.

3. Its heat of formation from the elements —

C6 + H6 + N4 + 07 = C6H6N407,

is equal to -f 801 Cal. ; or, for 1 kgm., 326 Cal.

4. Its total combustion needs an excess of oxygen —

C6H6£T407 -f 08 = 6C02 -f 3H20 + 2N2,

and liberates + 690*4 Cal. (liquid water), or -f 6604 Cal.
(gaseous water).

5. The equation of the explosive decomposition has not been
studied.

6. Only the combustion by a combustive agent, such as
potassium nitrate, will be examined —

5C6H6£T407 + 16KN"03 = 8K2C03 + 22C02 + 15H20 + 36N,
or

16KHC03 -f 14C02 + 7H20 + 36N
after cooling.

The total weight is here 569*5 grms. per 1 kgm.; viz., 568
grms. of saltpetre and 432 grms. of picrate.

The heat liberated by the combustion of " nitrated " ammo-
nium picrate amounts to + 701 Cal. (liquid water, bicarbonate),
or to -f 631-5 Cal. (gaseous water) ; or, per 1 kgm., 1231 Cal., or

7. The reduced volume of the gases = 245 -5 litres (gaseous
water), or 174 litres (liquid water, bicarbonate), which makes
per 1 kgm. 431 litres, or 305 litres.

AMMONIUM PICRATE. 467

o mi 305 atm. ... ,,

8. The permanent pressure = - 7— rr> Wlt^ tne usual

n = 0'17

exception.

n mi .T_ ^ T 9050 atm.

9. The theoretical pressure = - - .

It is higher than that of potassium picrate, whether pure or
mixed with potassium chlorate or nitrate.

10. The Brugere powder is, in fact, formed of ammonium
picrate and potassium nitrate. It contains 54 parts of picrate
and 46 of saltpetre. Here combustion is not total, and the true
reaction is therefore imperfectly known.

This powder is only slightly hygroscopic: it is stable, and
makes little smoke. Its strength is double that of black powder
weight for weight.

11. On account of its fusing properties, ammonium picrate
can also be employed in fireworks.

For example, this salt mixed with barium nitrate gives green
fires.

DesignoUe powder ... 48

C Ammonium picrate ...... 25

Brugere powder < Barium picrate ......... 67

(Sulphur ............ 8

Mixed with strontium nitrate it gives red fires.

Ammonium picrate ............... 54

Strontium nitrate ............... 46

None of these proportions correspond to total combustion.

2H2

( 468 )

CHAPTER IX.
DIAZO COMPOUNDS AND OTHERS.

§ 1. SUMMARY.

WE shall give in this chapter the observations and calculations
relative to various explosive compounds, such as mercury
fulminate and diazobenzene nitrate, both belonging to the
group of diazo compounds, the acid mixtures formed of nitric
acid associated with an organic compound, which is generally
already nitrified, the perchloric ethers and mercury and silver
oxalates. This list might be made much longer in theory (see
p. 368 and the following), but experimental data and practical
applications would be wanting.

§ 2. MERCURY FULMINATE.

1. The analysis and mode of decomposition of this body have
been given (p. 297) —

2. This reaction liberates + 114-5 Cal. at constant pressure
for 284 grms. ; the mercury being supposed gaseous, + 99'1 Cal.;
or, for 1 kgm., 463 Cal. or 349 Cal.

3. The formation from the elements absorbs — 62 '9 Cal. for
284 grni., or - 221'5 Cal. for 1 kgm.

4. The total combustion by free oxygen —

C2N2Hg02 + 02 = 2C02 + Hg + N* liberates + 250'9 Cal.,

or, the mercury being gaseous, + 235 -5 Cal.

5. The density is equal to 4 -43.

6. Pure fulminate may be kept for an indefinite length of
time. Water does not affect it. It explodes at 187°, and also
on contact with an ignited body.

It is very sensitive to shock and friction, even that of wood
upon wood. When used in a cannon, it bursts it, without the
projectile having time to displace itself. However, it may be

MERCURY FULMINATE. 469

employed for discharging bullets in saloon arms. If placed in
a shell, and the latter can be projected by the aid of some
artifice of progressive expansion, the shell bursts at the striking
point, owing to the shock and heating resulting from the sudden
stoppage of the projectile. A hollow projectile is broken by.
fulminate into a multitude of small fragments, much more
numerous than those produced by powder, but which are not as
widely scattered.

Its inflammation is so sudden that it scatters black powder
on which it is placed without igniting it ; but it is sufficient to
place it in an envelope, however weak, for ignition to take
place. The more resisting the envelope the more violent is the
shock, a circumstance which plays an important part in caps
and detonators.

The presence of 30 per cent, of water prevents the decom-
position of finely powdered fulminate by friction or shock.
With 10 per cent, of water, it decomposes without explosion ;
with 5 per cent., the explosion does not extend beyond the part
struck. But these results are only strictly true for small
quantities of the substance, and it would be dangerous to attach
too much importance to them.

Moist fulminate slowly decomposes on contact with the
oxidisable metals.

7. The reduced volume of the gases produced by the decom-
position is 66'96 litres per 284 grms., or 235*6 litres per 1 kgm.

If the mercury be supposed in the gaseous state, at a suitable
temperature t, we shall have 89*28 litres (1 -f at) per 1 equiv.,
or per 1 kgm., 3141 litres (1 + at).

235-6 atm.

8. The permanent pressure = r~rr~-

n — 0'05

6280 atm.

9. The theoretical pressure = •

n

The experiments which the author made with the crusher, in
common with M. Vieille, gave —

Density of charge O'l 480 kgm.

„ ' „ 0-2 1730 „

„ 0-3 2700 „

9000 atm.
We should, therefore, have for high densities about - —

71

But these figures should be reduced, in accordance with a more
exact estimation of the force of calibration (see p. 23). The
corrected calculation gives results very closely agreeing with
theory (p. 27), and leads to a specific pressure equal to

— '-. At the density 4'43, that is to say, the fulminate

n

exploding in its own volume, we should have, therefore,

470 DIAZO COMPOUNDS AND OTHERS.

28,750 kgm. according to the theoretical formula, or 27,470 kgm.
according to the indications of the crusher — values higher than
those of all known explosives. In fact, nitroglycerin gave only
12,376 kgm., and gun-cotton 9825 kgm.

It is the immensity of this pressure, combined with its sudden
development, which explains the part played by mercury
fulminate as priming.

Silver fulminate presents very similar properties ; but it is
much more sensitive, and therefore more dangerous.

§ 3. MERCURY FULMINATE MIXED WITH NITRATE.

1. Suppose, now, mercury fulminate mixed with potassium
nitrate, the mixture corresponds to the formula —

that is to say, 284 grms. of fulminate to 84*2 of saltpetre ; in
all, 368*2 grms. Or for 1 kgm. of the mixture, 229 grms. of
saltpetre and 771 grms. of fulminate.

In practice a third of saltpetre, that is, an excess, is em-
ployed. Antimony and lead sulphides are also added.

2. The heat liberated is -f- 224 Gal, the mercury liquid ;
4- 209'6 Cal., the mercury gaseous ; or, for 1 kgm., + 609 Gal.
or + 567 Cal.

3. The reduced volume of the gases = 64'9 litres, or (gaseous
mercury) 87*2 litres for 1 equiv. ; or, for 1 kgm., 176 litres or
257 litres.

4. The permanent pressure = - - — -i, with the usual re-

servation.

, ™ ., a , 4380 atm.

5. The theoretical pressure = -- TTTTT*

n — O'lJ

It will be seen that it is less by about a third than the
pressure corresponding to pure fulminate. Further, the presence
of the nitrate diminishes the rapidity of inflammation and the
violence of the shock. On the other hand, it gives more
expansion to the flame.

§ 4. MERCURY FULMINATE MIXED WITH CHLORATE.

1. The reaction is the following (exact combustion) —
3C2N202Hg + 2KC103 = 6C02 + 3N2 + 3Hg + 2KC1;

that is to say, 284 grms. of fulminate to 81'7 of chlorate ; in all,
356*7 grms. ; or, for 1 kgm. of the mixture, 223 grms. of chlorate
and 777 grms. of fulminate.

2. The heat liberated is + 258'2 Cal. for 1 equivalent, or

DIAZOBENZENE NITRATE. 471

-f 242-8 CaL (gaseous mercury); or, for 1 kgm., 706 Cal., or
663 Cal.

3. The reduced volume of the gases = 67 litres, or 89*2 litres
(gaseous mercury) ; or, for 1 kgm., 183 litres, or 244 litres.

4. The permanent pressure - - — —' with the usual reserva-
tion.

K m, .* A , 6830 atm.

5. The theoretical pressure = — — •

n — Oil

It is very nearly the same as that of pure fulminate. Potas-
sium chloride lessens the effects of the shock ; but potassium
chlorate renders the mixture very sensitive. Accidents, there-
fore, frequently ocour in factories when this mixture is being
prepared.

§ 5. DIAZOBENZENE NITRATE.

1. The properties and analysis of this body, as well as the
study of its explosive decomposition, have been set forth (p. 291).
We shall only repeat the following figures.

The formula is —

C6H4N2NO3H = 167 grms.

2. The formation from the elements absorbs — 47'4 Cal.

3. The total combustion —

2C6H4N2N03H + 230 = 12C02 + 5H20 + 6N,

liberates -f 782 '9 Cal. at constant pressure (liquid water). No
attempt has been made to study the effects of the combustion of
diazobenzene by the oxidising bodies. The mixture with these
bodies would, moreover, present great difficulties, owing to the
sensitiveness of the dry substance and its immediate decomposi-
tion by water.

4. The explosive decomposition yields complex products,
which vary with the conditions. They have been noticed
(p. 293).

5. Heat liberated. The decomposition having been effected
by the incandescence of a platinum wire, at a low density of
charge, it liberated -f 114'8 Cal. at constant volume for 1 equiv.,
or 687'7 Cal. for 1 kgm.

6. Gaseous volume. There was produced at the same time
136'6 litres of gas (reduced volume) for 1 equiv., or 817'7 litres
for 1 kgm.

7. The theoretical pressure = - 7- • It is higher than

n — O'Oo

that of the fulminate at the unit of weight, and approaches that
of the most powerful substances.

472 DIAZO COMPOUNDS AND OTHERS.

8. Let us compare these theoretical results with the experi-
mental measurement of the pressures. M. Vieille and the
author obtained with a crusher the following figures :—

Density of Weight of Pressures in k^ms.

charge. the charge. per square centimetre.

0-1 ... 2-37 grins. ... 990 kgm.

0-2 ... 4-74 „ ... 2317 „

0-3 7-11 „ ... 4581 „

In the last experiment made with diazobenzene nitrate, this
body filled the whole of the vacant space, and the steel tube
was cracked. This points to local effects, which may have
slightly affected the results. The recent researches of Sarrau
and Vieille on the calibration of the " crushers " tend to
reduce by half the absolute value of the pressures for substances
having so sudden an explosion, but without changing the
relations.

In any case, the pressures of diazobenzene nitrate are far higher,
actually and theoretically, for the same density of charge, than
those developed by the explosion of mercury fulminate. On
the contrary, the fulminate exploding in its own volume would
develop a far greater pressure (28750 kgm., instead of 7500
kgm.), owing to its great density.

The great activity of diazobenzene nitrate, in any case, renders
it more dangerous.

§ 6. NITRIC ACID ASSOCIATED WITH AN ORGANIC COMPOUND.

1. It has been seen in Chapter III. (p. 396), how the liquefied
oxygenated gases, especially nitrogen monoxide and nitric per-
oxide, when mixed with combustible liquids, form explosive
substances of a very special character. It has been proposed
to prepare similar substances by mixing nitric acid with com-
bustible organic substances. In case of need the mixture may
be made on the spot, the ingredients being separately conveyed ;
it is exploded by a fulminate cap. This is the principle of
Sprengel's acid explosive. In practice, the substances capable
of being mixed with nitric acid are few in number, owing to the
violent oxidising action exerted by this acid on the greater
number of organic substances. Few liquids can be mixed with
it without being attacked, and the pastes formed by imbibition
are also subject to reaction.

In fact, only two mixtures of this kind have been employed,
or rather, specially prepared — the mixture of picric acid (solid)
and nitric acid, which forms a paste ; and the mixture of nitro-
benzene and the same acid, bodies which dissolve each other
reciprocally. It will be seen that it is two already nitrified
bodies which serve as base to the mixtures ; further, that the
second would soon be transformed into crystallised dinitro-

HTTBIC ACID AND yiTBOBEHZEXE. 473

benzene. We will now give the theoretical calculations for the
combustion of these two mixtures, including dinitrobenzene.

§ 7. NITRIC Aero ASD PICRIC Aero.

1. The reaction corresponding to total combustion is,
SCtH^NOJaQ + 13HNO3 = 30CO, + 14H10 + 28N.

2. The proportions by weight are, 229 grms. of picric acid to
164 grms. of nitric acid ; in all, 393 grms. ; or, per kilogramme,
583 grm. of picric acid and 417 grms. of nitric acid.

3. The heat liberated will be, for 1 equiv., 318 CaL (liquid
water), or 290 CaL (gaseous water) ; or, for 1 kgm., 809 litres, or
738 CaL

4 The reduced volume of the gases, for 1 equiv., 500 litres, or
659 litre*,

5. The permanent pressure = - rr, with the usual reser-

n — 0*1.3

vation of the limit of liquefaction of carbonic acid.

9450 atm.

6. The theoretical pressure = -- •

n

No experiment has been made with the object of directly
measuring the heat, the volume of the gases, or the pressure ;
a remark which is equally applicable to the following mixtures.

§ 8. NITRIC ACID ASD

1. The reaction of total combustion i

C«H*NO2 + 5HNO, = 6CO, + 5H,O + 6X.

2. The proportions by weight are, 123 grms. nitrobenzene to
315 grms. nitric acid ; in all, 438 grms. ; or, for 1 kgm., 719 grms.
acid and 281 grms. nitrobenzene. It will be borne in mind
that the nitrobenzene is liquid.

3. The heat liberated1 will be, for 1 equiv., 415 CaL (liquid
water), or 365 CaL (gaseous water); or, for I kgm., 947 CaL, or
834 CaL

4 The reduced volume of the gases, for 1 equiv., 201 litres
(liquid water), 373 litres (gaseous water) ; or, for 1 kgm., 459
litres, or 714 litres.

5. The permanent pressure = - 7r7rrf with the usual reser-

n —
vation.

a „, ., _ ^ . 10700 atm.

6. The theoretical pressure =

474 DIAZO COMPOUNDS AND OTHERS.

§ 9. NITRIC ACID AND DINITROBENZENE.

1. The reaction of total combustion is —

C6H4N204 + 4N03H = 6C02 + 4H20 + 3N2.

2. The proportions by weight are, 168 grms. of dinitrobenzene
to 252 grms. of acid ; in all, 420 grms. ; or, for 1 kgm., 400 grms.
of dinitrobenzene and 600 grms. of acid. Note that the
dinitrobenzene is crystallised.

3. The heat liberated will be, for 1 equiv., 387'4 Cal. (liquid
water), or 347'4 Cal. (gaseous water); or, for 1 kgm., 899 Cal.,
or 827 CaL

4. The reduced volume of the gases, for 1 equiv., 201 litres
(liquid water), or 290 litres (gaseous water); or, for 1 kgm.,
479 litres, or 690 litres.

479 atm.

5. The permanent pressure = - .

n — U'lo

10800 atm.

6. The theoretical pressure = > with the usual

n

reservation.

It is nearly identical with that of nitrobenzene. This is as
it should be, the heat liberated and the reduced gaseous volume
being nearly the same for equal weights. With picric acid the
difference is also slight. On the whole, all these mixtures are
very inferior in theory to nitroglycerin or gun-cotton. The
corrosive properties of nitric acid must, moreover, render
difficult the transport of mixtures made beforehand. Lastly,
the stability of such mixtures is more than doubtful. But they
have this advantage, that they can be prepared on the spot and
instantaneously.

§ 10. PERCHLORIC ETHERS.

1. The ethers of the highly oxygenated acids are probably
explosive, but the only ones which have so far been prepared
are the perchloric ethers. These are, in fact, eminently explosive
bodies. The thermal and mechanical properties of methyl-
perchloric ether, the only one corresponding to a total com-
bustion among the ethers of monatomic alcohols, will be given.

2. The formula for methylperchloric ether is the following : —

CH2(C104H).
It corresponds to the equivalent 114*5.

3. The explosive decomposition will be —

CH2(C104H) = C02 + H20 + HC1 + 0.

It will be seen that it sets free an excess of oxygen, like
nitroglycerin and nitromannite.

4. The heat of formation of methylperchloric ether, from the

SILVER OXALATE. 475

elements, may be calculated, granting the formation of this

ether from the acid and alcohol, both dilute.

CH40 (dilute) + C104H (dilute) = CH2(C104H) (dilute) + H20

absorbs - 2'0 CaL,

a value found in general for organic oxacid ethers, and even for
nitric ether itself.
We further have —

C + H4 + 0 + water = CH40 (dissolved)

Cl + 04 + H + water = C104H (dilute)

Reaction ...

66-9

Supposing the solution of the ether in the water to have
liberated + 2*0, the formation of the pure ether then corresponds
to + 65 Cal.

Now, the formation of H20 produces + 69-0. We finally
obtain —

C + H3 + 04 + Cl = CH2(C104H) (dilute) - 4-0 Cal.,
approximately.

5. The explosive decomposition will liberate1 + 175 Cal.
(gaseous water) ; or, for 1 kgm., + 1529 Cal.
, 6. It will produce 781 litres, or, for 1 kgm., 682 litres.

7. The permanent pressure would be calculated from this
figure if reaction did not take place between water and the
acid during cooling (see note).

0 rm. i* 1 i7730 atm-

8. Theoretical pressure = •

n

9. From these numbers the heat liberated is nearly that of
nitroglycerin (1480 Cal. for 1 kgm. and gaseous water). The
gaseous volume is also nearly the same.

It is therefore easy to understand that the theoretical
pressure must also be nearly the same as that of nitroglycerin.
We should have a still more powerful effect by mixing 3
equiv. of methylperchloric ether with 1 equiv. of ethylperchloric
ether, so as to obtain an exact combustion of both ethers. On
the whole, the explosive properties of the perchloric ethers
correspond to those of nitroglycerin and the most powerful
substances. It is this that has led the author to mention here
this class of compounds.

§ 11. SILVER OXALATE.

1. It has been shown (p. 366) that this compound is explosive,
and explodes by shock, or heating, towards 130°. It is even a
shattering body.

1 HC1 and H20 being supposed separated from each other in the gaseous
state. In reality there will be a partial reaction during cooling with formation
of hydrate and corresponding liberation of heat.

476 DIAZO COMPOUNDS AND OTHERS.

2. The following reaction —

C2Ag04 = 2C02 + Ag2,

corresponds to 304 grins, of the substance.

3. It liberates + 29'5 Cal. for 1 equiv., or + 97 Cal. for 1 kgm.

4. The reduced volume of the gases is 44'6 litres for 1 equiv.,
or 114 litres for 1 kgm.

114 atm.

5. The permanent pressure = fwv?* W1^n * e usuai

reservation.

712 atm.

6. The theoretical pressure = T-TJ.

n — 0*06

This pressure is much less than that of the explosives
hitherto examined. However, owing to the great density of
the salt, it would be nearly quadrupled, if the latter exploded
in its own volume, which accounts for the shattering character
of the compound.

§ 12. MERCURY OXALATE.

1. This is a white, heavy, hard powder, which does not
explode by shock, but which explodes feebly by heating.

2. The reaction —

C2Hg04 = 2COS + Hg,

corresponds to 288 grms. of matter.

3. It liberates + 17 '3 Cal. per equivalent (liquid mercury),
or + 1-9 Cal. (gaseous mercary) ; or. for 1 kgm., + 60 Cal., or
6-6 Cal.

4. The reduced volume of the gases is, for 1 equiv., 44'6 litres
(liquid mercury), or 6 6 '9 litres (gaseous mercury) ; or, for 1 kgm.,
155 litres, or 227 litres.

5. The permanent pressure = -s^r-j with the usual

n - 0'05
reservation.

6. Theoretical presure =300 atm'

n

This pressure is very small compared with the other explosive
substances, which explains why mercury oxalate explodes so
feebly, and why the mixture of mercury oxalate with the
fulminate, which is produced when the manufacture is
defective, greatly lessens the properties of the fulminate.

( 477 )

CHAPTEK X.

POWDEES WITH A NITRATE BASE.
§ I-

1. BLACK powder consists of a mixture of saltpetre, sulphur,
and charcoal. According to the relative proportions of these
three ingredients, there is obtained — service powder, in which the
greatest possible strength is sought for ; sporting powder, in
which facility of inflammation and combustion are aimed at ;
and blasting powder, for which the most copious production of
gas is desired. Even the proportions of the ingredients of each
of these powders vary with different nations between very wide
limits.

Few substances have been more studied than powders of
this kind, and there is a copious literature on this subject. It
is not intended to give here a detailed examination of them,
which may be found in a more complete manner in the
" Treatise " by Piobert, in the " Traite sur la Poudre, par
Upmann et Meyer " (traduit et augmente par Desortiaux), as
well as the long and important pamphlets written by Bunsen
and Schisckhoff, Linck, Karolyi, and especially by Noble and
Abel, Sarrau, Vieille, Sebert, etc. Here we shall confine our-
selves to examining the various powders from the point of view
of the chemical reactions developed by their combustion, as
well as the heat liberated, and the volume of the gases pro-
duced by these reactions. The results of theory with those of
experiment will be compared, as far as is permitted, by the
following circumstances, which are difficult to introduce into a
precise calculation : —

1st. The charcoal employed is not pure carbon. It contains
only 75 or 80 per cent, of this element, 2 per cent, of
hydrogen, 1 or 2 per cent, of ash, and 15 or 20 per cent, of
oxygen.

2nd. Powder contains a little moisture, the quantity of which
varies, being, however, generally nearly 1 per cent.

478 POWDERS WITH A NITRATE BASE.

3rd. The mixture of sulphur, saltpetre, and charcoal is never
absolutely intimate, and undergoes continual variations during
the course of the operations.

4th. The combustion is never total, small quantities of nitre
and sulphur principally escaping the reaction owing to the lack
of homogeneousness. The saltpetre itself, under the influence of
the high temperature of explosion, tends at first to yield the
nitrites, then more and more stable compounds (hyponitrites,
potassium peroxide, etc.) still imperfectly known.

5th. The metallic vessels (iron, copper), in which the opera-
tions are carried out, are attacked, with the formation of
metallic sulphides, single and double sulphides resulting from
the association of the former with potassium sulphide. Never-
theless the theoretical calculations, however imperfect their
relation to practical conditions may be, offer the advantage of
indicating the maximum limit of the effects which we may
hope to attain, and the direction which should be given to
experimental inquiry for this end. In order to explain more
clearly the chemical phenomena, the fresh experiments will be
given which the author has lately made on various questions
relating to the theory of the reactions developed during the
explosion of service powder, such as the reactions between the
sulphur, the carbon, their oxides and salts (§ 2). The decompo-
sition by heat of the alkaline sulphides (§ 3). The decomposi-
tion by heat of the alkaline hyposulphites (§ 4). The
measurement of the heat of combustion of the charcoal
employed in the manufacture of powder (§ 5).

These preliminary notions having been gained, we shall
study —

1st. Powders corresponding to an exact combustion (§ 6).

2nd. Powders with an excess of combustible, such as service
powder properly so called, sporting and blasting powder (§ 7).

3rd. Powders formed of nitrates other than potassium,
which are employed for industrial purposes in particular
cases (§ 8).

§ 2. EEACTIONS BETWEEN SULPHUR, CARBON, THEIR OXIDES AND

SALTS.

1. The study of the products of the explosion of powder led
the author to make some observations on the reciprocal action
of sulphur, carbon, their oxides and salts. The operations were,
in some cases, carried out by means of the electric spark, and
in others by means of a red heat. In both cases there are
foreign energies which intervene in the chemical actions
properly so called, energies developed by electricity or heating,
especially successive decompositions, dissociations, and changes
of molecular states (polymerised carbon changed into gaseous

DECOMPOSITION OF SULPHUROUS ACID. 479

carbon, gaseous sulphur reduced to its normal molecular weight,
instead of sulphur of triple density volatilisable towards 448°).

It may be first noted that sulphur burning in dry oxygen
produces sulphurous acid, mixed with a considerable proportion
of anhydrous sulphuric acid. Sulphur vapour directed upon
charcoal at a red heat combines with it, producing carbon
sulphide.

Carbon burnt in oxygen produces carbonic acid, always
mixed with a little carbonic oxide.

Carbonic acid directed upon red-hot charcoal is changed into
carbonic oxide ; but the transformation is never complete.

2. Decomposition of the sulphurous gas. A series of electric
sparks decompose sulphurous acid gas into sulphur and sulphuric
acid (Buff and Hoffman)—

3S02 = 2S03 + S.

Operating in a sealed tube without mercury, with platinum
electrodes, several hours are needed to decompose the half of
the gas, and decomposition ceases at a certain point, as was
observed by Deville. It does not yield free oxygen, but a
portion of the sulphur unites with the platinum.

The greater portion of the sulphur forms with anhydrous
sulphuric acid a special viscid compound, which, moreover,
absorbs a certain quantity of sulphurous gas. This compound
is the real medium of the reaction. Being inversely decom-
posable, the tension of sulphurous and sulphuric gases which it
gives off limits the reaction.

3. Decomposition of the carbonic oxide. Carbonic oxide under
the influence of the spark, or even of a white heat, partly
decomposes into carbon and carbonic acid —

2CO = C02 -f C.

But the reaction is limited to a few thousandth parts. It was
found that it takes place at a bright red heat, and even at the
temperature of the softening of glass. The carbon is deposited
at the point where the porcelain tube issues from the furnace,
and undergoes a lowering of temperature, even without having
recourse to the artifice of the hot and cold tube. It may be
still better shown by placing fragments of pumice-stone in this
region of the tube. A trace of carbonic acid produced at the
same time may be observed in the gases collected by adopting
certain precautions.

Though so slight and inappreciable, this reaction is, never-
theless, of great importance ; for it intervenes, together with the
dissociation of the carbonic gas into carbonic oxide and oxygen,
in the reduction of the metallic oxides and in a great number of
other reactions, brought about by heat. We will now place
sulphur and carbon, whether free or combined, together.

4. Sulphurous acid gas and carbon (baker's embers calcined

480 POWDERS WITH A NITRATE BASE.

beforehand for several hours at a white heat, in a current of dry
chlorine, then cooled in a current of nitrogen). Operating in a
porcelain tube at a clear red heat, a gas was collected, formed of
carbonic oxide, carbon oxysulphide, and disulphide in the
following proportions : —

4S02 + 90 = 600 + 2COS + CS2,

a small quantity of sulphur being sublimed at the same time.
All this is intelligible, on the supposition that the carbon took
the oxygen,

S02 + 20 = 200 + S2,

and that the gaseous sulphur, being set free, combined for its
own part partly with the carbon and partly with the carbonic
oxide.

In these experiments the carbon contained in the tube
becomes covered with a sort of sooty coating, and undergoes a
remarkable disaggregation, which divides it into small frag-
ments, according to three rectangular planes ; phenomena which
appear to be due to the state of dissociation peculiar to
disulphide, which is partly destroyed at the same temperatures
at which it is formed according to former observations.1

5. Carbonic add and sulphur. The experiment was carried
out at two different temperatures.

1st. The sulphur is raised to the boiling point in a glass
retort, and a slow current of dry carbonic gas is passed through
it. This reaction has been given as producing carbon oxysul-
phide. This is not the case, as the author has assured himself
by most careful tests. What may have occasioned the error
are the traces of sulphuretted hydrogen, which even the best
purified sulphur always liberates when heated.

In reality, sulphur in a state of ebullition is without action
on dry carbonic gas.

2nd. If carbonic gas mingled with sulphur vapour be passed
through a porcelain tube at a clear red heat, a reaction, very
slight it is true, but unquestionable, may, on the contrary, be
observed.

Thus the gas liberated contained, out of 100 volumes, 2'5 vols.
of gases other than carbonic acid, viz. —

1 vol. COS; 1 vol. CO ; 0-5 vol. S02.

These small quantities seem to be attributable not to the action
proper of sulphur on carbonic acid, but to the previous dissocia-
tion of the latter into carbonic oxide and oxygen ; a dissociation
which, moreover, is but slight under these conditions, but which
the presence of sulphur, which unites at one and the same time
with the oxygen and carbonic oxide, tends to render manifest.

1 " Annales de Chimie et de Physique," 4" s£rie, torn, xviii. p. 169.

SULPHUROUS ACID GAS AND CARBONIC OXIDE. 481

6. Carbonic acid and sulphurous acid gas. The two gases were
mixed in equal volumes, passed into a glass tube, which was then
sealed. After two hours and a half of strong sparks the author
observed —

Diminution of volume 19 vols.

S02 31 „

C02 30 „

CO ... ... 20 „

Each of the gases was decomposed for its own part. The
oxygen resulting from the dissociation of the carbonic acid was
condensed, uniting with the sulphurous acid under the form of
sulphuric acid.

The sulphurous acid gas here seems more stable than the
carbonic acid gas, contrary to what might have been expected.

7. Sulphurous acid gas and carbonic oxide. 1st. The mixture
made in equal volumes was slowly passed through a very small
porcelain tube at a clear red heat. There was collected —

Intermediate gas. Final gas.

S02 47 vols. ... 37 vols.

C02 9 „ ... 20 „

CO ... ... 44 „ ... 43 „

Sulphur was formed. Neither carbon oxysulphide nor carbon
disulphide was present in any considerable proportion.

Thus the carbonic oxide reduced the sulphurous acid gas —

2CO + S02 = 2C02 + S.

But the reduction remained incomplete, as the experiment
made with carbonic acid permitted of foreseeing.

2nd. Two vols. of carbonic oxide and one vol. of sulphurous
acid gas were mixed and passed into a glass tube provided with
platinum electrodes, the tube being then closed. A series of
sparks was passed through it. The following are the results
of both trials : —

After After

half an hour. two hours.

Diminution 14 vols. ... 28 vols.

S02 20 „ ... 6 „

C02 18 „ ... 9 „

CO 48 „ ... 57 „

No sulphur nor carbon oxysulphide. Here again we see the
reduction of the sulphurous acid by the carbonic oxide. But,
and it is a remarkable circumstance, a considerable portion of the
former gas is destroyed for its own part without yielding its
oxygen to the carbonic oxide, and giving the same compound of
sulphur, sulphurous acid, and sulphuric acid already described,
and which condenses on the walls of the tube.

3rd. The same experiment, repeated over mercury, with
strong sparks, in the space of four hours caused the total

2 I

482 POWDERS WITH A NITRATE BASE.

destruction of the sulphurous acid gas, producing a final mixture
containing —

CO, 24vols.

CO 75 „

0 1 „

Under these conditions the mercury absorbs the anhydrous
sulphuric acid, and eliminates it, forming a sub-sulphate.

8. Saline compounds. All the alkaline oxysalts of sulphur
being reduced to the state of sulphate and sulphide towards
a red heat, special attention was paid to these two salts, together
with potassium carbonate, and they were allowed to act at a red
heat on sulphur, carbon, and their gaseous oxides. The salts
were contained in elongated vessels arranged in a porcelain
tube.

9. Potassium sulphate and carbonic acid. At a 'bright red
heat, no action took place. At a higher temperature it would
doubtless be important to take into account the dissociation of
the sulphates observed by Boussingault.

10. Potassium sulphate and carbonic oxide. At a bright red
heat the sulphate was charged into sulphide, or rather into
poly sulphide,1 containing some flakes of carbon, and a mixture
of carbonic acid and carbonic oxide was collected, the relative
proportion of the former gas varying between four-fifths and
the half, according to the speed of the current and the tempera-
ture.

The principal reaction here is —

S04K2 + 400 a K2S + 4002.

There is a trace of carbonate.

11. The reducing action of carbon on potassium sulphate is
so well known that it was not deemed necessary to reproduce it.

12. Potassium sulphate and sulphurous acid. There is no
action at a bright red heat.

13. Potassium sulphate and sulphur. Sulphur may be evapo-
rated in presence of potassium sulphate, provided the tempera-
ture be carefully kept below a red heat.

On the contrary, in a red-hot porcelain tube, sulphur vapour
reduces potassium sulphate, producing polysulphide and
sulphurous gas —

S04K2 + 4S = K2S3 + 2S02.

This transformation was never total. It seems, moreover, to
represent the last term of a series of changes, in which the lower

1 The constant formation of polysulphide in the actions caused by heat
which yield sulphur, has been remarked by Gay-Lussac, Berzilius, and Bauer.
It is connected with some imperfectly known reaction, such as the formation
of an oxy sulphide of potassium.

VARIOUS DECOMPOSITIONS OF POTASSIUM CARBONATE. 483

oxysalts of sulphur intervene ; compounds of which, in fact,
traces may be found by moderating the action.

The well-known reaction of carbon disulphide on potassium
sulphate, which it changes into sulphide, may be roughly
regarded as the sum of those of sulphur and carbon. But
according to Schone it is also preceded by intermediate com-
pounds, such as sulphocarbonate.

14. Sulphur and potassium carbonate. This is among the
number of reactions which have received the greatest amount
of investigation. At a red heat it yields polysulphide, sulphate,
and carbonic acid —

4C03K2 + 168 = 3K2S5 + S04K2 + 4C02.

But these are also the extreme terms of successive reactions,
hyposulphite, for instance, forming at 250°, according to
Mitscherlich.

15. Carbon and potassium carbonate. This reaction yields at a
red heat carbonic oxide and potassium oxide, not without there
being formed various secondary compounds, such as the acety-
lides. The dissociation of potassium carbonate also intervenes
(Deville).

16. Potassium carbonate and sulphurous acid. If the gas
passes rapidly, the red-hot salt changes into sulphate, with only a
trace of sulphide. If the current is slow the sulphide increases.

17. Carbonic acid and sulphite. Sulphate, polysulphide,
and a little carbonate are formed. Metasulphite (anhydrous
bisulphite) gives the same products.

18. Carbonic acid and potassium polysulphide. In a red-hot
tube some sulphur is sublimed, and the gas liberated contains
about three per cent, of a mixture of carbonic oxide, sulphurous
acid, and oxysulphide. It is the same reaction as that of sulphur
on carbonic acid, which is attributable to the dissociation of
the latter compound. A small quantity of alkaline carbonate
appears also to result from this dissociation ; the oxygen
supplied by the latter concurring with the excess of carbonic
acid to displace the sulphur.

19. From these facts, there result several consequences con-
nected with the study of the reactions produced during the
explosion of powder.

For example, if potassium carbonate subsists in any consider-
able amount in presence of sulphur resulting from the dissocia-
tion of the simultaneously produced polysulphide, it is apparently
because both salts do not form at the same spot of the substance
in ignition. The same sulphur would also attack the potassium
sulphate if both bodies were kept together at the same point.
The carbonic oxide would also destroy the sulphate if it were
formed at the same spot, or if it remained for some time in
contact with the melted salts, etc.

2i2

484: POWDEKS WITH A NITEATE BASE.

Hence we see how the more or less homogeneous character
of the initial mixture, the greater or less duration of combustion,
and the varying rapidity of cooling may cause the nature of the
final products to vary within very wide limits. There will be
occasion to return to these problems, which have a great im-
portance in practice.

20. Hitherto we have examined the final products of reactions
taking place at a red heat. In these reactions neither sulphite
nor hyposulphite is found, because both these classes of salts
are decomposed below this temperature.

§ 3. DECOMPOSITION OF THE ALKALINE SULPHITES BY HEAT.

1. We shall distinguish between the neutral sulphites and
the metasulphites formerly called anhydrous bisulphites.

The neutral potassium sulphite may be decomposed into
sulphate and sulphide, according to theory —
4S03K2 = 3S04K2 + K2S.

2. A special study was made of this decomposition, which
forms one of the most striking distinctions between normal
sulphites and metasulphites.

It was found that the accurate analysis of the products
verifies the above equation in the most precise manner, when
dry sulphite is brought to a dull red heat in an atmosphere of
nitrogen.1 Several estimations by iodine, made with the
requisite precautions, absorbed, for instance, 31'5 c.c., 32'5 c.c.,
30*8 c.c. of the iodine solution ; while the original salt took up
126 c.c. The quarter of the latter figure is just 31/5.

No sulphurous acid is liberated, contrary to an assertion
made by Muspratt, which would require an inexplicable setting
free of potash.

The decomposition of the sulphite does not commence at
450°, the salt remaining intact till towards a dull red heat, and
even at that temperature needing a certain time to be completely
transformed.

3. It is well known that two series of sulphites are dis-
tinguished: the neutral, and the acid sulphites, supposed to
correspond to the composition of a dibasic acid ; viz. S02K2O,
and S02KHO, salts which have been studied by Muspratt,
Rammelsberg, and De Marignac.

These investigators have further discovered an anhydrous
bisulphite: (S02)2K20.

In following up his researches on the products of the explo-
sion of powder, the author has been led to measure the heat of

1 Only the sulphite contains, as is always the case, some small amount of
a red poly sulphide, a compound which is met with under all the conditions
in which the monosulphide alone should be formed.

POTASSIUM METASULPHITE. 485

formation of these various potassium sulphites, and has found,
not without surprise, that the so-called anhydrous bisulphite,
far from belonging to the same type as the other sulphites,
constitutes in reility, by its chemical reactions and thermal
properties, a distinct and characteristic type of a new saline
series, viz. the mztasulphites, as distinct from the sulphites
properly so called, as the metaphosphates and pyrophosphates,
for example, are from the normal phosphates.

Pure potassium metasulphite is obtained by saturating with
sulphurous acid gas a concentrated solution of potassium
carbonate, either warm or even cold, and by drying at 120° the
salt which separates by crystallisation. The anhydrous salt
already described under the name of anhydrous bisulphite by
Muspratt and Mirignac corresponds to the formula S2O5K2.1
This salt is distinguished by its heat of formation, its stability,
its tendency to form hydrates, and even solutions distinct from
those of the normal bisulphite, and, finally, by its decomposi-
tion by heat. In leed, the normal bisulphite prepared in dilute
solutions by the saturation of the neutral sulphite by sulphurous
acid soon change 5 state in the liquid itself. It is dehydrated,
and becomes metasulphite, liberating +2*6 Cal., a fact which
accounts for the preponderance of the metasulphite and its
definite formation in solutions.

The dissolved potash, moreover, reduces the metasulphite to
the state of neutral sulphite.

Without dwelling any further here upon the characteristics of
the metasulphites, we shall describe the action which heat has
upon this one as entering into the scope of the present work.

4. Decomposition of metasulphite by heat. The action of heat
forms one of the most striking characteristics of potassium
metasulphite. In fact, dry metasulphite does not lose sulphurous
acid even at 150°.

However, if it be brought to a dull red heat, it liberates
sulphurous acid, but without regenerating a corresponding
amount of neutral sulphite, and even changing in a well-defined
and entire manner into potassium sulphate and sublimed
sulphur when the reaction is carefully carried out —

2S205K2 = 2S04K2 + S02 + S.

This equation has been verified by accurate measurements.
These are characteristic. Sulphurous acid is actually liberated.
The volume of this gas indicated by the above formula should
be the half of that corresponding to the normal reaction of a
bisulphite, such as

S205K2 = S03K2 + S02.

1 " Comptes Rendus des stances de 1'Acade'mie des Sciences," torn. xcvi.
p. 142, and especially p. 208.

486 POWDERS WITH A NITRATE BASE.

Further, the neutral sulphite should be decomposed in its
turn into sulphate and sulphide.

'Now, the author has ascertained, operating in a very confined
space filled with dry nitrogen, with a progressive heating, and
collecting the gases as they were formed, to prevent their
further reactions on the remaining salts —

1st. That the volume of the sulphurous gas is exactly the
half of the volume required by the second formula (normal
bisulphite).

2nd. That the salt residuum consists of almost pure sulphate,
only exercising an insignificant action on an iodine solution.

The transformation is perfectly definite when the metasulphite
alone is heated. In a current of an inert gas, such as nitrogen,
or even in a considerable space filled with this gas, metasulphite
commences to be decomposed into sulphurous acid, which is
carried off, and neutral sulphite, which afterwards yields a
certain amount of sulphide. But these complications may be
avoided by operating as has been described. These reactions
characterise metasulphite most distinctly.

§ 4. DECOMPOSITION BY HEAT OF THE ALKALINE HYPOSULPHITES.

1. On the occasion of the discussion which was raised some
years since on the composition of the products of explosion of
powder, the author showed that potassium hyposulphite, shown
by former analyses to the extent of 34 per cent., does not in
reality pre-exist in any appreciable proportion among these
products, but is introduced during the analytical manipulations.
This demonstration is based on the fact that potassium hypo-
sulphite is entirely destroyed near 500°, a temperature far lower
than that of the explosion of powder. It was finally accepted,
not without opposition at the outset, by Noble and Abel, after
the experiments of Debus, who proved that the hyposulphite
found in the analysis resulted from the use of cupric oxide to
eliminate the alkaline polysulphides.

The author since proved the same with zinc oxide. This
oxide, acting on potassium polysulphide, yields besides zinc
sulphide some hyposulphite, sulphate, and hyposulphate, the
relative proportion of sulphur contained in the three latter
bodies being 1118 and 8 in one experiment. The presence of
the hyposulphite in particular had escaped notice previously ;
it is probable that this body is produced also with cupric oxide.
It is even formed, though only in small quantities, when
polysulphide is destroyed by zinc acetate.

2. These facts being ascertained, it seemed desirable to
determine more accurately the temperatures of decomposition
of the alkaline hyposulphites. The experiments were made

DECOMPOSITION OF ALKALINE HYPOSULPHITES. 487

with salts dried in a progressive manner, at first, in vacuo,
then at 150°, conditions under which they undergo no altera-
tion.

If, on the contrary, they are suddenly heated to 200°, decom-

rition begins under the influence of the water vapour supplied
the hydrates.

When they are further heated, it is necessary to operate in an
atmosphere of pure and dry nitrogen, the least trace of oxygen
causing an oxidation and sublimation of sulphur. The decom-
position of the hyposulphites is shown by analysis by means of
iodine, which should be reduced to the half, according to the
theoretical formula —

4S203K2 = 3S04K2 + K2S5.

The first body takes I2, the second body only I.

The operations were carried out in an alloy bath, the tempera-
tures being given by an air thermometer. With standardised
solutions containing a known weight of iodine, the following
results were obtained : —

Amount of standard
iodine used.
Div.

S203K2 according to theory 323

„ dried in vacuo 323

„ heated to 255° 325

„ 310° ten minutes 320

„ 310° an hour 323

„ 430° for a short time 320

„ 470° 160

490° 161

S203Na2 theoretical (another standardised solution) 632

dried at 150° 632

„ 200° 634

„ 255° 634

„ „ 331° ten minutes 633

331° an hour 633

„ 358° 632

„ 400° 569

„ 470° 375

,, 490° ... ... 381

It results from these analyses that the potassium and sodium
hyposulphites resist without alteration up to about 400°. The
soda salt commences to alter at this temperature ; the potash
salt resists a little longer, up to about 430°, at least if the dura-
tion of the heating be not prolonged too much, otherwise it
commences to change. At 470° the decomposition is total. It
is strictly theoretical in the case of the potash salt. In that of
the soda salt there occurs partial sublimation of sulphur, and the
strength found is too high by about 8 per cent, (on 50).

488 POWDERS WITH A NITEATE BASE.

§ 5. ON THE CHARCOALS EMPLOYED IN THE MANUFACTURE OF

POWDER.

1. In equations relative to the combustion of powder, pure
carbon is usually considered ; but in reality the charcoal should
be taken with its true composition, for the results calculated on
the supposition that the oxygen is in the state of water whilst
carbon and hydrogen would be free, are not certain, owing to
the complex composition of charcoal and the thermal excess
which it liberates in its total combustion.

2. It might be imagined that, in order to take this fact into
account in calorimetric calculations, it would be sufficient to
calculate the formation of carbonic acid and carbonic oxide
from amorphous carbon —

C + 02 = C02 liberates + 48'5 Cal.,

instead of -f 47 Cal. for diamond carbon.

But even this way of reckoning gives figures which are too
low, because the charcoal used in the manufacture of powder is
not pure carbon, but contains hydrogen and oxygen nearly in
the proportions of water. For instance, the charcoal of the
powder studied by Bunsen contained in 11*0 parts —

0= 7'6; H = 0-4; 0 = 3-0.

Now, the combustion of the hydrocarbons yields more heat
than that corresponding to the carbon they contain, the hydro-
gen and oxygen being supposed in the state of pre-existing
water, that is to say, no longer contributing to the production
of heat. Thus Favre and Silbermann,1 burning bakers' embers
(which contained to 1 grm. of carbon 0*027 grm. of hydrogen),
found 52,440 cal., instead of 47,000 for 6 grms. of carbon burnt,
which makes an excess of more than a ninth, or 906 cal. per
gramme.

3. This is intelligible if it be noted that calcined charcoal is
derived from a carbohydrate, and that the carbohydrates, as the
author pointed out many years ago, yield by their combustion
more heat than the carbon which they contain, deduction being
made of the oxygen and hydrogen in the form of water.

The heat- of combustion of a carbohydrate of the formula
(C6HpOp) is, according to experiment, generally 709 Cal. to
726 Cal. for 72 grms. of carbon.

This would make for the heat of combustion of C = 6 grms.
59 CaL to 61*6 Cal., that is, an excess of more than a fourth of
the heat of combustion of the real carbon of the substance.
When the carbohydrates are dehydrated by heat, a portion of

1 " Annales de Chimie et de Physique," 3e serie, torn, xxxiv. p. 420. 1852.

COMPOSITION OF A CHARCOAL. 489

this thermal excess, that is, a portion of this excess of energy,
remains in the residual carbon.1

Further this latter carbon sometimes retains an excess of
hydrogen which yields, weight for weight, four times as much
heat as carbon.

4. It is hardly possible accurately to estimate the influence of
these complex circumstances, unless by very special analysis
and calorimetric determinations made on the charcoal employed
in the manufacture of a given powder. But it is clear that they
tend to reduce the error committed by assuming in the calori-
metric calculations the weight of the charcoal employed as
equal to the weight of pure carbon, than which it is really lower
by about a fifth. This compensation extends itself even to the
volume of the gases ; since the deficiency in volume of carbonic
acid produced is almost entirely replaced at the moment of the
explosion by the volume of water vapour, resulting from the
hydrogen and oxygen contained in charcoal.

5. With a view to rendering these notions clearer, we shall
give some observations made on the composition of a charcoal
derived from pure lignite. Having had occasion to see, in the
powder factory at Toulouse, some spindle-tree charcoal, pre-
pared with the ordinary precautions, that is to say protected
from the air and at a relatively low temperature, from young
branches containing a considerable quantity of pith, it was
deemed of interest to examine the carbonaceous portion derived
from this pith, a pure and homogeneous substance.

Further, the central position allows of the decomposition of
the substance by heat taking place outside the influence of the
the air and the gases formed by secondary reaction in the dis-
tilling apparatus. Some of the carbonised branches were
obtained, and the charcoal contained in the medullary channel
extracted and examined.

It retained exactly the appearance and structure of the
original pith, except, of course, its colour. In order to analyse
it, it was dried in an oven, and burnt in a current of oxygen,
completing the combustion of the gases by a column of cupric
oxide. Eesults :

(1) Loss at 100° 9-0

This loss is due to water, which can be absorbed by sulphuric
acid. However, there is also produced a trace of carbonic acid,
as was proved, which is doubtless produced by oxidation on
contact with the air, which is worthy of notice, from the low
temperature of the experiment (100°). But the weight is less
than the one-thousandth part, from direct measurements.

(2) Ash 3-5

1 See also the works of M. Scheurer-Kestner, who has found an analogous
excess in the combustion of certain kinds of coal.

490 POWDEKS WITH A NITKATE BASE.

(3) The combustible substance, dried at 100°, contained —

Carbon 73-6, that is, including the saline carbon of the ash 73-9

Hydrogen ..................... 2-2

Potassium ..................... 2-1

Oxygen ..................... 21-8

These numbers may be represented by the following empirical
proportions, C^B^KO^, which require —

C ............ 73-7

H ............ 2-2

K ... ................ 2-0

0 ..................... 22-1

Of course it will be understood that it is not the question here
of a formula properly so called.

These proportions, compared with those expressing the com-
position of cellulose, C^H^xAoo, show that distillation deprives
this substance not only of an excess of water, but also of an
excess of hydrogen, which corresponds with the formation of
methane, CH4, acetone, C3H60, and analogous products. The
charcoal of the pith is therefore not a simple carbohydrate, but
contains a proportion of oxygen higher than that which would
correspond to such a composition.

The proportion of oxygen contained in this charcoal, viz.
22 per cent., is very remarkable, on account of the physical
properties of the substance. We are here in presence of special
compounds having a very high equivalent, but the insolubility
and amorphous state of which prevent their being properly
determined. The author has elsewhere maintained the existence
of these moist and carbonaceous compounds formed by successive
condensations, and of which the various carbons represent the
extreme limit.1

§ 6. TOTAL COMBUSTION POWDERS: SALTPETRE AND CHARCOAL.

1. Two combustible elements being associated with the
combustive, it is easy to imagine an unlimited number of
powders of this kind. We shall consider the three following
cases : — .

(1) Mixture of saltpetre and charcoal.

(2) Mixture of saltpetre and sulphur.

(3) Mixture of saltpetre with sulphur and charcoal in equal
proportions.

1 " TraitS de Chimique organique," p. 384 (1872) ; 2e Edition, torn. i. p. 456
(1881) ; " Annales de Chimie et de Physique," 4e se'rie, torn. xix. p. 143, and
torn. ix. p. 475. The analogy (torn. ix. p. 478) of these compounds with the
metallic oxides obtained by a more or less intense calcination, and which
r epresent products of successive condensation, was also maintained.

SALTPETRE AND SULPHUR. 491

(1) Saltpetre and charcoal. The equation is the following —

4KN03 + 50 = 2K2 C03 + 3C02 + 4K

It corresponds to" 101 grms. of nitre and 15 grms. of carbon; in
all, 116 grms. ; or, for 1 kgm., 129 grms. of charcoal and 871
grms. of nitre.

1. This being admitted, the heat liberated will be, for 1 equiv.
of potassium nitrate employed to burn carbon, at constant
pressure + 90*7 Cal., or + 91-2 Cal. at constant volume; or, for
1 kgm., 782 Cal. at constant pressure, or 786 Cal. at constant
volume.

2. The reduced volume of the gases = 27*9 litres ; or, for
1 kgm., 240-5 litres.

240-5 atm.

3. Permanent pressure = - • , with the usual reservation

n — 0*27

relative to the liquefaction of carbonic acid.

(2) Saltpetre and sulphur. The equation is the following : —

2KN03 + S2 = K2S04 + S02 + 2K

It corresponds to 101 grms. of nitre and 32 grms. of sulphur ;
in all, 133 grms. ; or, for 1 kgm., 241 grms. of sulphur and 759
grms. of nitre. The sulphur may be considered as pure, in
practice.

1. The heat liberated will be, for one equivalent, 87'0 Cal.
at constant pressure, 8 7" 5 Cal. at constant volume; or, for
1 kgm., 654 Cal. at constant pressure, 658 Cal. at constant
volume.

2. Keduced volume of the gases = 22*3 litres for the equi-
valent ; or 168 litres for 1 kgm.

168 atm.

3. The permanent pressure = — — , with the reservation

n —0-25

of the liquefaction limit of sulphurous acid.

4. Theoretical temperature at constant volume, 3870°.

2545 atm.

5. Theoretical pressure = — — .

n — 0'2o

Note that under the conditions attending the use of black
powder the sulphurous acid shown by the above equations does
not appear.

(3) Saltpetre, sulphur, and carbon, the latter in equal weights
(black powder with excess of nitre.) The equation of the reaction
is —

10KN03 +3S + 8C = 3K2S04 + 2K2C03 +6C(V
It corresponds to 505 grms. of nitre, 48 grms. of sulphur and

1 Admitting the following specific molecular heats : — C02 = 3'6 ; N = 2-4 ;
C03K2 = 151-0 ; S04K2 = 16-6 ; S02 = 3-6 (see p. 141).

492 POWDERS WITH A NITRATE BASE.

48 grms. of carbon ; in all, 601 grms. ; or, for 1 kgm., 840 grms.
of nitre, 80 of sulphur, and 80 of charcoal.

1. The heat liberated will be for the equivalent weight,
479-6 CaL at constant pressure, or 481'2 Gal. at constant
volume; or, for 1 kgm., 798 Cal. at constant pressure, and
801 Cal. at constant volume.

2. Eeduced volume of the gases = 66'9 litres for the equiva-
lent weight ; or 111/3 litres for 1 kgm.

3. The permanent pressure = _ ', with the reservation of

Ti-0'27

the limit of liquefaction of carbonic acid.

4. Theoretical temperature, 4746°.

2046 atm.

5. Theoretical pressure = — — .

n — 0'27

6. The heat produced slightly exceeds that of sporting and
service powder. But the volume of the permanent gases
developed by the latter is double that corresponding to a
complete combustion. Hence the pressure is far lower for
powder with excess of nitre than for -sporting and service
powders.

The complete combustion effected by an excess of nitre is
therefore not advantageous from the point of view of the effects
developed by the pressure of powder. This inferiority of
powder with an excess of nitrate had already been discovered
in practice.

7. However, it is worthy of remark that the compounds
which are formed by the complete combustion of a powder with
an excess of nitre, viz. potassium sulphate and carbonate, are
also noticed by writers on the subject as principal products
in the deflagration of sporting and service powder, as well as in
that of powders the most different in appearance, such as
blasting powder, which is very rich in sulphur, and powder
with an excess of charcoal. Although the products vary a little
with the conditions of deflagration, potassium sulphate and
carbonate have almost always been observed, and this is the more
important, as these two salts do not figure in the theoretical
equations formerly admitted.

§ 7. SERVICE POWDERS.

1. We shall divide the study of service powders into four
sections, comprising —

(1) The general properties of powder.

(2) The products of combustion of powder.

(3) The theory of combustion of powder.

(4) The comparison between theory and observation.

SERVICE POWDER. 493

(1) General Properties of Powder.

1. The proportion, " six, ace and ace," that is to say—

Saltpetre 75'0

Sulphur 12-5

Charcoal 12-5

has never been far departed from in France. An excess of char-
coal and of nitre increases the strength ; an excess of sulphur
has been found favourable to the preservation of powder. The
presence of sulphur, moreover, lowers the initial temperature of
the decomposition of the substance and regulates it. The
actual proportions in France are —

Nitre. Sulphur. Charcoal.

Ordnance powder 75 12'5 12-5

Old coarse grained powder ... ... 75 10 15

Rifle powder, Class B 74 10'5 15-5

Rifle powder, Class F 77 8 15

Austria 75'5 10 14-5

United States, Switzerland 76 10 14

Holland 70 14 16

China 61-5 15-5 23

Prussia 74 10 16

England, Russia, Sweden, Italy 75 10 15

The composition 75, 12*5, 13*5, corresponds practically to the
relations

2KN03 -f S + 30,

or 101 -f 16 +18 ; in all, 135 grms. ; or, for 1 kgm., 748 grms. of
saltpetre, 118*5 grms. of sulphur, and 133*5 grms. of carbon.

2. The temperature of inflammation of powder was fixed at
316° by Horsley. This temperature varies with the process of
heating. It may fall to 265°, according to Violette.

If the heating takes place slowly, the sulphur melts, causes
the aggregation of the grains, then gradually vaporises and
may even be almost entirely sublimed. The nature of the
charcoal has great influence in this case ; some wood charcoals
yielding carbonic acid on contact with the air at 100° and even
below p. (489).

It is, therefore, natural to suppose that such charcoals, if their
surface be not completely covered by sulphur and saltpetre
through a very intimate mixture, may become more and more
rapidly oxidised at a temperature which moreover goes on
increasing owing to the oxidation. They may even take fire,
especially if the mass be so large that the heat produced by
this oxidation has not time to dissipate itself. We may in
this way account for certain accidents caused by spontaneous
inflammation of heaps of powder dust.

3. The inflammation of powder is caused by the shock of
iron on iron, iron on brass or marble, brass on brass, quartz on
quartz, less easily by iron on copper, or copper on copper. It

494 POWDEKS WITH A NITRATE BASE.

is caused even by lead on lead, or lead on wood ; seldom by
copper on wood, never by wood on wood, without of course the
interposition of gravel.

4. Powder absorbs a certain amount of moisture, principally
owing to the hygroscopic properties of the charcoal and the
impurities of the saltpetre; this amount varying from 0'5 in
dry magazines to 1*20 in damp magazines. The proportion of
water thus absorbed may rise to seven per cent, in a saturated
atmosphere, the temperature of which undergoes alternate
changes. When it exceeds a certain limit it causes the
separation of the saltpetre by eventual efflorescence, thus
destroying the powder.

5. The density of powder has been considered from three
points of view :

1st. The absolute density, denned in the sense in which it is
employed by physicists.

2nd. The apparent density of the isolated grains, called real
density.

3rd. The apparent density of unrammed powder, called
gravimetric density (weight of powder at the unit of volume).
The gravimetric density varies from 0'83 to 0'94, according to
the coarseness of the grain.

The so-called real density is found by plunging a given
weight of powder into a given medium of which the variation
of volume is observed. The following substances have been
used: — lycopodium, a solid body in a very fine powder,
essence of turpentine, water saturated with saltpetre, absolute
alcohol, and mercury ; the latter being the only liquid which
can be considered as exercising no dissolving action.

In the tests, it is subjected to a fixed pressure (2 atm.)
during the operation. The results obtained in this way have
only a relative significance.

The following have been found in this manner : —

Ordnance powder 1-56 to 1-72

Rifle powders 1-63 to 1-82

Sporting powder 1-87

The absolute density, measured by the volumometer, is 2*50.

(2) Products of Combustion of Powder.

1. These products are those of the combustion of charcoal
and sulphur by oxygen, modified by the presence of nitrogen
and the reaction between these products and potassium, proceed-
ing from the saltpetre, at the high temperature of combustion.

2. The proportions of the composition of powder are not
those of total combustion, oxygen being wanting ; they therefore
do not correspond to the greatest heat which might be liberated
by the oxidation of the sulphur and carbon by a given weight

PRODUCTS OF COMBUSTION. 495

of saltpetre. On the other hand, they yield a much greater
volume of gas, which compensates, so that the strength of such
a powder is after all superior to that of a total combustion
powder. It will be seen that this fact must introduce some
complication into the chemical reactions.

3. The latter, moreover, change greatly in character with the
pressure, when operating in a closed vessel. They are also
modified during the discharge of firearms owing to the rapid
expansion of the gases. But the analytical experiments then
become very delicate owing to the difficulty of collecting the
products, and preventing them from undergoing at this moment
the oxidising action of the air, which is the more to be appre-
hended, the more divided the pulverulent products are.

4. Let us now go into detail. Observation shows that the
combustion of powder produces as principal products the follow-
ing bodies (neglecting certain accessory substances to which we
shall return later on) : —

Potassium carbonate, sulphate, and sulphide, or rather,
poly sulphide, carbonic oxide, and nitrogen.

There subsists no sulphurous acid, nor carbon, nor oxygenated
compounds of nitrogen, whether free or in the saline form
(except sometimes some nitrite).

5. These results are accounted for in the following way. At
first the salts of the lower oxygenated acids of sulphur and
nitrogen are all decomposed by the high temperature of the
explosion. As for sulphurous and hyponitric acids, they are
reduced by the carbon and carbonic oxide (see p. 480).

6. Nevertheless, some traces of accessory products are obtained,
such as water, ammonium carbonate, potassium hyposulphite, and
sulphocyanide, sulphuretted hydrogen, hydrogen and methane ;
all these bodies being due to secondary reactions, or reactions
developed during cooling. We shall presently return to them.

7. We have now to examine the relative proportions of the
various products. We shall first define the initial state.

8. Initial state. The analyses were made on powders, the
composition of which was nearly the following : —

Saltpetre 74-7

Sulphur 10-1

Charcoal 14-21

Water 1-0

These numbers, taken roughly, approach the following relations :

16KN03 + 21C + 7S,
in the vicinity of which the composition of the powder of the

1 The charcoal used contained in 14-2 parts : pure carbon, 12'1 ; hydrogen,
0'4 ; oxygen, 1-45 ; ashes, 0'2.

Nitre 772-5

Carbon 120'5

Sulphur 107

496

POWDERS WITH A NITRATE BASE.

principal nations would fluctuate, according to Debus. These
relations expressed in weights represent ; 1616 grms. of nitre,
252 grms. of carbon, 224 grms. of sulphur; in all, 2092 grms.,
which makes per kilogramme 772'5 grms. nitre, 120-5 carbon,
107 sulphur.

It should be observed that in this estimate three to four per
cent, of matter are neglected, represented by moisture (I'O), ash
(0*2 to 0-3), and especially by the hydrogen (04 to 0'5) and
oxygen (1'5 to 2*5) of the charcoal. The moisture and ash
have little influence ; but the hydrogen and oxygen of the
charcoal modify sensibly the volume of the gases. They increase
above all the heat liberated, to such a degree, that the difference
between the latter calculated from the weight of carbon sup-
posed pure and the real heat amounts at least to a tenth, and
with some kinds of charcoal might even rise to the fourth of
the former quantity (see p. 488).

9. Final state. We are indebted to Noble and Abel for
a long and important work on this question. They effected the
combustion of powder in a closed vessel ; a condition which is
not quite the same as that of the combustion of powder in fire-
arms, on account of expansion, and also of the action on the
walls of the vessels, with the formation of iron sulphide, a com-
pound which was produced in very considerable quantities in
their experiments. The mean density of the products of combus-
tion varied in their experiments from O'lO to 0'90. The following
are the proportions by weight of the products observed : —

Pebble Powder, W.A.

R.L.G. Powder, W.A.

F,G. Powder, W.A.

Mean.

Mean.

Mean.

C02

25-0 —27-8

26-8

24-8 —27-6

26-3

24-9 —28-9

26-9

CO

5-7 — 3-7

4-8

5-8 — 3-1

4-2

5-8 — 2-6

3-5

N.

11-0 —11-2

11-5

12-3 —10-5

11-2

11-7 —10-6

112

H.

0-06

0-4 — 0-03

o-i

o-i

0-07

H2S

1*8 _ o-7

1-1

1-8 — 0-8

1-1

1-5 — 1-0

0-8

CU4

1-14— O'O

0-06

0-17— 0-01

0-08

o-i

0-04

0 .

0

0

0-2

...

o-i — o

0-03

Total gaseous 1
products . /

43-2 —44-8

44-1

42-1 —43-7

43-0

41-5 —43-7

42-8

K2CO,

371 —29-8

37-1

38-0 —28-8

34-1

34-3 —25-1

28-6

K2S04

5-3 — 8-6

7-1

2-8 —14-0

8-4

10-4 —14-0

12-5

K2S .

12-5 — 6-7

10-4

10-9 — 6-2

8-1

12-1 — 4-7

10-0

S . .

6-2 — 2-3

4-4

7-2 — 2-7

4-9

5-8 — 2-3

3-8

KCyS.

0-3 — 0-003

0-14

0-2

o-i

0-15 ...

007

(NH4)4(C02)3

0-09— 0-03

0-05

0-08— 0-02

0-04

0-09— 0-01

0-09

Charcoal

...

0-08

0-4

0-04

...

...

KN03

6-27— 6-0

0-13

0-33

0-15

o'i'e— 6-05

0-09

K20 .

3-1

0-6

Total solid\
products . [

559—54-2

55-0

56-7 —55-2

55-9

57-0 —54-8

55-7

Water . .

0-95

1-1

...

1-5

...

POTASSIUM HYPOSULPHITE NOT FOKMED. 497

The variations are wider when we pass to powders in which
the proportion of nitre is different, such as sporting and blast-
ing powders, but we suppress these data in order not to unduly
extend our explanations.

10. These analyses give rise to various remarks. It should
in the first place be noted that the sulphur observed is not free
in reality, but combined, partly in the form of potassium poly-
sulphide, and partly as iron sulphide (or rather of double
sulphide of iron and potassium), resulting from the action on
the walls of the vessels. This phenomenon manifested itself to
the greatest extent in Noble and Abel's experiments, but it is
far less appreciable in firearms owing to the rapidity with
which the products are cooled by expansion and expelled.

11. For a long time potassium hyposulphite, which appears
in the analyses of Bunsen, Linck, Federow, and in the early
publications of Noble and Abel, as representing an amount
sometimes very considerable, had been admitted among the
products of the combustion of powder. The author had called
attention some years since to the fact that this compound could
not be an initial product of the combustion of powder, since it
is completely decomposed by heat towards 450° into sulphate
and poly sulphide (see p. 487). At the very most the presence
of some trace of it might be admitted, due to the secondary
reactions taking place during cooling. But the considerable
amounts observed by writers on the subject appeared attributable
to the alteration of the products produced both by contact with
the air and during the analytical manipulations.

Shortly afterwards Debus confirmed this opinion, and dis-
covered that the hyposulphite found was attributable chiefly to
the reactions of the potassium polysulphides on the copper oxide
employed in the analysis to separate the sulphur from the
alkaline sulphide. Thus at the present day hyposulphite has
disappeared from the list of the essential products formed
during the combustion of powder.

12. It will further be remarked that in exceptional cases
a small quantity of charcoal escapes combustion. A small
quantity of nitre up to three thousandth parts is almost always
found.

Lastly, some powders would yield free potash up to three per
cent. ; a sign of some dissociation of which the suddenness of
cooling or of solidification has preserved a trace; this potash
not having had time to unite with the carbonic acid of the
superposed atmosphere.

The free oxygen which would result from some analyses may
be attributed either to particles of nitrate remaining isolated in
the mass and decomposed by the high temperature of the explo-
sion, or more probably to the dissociation of the carbonic acid
(see p. 504), and the sudden cooling of the mass, which did not

2 K

498

POWDERS WITH A NITEATE BASE.

allow this oxygen to recombine with the excess of carbon or
sulphur.

13. Hydrogen and methane are unimportant products, due to
the complex composition of the charcoal.

The sulphocyanide appears to result from the action of the
sulphur on a small quantity of potassium cyanide, which may
be formed, as is well known, in the reaction of carbon in excess
on potassium nitrate.

A portion of this cyanide changed into cyanate by the
oxidising action, then decomposed by water vapour during
cooling, appears to be the origin of the ammonium carbonate.

The same reaction of water vapour and the co-existing car-
bonic acid on the alkaline sulphide explains the formation of a
small quantity of sulphuretted hydrogen.

14. Equivalent relations. If, for the sake of simplicity, the
accessory products (sulphuretted hydrogen, methane, hydrogen,
sulphocyanide, oxygen, ammonium carbonate, etc.) be neglected,
we find the following equivalent relation between the principal
products : —

POWDER.

Pebble.

R.L.G., W.A.

F.G., W.A.

CO,

mean.
1-22
0-34
0-80
0-54
0-08
0-19
0-28

deviation.
0-08
0-023
0-05
0-11
0-02
0'07
0-13

mean.
1-20
0-30
0-80
0-50
0-10
0-15
0-30

deviation.
0-06
0-07
0-08
0-08
0-06
0-05
014

mean.
1-22
0-25
0-80
0-41
0-14
0-19
0-24

deviation.
0-09
019
0-05

0-11

0-02

o-io

0-12

CO. . . .

N

K,CO, .

K-SO. . . . .

K^S . .
8

The general mean of the analyses would not differ greatly
from the following relation proposed by Debus : —

16KN03 + 21C -f 7S == 13C02 + SCO -f 5K2C03 +K2S04
+ 2K2S3 -f 16N.

15. Variations in, the composition of the final products. But
this mean does not take into account the variations amounting
in the case of carbonic oxide to from 2 '6 to 5'8 ; in that of
potassium carbonate from 251 to 38'0 ; the sulphate from 2'8
to 14'0 ; the sulphide from 4*7 to 12'5.

Generally speaking, the amount of carbonic acid and
potassium carbonate increases slightly (except F.G., W.A. for the
latter) with the pressure ; while the carbonic oxide tends to
diminish (except E.G., W.A.).

The potassium sulphate, sulphide, and carbonate must contain
the whole of the potassium. Hence no one of these three salts
can vary without the whole of both the others undergoing a

THEORY OF THE COMBUSTION OF POWDER. 499

complementary change. Similarly the carbon will be shared
between the potassium carbonate, carbonic acid, and carbonic
oxide, which are complementary. The variations in the
sulphur have less influence on the other compounds, owing
to the formation of the polysulphide, which absorbs a variable
excess of this element. The nitrogen becoming free almost in
its entirety does not enter into account.

The free carbonic acid changes but little.

But the variations in the carbonic oxide and carbonic acid,
combined with the potassium, are complementary to the more
or less advanced transformation of the sulphate into sulphide.

We are about to attempt to account, by a theory, for the
formation of the fundamental products, together with the fluctua-
tions observed in their relative proportions.

3. Theory of the Combustion of Powder. Simultaneous Equations.

1. In the case of powder, as well as in that of ammonium
nitrate (p. 5), and generally of the substances which do not
undergo total combustion, several simultaneous reactions are
produced, due to the diversity of the local conditions of combus-
tion in the unavoidable absence of homogeneousness in a purely
mechanical mixture of three pulverised bodies, and to the
rapidity of the cooling of the mass, which does not allow of the
reactions attaining their limits of definite equilibrium. If we
limit ourselves to the principal products these equations may be
reduced to the following : —

(1) 2KN03 + 8+30 = ^8 + 3C02 + 2N

(2) 4KN03 + 50 = 2K2CO, + 3C02 + 4N

(3) 2KN03 + 30 = K2C03 + C02 + CO + 2N

(4) 2KN08 + S + 2C = KjS04 + 2CO + 2N

(5) 2KN03 + S + C = K^ + C02 + 2N

By combining them with each other, two by two, three by
three, etc., we obtain systems of simultaneous equations repre-
senting all the analyses, the extreme as well as the intermediate
cases.

In this way equations less numerous, but more complicated,
are formed, which any one may combine so as to represent any
particular circumstance of the explosion to which he attaches a
special importance. But all these arrangements essentially
belong to an analogous conception. Kepresentations of this
kind are, moreover, indispensable, unless by an arbitrary fiction
we suppress the experimental variations, which it is precisely
the object of the simultaneous equations to express.

2. On the contrary, by devoting exclusive attention to the
variations, one would run the risk of falling into a blind

2 K 2

500 POWDERS WITH A NITRATE BASE.

empiricism, incapable of serving as guide for the perfecting of
the practical applications. We shall now apply these ideas in
detail. Take, for instance, the mean value given above, accord-
ing to Debus (p. 498) ; it corresponds to an equation which is
too complicated to be admitted as the general representation of
the phenomenon, but it is easy to see that it results from a
certain system of transformations in which a fourth of the salt-
petre has been destroyed according to equation (1) — the sulphide,
moreover, having been changed into polysulphide at the expense
of the excess of sulphur ; an eighth of the saltpetre has been
destroyed according to equation (5) ; three-eighths according to
equation (3) ; and a fourth according to equation (2).

On the other hand, the analyses which have given the
maximum of carbonate also correspond to the maximum of
carbonic oxide, and to a very small proportion of sulphate, all
these being correlative circumstances which may be expressed
by the system according to simultaneous equations, viz. equa-
tion (1) for a third of the saltpetre ; equation (3) for the half ;
equation (2) for about a sixth.

The opposite extreme is that in which the potassium sulphate
gives the maximum proportion, or a fifth of the potassium ;
while the carbonate retains the half of it, and the carbonic oxide
tends to disappear. These relations still show regular reactions,
always expressed by a certain system of simultaneous equations :
or equation (1) for a third of the saltpetre, and equation (2) for
nearly the half, which corresponds to the carbonate ; while the
formation of potassium sulphate would correspond for an eighth
of the substance to equation (4), and for a twelfth to equa-
tion (5).

3. The five simultaneous equations, therefore, represent the
extreme cases ; but it is easy to prove that their combinations
also represent in an approximate manner the intermediate cases.

Consequently, the system of equations expresses the chemical
transformation of powder, at least as regards the fundamental
products. Further, it represents the variations, which could not
be done by a single equation.

The transformation reduces itself definitively to five simple
reactions, which cause the formation of the potassium sul-
phate, sulphide, and carbonate, of carbonic acid, and carbonic
oxide.

4. It is also easy to prove that the combustion of any
powder may be represented by a certain combination of the
above five equations; the first members being taken in such
relative proportions that they represent the initial composition
of the powder under consideration, provided the more or less
abundant formation of the polysulphide, and of the deficit of
about a fourth compared with real carbon, which results from
the use of charcoal, be taken into account.

THEORETICAL TEMPERATURE AND PRESSURE. 501

5. The chemical transformation of powder being thus defined,
let us now calculate the heat liberated and the volume of the
gases produced, according to each of the five equations regarded
separately.

6. Equation (1),

2KN03 -f S + 30 = K2S + 3C02 + 2£T,

represents 135 grms. of matter ; or, for 1 kgm., 784 grms. of
nitre, 118*5 grms. of sulphur, 133-3 grms. of carbon. The
products being, 408 grms. K2S, 488 grms. C02, 104 grms. K.

The reaction liberates + 734 Cal. at constant pressure, 74'5
at constant volume, a quantity which the formation of the
polysulphide, K2S2, by an excess of sulphur during cooling
would raise to about 77 Cal.1 This figure itself is calculated
with the aid of data obtained at the ordinary temperature. At
the high temperature of the explosion it is modified by various
circumstances, such as the partial dissociation of the carbonic
acid, the state of fusion, or even of volatilisation, of the
potassium sulphide, the variation in the specific heats, etc.
But it is not possible in the present state of the science to take
these various circumstances into account; we shall therefore
confine ourselves to the calculation based upon the data
observed. These remarks apply likewise to the other equations.

Supposing, therefore, -f 73 '4 or 74*5 liberated by the trans-
formation (1) ; this quantity, referred to 1 kgm., becomes 544
Cal. at constant pressure, or 552 Cal. at constant volume.

The reduced volume of the gases is 44*6 litres, or, for 1 kgm.,
330-4 litres.

330-4 atm.

Permanent pressure = , with the usual reservation

n — 0*1^

of the liquefaction of carbonic acid for small values of n.
Theoretical temperature 2 = 3514°.

4592 atm. 5740 atm.
.Theoretical pressure 8= nTTo""' °r ' assuming

the total vaporisation of the potassium sulphide.

1 It is here supposed that C + 02 = C0a liberates + 47-0 Cal. See remarks
on page 488.
8 The following specific heats are taken —

C02 3-6 (at constant volume)

N 2-4

CO 2-4

ILS 8.0

C03K2 15-0

BOA 16-6

They are supposed constant for the sake of simplicity.

3 The real density of sulphide of potassium not being known, a density
nearly equal to 3 has here been taken.

502 POWDEKS WITH A NITRATE BASE.

7. These figures would be appreciably modified if we
assumed, as was formerly done, the total vaporisation of the
saline compounds at the moment of the explosion, which would
increase the volume of the gases by a fourth, while slightly
diminishing the heat liberated. But this hypothesis appears
to be abandoned by nearly all specialists at the present day.
It might, however, be true for potassium sulphide, a body
which is volatilised at a red heat. It should further be
observed that the theoretical temperature is too high, as in
all calculations of this kind, owing to dissociation and the
variation in the specific heats with the temperature. This
tends to lower the theoretical pressure. But there is, as we
have said elsewhere (p. 11), a certain compensation, due to
the fact that in greatly compressed gases the variation of
pressure with temperature is far smaller than would be indicated
by Mariotte's and Gay-Lussac's laws. All these remarks apply
equally to the other equations above set forth, and which we
are about to discuss.

8. If the substance used contained a certain proportion of
sulphur in excess and this sulphur were changed into iron
sulphide (p. 497), 11 '9 Cal. should be added per equivalent of
iron sulphide. The heat liberated will therefore be increased.
This increase represents one-eighth of the heat liberated ; but
the increase in the relative weight for an equivalent of sulphur
is nearly the same, which forms a compensation for the same
weight of matter.

These observations are equally applicable to the other
equations.

9. Equation (2),

4KN03 + 5C = 2K2C03 + 3C02 + 4N,

represents 116 grms. of matter, or, for 1 kgm., 129 grms. of
carbon and 878 grms. of nitre. The products being, 593*6 grms.
C03K2, 284-5 grms. C02, 120-5 K

The reaction liberates + 901 Cal. at constant pressure, or 90'8
CaL at constant volume; or, for 1 kgm., 777 Cal. at constant
pressure, or 783 Cal. at constant volume.

The reduced volume of the gases = 27'9 litres ;' or, for 1 kgm.,
240-5 litres.

Permanent pressure = — — ', with the usual reservation.

n — O'At

Theoretical temperature = 3982°.

T, , . , 3749 atm.
Ineoretical pressure = — .

DATA FOB EQUATIONS OF DECOMPOSITION. 503

10. Equation (3),

20T03 -f 30 = K2C03 + CO + C02 + 2N,

represents 119 grins, of matter; or, for 1 kgm., 106 grms. of
carbon and 894 grms. of nitre. The products being, 580 grms.
C03K2, 117-5 grms. CO, 117'5 grms. N, 185 grms. C02.

The transformation liberates + 801 Cal. at constant pressure,
+ 80 '9 Cal. at constant volume ; or, for 1 kgm., 673 Cal. at con-
stant pressure, 680 Cal. at constant volume.

The reduced volume of the gases = 33'5 litres ; or, for
1 kgm., 281-5 litres.

_, 281-5 atm.

Permanent pressure = Tr^r, with the usual reservation.

n — U'Jo

Theoretical temperature = 3458°.

3847 atm.

Theoretical pressure = ^r~'

n - 0-26

11. Equation (4),

2KNO3 + S -f 2C = K2S04 -f 2CO -f 2N,

represents 129 grms. of matter; or, for 1 kgm., 124 grms. of
sulphur, 93 grms. of charcoal, and 783 grms. of nitre. The
products being, 675 grms. S04K2, 217 grms. CO, 108 grms. N.

The transformation liberates 78'2 Cal. at constant pressure,
79-0 Cal. at constant volume ; or, for 1 kgm., 606 Cal. at con-
stant pressure, 612 Cal. at constant volume.

The reduced volume of the gases = 3 3 '5 litres ; or, for
1 kgm., 260 litres.

260 atm.
Permanent pressure = — - with the usual reservation.

Theoretical temperature = 3320°.

„_ •-. . 3422 atm.
Theoretical pressure = ?VOK"'

12. Equation (5),

2KN03 + S + C = K2S04 + C02 + 2N,

represents 123 grms. of matter ; or, for 1 kgm., 821 grms. of
nitre, 130 of sulphur, 49 grms. of carbon. The products being,
708 grms. S04K2, 178 grms. C02, 114 grms. N.

The transformation liberates + 99'4 Cal. at constant pressure,
+ 100 Cal. at constant volume.

The reduced volume of the gases = 22'3 Cal. ; or, for 1 kgm.,
181-5 litres.

Theoretical temperature = 4425°.
3122 atm.

504 POWDERS WITH A NITRATE BASE.

181-5 atm.

Permanent pressure = - TT^T, Wltn tne usual reservation.
n — U'zb

Theoretical tempera

Theoretical pressure

n — U'^o

The five foregoing equations are the only ones which it is
necessary to take into account in problems relating to service
powder where the whole of the charcoal disappears, as has been
said.

13. However, the study of blasting powder, which contains
an excess of charcoal, has led us to consider a fresh reaction,
that of charcoal on carbonic acid. This reaction appears due to
the previous decomposition of the latter producing free oxygen
capable of changing in its turn the carbon into oxide.

C02 = CO + 0 (partial dissociation) absorbs — 34*1 ;
C -f 0 = CO (oxidation) liberates + 12'9.

It already plays a part in two of our equations, for it allows
of passing from (3) to (2), and from (5) to (4).

Without dwelling on the intermediate cases, we shall con-
sider the hypothesis of a decomposition as far advanced as
possible, a hypothesis which never really applies to any but a
portion of the substance.

Take, therefore, the equation

2KN03 + S + 6C = K2S + 600 + 2N.

It represents 153 grms. ; or, for 1 kgm., 105 grms. of sulphur,
235 grms. of carbon, 660 grms. of nitre. The products being,
360 grms. K2S and 640 grms. CO.

The heat liberated is 9 '8 Cal. at constant pressure, or 114 at
constant volume ; or, for 1 kgm., 64 Cal. at constant pressure,
or 74*5 Cal. at constant volume.

The reduced volume of the gases = 66'9 litres ; or, for 1 kgm.,
437 litres.

437 atm.
Permanent pressure = - nTTf

Theoretical temperature = 501°.

1304 atm.
Theoretical pressure = - .

The heat liberated is very slight, and the theoretical
temperature so low that this reaction can hardly be regarded
as explosive.

14. If the foregoing results be compared from the point of

CHANGE IN CONSTITUTION OP SULPHUR "AND CAEBON. 505

view of the heat liberated and the volume of gases produced by
a given weight of nitre, we obtain the following table —

Equivalent
weight.

Heat at constant volume.

Volume of the gases.

Theoretical
pressures.

1 equi v.

1 kgm.

1 equi v.

1 kgm.

(1)

(2)
(3)
(4)
(5)
(6)

grins.

135
116
119
129
123
153

Cal.
74-5

91-4
80-9
79-0
100-0
11-4

552
783
680
612
813
74-5

litres.
446

27-9
33-5
33-5
22-3
66-9

330
240-5
281-5
260
181-5
437

atm.
4592

n - 012
3749
n - 0-27
3847

n - 0-26
3422

n - 0-25
3122

n - 0-26
1304
n - 0-11

15. Equation (5) would be that liberating the maximum
heat, if this maximum still subsisted at the temperature of
combustion, in spite of the variation in the specific heats.
Hence it seems that this reaction should take place to the
exclusion of the others. In any case it should be so with the
integral transformation of the oxygen by the carbon changed
into carbonic acid in accordance with equations (2) and (5).

16. But these preponderating productions are checked by
the following circumstances : —

(1) Dissociation, which does not allow either the whole of
the potassium sulphate or the whole of the carbonic acid to be
formed at the high temperature developed by the combustion.

(2) The change in the constitution of the sulphur at this high
temperature (see p. 27), a change which tends to increase in
an imperfectly known but certainly considerable proportion the
heat of formation of the compounds of this element. This fact
may play a specially important part in the way of increasing
the thermal importance of the polysulphides.

(3) The change in the constitution of the carbon at a high
temperature ; this element existing in the gaseous state, at least
for an instant, in the flames, and the heat of formation of carbonic
oxide being then increased so as to become equal to, or perhaps
higher than, that of carbonic acid for the same weight of oxygen.1

Owing to these circumstances the thermal maximum calcu-

1 " Annales de Chimie et de Physique," 4s serie, torn, xviii. pp. 162 and
175, 176 ; 1869. " Revue Scientifique/' Novembre, 1882, pp. 677-680.

506 POWDERS WITH A NITRATE BASE.

lated for the ordinary temperature may be very different from
the thermal maximum near 2000° or 3000°, temperatures ap-
proaching that of the explosion of powder.

(4) The rapidity of cooling is .too great for the products,
formed at the first instant to have time to react on one another
so as to reconstitute the most stable system.

The specific rapidity of each reaction (p. 40) plays here a
most important part, both at the moment of the initial forma-
tions which take place at the highest temperature, and during
the successive reactions.

It should further be noted that cooling is more rapid at the
point of contact with the walls of the vessels, when operating
in a closed vessel, than towards the centre of the mass. Hence
the composition is different at the various points of the mass,
apart from the reactions exercised by the substances of the walls
themselves, such as the formation of iron sulphide.

17. The rapidity of cooling is very different according as
combustion takes place in a closed vessel strong enough not to
be broken ; or in a shell which bursts suddenly, the fragments
being projected and a portion of the heat being transformed
into mechanical work; or again in a firearm, where the
expansion of the gases takes place according as the projectile is
thrust forward and the gases themselves are continually expelled
towards the cold portions of the metallic tube. The variation
in the chemical reactions which may result from these different
circumstances would be very interesting to study, but it has not
been fully examined.

18. We shall, however, note that according to thermo-
chemical principles the progressive reactions produced during
cooling must be such as to liberate increasing quantities of
heat.

In principle, when operating without changing the condensa-
tion of the substance — that is to say, at constant volume — it
cannot be admitted, in the author's opinion, that endothermal
reactions, such as dissociations, succeed during the period of
cooling to a total combination produced at the instant of
explosion. The dissociation must, generally speaking, be
regarded as being at its maximum at the outset, that is, at the
moment when the temperature is highest, and diminishing as
cooling progresses. This applies principally to reactions effected
in closed and resisting vessels.

It is only when expansion takes place at constant temperature,
owing to the increase in the volume of the gases, that dissocia-
tion, regarded as a function of the pressure, could increase ; the
possibility of this increase may even be conceived, strictly
speaking, in a case of this sort during a certain period of the
cooling.

But these are quite exceptional cases, and endothermal

COMPARISON BETWEEN THEORY AND OBSERVATION. 507

reactions cannot in general be admitted during the period of
rapid cooling succeeding combustion.

19. Let us now compare the volume of the gases liberated.
The reactions of powder, according to the table on page 506,
liberate a volume of gas greater in proportion as they develop
less heat.

The minimum of gaseous volume (22*3 litres) corresponds to
the thermal maximum (100*0 Cal.), and vice versa (66'9 litres
and 11-4 Cal.).

The gases may also vary from the single to the double, the
heats only changing by a fifth, with the exception, however, of
the transformation (6).

20. Hence follows this interesting consequence, that the
theoretical pressure appears to be the greatest for the transforma-
tion liberating the least heat (except 6) ; it would, on the con-
trary, be the smallest for that liberating the most.

In fact, several transformations take place at the same time
owing to locarconditions of temperature, dissociation, and relative
rapidity of combination. The heat liberated, the volume of
gases, and therefore the pressure, will consequently remain inter-
mediate between these extreme limits.

4. Comparison between Theory and Observation.

1. Such are the general consequences of the theory. We are
about to show that observation confirms these consequences by
summing up the results of the experiments, especially of those
made by Noble and Abel, which have been carried out with
greater care than any others.

2. Take first the mean equation (p. 498) —

16KN03 + 210 + 7S = 13C02 + 3CO + 5K2C03 + K2S04

+ 2K2S3 + 16N.
Equation (1) x 2 + eq. (5) X 1 + eq. (3) x 3 + eq. (2) + 2.

Further, it is supposed that the excess of sulphur has been
changed into trisulphide, K^.

According to this mean equation, 964 grms. of matter would
have yielded 674'5 Cal. at constant volume, developing 290'1
litres ; or, for 1 kgm., 697 Cal. and 300 litres.

4350 atm.

The theoretical pressure would be TTT^r-

n — O'Zb

3. Take, now, the transformation observed which produced
the most carbonate and carbonic oxide, that is to say, the
following system : —

Equation (1) X J + eq. (2) X J + eq. (3) X £.

We should have had in this case, for 120 '8 grms. of imatter,
815 Cal. and 363 litres of gas ; or, for 1 kgm., 674'5 Cal. and
300*5 litres. These are practically the same figures as above.

508 POWDERS WITH A NITEATE BASE.

4. On the contrary, the transformations yielding the maximum
of sulphate, that is to say, the following system : —

Eq. (1) X £ + eq. (2) x i 4- eq. (4) X J + eq. (5) X ^

should have produced, for 1238 grms. of matter, 853 Cal., and
321 litres of gas ; or, for 1 kgm., 689 Cal. and 259 litres.

5. But the heat calculated as corresponding to the preceding
transformations is greatly too small. In short, we neglected in
the calculation —

1st. The change of the sulphide into trisulphide, which
liberates about -f 6 Cal. per equivalent.

2nd. The change of an appreciable portion of carbonic acid
into bicarbonate, under the influence of a portion of the water
(1 per cent.), contained in powder a reaction unnoticed in theo-
retical equations neglecting the presence of water.

This quantity, moreover, can hardly exceed 2 per cent. ; that
is to say, about an equivalent, being limited by the weight of
the water itself as well as by the quantity of the latter, which
produces sulphuretted hydrogen. Nevertheless this might add
further + 124 Cal. '

3rd. A portion of the sulphur, instead of producing potassium
trisulphide, was changed into iron sulphide, which liberates per
equivalent of sulphur —

F-f S = FeS,4-ll'9Cal.

If the whole of the excess of sulphur assumed this form, we
might therefore have a thermal excess of + 47'6 Cal., and even
more, owing to the formation of a double iron and potassium
sulphide. The real figure is lower, the sulphur being by no
means all changed into iron sulphide, but it is impossible to
determine it for want of data.

4th. The heat of combustion of carbon has here been calculated
supposing it pure, and even in the diamond state. In reality
the figure thus calculated is too low by an amount which may
be regarded as compromised between 1*5 Cal. . (pure carbon
derived from charcoal) and 5 '2 (bakers' embers), for 1 eq.
(6 grms.) of carbon. This makes for 964 grms. of powder a
thermal excess comprised between 31 '5 Cal. and 109'2 Cal. It
is true that this error is partly compensated, because we have
taken the weight of real carbon as equal to the weight of
charcoal, while it is less by about a fourth (see p. 488).

However this may be, we see from this that the error in the
number above calculated (674*5 Cal.) might amount in an
extreme case to —

109-2 + 47-6 + 12-4 = 169-2,

which would make in all 8437 Cal., or an excess of a fourth in
the number calculated.

The real excess, under the conditions of the experiment of

HEAT LIBERATED. 509

Noble and Abel and the other authorities, is certainly less. But
nothing can be definitely settled with regard to this point till a
special study has been made of the heat of combustion of the
charcoal used in the manufacture of each of the classes of
powder which have been the object of the thermal measurements
and the chemical analysis ; as also the real proportion of the iron
sulphide, and even of the double iron and potassium sulphide
formed during combustion in the interior of an iron vessel.

6. Heat liberated. These reservations having been made, we
shall give the figures found by the authorities who have measured
the heat liberated by the combustion of powder in closed
vessels. Bunsen and Schiskhoff found for 1 kgm., 619*5 Gal. ;
but this number, far lower than those of the other operators,
appears to be invalidated by some error.

Koux and Sarrau found for 1 kgm. at constant volume in a
bomb filled with air, of which the oxygen contributed to increase
the heat liberated —

Cannon powder 753 Cal.

Fine sporting powder 807 „

B rifle powder 731 „

Powder of commerce .' ... 694 „

Blasting powder 570 „

Tromeneuc found from 729 to 890 Cal., viz. —

Ordnance powder 840 Cal.

English powder 891 „

Blasting powder 729 „

Noble and Abel gave at first (dry powder) —

KLG 696 to 706 Cal.

FG 701 to 706 „

mean, 705 Cal. They since discovered that these figures were
slightly too low, and they supplied after correction the following
new mean values : —

QUANTITIES OF HEAT LIBERATED BY THE COMBUSTION OP
IJ&RM. OF POWDER SUPPOSED PERFECTLY DRY.

Pebble powder 721-4 Cal.

R.L.G., W.A. powder 725-7 „

F.G., W.A. powder

No. 6 Curtis and Harvey's powder

Blasting powder

Spanish spherical

738-3 „

764-4 (733 to 784) Cal.

516-8 Cal.

767-3 ,

In order to be able to compare these figures with the numbers
calculated, we must first take into account the ash, oxygen, and
hydrogen contained in the charcoal, and, finally, the nitre which
has escaped combustion. The weight of these various substances
is properly known only for Noble and Abel's experiments. It
amounts to about four per cent, of the weight of dry powder
(more than one per cent, of moisture in ordinary powder).

510 POWDEKS WITH A NITBATE BASE.

This being taken into account, the heat liberated amounts for
1 grm. of explosive matter really transformed to 750 Cal., a
figure which exceeds by 75 '5 Cal., or by a ninth, the theoretical
value 674-5 Cal.

This excess is evidently owing to the causes just described,
and principally to the use of charcoal instead of pure carbon,
and to the formation of iron sulphide. The calculation made
from the heat of combustion of the weight of pure carbon, as
extracted from charcoal, would give 706 Cal., a value which is
also too low. But the number 750 Cal. remains below the
possible difference, which amounts to 843 Cal., according to
what has been said above. For the powders studied by other
observers, the effective reaction being unknown, we cannot
carry out the thermal work with certainty. The values deduced
from our equations generally remain below the figures actually
found, which is attributable to analogous causes, and principally
to the excess of heat produced by the combustion of the charcoal
of powder. This excess will, moreover, vary with the constitu-
tion of this charcoal itself, which changes greatly in the different
countries and for the different kinds of powder.

7. Volume of the gases liberated. The uncertainties are less,
and consequently the discrepancies between theory and practice
more limited for the volume of the gases. For instance, the
volume of the gases obtained by Noble and Abel had a mean
value of 267 litres, with variations comprised between 285 and
232 litres.

The following table, according to these authors, expresses the
volume of the permanent gases produced by the explosion of
1 grm. of powder, supposed perfectly dry : —

W.A. pebble powder 278-3 c.c.

K.L.G., W.A. powder 274-2 „

F.G., W.A. powder 263-1 „

No. 6, Curtis and Harvey's powder 241-0 „

Blasting powder 360-3 ,,

Spanish spherical 234-2 „

Gay-Lussac assigned 250 c.c.1 at a low pressure. Bunsen and
Schiskhoff (sporting powder), 193 c.c. (at a low pressure) ; Linck,
218 c.c. (cannon powder) at high pressures ; Karolyi, 209 c.c.
(ordnance powder) and 227 c.c. (rifle powder); Yignotti, 231 c.c.
to 244 c.c., according to the nature of the charcoal — results,
the differences of which are attributable to the diversity of the
pressures and relative proportions.

The above formula indicates 300 litres, a figure which would
reduce itself to 288 litres were the foreign substances taken into
account. The change of a small quantity of carbonic acid into
bicarbonate would lower it still more, and bring it nearly to the
value found by Noble and Abel.

1 He gives elsewhere 449*5 cc., owing to some error in copying.

PRESSURES DEVELOPED. 511

8. The value of the permanent gases varies nearly inversely
with the heat developed in accordance with the equation on
p. 499 (see also p. 505), and as is shown by the table below —

Heat disengaged Volume of gases
Powder. per gramme of produced per

powder. gramme of powder.

Spanish pellet 767-3 Cal. 234-2 cc.

Curtis and Harvey No. 6 764-4 „ 241-0

F.G., W.A 738-3 „ 263-1 „

R.L.G.,W.A 725-7 „ 274-2 „

Pebble W.A 721-4 „ 278-1 „

Blasting 516-8 „ 360-3 „

9. In general the characteristic product, QV, is nearly con-
stant, as the author observed in 1871, for the various powders.
Now, this product measures the strength for explosive sub-
stances of which the specific heat is the same, which is
practically the actual case (p. 34).

The temperature of the combustion of powder has been
estimated by writers on this subject, from rather uncertain
tests, at 2200°.

10. The pressures developed during the combustion of powder
at constant volume have been observed by Noble and Abel
with the aid of the crusher. The following are their numbers : —

Powder.
Density of charge^

0-1
02
0-3
0-4
0-5
0-6
07
0-8
0-9
1-0

These results may be represented, according to these authors,

2193 2460 '

by the empirical formula, — -, or — - , formulae in

n — U'bo n — U'b

which they suppose that the products which are not gaseous
at the temperature of the explosion occupy 0*68 c.c., or, more
simply, 0'6, the same volume being calculated at the ordinary
temperature.

The theoretical formula on p. 499 gives pressures nearly the
same as the above for high densities of charge (1*0 and 0'9) ;
below which it gives results which are too high, amounting to
double the numbers found for the density 01. This difference
increases as the pressure diminishes; it may be connected
with the increase of dissociation.

! Pebble and R.L.G.
n r
Kgm.
231-3

F.<£
Kgm.

231-5

513-4

513-4

829-4

539-4

1220-5

1219-0

1683-6

1667-8

2266-3

2208

3006-5

2883

3944-2

3734-1

5112

4786

6567

6066-5

512 POWDERS WITH A NITRATE BASE.

§ 8. SPORTING POWDER.

1. Sporting powder is distinguished from service powders
principally by the surplus proportion of saltpetre and by the
choice of the charcoal.

The following are the proportions adopted in France : —

Saltpetre 78

Sulphur 10

Charcoal 12

2. The rapidity of inflammation of sporting powder is less,
according to Piobert, than that of service powder, being in
proportion to the coarseness of the grains. For a sporting
powder containing 30,000 grains to the gramme, the rapidity of
the inflammation was 0'30 m. per second ; while for a service
powder containing 259 grains to the gramme, the rapidity
amounted to 1-52 m.

The rapidity of combustion is also checked by the surplus
proportion of saltpetre. It amounted to 8 mm. to 9 mm. per
second in Piobert's experiments ; while for service powder this
writer found 10 mm. to 13 mm.

3. Brown charcoal tends to give powder shattering properties
because it increases the heat liberated owing to the special com-
position of this charcoal.

4. The surplus proportion of saltpetre also increases the heat ;
but it diminishes the volume of the gases, as is shown by the
figures on page 491, compared with those on page 503.

5. If the heat liberated be supposed proportional to the
weight of the saltpetre, which should not be far from the truth,
the heat will be greater by about a twenty-fifth for sporting,
than for service powder, weight for weight. Now, the experi-
mental data are not greatly at variance with this calculation.
On the other hand, the permanent gases will diminish, which
also agrees with Noble and Abel's results. Hence a certain
compensation is afforded by it. Owing to this fact there is
little difference between the strength of sporting and that of
service powder.

§ 9. BLASTING POWDERS.

1. Blasting powders present very varying proportions. The
principal object aimed at is to increase the volume of the gases ;
which is attained by the diminution of the saltpetre, and the
increase of the sulphur and charcoal. It is also sought to
dimmish the cost of this powder.

The following are the proportions adopted in France : —

Saltpetre 62

Sulphur 20

Charcoal 18

BLASTING POWDERS. 513

In Italy :—

Saltpetre 70

Sulphur 18

Charcoal 12

What is called export trade powder in France, or strong
Hasting powder, contains —

Saltpetre 72

Sulphur 13

Charcoal 15

2. There was formerly distinguished a class known as slow

blasting powder : —

Saltpetre 40

Sulphur 30

Charcoal 30

But the slowness of the reaction tended to diminish the effects
too much, and this powder is no longer in use. However, this
slowness may offer certain advantages for special uses, such as
the making of flying fuses, composed in the following manner : —

Powder dust 25O

Saltpetre 44-5

Sulphur 9-1

Wood charcoal 2*4

3. It was formerly supposed that blasting powder produces
a much greater volume of gases than that of service powder,
because it would be decomposed according to the following
equation : —

2KN03 + 6C + S = 600 + K2S + 2K
This equation would correspond to the proportions —

Saltpetre 65'5

Sulphur 10-0

Charcoal 24'5

But observation has proved that it must be rejected, at least as
the fundamental representation of the reaction.

It would produce, moreover, so little heat (74'5 Gals, per kgm.,
p. 504), that the reaction could hardly propagate itself.

4. Now, powder with an excess of charcoal deflagrates with
vivacity, and forms, like other powders, potassium sulphate and
carbonate, with a liberation of heat which is probably not far
remote from that of blasting powder for the same weight of
nitre consumed. A portion of the carbon tends, however, to
increase the proportion of carbonic oxide ; but a considerable
portion of the charcoal must remain intact.

5. Hence, in this case, as in the foregoing, the sudden transfor-
mation of the explosive substance has a tendency to form the
products liberating the most heat, a remark of capital impor-
tance, and without which it would be difficult to understand

2L

514 POWDERS WITH A NITRATE BASE.

the preponderating proportion of potassium sulphate and car-
bonate which is produced in every case. The production of
potassium sulphide and carbonic oxide is due to the secondary
reaction of the sulphur and charcoal on the above salts ; it plays
an essential part in the study of powder, as it contributes to
increase the volume of the gases.

6. This being established, it may in general be admitted that
the heat liberated by any powder is nearly proportional to the
weight of saltpetre which it contains. The heat liberated by
blasting powder will therefore be to that of service powder in
the ratio of 62 to 75 ; Eoux and Sarrau actually obtained 570 Cal.
instead of 751 Cal.

7. Sarrau and Vieille since found the volume of the gases
equal to 304 c.c. for French blasting powder at the density of
charge 0*6.

This volume is greater by a tenth than that developed by
service powder.

The pressures observed by them were —

Density of charge. Pressure.

0-3 800 kgm.

0-6 2730 „

4540

from which would result the pressure - — ; a formula in which

n

the volume of the solid substances is not taken into account.

8. These gases contained in 100 volumes —

C02 ... ... 49-4

CO ;.. 20-5

H 2-0 to 1-4

CH4 0-3 to l-4i

H2S 7-0 to 5-5

N20 21-3

The proportion of sulphuretted hydrogen is far larger than that
for ordinary powder (4 per cent.).

The carbonic oxide forms a fifth of the volume of the gases,
or 20 c.c., whilst with ordinary powder it amounts only on an
average to one-eighth, viz. 12*5 c.c. Hence it will be seen that
the volume of deleterious gases is nearly double, in the case of
blasting powder, the volume of the s.ame gases yielded by
ordinary powder.

Noble and Abel found also 7*0 of sulphurettted hydrogen;
but nearly equal volumes of carbonic oxide (33 '7) and carbonic
acid (321), which is still more disadvantageous. They obtained
less heat and more gas with blasting than with service powder ;
which affords a compensation from the point of view of
strength.

1 The proportion of methane increases with the pressure (see p. 288 and
464).

POWDEKS WITH SODIUM NITKATE. 515

9. On the whole, blasting powder offers hardly any other
advantage than its low price, due to the diminution in the
weight of nitre. It would certainly be preferable to employ a
less weight of ordinary powder, which would realise the same
economy. Moreover, the daily increasing use of dynamite tends
to limit the consumption of blasting powder.

§ 10. POWDERS WITH SODIUM NITRATE BASE.

1. Sodium nitrate lends itself as well as potassium nitrate to
the manufacture of powders ; it has been employed on a large
scale in the Isthmus of Suez works, and offers a marked
economy. It has also been employed in the mines of Freyberg
and Wetzlar.

Unfortunately this salt is very hygroscopic, and the keeping
of the powders into the composition of which it enters needs
special precautions.

2. Thermal theories increase the interest there may be in
overcoming these difficulties by showing that the powder with
sodium nitrate base develops a greater pressure, weight for
weight, than powder with potassium nitrate base, and that it
can effect a greater work.

3. Take, in fact, a composition equal to that of powder,
such as —

Saltpetre 75

Sulphur 10

Charcoal 15

It would correspond by weight to the following proportions : —

Sodium nitrate 71*8

Sulphur 11-3

Charcoal 16'9

4. Supposing the chemical reactions to be exactly the same,
the heat liberated and the gaseous volume would also remain
nearly the same at equal equivalents (p. 4). But at equal
weights there would be, on the contrary, an eighth more heat,
or for 1 kgm. 782 Cal. from the calculation (or 818 Cal. for
carbon derived from wood charcoal), there would further be a
volume of gas equal to 338 litres.

The resultant force would retain the same expression, but
it would be increased by about an eighth for a given density of
charge. Such are the results indicated by theory. But up to the
present no experiment has been made to study the true re-
actions.

5. In general, powders with sodium base will develop stronger
pressures and a greater quantity of heat, that is of work, than
the same weight of powders with potassium base and of equiva-

composition. Indeed, experiment proves that the substitu-

2 L 2

516 POWDERS WITH A NITRATE BASE.

tion of sodium for potassium in a defined salt, whether dissolved
or anhydrous, causes an almost constant liberation of heat,
whatever be the nature of the salt. Now, the alkaline metal
existing in the saline form, both in the powder and in the
products of combustion, its influence is eliminated in estimating
the heat liberated by combustion, that is when the heat is
estimated for equivalent weights of the sodium and potassium
salts. Weight for weight, on the contrary, much more heat
will be obtained with the sodium salts ; similarly, a larger
volume of gas will be obtained, since the equivalent of sodium
is lower than that of potassium. Various explosives proposed
for industrial purposes, such as Davy powder, pyronome,1 Espir
powder, may be classed with this one.
Take for example —

Sodium nitrate 63

Sulphur 16

Wood sawdust 23

This is a slow acting substance, employed in quarries, especially
to produce dislocations. It is not explosive either by heating,
ordinary shocks, or friction. It contains three to four per cent,
of moisture, a quantity which may increase to as much as 30
per cent, by its being in a damp place, but not without the
powder becoming deliquescent.

The following have been found as the tensions in a closed
vessel : —

Densities of charge 0'4 1613 kgms.

„ 0-5 2401 „

values differing but slightly from that of ordinary blasting
powder, which confirm the foregoing deductions.

7. The sodium nitrate powders have sometimes been mixed
with dry sodium sulphate, or dried magnesium sulphate, to
check the absorption of moisture. But the remedy is merely
temporary, and of little efficiency.

The potassium and sodium, and even barium nitrates, have
also been associated in the same explosive.

8. We shall further mention Violette's mixture : —

Sodium nitrate 62-5

Sodium acetate 37'5

This mixture corresponds to a total combustion —
10C2H3lsra02 + 16NaHT03 = ISCOaNTa, + 7C02 + 15H20 + 16K

The two salts may be melted together, which gives a very
intimate mixture. But if the temperature be raised slightly

1 Under the latter name variable mixtures containing as combustive
elements the alkaline nitrate and potassium chlorate. This confusion should
be avoided, the chlorate base powders being highly dangerous.

POWDERS WITH BARIUM NITRATE. 517

above the melting point, the mixture explodes towards 350°. It
is hygroscopic.

9. Lastly, the sulphur and carbon have been replaced by
a compound which contains both, such as potassium ethylsulpho-
carbonate, or xanthate (xanthine powders) —

Saltpetre 100

Xanthate 40

Wood charcoal 6

§ 11. POWDERS WITH BARIUM NITRATE BASE.

1. Barium nitrate has been introduced into the composition
of the complex powders with special objects. The equivalent
of this salt (130*5) being higher by nearly a third than that of
potassium nitrate, it will be necessary to employ more of it.
For instance, the following proportion —

Barium nitrate 80

Sulphur 8

Charcoal 12

will be equivalent to service powder.

2. With equivalent weight, always assuming the same
chemical reactions, we should have nearly the same quantity of
heat and the same gaseous volume. But it will be necessary
to take a weight of powder greater by a little more than a fifth.
Hence, weight for weight, the heat will be diminished by about
a fifth, together with the volume of the gases and the strength,
for a given density of charge.

3. The following mixtures, for example, have been proposed
— lithofracteur, or saxifragine :

Barium nitrate 77

Wood charcoal 21

Potassium nitrate 2

Similarly the Schultze powders, a mixture of pyroxylated wood
with potasium and barium nitrates (p. 459).

4. Barium nitrate is also employed in pyrotechny to produce
green fires.

5. Strontium nitrate equivalent (1057) differs but slightly
from potassium nitrate. It is hardly employed, save in pyro-
techny to produce red fires.

6. Lead nitrate equivalent (165*5) is capable of yielding for
equal equivalents a fifth more oxygen than the other nitrates ;
but the reactions which it develops are by this very fact all
different, since the lead is reduced to the metallic state, instead
of subsisting under the form of carbonates, as happens with
the alkaline nitrates. Besides, the high price of this substance,
and its high equivalent, hardly permit of its being used, except
for very special purposes ; for instance, by mixing it with red
phosphorus.

( 518 )

CHAPTEE XL

POWDERS WITH CHLORATE BASE.

§ 1. GENERAL NOTIONS.

1. BERTHOLLET, after having discovered potassium chlorate, and
recognised the oxidising properties so characteristic of this salt,
thought of utilising it in the manufacture of service powders.
He made several attempts in this direction, but immediately
suspended them after an explosion which happened during the
manufacture carried on at the Essonnes powder factory, an
explosion in which several persons were killed around himself.
The same attempt has been revived at various periods, with
certain variations in the composition.

But in every case explosions, followed by loss of lives — such,
for instance, as those which happened during the siege of Paris
in 1870, and at L'Ecole de Pyrotechnie in 1877 — happened
before long in the course of its manufacture.

It is thus clear that potassium chlorate is an extremely
dangerous substance, which is only natural, because its mixture
with combustible bodies is sensitive to the least shock or
friction. The catastrophe in the Eue Beranger (see p. 46),
produced by an accumulation of caps for children's play-
things, containing potassium chlorate, has helped to confirm
these ideas. Chlorate powders are, generally speaking, more
easily ignited, and burn with more vivacity than black powder.
They explode, like the latter, on contact with an ignited body.
They are" hardly used at the present day, except as fuses for
fireworks, or to produce shattering effects in torpedoes, for
instance. A powder of this kind has even been proposed in
America as motive agent of forge-hammers or pile-drivers. In
this case the cartridge is placed between the head of the pile and
the ram, when the explosion drives in the one and sends the
other upwards. Their strength is superior to that of nitrate
base powders, but less than that of dynamite or gun-cotton.

2. We shall first state the general properties of chlorated

DANGERS OF CHLORATE POWDERS.

compositions. Potassium chlorate, which is the essential
ingredient, is a salt fusible at 334°, and which decomposes
regularly at 352°. Nevertheless, it may become explosive by
itself under the influence of a sudden heating, or a very violent
shock (p. 406).

We have seen that it yields 39 '1 per cent, of oxygen and
60 49 of chloride of potassium —

C103K = KC1 + 03,

liberating, at the ordinary temperature, 11 Cal. for each
equivalent of oxygen (8 grms.) fixed; or T4 Cal. per gramme
of oxygen ; or 0'54 CaL per gramme of potassium chlorate.

These quantities of heat must therefore, generally speaking,
be added to those which would be produced by free oxygen,
when developing the same reaction at the expense of a com-
bustible body (p. 134). But the presence of the potassium
chloride, which acts as inert matter, tends to lessen this
advantage.

3. The extreme facility with which potassium chlorate
powders explode under the influence of the least shock is a
consequence of the great quantity of heat liberated by the com-
bustion of the particles which are ignited at the very outset
and their low specific heat ; this heat raises the temperature of
the neighbouring portions higher in the case of chlorate than
of nitrate powder, and it therefore more easily propagates the
reaction. The influence is the more marked the lower the
specific heat of the compounds,1 and as the reaction commences,
according to the known facts, at a lower temperature with the
chlorate than with the nitrate of potassium.

Everything, therefore, combines to render the inflammation of
the powder with chlorate base easier.

Therefore the substances of which they are formed should not
be pulverised or crushed together, but pulverised separately
and mixed by screening.

The drying in the stove of these powders is dangerous. The
presence of powdered camphor, so efficacious with gun-cotton,
does not lessen the sensitiveness of chlorate powders.

4. Not only is the chlorate powder more energetic and
inflammable, but its effects are more rapid; it is a shattering
powder. Theory again is able to account for the property. In
fact, the compounds formed by the combustion of chlorate
powder are all binary compounds, the simplest and most stable
of all, such as potassium chloride, carbonic oxide, and sulphurous
acid. Such compounds will undergo dissociation at a higher
temperature and in a less marked manner than the more com-

1 In fact, these two powders only differ by the substitution of the chlorate,
the specific heat of which is 0-209, for the nitrate, the specific heat of which
is 0-239.

520 POWDEKS WITH CHLORATE BASE.

plex and advanced combinations, such as potassium sulphate
and carbonate, or carbonic acid, which are produced by nitrate
powder. It is for this reason that the pressures developed in
the first instance will be nearer the theoretical pressures with
chlorate than with nitrate powder, and the variation in the
pressures produced during the expansion of the gases will be
more abrupt, being less checked by the action of the combina-
tions successively reproduced during the cooling.

5. The explanations just given apply not only to powders in
which potassium chlorate is mixed with charcoal and sulphur,
compared with analogous powders with nitre as base, but also
comprise all powders formed by the association of the same
salts with other substances. It can be shown that this is so,
without entering into special calculations, for which the exact
values would in the majority of cases be wanting.

Now, our comparisons are based on the following data, which
present a general character : —

1st. Both salts employed in equal weights supply to the
bodies which they oxidise the same quantity of oxygen. 122*6
grms. of chlorate yield 6 equiv. or 41 grms. of oxygen ; that is to
say, 8 grms. of oxygen for 20 grms. of chlorate ; whilst 101 grms.
of potassium nitrate yield only 5 equiv., or 40 grms. of available
oxygen, viz. 8 grms. of oxygen to 20*2 grms. of salt. Hence
it follows that both salts must be employed in equal weights
in the greater number of cases.

Now, one and the same weight of oxygen, 8 grms., yielded by
potassium chlorate liberates + 11 Gal. more than free oxygen ;
if it be yielded by the nitrate, it produces on the contrary
+ 8*3 Cal. less j1 which makes a difference of 19'3 CaL, or 6*95
Cal. per gramme of salt employed.

The formation of the same compounds will therefore liberate
more heat with the chlorate than with the nitrate, and the
excess will subsist, even in taking into account the union of
the acids of sulphur and carbon with the potash of the nitrate.

This greater quantity of heat will give rise to a higher
temperature, since the mean specific heat of the products is less
with the chlorate than the nitrate. The mean specific heat of
the products at constant volume may be calculated theoretically
by multiplying the number of atoms by 2 '4, and dividing the
product by "the corresponding weight. Now, the weight of the
combustible body being the same will require the same
respective weights of nitrate and chlorate, according to what
has just been said ; but the latter will correspond to a less
number of atoms, since the equivalent of chlorine is greater
than that of nitrogen.

2nd. The volume of the permanent gases is greater, or at the

1 Supposing it to act upon a carbonated body, the carbon of which is
changed into potassium carbonate.

POTASSIUM PERCHLORATE. 521

lowest equal, with potassium chlorate than with the nitrate,
because the potassium of the former salt remains in the form
of chloride, the whole of the oxygen acting on the sulphur and
carbon to produce gases ; whereas the potassium of the nitrate
retains a part of the oxygen, at the same time as it brings a
portion of the sulphur and carbon to the state of saline and
fixed compounds, the formation of the salts more than com-
pensating for the volume of nitrogen set free.

3rd. In the case where only the carbon or a hydrocarbon
burns, the compensation in the gaseous volumes is exactly
effected because each volume of nitrogen liberated from the
nitrate replaces an equal volume of carbonic acid combined
with the potassium yielded by the said nitrate. Nevertheless
the pressure will be increased, even in this case, with the
chlorate, because its temperature is higher.

4th. The compounds formed with the chlorate being in
general simpler than with the nitrate, dissociation will be less
marked, and consequently the action of the pressures will be at
once more extended, because the initial pressure is greater, and
more abrupt, because the state of combination of the elements
varies between narrower limits. Hence arise shattering effects
rather than those of dislocation or projection.

6. Potassium chlorate possesses another property which has
sometimes been utilised. Its mixture with organic substances,
or with sulphur or other combustible bodies, takes fire under
the influence of a few drops of concentrated sulphuric acid ;
which is due to the formation of chloric acid, which is immedi-
ately decomposed into hypochloric acid, an extremely explosive
compound and a very powerful combustive.

This property has been utilised to cause the ignition by
shock of torpedoes and hollow projectiles charged with
potassium chlorate powder. It is sufficient to place in them
a tube or glass balls, filled with concentrated sulphuric acid.

This artifice may even be employed to ignite chlorate fuses
for exploding dynamite or gun-cotton.

But all these arrangements are very dangerous for those who
put them into execution, and they have not been practically
adopted.

7. We have yet to say a few words about potassium perchlorate,
which is generally regarded as equivalent to the chlorate, but
by a mere theoretical generalisation, for it is a salt which is
expensive, difficult to prepare pure, and it has hardly formed
the object of real experiments as an explosive agent.

Weight for weight it yields a little more oxygen than the
chlorate; about a sixth, viz. 46 '2 per cent, instead of 391.

C104K = KC1 + 04.
But this liberation of oxygen absorbs heat; - 7 '5 Cal. per

522 POWDERS WITH CHLORATE BASE.

equivalent of salt, or - O9 Cal. per equivalent of oxygen, instead
of liberating it.

From this point of view, therefore, the perchlorate acts almost
like free oxygen, with the disadvantage of half of it being
useless inert matter.

Pure perchlorate is not explosive either by shock or inflam-
mation, as the chlorate. Further, its mixtures with organic
substances are far less sensitive to shock, friction, the action
of acids, etc. They ignite with more difficulty and burn
slower.

§ 2. CHLOEATED POWDERS PROPERLY so CALLED.

1. Potassium chlorate powder was formerly manufactured in
the following proportions : —

Chlorate 75-0

Sulphur 12-5

Charcoal 12-5

This powder is extremely shattering and easy to ignite ; its
preparation has occasioned terrible accidents, but the true
reaction which it develops is not well known. The above
proportions correspond to the following weights : —

3C103K + 2S + 50,

assuming the weight of pure carbon equal to that of charcoal,
which however is not exact (see p. 488).

It was first supposed that the reaction consists in the trans-
formation of this system into the following bodies : —

3KC1 + 2S02 4- 500.

The presence of sulphurous acid is unquestionable at any
rate, but carbonic acid is also produced, which the equation
does not take into account.

The same uncertainty prevails concerning the numberless
mixtures formed by potassium chlorate, whether pure or
mixed with nitrate, these bodies being associated with com-
bustible substances, such as charcoal, sugar, ferrocyanide, tan,
wood sawdust, gamboge, benzene, sulphur, carbon disulphide,
antimony sulphide, and the metallic sulphides, phosphorus and
the phosphides, etc., all these being mixtures which have been
proposed or patented of late years, both as explosives and fuses.
We shall give the theoretical calculations only for the total
combustion mixtures formed by the association of potassium
chlorate with carbon, sulphur, sugar and yellow prussiate, for
the sake of comparison between them, and the analogous
mixtures formed by potassium nitrate.

CHLORATE MIXTURES. 523

2. Take first the chlorate mixed with carbon supposed pure —

2C103K + 30 = 3C02 + 2KC1.

The equivalent weight is 140 "6 grms., and there is formed
66 grms. carbonic acid and 74*6 potassium chloride, which
makes for 1 kgm. 872 grms. chlorate, 128 grms. carbon, with
the production of 469 grms. carbonic acid.

The heat liberated amounts to -f 152 Cal. at constant pres-
sure, 4- 153*5 at constant volume ; or, for 1 kgm., 1010 Cal.
at constant pressure, 1092 at constant volume.

Eeduced volume of the gases, 33'5 litres; or, for 1 kgm.
238 litres.

_ 238 atm. . _ _ .

Permanent pressure = — — , with the usual reservation.

n — (j'2ii ,

. u 5950 atm.
Theoretical pressure = 0^7"

3. Take again chlorate mixed with sulphur —

2C103K -f 3S = 3S02 +2KC1.

This mixture ignites at 150°.

The equivalent weight is 170'6 grms., and there is formed
96 grms. sulphuric acid and 74'6 grms. potassium chloride.
This makes for 1 kgm. 719 grms. chlorate, 281 grms. sulphur,
with the production of 563 grms. sulphurous acid. The heat
liberated amounts to 124*8 Cal. at constant pressure, 126*3 at
constant volume ; or, for 1 kgm., 731 Cal. at constant pressure,
740 Cal. at constant volume.

Eeduced volume of gases, 33-5 litres ; or, for 1 kgm., 196*4
litres.

196-4 atm.

Permanent pressure = - — , with the usual reservation.

ti — 0*22

4120 atm.

Theoretical pressure = .

n - 0'22

4. Chlorate mixed with equal weights of sulphur and carlon
(total combustion) —

22C103K -f 9S + 24C = 9S02 + 24C02 + 22KC1.

The equivalent weight is 1637 grms. and there is formed 288
grms. sulphurous acid, 528 grms. carbonic acid, and 821 grms.
potassium chloride ; which makes for 1 kgm. 824 grms. chlorate,
88 grms. sulphur, 88 grms. carbon, with the production of 176
grms. sulphurous acid and 322 grms. carbonic acid.

The heat liberated amounts to 1560 Cal. at constant pressure,
1576 at constant volume; or, for 1 kgm., 953 at constant
pressure, 963 at constant volume.

524 POWDERS WITH CHLORATE BASE.

Volume of the gases, 368 litres ; or, for 1 kgm., 225 litres.

225 atm.

Permanent pressure = 7—3, with the usual reservation.

n — 0'25

. 5170 atm.

Theoretical pressure = . nr .

7i — 0*25

5. Chlorate mixed with cane sugar —

8C103K + CjaHaOu = 12C02 + 11H20 + 8KC1.

The equivalent weight is 661 grms. There is formed 264
grms. carbonic acid, 99 grms. water, and 298 grms. chloride,
which makes for 1 kgm. 742 grms. of chlorate, 258 grms.
sugar, with the production of 400 grms. carbonic acid, 150
grms. water.

Heat liberated : + 766 Cal. liquid water * at constant volume,
4- 726 Cal. gaseous water ; or, for 1 kgm., 1159 CaL liquid
water, 1098 Cal. gaseous water.

Volume of the gases, 134 litres liquid water, 257 litres
gaseous water.

134 atm

Permanent pressure = — — ^, with the usual reservation.

n — (j'26

5400 atm.

Theoretical pressure = •

n _ 0-23

6. Chlorate mixed with potassium ferrocyanide (yellow
prussiate), supposed dry —

46KC103 + 9K4FeC6N6 = 36C02 + 18K2C03 + 54N + 3Fe304

+ 46KC1.

This makes by weight, 1880 grms. of chlorate and 1105 grms.
prussiate ; in all, 2985 grms. ; or, for 1 kgm., 630 grms. chlorate
and 370 prussiate.

There is formed 528 grms. carbonic acid, 828 grms. car-
bonate, 232 grms. nitrogen, and 323 grms. magnetic oxide.

The heat liberated amounts to 2700 Cal. at constant pressure,
2711 Cal. at constant volume ; or, for 1 kgm., 904 Cal. at
constant pressure, 908 Cal. at constant volume.

Volume of the gases, 468 litres ; or, for 1 kgm., 157 litres.

Permanent pressure = — — ", with the usual reservation.

n — 0*o4

3120 atm.

Theoretical pressure = — — .
n - 0-34

7. We shall first compare with each other the results
1 Neglecting the dissolving action of the water on the chloride.

OOMPAEISON OF CHLORATE AND NITRATE MIXTURES, 525

obtained by the total combustion of various bodies by potassium
chlorate.

Weight of
the chlorate.

Heat libe-
rated by
1 kgm. of
the mixture.

Gaseous
volume.

Theoretical
pressure.

Chlorate and carbon ....
Chlorate and sulphur . .
Chlorate with sulphur and carbon
Chlorate and sugar ....
Chlorate and prussiate . . .

872
719
834
742
630

Cal.

1092
740
963
726
931

Litres.
238

196
225
257 l
157

5950
n - 0-27
4120

n - 0-22
5170

n - 0-25
5400

n - 0-23
3120

w-0-3

From this we see that the mixture of chlorate and carbon is
the most advantageous, weight for weight ; but that the mixture
of chlorate and sugar develops a nearly equal pressure, with
a relative weight of chlorate less by a seventh. The mixture
of chlorate and prussiate is not advantageous, the iron acting
as an almost useless inert component, that is to say liberating
a relatively small amount of heat.

8. Let us now examine the results obtained with chlorate
and the analogous data relating to the mixtures formed by salt-
petre, for equal weights, such as 1 kgm. of the mixtures, always
considering total combustion.

Heat.

Volume of
the gas.

Theoretical
pressure.

(Chlorate and sulphur ......

Cal.
740

Litreg.
196

4120

n — 0-27

Nitrate and sulphur

fi*tt

168

2550

n - 0-25

(Chlorate and carbon • • . . .

1092

232

5950

n - 0-22

Nitrate and carbon

786

245

3430

« - 0-27

QflQ

225

5400

j

n — 0-25

Nitrate, sulphur, and carbon ....

801

111

2060

n-0 12

1 Gaseous water.

526 POWDEKS WITH CHLORATE BASE.

It will be seen that in general the values for the chlorate base
powders are much greater than those for the corresponding
nitrate base powders.

The pressures exerted by the former are greater, for the two-
fold reason that the quantities of heat developed are greater, and
the gaseous volumes equal or greater. Hence these powders
will produce effects, both of dislocation and projection, superior
to those of the nitrate base powders.

These conclusions agree perfectly with the known facts,
and it seems that they may be extended to incomplete combus-
tion powders.

But, on the other hand, all the numbers given are far inferior,
with regard both to heat and gaseous volume, to those of gun-
cotton and dynamite (pp. 425 and 451). This inferiority will not
disappear, even for the greatest gaseous volumes which result
from incomplete combustion.

From this point of view, therefore, the chlorate powders
do not exhibit any superiority over the new explosive sub-
stances sufficient to compensate for the exceptional dangers in
manufacturing and handling them. It is only as fuses that
their easy inflammation may offer certain advantages.

( 527 )

CHAPTER XII.

CONCLUSIONS.

WE have now reached the end of our task. We have submitted
a general theory of explosive substances, based on the know-
ledge of their chemical metamorphoses, and of the heat of
formation of the compounds which contribute thereto, that is to
say, entirely deduced from thermo-chemistry. We will sum-
marise the fundamental results of this study, both as regards
general notions and as regards the particular definition of ex-
plosive bodies.

Meanwhile, industry, in this respect, as in many others, has
received an unexpected impetus as a consequence of the
theoretical discoveries of organic chemistry; discoveries which
have facilitated the manufacture at will of a multitude of ex-
plosive substances hitherto unknown, and whose properties
vary ad infinitum.

Empiricism, however, was still the only guide in forecasting
with accuracy the properties of each of these substances at the
time when thermo-chemistry came to our aid, enabling us to
establish the general principles which define new explosive
substances according to their formulae and their heat of forma-
tion. Thermo-chemistry thus marks the limits which we can
hope to reach in practice, and it lends the light of rational rules,
by which alone the subject is capable of being fully developed.

It is this transformation of the empirical study of explosive
substances into a strict science, based on thermo-chemistry, that
the author has been pursuing since 1870, and of which the
present work is the most advanced expression at the present
state of our knowledge.

§ 1. SUMMARY OF THE WORK.— BOOK I.

1. The sudden development of a considerable expansive force
characterises explosive substances. By this means they effect
enormous mechanical work, which industry would be unable to

528 CONCLUSIONS.

accomplish otherwise, except by the aid of complicated, bulky
machinery, necessitating considerable hard labour and expendi-
ture. By this means also we have replaced with unspeakable
advantage the energy afforded by the old war appliances based
on the use of the lever and the sling, while at the same time
the range and the power of the new weapons are extended far
beyond the dreams of former days.

Such mechanical effects are produced by the act of explosion
and by the energy of gaseous molecules, and even this energy
results from chemical reactions, these latter, in fact, determining
the volume of the gases, the quantity of heat, and consequently
the explosive force.

2. Two orders of effects should here be distinguished: the
one due to pressure, the other to the work developed. Thus the
rupture of hollow projectiles and the dislocation of rocks is due
more especially to pressure ; whereas the clearing away of
materials in mines and the projection of missiles in firearms
represent more especially work due to expansion. Now, pressure
depends both on the nature of the gases formed and on their
volume and temperature. Work, on the contrary, depends
especially on the heat liberated, which is the measure of the
potential energy of the explosive substance.

The time necessary for the realisation and the propagation of
chemical reactions plays an essential part in the applications, as
the terms shattering powders, slow powders, and rapid powders
themselves indicate. These various characters do not depend
merely on the structure of the powders and of the nature of the
reactions ; but we may observe, even with the same explosive
substance, taken in an identical form, extremely unequal
durations of combustion, and consequently of its effects.

This, for instance, is what is exemplified in dynamite. Such
diversities are observable in a substance which is identical in its
chemical composition and in its physical structure. They result
from the establishment of two very different laws : the law of
ordinary combustion slowly communicated, and the law of
detonation, that is to say, the law of the explosive wave which
propagates itself with a lightning-like velocity.

These notions on the velocity of the propagation of pheno-
mena, added to the knowledge of the heat liberated and of the
volume of gases, characterise the comparison which may be
made between the old black powder and the new substances
now practically used, such as dynamite and gun-cotton.

From this it follows that, in order to define the force of an
explosive substance, we should know the following data : first,
the nature of the chemical reaction which determines the heat
developed and the volume of gases, and secondly, the rapidity of
the reaction.

3. Chemical reaction is characterised by the initial composi-

EFFECTS OF DISSOCIATION, 529

tion of the explosive substance and by the composition of
the products of explosion. These are further denned, a priori,
in the case of a total combustion, that is to say, when the
substance contains a sufficient quantity of oxygen. This is the
case with nitroglycerin and nitromannite, where carbon and
hydrogen are entirely transformable into water and carbonic
acid.

If, on the other hand, oxygen be deficient, the products vary
with the conditions, and several reactions are often produced
simultaneously, as is the case with ammonium nitrate, with
gun-cotton, and also with service powder. This last, for
instance, does not only produce carbonic acid, potassium sul-
phate, and carbonate, the results of a complete explosion, but
also carbonic oxide and potassium sulphide, due to an imperfect
reaction.

In both cases it must be borne in mind that the products
developed at the moment of the explosion, and at the high
temperature of such explosion, are not necessarily the same as
the products observed after cooling. A part of the water, for
instance, may be found decomposed into oxygen and hydrogen,
a part of the carbonic acid into oxygen and carbonic oxide.
Such are the effects of dissociation; it tends to diminish the
pressure of the system at the moment of the explosion, owing to
the lesser amount of heat developed, but heat is regenerated,
even during the process of cooling ; and it is this which
moderates the expansion and brings the total amount of work
to the same value as if dissociation had not taken place.

4. The liberated heat is calculated from our knowledge of the
products of the reaction, either under constant pressure or under
constant volume ; it is calculated, that is, if the reaction is not
accompanied by any mechanical work. Otherwise, there is a
transformation of a part of this heat into work. Now, it is
precisely this transformation which it is proposed to effect by
the use of explosive substances. It never takes place except
fractionally, as we see in all transformations of this kind in
mechanics. The fraction available in principle amounts almost
to one-half in ordinary gunpowder; in practice we have not
obtained more than one-third. This figure defines the maximum
results which have been observed for this substance, constantly
employed in artillery.

5. The volume of gases also results from chemical reaction ;
it is easily found from the equation which expresses this re-
action. It may be calculated either at a temperature of 0°, and
under normal pressure, or at any temperature or pressure. It
should be observed that in making this calculation it is neces-
sary to add to the permanent gases the volume of the bodies,
such as water or mercury, which are susceptible of acquiring
the gaseous stage at the explosive temperature. Water, in fact,

2 M

530 CONCLUSIONS.

hardly plays any part in the case of service powder, which
barely contains one per cent, of its weight of water ; but water
is, on the other hand, a very important factor in gun-cotton,
nitroglycerin, and in the majority of organic explosive sub-
stances.

6. Having thus defined the volume of the gases we deduce
from it the pressure which they should exercise at the tempera-
ture developed by the explosion at constant volume, and even
at any volume. This calculation rests on the ordinary laws of
gases, laws whose application to these conditions requires the
greatest caution. Thus it is preferable, in practical application,
to measure the pressure of the gases direct from some of their
given mechanical effects, and particularly from the crushing of
small copper or leaden cylinders.

The results should be referred to the weight of the water
contained in the unit of volume. Now, experience shows that
the pressure of the unit of weight for the unit of volume tends
to a constant value ; this is what we term specific pressure, and
this can be taken as a certain measure of force. Here we may
note a remarkable circumstance : the pressures found by experi-
ment are similar to the figures calculated by the ordinary laws
of gases, whether for solid or liquid explosive compounds ; at
least, for those which, in becoming transformed, give rise to pro-
ducts which cannot be dissociated, such as nitrogen sulphide
and mercury fulminate.

On the other hand, in the case of gaseous explosive mixtures,
systems whose density for the unit of volume is low, we find a
considerable difference ranging from the single to the double,
and even beyond this. This difference may be attributable
either to dissociation, or to uncertainty as to the real laws of
gases, which would be applicable under these extreme con-
ditions.

The maximum effort of an explosive substance evidently
applies to that case in which it explodes in its own volume.
Owing to this the effect will be all the greater in proportion to
the density of the substance. Such is the circumstance which,
added to the suddenness of the chemical decomposition, appears
to confer on mercury fulminate the pre-eminence over all other
bodies use^ as primings. The density of the fulminate is, in
fact, alrnostifive- times as great as that of nitroglycerin. This
allows mercury fulminate to exercise an effort which seems to
attain 27,000 kgm. per square centimetre, being almost triple
the effort exercised by the other known substances.

Here we have the total consequences deducible from the
mere knowledge of chemical reaction. But in order to com-
pletely define an explosive substance it is also desirable to
know, as we have said above, what is the duration of its
transformation.

MOLECULAR VELOCITY OF REACTIONS. 531

7. This is a new datum in the problem, and one of the most
important, since it determines the real effects of explosive sub-
stances in their various applications, such as the velocity com-
municated to projectiles in fire-arms, the division and the
projection of fragments of bombshells, and, in fine, the various
results developed in blasting at the expense either of the rocks
required to be dislocated or removed, or of any obstacles which
it is proposed to crush or overturn.

8. The origin of explosive reactions, that is to say, of the
preliminary work which determines their beginning, appears to
correspond in all cases to an initial heating, which raises the
substance to its decomposing temperature, and from which re-
action propagates itself. In order for this heating to be
efficacious, the heat developed by the decomposition must attain
a sufficient intensity to raise gradually, and up to the same
degree, the temperature of the adjacent portions ; it is necessary,
also, that the heat should not become dissipated meanwhile by
radiation, by conduction, or by the expansion of the compressed
gases. In other words, the molecular velocity of the reaction
in the system regarded as homogeneous, and raised to a uniform
temperature throughout, must be sufficiently great, otherwise
there would be no explosion. This is noticeable when decom-
posing cyanogen by means of the electric spark, or when
changing acetylene into benzene by heating. The heat liberated
by this last reaction is enormous, and for equal weights is four
times that of the explosion of gunpowder, but it is so slowly
disengaged that dissipation takes place gradually.

9. The molecular velocity of a reaction is therefore a main
element in the question. Let us summarise the laws which
characterise it.

It increases with the temperature according to a very rapid
law.

It increases also with the condensation of the substance, that
is to say, with the pressure in the gaseous systems.

On the other hand, its action is retarded by the presence of
an inert body which lowers the temperature at the same time
as it lessens condensation. In this way we can at will modify
the character of an explosive substance. For instance, black
powder, mixed with sand, will fuse instead of detonating ;
dynamite, which is a mixture of silica and nitroglycerin, is less
shattering than nitroglycerin ; besides, the shattering character
due to the nitroglycerin decreases rapidly in proportion as the
quantity of silica is increased.

10. The velocity of the propagation of reactions developed
in consequence of ignition or of a local shock, represents a
phenomenon totally distinct from the molecular velocity which
we have just defined ; for it expresses the requisite time for the
physical conditions of temperature, etc., which have caused the

2 M 2

532 CONCLUSIONS.

phenomenon at one point to reproduce themselves successively
at all points of the mass. This is what has been illustrated by
the works of artillerists on the velocity of the combustion of
ordinary powder, a velocity which is variable with the physical
structure of powders and their chemical composition. This
velocity varies exceedingly with the pressure ; gunpowder, for
instance, does not explode in a vacuum, because the heated
gases which combustion has caused, escape, and are dispersed
before having had time to communicate the heat to the adjacent
particles.

Here considerations of an entirely novel character intervene.

Formerly, it was thought it was sufficient to inflame an
explosive substance, no matter how, since the effects of the
ensuing explosion did not appear to depend on the initial pro-
cess of inflammation. But nitroglycerin and gun-cotton have
manifested a peculiar diversity in this respect. Thus, for
instance, according to the process employed in ignition, dynamite
can decompose quietly and flamelessly, or it may burn with a
flame, or again, it may give rise to explosion properly so called ;
this explosion may further be either moderated or accompanied
by shattering effects. Mercury fulminate used as a priming is
particularly apt to cause these latter effects ; it is the detonating
agent par excellence.

11. It has been shown how thermo-dynamic theories and the
suitable analysis of the phenomena of shock will explain this
diversity ; the energy of the shock transforming itself into heat
at the point acted on, and raising the temperature of the parts
first struck, up to the degree of explosive decomposition, their
sudden decomposition produces a fresh shock more violent than
the first on the adjacent parts ; and this regular alternation of
shocks and of decompositions transmits the reaction from layer
to layer throughout the whole mass, developing a real explosive
wave, which progresses with a velocity incomparably greater
than that of simple inflammation.

12. By this we see the all-importance of primings, hitherto
looked upon as simple igniting agents. Here also we note the dis-
tinction between progressive combustion and the almost instan-
taneous detonation of explosive substances, extreme phenomena
among which we observe a series of states and of intermediate
reactions, which explain the variety of the effects produced by
the same agent. In fact, there exists in chemistry a certain
number of endothermal combinations, that is to say, those
which are susceptible of liberating heat by their decomposition ;
these are acetylene, cyanogen, and arseniuretted hydrogen, etc.
Yet these gases do not detonate either by heating or by the
electric spark. The author has now shown that these same
gases do, on the contrary, detonate and resolve themselves into
elements, and with peculiar violence, under the influence of

SYNCHRONOUS VIBRATIONS AND THE EXPLOSIVE WAVE. 533

the sudden shock produced by the explosion of mercury
fulminate.

13. Hence we are led to account for explosions by influence,
peculiar phenomena which have singularly attracted the atten-
tion of artillerists and engineers.

. It has been seen, for instance, that a cartridge of dynamite or
gun-cotton, exploded by means of a fulminate priming, causes
the explosion of the neighbouring cartridges even when placed
at considerable distances, and without the detonation being
followed by a direct propagation of the inflammation. Torpedoes
charged with gun-cotton and submerged will also explode
under the influence of strong cartridges of the same agent
placed in the vicinity. In the present work it has been shown
how these phenomena explain themselves by the development
of the explosive wave in the detonating substance, and by the
violence of the sudden shock which results therefrom, and
which the surrounding medium transmits to the second
cartridge.

Here is recalled to mind, though the author does not adopt
it, the ingenious theory of synchronous vibrations, according to
which the determining cause of the detonation of an explosive
body consists in the synchronism between the vibrations of the
body which causes the detonation and that which would be
produced by the body acted upon. It is shown that this theory
does not in reality explain the facts observed, and the chemical
stability of matter in sonorous vibration is proved by direct
experiment ; these experiments have been made with the most
unstable substances, such as ozone, arseniuretted hydrogen,
persulphuric acid, oxygenated water, etc.

The sonorous waves, properly so called, are not therefore the
real agents propagating chemical decompositions and explosions
by influence ; their energy and their pressure are too slight to
provoke such effects. But propagation takes place in conse-
quence of the explosive wave, a phenomenon of quite a different
nature, and in which the pressure and energy are incomparably
greater, and are incessantly regenerated throughout the wave by
chemical transformation itself.

Thus, according to the new theory, explosive matter detonates
by influence, not because it transmits the initial vibratory
movement by vibrating in unison, but, on the contrary, because
it stops it and appropriates to itself the energy thereof.

14. Let us examine somewhat more closely the characteristics
of this explosive wave which we have been led to discover, and
of which we avail ourselves, in order to explain the detonations of
dynamite and gun-cotton. Its discovery, as well as the study
of it, constitute one of the most interesting chapters in the
present work.

It is in gaseous media that the study of it is at one and the

534 CONCLUSIONS.

same time the easiest and the strictest, and it is then that the
results it offers are most far reaching, theoretically speaking.
This study enables us, in fact, to show the existence of a new
kind of undulatory movement of a compound order, that is to
say, produced in virtue of a certain concord of physical and
chemical impulses, within a substance under transformation.
In the sonorous wave the energy is weak, the excess of pressure
stands at the minimum, and the velocity is determined by the
mere physical constitution of the vibrating medium. On the
other hand, it is the change in the chemical constitution which
propagates itself in the explosive wave, and which communicates
to the system an enormous energy and considerable excess of
pressure. Like phenomena may become developed both in
solids and in liquids.

This wave propagates itself uniformly with a velocity depend-
ing essentially on the nature of the explosive mixture, and
which is almost independent of the diameter of the tubes, except
when these latter are capillary. It is equally independent of
pressure, a fundamental property which determines the general
laws of the phenomenon.

Finally, the energy of the translation of the molecules of the
gaseous system produced by the reaction, and containing all the
heat developed by such reaction, is in proportion to the energy
of the gaseous system itself, containing merely the heat which
it retains at zero. This is an essential detail which experience
has confirmed, and which enables us to calculate the velocity of
the explosive wave in the most diverse mixtures.

It appears that in the act of explosion a certain number of
gaseous molecules among those which form the inflamed sections
at the outset, are hurled forward with all the velocity corre-
sponding to the maximum temperature developed by the
chemical combination. Their shock determines the propagation
of this latter through the neighbouring sections, and the move-
ment is reproduced from section to section with a velocity
which may be compared to that of the molecules themselves.

It is in this way that observations were made of the propaga-
tions of explosions, with velocities of 2480 metres per second
in a mixture of oxygen and hydrogen, of 2480 metres in a
mixture of oxygen and acetylene, and of 2195 metres in a
mixture of 'cyanogen and oxygen, etc. This velocity constitutes
a genuine specific constant for every gaseous mixture.

The propagation of the explosive wave is a phenomenon
altogether distinct from ordinary combustion. It only occurs
when the inflamed section exercises the greatest possible
pressure on the adjoining section ; that is to say, when the
inflamed molecules preserve almost in its entirety the heat
developed by chemical reaction. This state constitutes the law
of detonation.

THERMO-OHEMICAL RESEARCHES. 535

On the other hand, the law of ordinary combustion answers
to a system in which heat is to a great extent lost by radiation,
conduction, expansion, contact with surrounding bodies, etc.,
with the exception of the very small quantity indispensable for
raising the adjacent parts up to the temperature of combustion ;
the excess of heat here tends to reduce itself to zero, and con-
sequently the excess of the velocity of translation of the mole-
cules, that is to say, the excess of pressure of the inflamed
section on the adjacent section.

After having shown in Book I. the general characteristics of
explosive phenomena, it is now desirable to define the funda-
mental circumstance which determines their energies, that is,
the heat liberated by chemical transformation. This is the
object of Book II.

BOOK II.

Any theoretical study of explosives demands a general
knowledge of the principles of thermo-chemistry, namely, of its
methods and of its results ; we have deemed it fitting to sum-
marise these notions at the opening of Book II. The reader
will there find more especially the description of the author's
ordinary calorimeter and of the calorimetric bomb which he
used in studying the heat of detonation of a large number of
gases. Some extensive tables will be shown in this summary,
showing the heat of formation of the principal combinations in
various stages, as also the specific heats and densities of the
various compounds likely to intervene in the study of explosive
substances.

We have devoted ourselves principally to the heat of forma-
tion of those fundamental compounds which help to form these
substances, namely, oxygenated compounds of nitrogen and
their salts, the hydrogenated compounds of nitrogen, cyanic
compounds, carbonated derivatives of nitrogen, nitrogen
sulphide, hydrocarbon nitric derivatives, such as nitric ether of
alcohol, nitroglycerin, nitromannite, gun-cotton ; the nitrated
derivatives, such as nitro-benzene, picric acid, etc. ; the azoic
derivatives, such as diazobenzene and mercury fulminate. We
have also studied the results derived from the oxacids of
chlorine and the explosive oxalates.

This study, which has been lengthy, difficult, and sometimes
even fraught with danger, is almost entirely the result of the
author's own personal experiments.

Hence it has been thought advisable to set down here the
amplified statement of methods and results, and thus to place
before the readers all the data on which the thermo-chemistry of
explosive compounds is based.

536 CONCLUSIONS.

BOOK III.

1. It now remains merely to define the force of the various
explosive matters, regarded individually, in accordance with the
general principles set down in the first two portions of the work.
This is the object of Book III.

2. In practice, a system susceptible of a rapid transformation,
accompanied by a marked development of gas and by great
development of heat, may be utilised as an explosive agent.
These systems belong, in fact, to eight distinct groups,
namely : —

The explosive gases (ozone, oxacids of chlorine) formed with
absorption of heat, that is to say, containing an excess of energy
(acetylene, cyanogen, etc.).

Detonating gaseous mixtures — such as hydrogen, carbonic
oxide, and hydrocarbons, mixed with oxygen, chlorine, and
oxides of nitrogen.

Explosive mineral compounds — nitrogen sulphide and
chloride, fulminating metallic oxides, ammonium nitrate, etc.

Explosive organic compounds — nitric ethers, nitric derivatives
of hydrocarbons, nitro derivatives, diazoic derivatives, fulmi-
nates, perchloric ethers, salts of metallic oxides easily re-
ducible.

The mixtures of explosive compounds with inert bodies.

The mixtures formed by an explosive oxidisable compound
and a non-explosive oxidising body — gun-cotton mixed with
nitrate, picrate mixed with chlorate, mixtures of nitric acid or
hyponitric acid with nitrated and other bodies.

Mixtures with an explosive oxidising base — such as charcoal
dynamite, and blasting gelatin.

Mixtures formed by oxidising bodies and by oxidisable bodies,
none of which is explosive separately — such as powders with a
nitrate or chlorate base.

3. The theoretical and practical data which characterise
explosive substances having been generally enumerated, as also
the practical questions relative to the use, manufacture, and
preservation of the same, as well as the proofs of their stability,
we have now come to the special study of these matters.

4 We at first treated of gases and detonating gaseous
mixtures, beginning with the figures relative to the heat of
transformation, at the theoretical gaseous volume and pressure
in regard to explosive gases properly so called. Thus, at page
387, we have given the table of the characteristic data respect-
ing the chief gaseous mixtures.

This table indicates that the potential energy of gaseous
compounds at unit weight only varies from single to double in
the case of gases containing carbon and hydrogen mixed with
oxygen. It is also the same in the case of the various hydro-

GREAT ENERGY OF GASEOUS MIXTURES. 537

carbon gases. But it is far beyond that of all solid or liquid
compounds. For instance, in the case of hydrogen and oxygen,
the potential energy is four times what it is in ordinary gun-
powder, and double what it is in nitroglycerin. In most of the
hydrocarbons associated with oxygen it scarcely attains to two-
thirds of the energy of an oxyhydric mixture ; acetylene alone
approaches hydrogen.

But these advantages are discounted by the considerable
volume of gaseous mixtures and by the necessity for preserving
them in strong receptacles.

We have given the theoretical pressures and the pressures
observed for these different mixtures. By comparing these we
may observe that the theoretical pressures exceed the real pres-
sures by double and sometimes even more, probably owing to
the dissociation of the compounds, water and carbonic acid, and
to the increase in the specific heats with the temperature.

In fact, the pressures observed with total combustion mixtures
have not exceeded 20 atm., and in most cases they were con-
siderably below the figure. These pressures are very far
inferior to those of solid or liquid explosive substances, this
inferiority being due to the lesser condensation of the sub-
stance.

In the case of liquefied gases, or of analogous bodies, such as
hyponitric acid, we obtain a nearer approach to solid substances.
The table at page 398 furnishes a certain number of details on
-this point.

Finally, we have examined the mixtures of gases and com-
bustible dusts to which numerous accidents in mines have been
attributed, and we have briefly summarised both the theoretical
data and the facts which have come under notice.

5. We now come to liquid or solid explosive compounds.
In the case of each of these we have given the physical
properties, the temperature of decomposition, the heat liberated,
the volume of gases, the permanent pressure, the theoretical
pressure at the moment of the explosion, in fact, the results
of experiments made recently in order to measure the real
pressures and the time necessary for the propagation of the
explosion.

6. All these particulars are shown in the following table,
which summarises the characteristic details of the principal
explosive substances (see next page).

According to this table, gaseous mixtures, such as hydrogen
and oxygen, or acetylene and oxygen, represent those systems
whose potential energy is the greatest ; nitroglycerin and nitro-
mannite, which are the most powerful among solid or liquid
powders, do not attain the half of the proportions referred to
gases; gun-cotton one-third; potassium picrate slightly over
one-fourth, and black powder does not even reach one-fourth.

538

II

• Id

I!"

19*

III

I

I!

CONCLUSIONS.
1

2 .S £

00 • O <M
0} IH 04

CM CM CN

§ £

r o o

A O O

•* 00
i—t rH
CO 00

O CO ^COCDt^ (M CO

00 O (Mi-c±i<M 10 ^

w w

o o o

1

II

I T3-- 'd

1 M §

111

g.2 §

t^>

3

83

ill

x

| g
2 -S

l il

i I S
a -s I

-S &J

a

n

III

s s a

i £

S 3

l!
^S

•*3 V <1

1 3 II

3 1 II

> 2

-8

2

«.

1 ill in"

£ «g S|J

ilrlr*

'Il"l"l

MERCURY FULMINATE GIVES GREATEST PRESSURE. 539

But this inequality is redeemed in practice by the impossi-
bility of raising gaseous mixtures to densities of charge compar-
able with those of other explosive substances. This observation
applies equally to the comparison of the gaseous volumes
developed by the two orders of substances. The absolute volume
of gases produced by one kgm. of matter is the maximum for
hydrogen mixed with oxygen; the other gaseous mixtures
scarcely attain the half of this. Among solid or liquid com-
pounds, gun-cotton and diazobenzene nitrate are those which
furnish the largest volume of gas, namely, two-fifths of the
volume produced by the oxyhydric mixture ; nitroglycerin is less
by one-sixth ; service powder does not attain to one-fourth the
volume furnished by the oxyhydric mixture, and is about one-
third the volume developed by nitroglycerin or gun-cotton.

Any advantage, however, which gaseous mixtures appear to
offer according to these figures is not founded on the actual
measurements which have been made of specific pressures. In
fact, the most energetic mixtures, such as oxygen and hydrogen,
and methane and oxygen, barely attain the same pressures at a
given density of charge as nitroglycerin, nitromannite, and
gun-cotton, which substances are very similar to one another in
this respect.

In truth, the specific pressures are deduced from experiments
made with gaseous mixtures at very small densities of charge.
Probably, if experimenting with gases compressed beforehand
so as to bring them up to densities comparable to those of
liquids, we might arrive at much higher specific pressures. At
all events the fact is one worth noting.

The specific pressure of black powder under a density of
charge equal to unity would exceed the foregoing by about
one-half. Mercury fulminate does not go beyond this at this
density of charge. But its great specific weight (443) allows it
to attain four times this pressure when it detonates in its own
volume ; pressures to which no known body approaches. We
have said already that this circumstance plays a leading part
in the use of fulminate as a priming.

In order to complete these ideas and to fully characterise
explosive bodies, we must further know the duration of the
decomposition in each of the substances, that is to say, the
specific velocity of their explosive wave. This velocity has, in
fact, been found equal to 2840 metres per second in oxyhydric
mixtures, and to 2400 metres in acetylene mixed with hydrogen.
The other combustible gases give similar velocities, with the
exception of carbonic oxide mixed with oxygen, which falls to
1089 metres. With solid or liquid substances similar data are
for the most part wanting, nevertheless velocities of 5000 metres
have been observed with dynamite, and 5000 to 6000 metres
with gun-cotton. These velocities are ample to account for the

540 CONCLUSIONS.

shattering effects produced by these substances. In order to
attenuate these effects it is well to dilute the bodies with an
inert matter ; this tends to change the detonation into a pro-
gressive combustion, a phenomenon of quite another character,
and in which mechanical actions are exercised more slowly;
this kind of combustion is the only one known with any
certainty in connection with black powder.

Such are the general results of the comparison of different
explosive substances. In this work will be found the theoretical
volumes calculated for a great number of other mixtures ; but
in the above table we have limited ourselves to facts resulting
from experiments.

7. Among the interesting conclusions which we have had
occasion to develop, attention may be called to the study of
the manifold decompositions of the same explosive substance,
such as ammonium nitrate ; the examination of the properties
of nitrogen chloride, of potassium and ammonium chlorate, and
of ammonium bichromate ; the decomposition of the nitro-
ethylic and nitro-methylic ethers ; the classification of the
various kinds of dynamite and the theoretical discussion of their
properties ; the study of gun-cotton properly so called, and that
of wet, paraffined, and " nitrated " gun-cotton ; the examination
of picrates, of mixtures formed with nitric acid, associated with
an organic matter, and the examination of perchloric ethers, and
lastly of oxalates.

8. The study of powders with a nitrate base has led to
special developments, both practically and theoretically, owing
to the importance of this class of powders.

The chemical reactions which take place between sulphur,
carbon, their oxides and their salts, have been carefully studied,
as also the decomposition of sulphites and of hyposulphites,
and the study of certain charcoals used in the manufacture of
gunpowder, and which retain an excess of the original energy
of the hydrocarbons from which they are derived. This excess
plays a very important part in the explosive properties of
gunpowder.

Then the different mixtures of nitre, sulphur, and charcoal
which answer to total combustion were examined; the only
mixtures in which chemical reaction can be foreseen a priori.

Service powders are first studied, taking the products of their
combustion such as are known by analysis. After having
summarised these analyses and carried them to the fundamental
products and to the equivalent relations, the fluctuations
observed between these relations are considered, and a theory
founded on the existence of five simultaneous equations is
established, in accordance with which the metamorphosis is
developed in a direction and relative proportion determined by
the local conditions of mixture and of inflammation. The

DISADVANTAGES OF CHLORATE POWDERS. 541

characteristic data of each of these equations is estimated, and
it is shown that they represent all the observed phenomena.

In the case of blasting powder we must also consider the
transformation of carbonic acid into carbonic oxide.

Powders with a sodium or barium nitrate base are then
considered, but bearing in mind this circumstance, that chemical
reactions referred to equivalent weights ought to liberate
approximately the same quantities of gas- and heat as powder
with a base of potassium nitrate, yet that at the same weight
sodium nitrate is superior, whereas barium nitrate would be
less favourable.

9. We conclude by the examination of powders with a
potassium chlorate base, and we show how these powders
possess a force superior to those with a base of nitrate, seeing
that they liberate more heat and at least an equal volume of
gas, but they are very inferior to dynamite and gun-cotton.

They are besides much more dangerous, owing to the
extreme facility with which they inflame under the influence
of shock or friction, and on account of their shattering proper-
ties ; the theory of all of which circumstances is accounted for,
and which circumstances explain the numerous accidents pro-
duced in manufacturing experiments, and the use of chlorate
powders made at different periods. Such powders, being also
surpassed by dynamite and gun-cotton, do riot offer any special
advantage to compensate for the exceptional dangers attending
their preparation and application.

( 542 )

TABLE GIVING WEIGHT or A LITRE OF THE PRINCIPAL GASES.

Names.

Formulae.

Equivalent
weight.

Weight of a litre.

o

8

/1-433 (Theory)

H

1

\1'430 (Regnault)
0-08958

N

H

f 1-254 (Theory)

Cl

35-5

\l-256 (Regnault)
3-18

Br

80

7-16

I

127

11-18

s

16

2-87

P

31

2-78

H°-

100

8-96

Hydrochloric acid ....
Hydrobromic acid ....
Hydriodic acid
Hydrofluoric acid ....
Water vapour
Hydrogen sulphide ....

HCI
HBr
HI
HF
H20
H26
NH,

36-5
81
128
20
9
17
17

1-635'
3-63
5-73
0-896
0-806
1-523
0761

Hydrogen phosphide . . .
Nitrogen monoxide ....

PH,
N20
NO

34
22
30

152
1-971
1 343

Nitrogen trioxide ....

NiO,

NO2

38
46

3-40
2-06

SO,

32

2-87

Carbonic oxide
Carbonic acid
Hypochlorous oxide ....

CO
CO2

C120
CLO,

14
22

43-5
59-5

1-254
/1-971 (Theory)
\1 9774 (Regnault)
3-90
5-33

Chlorine tetroxide ....
Carbon oxy sulphide ....
Carbon oxychloride ....

Acetylene

C1204
COS
COC12
/CHor

67-5
30
49-5

131

3-024
269
4-43

1-165

1C2H2
/CH2 or

26/
14}

T254

Ethane . . . .

\C2H4
/CH3 or

28J
16)

1'343

Methane . . .

\o*

SO/
16

0-716

C* TT

42

1-881

/CNor

26\

2-330

Hydrocyanic acid . . .

\C2N2
HCN

52/
27

1-210

( 543 )

APPENDIX.

MM. BERTHELOT and Vieille continued their researches on deto-
nating gaseous mixtures,1 and their experimental results and their
conclusions are embodied in a series of papers published in the
*' Annales de Chimie et de Physique," 6e serie, torn. iv. pp. 3-90 ;
but it is impossible in the space available to give here more than
a brief indication of the general character of the communications.

The first is " On the calculation of the temperatures of combustion,
specific heats, and dissociation of detonating gaseous mixtures." This
is essentially theoretical in character. The second paper is entitled
" Experimental determinations of pressures" and the third relates
to the " Relative rapidity of combustion of various gaseous mixtures."
The fourth is on the " Influence of the density of gaseous mixtures on
the pressure, and isomeric mixtures. The experiments were made
both with . gaseous mixtures compressed beforehand and with
isomeric mixtures.

The remaining four papers are theoretical, and treat of the
Calculation of the temperatures and specific heats of gaseous mixtures ;
the specific heats of gaseous elements at very high temperatures;
the specific heats of water and carbonic acid at very high tempera-
tures ; and finally, in the last paper, M. Berthelot examines the
manner in which the consequences which result from the experi-
ments affect two fundamental questions — the scale of temperatures
and that of the molecular weights.

The following results are taken from the second paper, on the
determinations of pressures : —

FIEST GROUP. — HYDROGEN MIXTURES.

I. Hydrogen and oxygen. Pressures.

atm.

(1) H2 + O... 980

(2) H2 + O + H2 8-82

(3)H2 + O+2H3 8-02

(4)H2 + 0 + 3H2 7-06

(5)H2+0 + 02 8-69

(6)H2 + O + 303 6-78

» p. 383.

544 APPENDIX.

II. Hydrogen, nitrogen, and oxygen. Pressures.

atm.
(7)H2+0+iN .................. 9-16

(8)H2 + 0 + N2 .................. 8-75

(9) H2 + O + 2N2 .................. 7-94

(10)H2 + O + 3N2 .................. 6-89

III. Hydrogen and nitrogen monoxide.

(11)H2+N20 .................. 13-60

(12) H, + N,O + N, ............... 11-08

SECOND GROUP. — OXYCARBONIC MIXTURES.

I. Carbonic oxide and oxygen.

(13)CO + O .................. 1012

II. Carbonic oxide, nitrogen, and oxygen.

(14) CO + N + O .................. 9-33

(I5)CO + N2 + 0 .................. 8-77

(16)CO + 5N + O .................. 7-05

III. Carbonic oxide and nitrogen monoxide.

(17)CO + N20 .................. 11-41

IV. Varied mixtures.

(18) 2CO + H3 + O8 ............... 9-81

(19) 2CO + H4 + 04 ............... 8-79

(20) 2CO + He + 05 ............... 9-44

(21) 2CO + H8 + O6 ............... 9-61

THIRD GROUP. — CYANOGEN.

I. Cyanogen and oxygen ; total combustion,

(29)C2N, + 04 .................. 20-96

II. Cyanogen, nitrogen, and oxygen ; total combustion.

(30) 2C2N2 + 2N2 + O, ............... 17-70

(31) C2N2 + 2N2 + 04 ............... 14-74

(32) C2N2 + 4N2 + 04 ............... 12-33

III. Cyanogen, nitrogen, and oxygen ; incomplete combustion.

(33)C2N2 + O2 .................. 25-11

(34) C2N2 + 1JN + 02 ............... 20-67

(35) C2N2 + 2N2+02 ............... 15-26

(36)C2N8+#N2+0, ............... 11-78

IV. Cyanogen, carbonic oxide, and oxygen ; incomplete combustion.

(37) 2C2N2 + l^CO + 04 ............... 21-24

(38) C2N2 + 2CO + 02 ....... ........ 15-46

V. Cyanogen and compound combustive gases ; total combustion.

(39)C2N2 + 4NO , ................. 16-92

(40) C2N3 + 4N2O ... ....... ... ........ 22-66

VI. Cyanogen and compound combustive gases ; incomplete combustion.

(41)C2N2 + 2NO .................. 23-34

(42) C2N2 + 2N2O ......... ... ...... 26-02

This last is the greatest pressure which has been obtained with
gaseous mixtures taken at the normal pressure.

APPENDIX. 545

FOURTH GROUP. — HYDROCARBONS.

I. Pure gases. Pressures.

attn. .

(22) Acetylene, C2H2 + O5 15-29

(23) Ethylene, C2H4 + O8 1613

(24) Ethane, C2Hfl + O7 16-18

(25) Methane, 2CH4 -f O8 16-34

II. Varied mixtures..

(26) Ethylene and hydrogen, C2H4 + H2 + O7 ... 14-27

III. Gases containing oxygen.

(27) Methylic ether, C2H6O + O6 1991

(28) Ordinary ether, C4H10O + OJ2 16-33

In regard to the relative rapidity of combustion of various deto-
nating gaseous mixtures, the authors found that in the total
combustion of hydrogen, carbonic oxide, cyanogen, and hydro-
carbons containing much hydrogen, by oxygen and nitrogen
monoxide, the rate of combustion was much slower with carbonic
oxide than with hydrogen. The use of nitrogen monoxide in place
of oxygen retarded the action, and the rapidity of combustion of
cyanogen and the hydrocarbons was little different from that of
hydrogen.

In the case of incomplete combustion of cyanogen the rate was
more rapid than when the combustion was complete.

Experiments on the influence of an excess of one of the com-
ponents, hydrogen or oxygen, showed that in both cases the com-
bustion was retarded, the retarding effect of the oxygen, however,
being nearly double that of the hydrogen for equal volumes.

The presence of products of combustion also caused great
retardation, the rate being three times slower for an equal volume
of carbonic acid, and six times for carbonic oxide. An inert gas,
such as nitrogen, retards the combustion of hydrogen more than
that of carbonic oxide. This shows that the phenomenon is not
only due to the lowering of temperature, which is approximately
the same in both cases, but also to the greater inequality between
the velocities of translation of the gaseous molecules.

Combustion proceeds more slowly in the less condensed isomeric
systems.

When two combustible gases, such as hydrogen and carbonic
oxide, are burned with oxygen, the rate is in no case the mean of
that of the two gases. They appear to burn separately, each with
its own rapidity.

The fact that the rapidity of combustion of hydrocarbons rich
in hydrogen is nearly the same as that of hydrogen appears to
indicate that the hydrogen burns before the carbon, even in total
combustions.

From their experiments on the influence of the density of
detonating gaseous mixtures on the pressure the authors find that
the results do not differ much from those calculated according
to the ordinary laws of gases, but have the advantage of being
independent of the laws themselves. They conclude that at

2N

546 APPENDIX.

about the highest temperatures known, 3000°-4000° on the air
thermometer :

(1) The same quantity of heat being supplied to a gaseous
system, the pressure of the system will vary in proportion to its
density.

(2) The specific heat of gases is practically independent of the
density as well at high temperatures as at 0°.

(3) The pressure increases with the quantity of heat supplied
to the same system.

(4) The apparent specific heat increases with this quantity of
heat.

Referring to temperatures deduced from the expansion of a
given volume of air, M. Berthelot points out that the scale of
temperatures defined by the variations in volume at constant
pressure (or by the variations in pressure at constant volume) and
the scale of temperatures defined by the quantities of heat absorbed
will correspond between 0° and 200°, but will diverge more and
more as the temperature increases until when the temperature
deduced from the expansion indicates 4500°, that calculated from
the heat absorbed will be 8815°.

Further, he says that the indications of an air and of a chlorine
or iodine thermometer differ greatly at high temperatures, and
that there is no valid reason for preferring the indications of an
air thermometer to those of a chlorine thermometer in the defini-
tion of temperatures.

The rapidity of propagation of detonation in solid and liquid
explosives.

In continuation of the experiments made with gaseous mixtures
while studying the explosive wave (p. 88), M. Berthelot, with
the assistance of members of the French Explosive Commission,
has extended his experiments to solid and liquid explosives. Full
details are to be found in " Annales de Chimie et de Physique," 6°
serie, torn. vi. pp. 556-574. Trials were made with gun-cotton and
"starch powder" compressed in metallic tubes, and at different
densities of charge ; also on granulated gun-cotton, dynamite,
liquid nitroglycerin, and panclastite, a mixture formed of equal
parts of carbon disulphide and liquid nitric peroxide.

I. COMPRESSED GUN-COTTON.

(1) In a former series of experiments the velocity in leaden tubes
4 mm. exterior diameter, and about 100 m. in length, varied

from 3903 — 4267 m. per second,
and from 4818 — 6238 m. per second

in tin tubes of the same size. The density of charge, however, was
1'4 in the tin tubes, and varied from 0'9 to 1'2 in the lead tubes.
This may have occasioned the variation in velocity.

(2) In a second series of experiments, made a few years after-
wards on similar gun-cotton, at densities of charge varying from

APPENDIX. 547

1 to 1'2, contained in leaden tubes 4 mm. external diameter, and
about 100 m. in length, the average velocity varied from

4952 m.— 9500 m. (9 experiments),

and from 4749 m. — 5133 m. for similar tubes covered with
plaited string.

In a similar tube the velocity measured at successive intervals
of 25 m. varied from

4671 m.— 5980 m.,
being least at the beginning.

The general average of the velocities is

5200 m.

The irregularity of the results appears to be due to the difficulty
of obtaining leaden tubes of uniform internal diameter. The
duration of the phenomenon may also be influenced by the time
necessary to destroy the tubes. To get perfectly regular results it
would be necessary to have tubes which would not burst. This
has only been accomplished when working with gaseous systems.

(3) With the same product contained in a tin tube, the density
of charge being slightly higher, that is over 1'2} the average
velocity varied

from 5736 m. — 6136 m. for tubes of 4 mm. external diameter,
and from 5845 m. — 6672 in. for tubes of 5'5 mm. diameter.

II. NlTROHYDROCELLULOSE:

Velocity per second.

(25 m.) from 2nd to 3rd interrupter 6389m.

(25m.) „ 3rd to 4th „ 5932 „

(25m.) i, 4th to 5th ,, 6435 „

Mean velocity 6242 m.

Experiments were also made in a tin tube, consisting of two
parts, one 4 mm., the other 5'5 mm. in diameter. The general
average velocity in the 4 mm: tube was

4919 m.

and in the 5'5 mm. tube 6100 m.

Apparently the velocity was rather more rapid in the tin tubes
(5916 m.) than in the leaden ones (5200 m.) ; perhaps because the
former metal resists longer than the latter the explosive effort
which destroys the tube.

III. GRANULATED GUN-COTTON.

At high density of charge, 1'17, in a tube 2 mm. internal
diameter, the average velocity was

4770 m.

In a tube 3-15 mm; internal diameter, density of charge 1'27,
the mean velocity was

5406 m.

This greater velocity was due to the greater diameter of the tube
and density of charge.

2N2

548 APPENDIX.

At low densities of charge, 0'67 and 0'73, the mean velocities
varied from

3767 m. to 3795 m.

This reduced velocity is evidently occasioned by the greater
discontinuity of the explosive resulting from the diminished
density of charge.

Abel, operating with dry compressed gun-cotton placed in
continuous trains in the open, observed velocities of

5320 m. to 6080 m.
with gun-cotton containing 20 per cent, of water —

6090 m.
with "nitrated " gun-cotton —

4712 m. and 4865 m.

With charges of gun-cotton placed in an iron tube and separated
by spaces of 1 mm., he found

1800 m.,
the transmission being retarded on account of the discontinuity

IY. "STARCH" POWDER.

The average velocities observed with this powder, density of
charge about 1'2 in a tin tube 4 mm. external diameter, were in
two experiments 5222 m. and 5674 m.

In a tin tube 5'5 mm. external diameter, the velocity was 5816 m.

In a leaden tube, for density of charge between I'l and 1*2, the
average velocity was 5006 m., and for 1'35 density 5512 m. All
other things being equal, the velocity increases with the density of
charge. The process employed for making these tubes does not
permit of the interior diameter being sufficiently guaranteed to
authorise definite conclusions being drawn from the difference in
velocities observed in tubes of tin and lead.

Y. NlTROMANNITE.

Compressed pulverulent nitromannite fired in leaden tubes 4 mm.
external diameter, density of charge 1'58 and 1'53, gave average
velocities of 6911 m., 7082 m., and 6965 m. At higher density of
charge, 1*9, the average velocity was 7705 m. ; and this is the
highest average velocity which has been observed.

VI. NlTROGLYCERIN.

Liquid nitroglycerin detonates with difficulty in narrow tubes
at low temperature, it having been found impossible to detonate it
in a leaden tube of less than 3 mm. diameter at 12° to 13°.

In tubes of lead or Britannia metal of 3 mm. to 4 mm. diameter,
temperature 14°, when placed in the shade, the detonation was only
transmitted a short distance ; but when the tubes were placed in
the sun, and thereby heated to 18° to 20°, the detonation was
transmitted the whole length of the tube.

APPENDIX. 549

This difference is apparently due to the greater viscosity of the
liquid at lower temperatures.

Average velocities of 1310 m., 1015 m., and 1286 m. were
observed in lead, Britannia metal, and tin tubes 3 mm. in diameter.
A Britannia metal tube 9 mm. in diameter gave 1386 m. Abel
found 1672 m. under slightly different conditions.

VII. DYNAMITE.

Velocities of 2333 m. and 2753 m. were observed in Britannia
metal tubes 3 mm. internal diameter, while in tubes of the same
metal or lead 6 mm. diameter the average velocity was 2668 m.

Abel found 5928 m. to 6566 m. for a train of dynamite cartridges,
30 mm. in diameter, placed end to end and fired in the open air.
These much higher velocities are no doubt due to the much
greater diameter of the explosive cylinders.

VIII. PANCLASTITE.

Owing to the extreme volatile nature of this mixture, bubbles of
gas formed in the interior of the tube, and caused irregularity in
the results. A mixture of equal parts of liquid nitric peroxide and
carbon disulphide, contained in a leaden tube 3 mm. internal
diameter, gave 4685 m. velocity ; another similar experiment gave
5470 m. in the first half of the tube, and 6658 m. for the total
length. On the whole, these figures are similar to those found for
gun-cotton.

To sum up, principally from the experiments made with gun-
cotton —

The velocity increases with the density of charge. It increases
with the diameter, at least within the limits of the very narrow
tubes experimented with.

It appears to increase with the resistance of the envelope (the
latter being pulverised by the explosion).

Finally, comparative measurements made with a tube of 200 mm.
very much curved, and a similar but straight tube, gave practically
the same velocity.

These experiments should be regarded as applicable to practical
conditions comparable to those under which they were made,
although the indications of the correlation between the velocity
and the density of charge or the resistance of the envelopes
appear conformable to theory.

To further develop this study, experiments were made with a
homogeneous and very mobile liquid explosive, methyl nitrate
(p. 420), contained in tubes of caoutchouc, glass of different thick-
nesses, Britannia metal, and steel.

The details of these experiments are to be found in a com-
munication on the " Explosive Wave," by M. Berthelot, " Annales
de Chimie etde Physique," 6* serie, torn, xxiii. pp. 485-503 (1891).

1. Canvas-covered caoutchouc tubes. — The tube had an internal

550 APPENDIX.

diameter of 5 mm., the external diameter being 12 mm. and
the length 39'8 m. The velocity was found to be 1616 m. per
second. The tube after the explosion was rent in long plates in
the direction of the length of the tube.

2. Glass tubes. — Numerous experiments were made, but the
results were not very concordant. The following are extreme
numbers : —

Internal diam. Thickness. Velocity per second.

3 mm. 4'5 mm. 2482 m.

3 „ 2-0 „ 2191 „

3 „ 1-0 „ 1890 „

The thinnest glass tube resisted longer than the canvas-covered
caoutchouc, but the glass tubes were pulverised in every case by
the explosion.

3. Britannia metal tubes, — Experiments made with tubes 3 mm.
internal and 6 mm. external diameter, and about 50 m. long, all in
one piece, showed an average velocity of 1217 m. This metal
offers less resistance and breaks more quickly than the thinnest
glass and canvas-covered caoutchouc.

4. Steel tubes. — Specially drawn steel tubes, in uniform lengths
of 5 m., were obtained which had been very carefully annealed
by heating in a closed vessel for 42 hours, in order to prevent all
crystalline structure. The internal diameter was 3 mm, and the
external 15 mm.

Experiments were made in tubes about 20 m. long, formed of
four lengths carefully joined together in a special manner.

Average velocities of 2155 m, and 2094 m. were observed. All
the steel tubes operated with, opened during the explosion, and
were split into long plates as in the caoutchouc tubes.

The fracture of such thick steel tubes shows that there is no
hope of being able to detonate a liquid explosive in a metallic
tube in its own volume without breaking it, whatever be the
thickness of the tube. This is explained by the fact established
by the theory of elasticity, that the resistance of a metallic tube
does not increase indefinitely with its thickness. The resistance
tends towards a certain limit beyond which the walls of the vessel
tear whatever be the thickness. Now explosive liquids, like
methyl nitrate, offer this remarkable property — that the volume
defined by their density is less than the volume limit below which the
gases or liquids produced by the explosion are susceptible of being
reduced by the pressure developed in the limits of the experiments.

It is known that gases cannot be indefinitely reduced in volume
by compression, their compressibility diminishing beyond certain
limits. This is still more the case with solids and liquids, the
volumes of which cannot be materially altered by pressure.
Suppose, for instance, that the gases produced by the explosion
of methyl nitrate — carbonic acid, carbonic oxide, nitrogen, gaseous
water — at about 3000°, the temperature developed by the explosion,
tend towards a density near unity; then the possible volume of
the gas will be about one-fifth greater than that of the methyl
nitrate (density 1*182). Consequently the vessel will necessarily
be ruptured before the whole of the matter has detonated, and

APPENDIX. 551

this will take place at a moment which will vary with its own
instantaneous resistance. This resistance is quite different from the
static resistance of the vessel, which can be measured by hydraulic
pressure.

Let us examine what actually happens when an explosive
detonates in a tube, the detonation being provoked in the first
instance by the violent shock of the mercury fulminate, which
immediately raises to the extreme limit the initial pressure, the
heat which it disengages, and the chemical reactions developed from
layer to layer which arise from it.

No regular state of affairs corresponding to the explosion of the
matter in its own volume can be established, since the tube is
necessarily broken. However, if it be homogeneous, so that the
pressures and reactions can be propagated in a uniform manner,
then the tube will be regularly and progressively ruptured in pro-
portion as the pressure propagated attains a certain limit, and
thus a special regime of detonation may be established which will
depend on the conditions realised in the system. A velocity
of propagation fairly uniform for each given system will then be
observed, but very variable between different systems even when
the same explosive has been used, as shown by the experiments
with methyl nitrate and the tubes of different material.

This regime of detonation depends on the structure of the
explosive as well as on the nature of the envelope. Thus nitro-
glycerin gives a lower velocity than dynamite, it being a viscid
liquid which transmits the shock which determines detonation
more irregularly than the silica uniformly impregnated with it.
Dynamite made with mica gives still higher velocities, which is
accounted for by the crystalline structure of the mica, this body
being more rigid than the amorphous silica. This view is also
confirmed by the observations made with nitromannite, a solid
crystalline body, which appears more apt to transmit the detona-
tion than liquid methyl nitrate, having given a velocity of 7700 m. ;
picric acid, another crystalline body, has given 6500 m. This
contrast between liquid methyl nitrate and crystallised nitro
compounds is thus in accord with what has been observed between
nitroglycerin and dynamite.

On the other hand, in certain pulverulent systems in which
complete continuity has almost been attained by compression,
experiment has proved that there is a limit of compression beyond
which the mass cannot be exploded by a fulminate detonator.
This has been observed with certain powders formed of potassium
chlorate and tarry materials.

A few further observations with gun-cotton may be given as
showing the influence of the envelope.

Compressed gun-cotton, density of charge 1 and 1*27, gave
velocities of 5400 in. in leaden tubes of 3'15 mm. internal
diameter; while with density of charge 0'73, in a leaden tube
3*77 mm. internal diameter, the velocity observed was 3800 m.,
the inequality being evidently due to the less continuity of the
material. The feeble resistance of the envelope may be com-
pensated by the mass of the explosive, which prevents, especially

552 APPENDIX.

in the centre, the instantaneous escape of the gases. This is
shown by Abel's experiments, already referred to, when he
observed velocities of 5300 m. to 6000 m.

The facts set forth in this paper show that the explosive wave
only exists with its simple characteristics and definite laws, in the
detonation of gases. These laws and characteristics only partially
hold in the detonation of liquids and solids while remaining
subject to the same general notions of physico-chemical dynamics.

On the different modes of explosive decomposition of picric acid and
nitro compounds.1

Considerable diversity of opinion has existed as to whether
picric acid can be exploded by simple heating. It is indeed much
less explosive than nitric ethers like nitroglycerin and gun-cotton,
for if a fairly large mass be heated gradually in a capsule or flask
it melts and emits vapours which catch fire and burn with a
fuliginous flame, but without giving rise to any explosion. A
very small quantity carefully heated in a glass tube may even be
volatilised without decomposition.

But it is a mistake to believe that picric acid is incapable of
exploding by simple heating. Now this body, when submitted
to a high temperature, decomposes with disengagement of heat,
oxidising itself at the expense of the nitrous vapours it contains.
The author has experimentally proved that when a reaction
liberates heat its rapidity increases, on the one hand, with the
condensation of the matter for the same temperature, and, on the
other hand, with the temperature for the same condensation.
The latter increase takes place very rapidly, according to a law
expressed by an exponential function of the temperature. This
tends to render the reaction explosive.

When a closed vessel is used the heat disengaged by the
reaction helps further to elevate the temperature, and consequently
to accelerate the phenomena.

In conformity with these principles picric acid may be caused to
detonate violently in an open vessel at the ordinary pressure,
when it is suddenly heated in a vessel which has been previously
raised to a high temperature, and the mass of which is such that
the introduction of a small quantity of picric acid does not
appreciably modify the general temperature.

The experiment may be made in the following way : — A glass
tube is taken, closed at one end, and about 25 mm. or 30 mm. in
diameter, placed vertically over the flame of a gas burner, and
heated to visible redness, without, however, melting the tube.
Two or three crystals of picric acid, not exceeding a few milligrams
in weight, are projected into the bottom of the tube, when they
immediately explode violently, before having had time to be
reduced to vapour, a very bright white light and characteristic
noise being also produced.

An experiment was made in an atmosphere of nitrogen, and
only a few flakes of carbon remained.

1 "Annales de Chimie et de Physique," 6C serie, torn, xv. pp. 21-25.

APPENDIX. 553

If a larger quantity, not more, however, than a few centigrams,
be used, the bottom of the tube may be sufficiently cooled to pre-
vent immediate detonation, but the picric acid will be vaporised
and a less violent explosion, accompanied with flame, will soon
take place.

With a decigram of the acid the action is slower, but the sub-
stance soon fuses and deflagrates with vivacity. Finally, if the
quantity be still further increased, the acid decomposes without
deflagration.

Similar experiments were also made on other nitro compounds,
and it was found that nitrobenzene, dinitrobenzene, mono-, di-, and
trinitronaphthalene all detonated under the conditions of the
experiments, while giving rise to different modes of decomposition
when the quantities were increased.

These varied modes of decomposition l depend on the initial
temperature of decomposition.

If the surroundings be of sufficient mass to absorb the heat
produced there will be neither deflagration nor detonation. If,
however, a large mass of a nitro compound like picric acid while
burning heats the walls of the vessel containing it sufficiently to
start deflagration, this will still further heat the containing vessel,
and the phenomenon may develop into detonation.

It would suffice, if this happened at an isolated point, through a
fire or any local superheating, to start the explosive wave, which
would be propagated through the entire mass, and thus give rise
to a general explosion.

1 Compare the different modes of decomposition of ammonium nitrate, p. 5.

( 555 )

INDEX.

ABEL, memoirs of, 16

on ballistics, 20, 21

on explosion of black powder, 31

on synchronous vibrations. 80

on deflagrating gun-oottou, 52

on primings, 59

on explosions in water, 77

on explosives, 374, 401

and gun-cotton, 444

on hyposulphite, 486

on final state, powders, 496, 497

on gulphur etiects on arm?, etc.,

497

on heat liberation, 509

on gas liberation, 512

on blasting, etc,, powders, 514

ABEL'S glyoxyliu, 433, 445

powder, 461

Absorbents, 434
Absorption of energy, 2
Accelerograph, 21
Accel eronieter, 21
Accidents at Aspinwall, 431

at Bouchet, 444

at Hamburg, 431

at L'Ecole Poly technique, 518

at Paris, 518

, Place de la Sorhonne, 463

at Rue Be'ranger, 46, 77

at Parma, 436

at Quenasr,431

at St. Denis, 281, 420

at San Francisco, 431

at Simmering, 444

at Stockholm, 431

at Stowmarket, 444

on the Thames, 46

at Vanves, 46

at Vincennes, 444

at Wiener Neustadt, 444

Acetone, 136, 140

Acetylene, 66, 69, 71, 73, 136, 184,

227, 269, 301, 385

Acetylene, heat liberated, 385
Acetate of soda, 516

of ammonia, 318

Acetates, 127, sqq.
Acetylide, 317, 369
Acid, organic, 138
Aoids, 137

and salts, 119

, complex, 317

, explosive, and salts, 268

, free, trace of, 423

, fumes of, 448

, highly oxygenated, 368, 369

, medium, 211, 212

, non -oxygenated, 329

, strong, 212

, vapours of, 381

, weak, 118, 212

Affinities, chemical, 114

, thermal balance of, 123

Agglomeration, 378

Air, detonation in the open, 14

, incomplete combustion, 1G

and transmission, 76, 77

and the explosive wave, 78, 1 07

and alkalis, 207

and nitrogenous compounds, 207

and oxidation, 211

and stability, 381

, combustion by, 390

and charcoal, 399

and starch, 399

and sulphur, 400

and firearms, 495

Albumenoids, 217
Alcohol, 139,265,494

, oxidised, 216

Alcohols and nitric ethers, 279

, complex, 285

Aldehyde, 125, 136, 140

, formation of, 138

Algae, microscopic, 234
Alkali and oxidation, 210, 211
Alkaline sulphites, 484-486
hyposulphites, 486, 487

556

INDEX.

Alkalinity, 209
Alkalis and acids, 215

, organic, 253

Alum, calcined, 432
Alumina, 130, 144
Aluminium, 143
Amarantaceae, 207
Amides, 258, 315
Ammonia, 237, 250

, heat of formation, 237

, combustion of, 240

Ammoniacal salts, 243
Ammonium bichromate, 41 5

, composition, 415

bromide, 317

chloride, 317

cyanide, 317

nitrate, 409

nitrite, 165, 167, 203, 409

, decomposition, 409

perchlorate, 414

picrate, 466

in fireworks, 467

with strontium nitrate, 467

Amorces, 46, 77, et passim.
ANDRE, 124
ANDREWS, 124

on heat measurement, 160

on the electric spark, 191

Aniline, 291

Atmospheric electricity, 233

Authorities, list of, 124

B

Bacteria, 208

BALARD on ammonium hypochlorite,

240

Ballistic pendulum, 21
BARBE on " La Dynamite," 426
Barium, specific heat of, 143

carbonate, 142

chlorate, 133, 345

dioxide, 169, 173, 174, 179

nitrate, 134, 182

nitrite, 164, 167

, formation of, 182

perchlorate, 133

sulphate, 142

BARRAL, analysis of rain, 222
Baryta, 130, 173

and acids, 144

and solid salts, 126, 127, 133

Bases and salts, 119

, weak, 119

, strong, 212

Benzene, 125, 136, 138, 139, 140, 271

, combustion of, 154

and sulphuric acid, 276

Benzoates, 127
BERTHOLLET on ignition, 61

, discoveries of, 518

BERZELIUS on ignition, 416
BERZELIUS on polysulphide, 482

BIANCHI on gunpowder in a vacuum.

48

Bismuth, 139, 141, 142, 370
Blasting gelatine, 433, 441
Bodies, inert, 43
Borlinetto powders, 461
BOULENGE, LE, dynamometer of, 21

, monograph of, 21

, chronograph of, 93, 95, 96

BOUSSINGAULT on gaseous states, 31

on mould, 232

on absorption of free nitrogen,

234

on dissociation of sulphates, 482

BOUTMY on nitroglycerin, 423
BOYLE on powder in a vacuum, 48
BRACONNOT on nitric compounds, 444
Bromic acid, 357
Bromine, liquid, 176

, oxygenated compounds of, 357

BROWN on gun-cotton, 444
Brugere powder, 46 1

, analysis of, 467

BUFF on nitrogen monoxide, 191

on nitric oxide, 192

on sulphurous gas, 479

RUIGNET on hydrocyanic acid, 310
BUNSEN on velocity of combustion, 49 ,

56

on the regime of combustion,]! 13

, ice calorimeter of, 145

on measurement of heat, 160

elements, 191, 193

on pressures, 389

on temperature, 390

on powders, 477

on charcoal powders, 488

on potassium hyposulphite, 497

on heat liberation, 509

on gas liberation, 510

BUSSY on hydrocyanic acid, 310

Cadmium, 318

Calibration of crusher gauge, 29X

Calorie, 14, 145

Calorimeter, the, 145-148.

, ice, water, and mercury, 145

Calorimetric bomb, 148-159

Cane sugar and chlorate, 524

Carbon burnt in oxygen, 479

and potassium carbonate, 483

and sulphurous acid gas, 479

Carbonates, 126

Carbonic acid and sulphur, 480

oxide, decomposition of, 479

, combustion of, by oxygen,

162

, combustion by nitrogen mon-
oxide, 162

CARIUS on ozone, 218

CASTAN on combustion velocity, 49

CAVALLI and the ballistic pendulum, 21

INDEX.

557

CAVENDISH, celebrated experiment of,

222
Celluloid, 452

• , explosion of, 73

CHABRIER on nitrites and nitrate, 213
CHAMPION on synchronous vibrations,

80

on nitroglycerin, 281

CHAPPUIS on ozone, 384
Charcoals in powder, 488
Chili, nitrate mines in, 207
Chlorate mixtures, 523-526

powders, dangers of, 518

Chlorated powders, 522

, disadvantages of, 541

Chlorates, 133, 143, 344

, formation of, 348

Chloric acid, 344
Chlorine, 73, 344

on ammonia, 239

Chromates, 134
Chronograph, the, 93
CLAUSIUS on specific heats, 10

on molecules, 90

CLOEZ on porous bodies, 218

on rust, 219

Collodion cotton, 442, 446
Combination, chemical, 11
Combustible bodies, list of, 159
Combustion and detonation, 55

, modes of, 52

of bodies, 153

Compounds, see tables, 125

through nitric acid, 265

COVILLE on dynamite cartridges, 75
Curtis and Harvey powder, 510
Cyanides, double, 325

, mercury and potassium, 325

, silver and potassium, 326

Cyanogen, 66, 71-73, 300

compounds, 132

series, formation of, 299

, combustion of, 161, 300

witli chloride, 320, 333

with chlorine, 337

with iodine, 340

iodide, 339

Daniell cells, 231, 233

D ALTON on oxides of nitrogen, 190

DAVY, H., on ignition, 61, 394

Davy powder, 516

DEBUS on hyposulphite, 486, 497

on equivalent relations, 498, 500

Decoin position in a Closed vessel, 15

, spontaneous, 45

, modes of, 57

Deflagration, propagation of, 53
DEHERAIN on nitrites and nitrates, 2Q9,

213

DEMONDESIR on mixture limits, 393
Densities, table of, 144

DEPREZ, MARCEL, manometric balances

of, 21

, accelerograph of, 21

, accelerometer of, 21

on the blow of a hammer, 53

Designolle powders, 461

, analysis of, 467

DESAINS on nitrogen pentoxide, 183

DESORTIAUX, 20

Detonating gaseous mixtures, 383, 386,

543
Detonation, 55

, limit of, 107, 110

, conditions of, 103

, rapidity of, in solid and liquid

explosions, 546
Detonators, 148-159

, the most powerful, 55

DEVILLE on potassium nitrate, 160

on nitrogen pentoxide, 181

on sulphur-density, 402

on nitrogen chloride, 405

on sulphurous gas, 479

Dextrine, 231

Diazobenzene nitrate, 291, 471
Diazo compounds, 290, 468
Dinitrobenzene, 272, 553
Dinitroglycolic ether, 421
Dispersion of explosives, 56
Dissociation, 8

, influence of, 11

, phenomena of, 12, 13

, annulled influence of, 111

, effects of, 529

DITTE on hydrated iodic acid, 361
DIVERS discovers silver hyponitrite, 185

on ammonium chloride, 190

Dynamite, 2, 54

with ammonium nitrate base, 439

with nitrocellulose base, 444

Dynamites, 431

, classes of, 431-433

, designation of, 433, 436

, practical needs of, 433

, general notions of, 434

DULONG on specific heats, 115, 121

with the water calorimeter, 124

on oxides of nitrogen, 1 90

on nitrogen trioxide, 196

on cyanogen, 301

Dusts, gas and combustible, 399

, air and charcoal, 399

, and starch, 399

, and sulphur, 400

Electricity and nitrous oxide, 1 93

in general, actions of, 219

, low tension of, 230

, atmospheric, 233

Empiricism, 527
Endothermal, 66, 115
Energy, absorption of, 2

558

INDEX.

Energy, potential, of an explosive, 17
Eprouvette, the, 22
Espir powder, 516
Ethane, combustion of, 155
Ethers, 137

from alcohols, 279

, nitric, 418

Ethylamine, 254
Ethylene nitrate, ?85
Exothermal reactions, 115
Expansion of gases, 17
Explosions by influence, 75-87
Explosive compounds, 402, 536
, decomposition by heat of

picric acid and nitro compounds, 552
, multiple decomposition of,

135

gases, 536

mixtures, table of, 387, 536

substances, table of, 538

, remarks on, 537, 539, 540

wave, theory of the, 77-82

, general characteristics of, 88

, establishment of, 108

, real nature of the, 533

Explosives, force of, 1-4, 367

, high, low, pure, mixed, 2, 3

, effects of decomposition of, 7

, list of, 368

, data as to employment of, 371

, practical question as to, 376

, as to the manufacture of, 379

, as to pressure by, 379

, stability, tests of, 381

FAUCHER on saltpetre, 207

on nitroglycerin, 423

Faversham u cotton powder," 459
FAVBE, mercury calorimeter of, 145

on potassium nitrate, 160

on nitric oxide, 161

on nitrogen trioxide, 167

monoxide, 183

on absorption of chlorine, 237

on hydrochloric acid and am*

rnonium chloride, 238

• on phenol, 277

on chloric acid and chlorates, 344

on bakers' embers, 488

FEDEROW on potassium hyposulphite,

497

Ferment, the nitric, 209
Ferrocyanide, 327-329

, heat of formation of. 329, 332

, liberated by, 330

FILHOL on nitric acid from rain, 222

Firearms, theory of, 16

Fontaine powder, 461

FORDOZ on nitrogen sulphide, 261

Formamide, 259, 414

reaction, 260

Formate, 317

Formates, 127

Formation, tables of, 125-139 .

Formic acid, 259

Forrnonitril, formation of, 314

FRANKLAND on chloric acid and

chlorates, 344

FRITSCH on nitroglycerin, 431
FRITZSCHE on nitrogen trioxide, 196
Fuses on high mountains, 49
Fusion, table of heat of, 139

G

GALLOWAY on combustible dusts, 401
Gas and combustible dusts, 399
Gaseous mixtures, detonating, 383,

536, 543

, pressure by, 388, 395

with an inert gas, 392

, inflammability of, 392, 394

, energy of, 536

Gases, volume and increase of, 8

, pressure of, 20

, temperature of, 18

, specific heat of, 19, 141-143

, formation of, 125

, explosive, 383

Gauge, crusher, 20

GAY-LUSSAC, law of, on pressure, 9, 11.

28, 373, 383

on nitric oxide, 192

on oxygen and nitrogen, 195, 196

on hydrocyanic gas, 313

on polysulphide, 482

on powder-explosion, 510

GAYON on nitrites and nitrates, 209, 213
GELIS on nitrogen sulphide, 261
GERNEZ on nitrogen sulphide, 240
GILBERT, agricultural experiments of,

235
GIRARD on nitroglycerin, 51

on explosion, 439

Glucose, 73, 123

Glycol, heat of combustion of, 285

GOPPELSHODER on nitrites in stables,

213 7

GRAHAM on potassium nitrate, 160
Griess's process on aniline nitrate, 292
GROVE on nitrogen monoxide, 191
Gunj-cgiton, 286, 444, 447, 546

, non-compressed, 49

, pulverulent compressed, 55, 56

, mode of explosion, 59

compounds, 267

, preparation of, 287

, heat of detonation, 288, 450

and dynamite, 445

and paraffin, 445, 454

, compressed, 444

and water, 54, 445, 453

, properties of, 447-449

, comparison with other explosives,

452
, nitrated, 453

INDEX.

559

Gun-cotton with ammonium nitrate,
453

with potassium nitrate, 456

and chlorate, 460

H

Halogen elements, 344

and alkalis, 362, 363

Haloid salts, 131

HASENBACH on nitrogen trioxide, 196
HAUTEFEUILLE on potassium nitrate,
160

on ozone, 384

on nitrogen chloride, 405

HAWKSBEE on powder fusing, 48
Heat, absorption, 3

hi reaction, 10, 11, 14

, summation of, 13

, explosive measure of, 14

and oxidising agents, 134

of organic compounds, 134

of fusion, 139

, volatilisation of, 140

, specific, 19, 141-143

HEEREN on nitroglycerin, 48
HENRY on ignition, 62, 65

on impact, 21

HESS on potassium nitrate, 160

on blasting gelatin, -143

HIRN on specific heat, 19

on mixed pressure, 391

HOFFMAN. See BUFF
Holz machine, 223, 229
HORSLEY on temperatures of powder,
493

HUGONIOT on powder, 16, 17

Hydrates, 134

Hydrations of organic compounds, 260

Hydrocarbons, 136, 138

Hydrochloric and hydrocyanic acids,
320

Hydrocyanic acid, 329, 341

, heat of formation, 308, 313

, conversion of, 302

, vaporisation of, 309

Hydroferrocyanides, 317

Hydrogen, arseniuretted, 72, 73

, vibrations of, 85, 86

Hydroxylamine, 245

, derivation of, 245

, apparatus for, 245

, decomposition, 248

, reactions, 249

, constitution, 252

, temperature, 253

Hypobromite, 358, 359

Hypobromous acid, 358

Hypochlorous gas, detonation of, 67

Hypoiodite, 359

, formation of, 359, 360

, final product of, 360

Hyponitrites, 185

Hyponitrous acid, formula of, 185

Hyponitrous acid, formation of, 186
— , heat of formation of, 188
— , heat of neutralisation of, 189

HUYGHENS on ignition of powder, 48

I

Iboz dynamite, 436
Inflammation, propagation of, 56

— , influence of initial, 109
lodates, 126, 358
lodic acid, 360
Isomeric mixtures, 105

JOANNIS on cyanides, etc., 300
Joule's law, 19, 121

K

KAROLYI on gun-cotton, 350

on powders, 477

• — — on ordnance and rifle powder, 510
Kieselguhr, silicious earth, 436, 441
KOENIG on ozone vibrations, 85
KOPP on specific heats, 143

on nitrobenzene, 270

KUHLMAKN on oxidation, 208

LAMBERT on dynamite cartridges, 82

LAPLACE, ice calorimeter of, 145

on heat relation, 115

Latent heat, 140

LAVOISIER on heat relation, 115

, ice calorimeter of, 145 -

on oxygen. 160

LAWES, agricultural experiments of,
235

LE CHATELIEK on combustion velocity,
50

on the regime of ordinary com-
bustion, 113

on gaseous pressures, 389, 391

on combustible dusts, 401

Lead, explosions in a block of, 374,
434

Leclanche cells, 231

LINCK on gun-cotton, 444

on powders, 477

on potassium hyposulphite, 497

on cannon powder, 510

Liebig tube, 157

Liquefied gases, etc., 396

, table of, 398

LONGCHAMP on absorption of nitrogen,
218

LOUGUININE on nitroglycerin, 282

on glycol, 285

Lycopodium, 494

560

INDEX.

M

MALLARD. See LE CHATELIER.
Manometer, 389
Manometric balances, 21
MAQUENNE on nitrites and nitrates, 209,

213
MARIGNAC on sulphites, 484

on anhydrous bisulphite, 485

Mariotte's law, 27, 28, 403

. See GAY-LUSSAC

MARSH on arseniuretted hydrogen, 68
MASCART on fixation of nitrogen, 233
Measurements, 20, 89
MEILLET on cyanide, 327
MELLONI on Soliatara, 393
Mercuric cyanide, 302, 310, 318, 322
, formation from the acid and the

oxide, 318
, formation from the elements, 298,

319

, absorption of heat of, 319, 320

Mercury fulminate, 2, 48, 91, 297, 468
, point of deflagration of, 292,

469

, analysis of, 297

, heat of formation of, 297, 468

, heat of decomposition of, 298

oxalate, 476

Metallic cyanides, 318

oxalates, 364

, heat of formation of, 365

oxides, 130

sulphides, 131

Metalloids, 128, 129

Methyl nitrate, 284, 549. See also

Nitro-methylic ether
Methylperchloric ether, 474
MEUDON, apparatus of, 20
MEYER. See UPPMAN
Microbes, 208
Microderms, 208
MILLON on barium chlorate, 350
MILLOT on nitroglycerin, 439
MITSCHERLICH on hyposulphite, 483
Mixtures, chlorate and nitrate, 525
Molecular volumes, table of, 144
MONTLUISANT, apparatus of, 20
MULDER on earthy substances, 219
MUSPRATT on sulphites, 484, 485
MUNTZ on nitre, 208
on the oxidation of ammonia, 214

N

Naphthalene, 136, 139
Neumann, ballistic pendulum, 21
Nitrate, diazobenzene, 291
Nitrates, 126

, formation of the, 126, 1 27, 133, 1 35

, heat of formation, 202

, general remarks, 204

, origin of the, 207

Nitrates, queries on, 208
Nitric acid, 200, 213, 293
, heat liberated by, 200

, formation of, 213

, reproduction of, 296

, heat of formation, 173

and dinitrobenzene, 474

and nitrobenzene, 473

and organic compounds, 472

and picric acid, 473

derivatives, 139

ether, 279

, formation of, 279-281

, heat of formation, 280

, decomposition of, 281

, heat of total combustion, 28 1

ethers, 279 418

oxide, 160, 192

, decomposition, 192

, heat of formation, 1 60

, heat liberated by, 200

, want of stability, 194

peroxide, 167, 179, 198

, heat of formation, 167-172

, heat liberated, 172, 199

Nitrides, 368
Nitrification, chemical, 210

, natural, 207

, thermal, 215

Nitrites, 124, 163-167

, ammonium, 165

, barium, 164

, silver, 166

Nitrogen, fixation of, 217

chloride, 404

iodide, 406,,

monoxide, 162, 195

pentoxide, 181-184

— — selenide, 263

, an endothermal compound,

263

sulphide, 261

trioxide, 163, 167, 171, 195

Nitro compounds, 137
Nitrobenzene, 269, 553
Nitrobenzoic acid, 275
Nitrogenous compounds, 217, 290
Nitrocelluloses, 444, 446
Nitro-ethylic ether, 418
Nitroglycerin, 14, 59, 281, 422, 548

, heat of formation, 282, 423

, inflammation of, 423

, poisonous, 423

, pressures, 424, 425

, summation, 426, 427

Nitrohydrocellulose, 448, 547
Nitromannite, 283, 428, 548

, maximum work, 430

Nitro-methylic ether, 420
Nitro-starch (xyloidin), 286, 548
Nitrous acid, 195, 212

, formation of, 212

, from ammonia, 212

NOBEL on detonators, 52

INDEX.

561

NOBEL on explosion, 68 1

on dynamite, 431, 435

uses Kieselguhr, 436

on mixtures, 439

invents blasting gelatin, 441

NOBLE. See ABEL

Notes, influence of musical, 82

O

OQIBB on temperature, 61

on arseniuretted hydrogen, 63,

on silver hyponitrite, 185

, combustion vessel, 241

Organic compounds, forms of, 138
Oxalates, 127

, metallic, 364

Oxalic acid, 365, 366

Oxamide, formation of, 258, 259

Oxidation by nitric acid, 200

Oxyammonia, 245

Oxysalts, 127

formation of solid, 133, 134

PAMABD on dynamite, 76
Panclastite, 375, 397, 549
PAPIN on motive power, 396
PELIGOT on nitric peroxide, etc., 190,

196

PELLET on vibrations, 80, 82
PELOUZE on nitric compounds, 444
Perchlorates, 351
Perchloric acid, 351

ethers, 474

Pernitro-cellulose, 286

Phosphates, 126

Phosphorus, 64

Picrate, potassium, 463, 465

Picrates, 127, 277, 461

Picric acid, 277, 471

PIOBERT on combustion of powder, 47-

49, 512

on gun-cotton and powder, 449

, empirical powder-ratio of, 452

, Treatise of, 477

PIBIA on the Solfatara, 393

PLAATS, VAN DEB, on silver hypo-

nitrite, 185

Porous bodies, function of, 218
Potassium carbonate and sulphurous

acid, 483

chlorate, 407, 520-522

, tables of total combustion

by, 525

cyanate, 340

cyanide, 315, 342

. ferrocyanide, 327

hyposulphite, 497

iodate in solution, 360, 361

nitrite, 167

perchlorate, 521

Potassium permanganate, 177

picrate, 278

sulphate, 482

POTHIEB on slow action, 381
Powder, black, 477

, service, 477

, sporting, 477, 512

, blasting, 477, 512

, properties of, 494

, densities of. 494

, products of initial state, 495

, final state, 496

, , final, variations in, 498

, combustion theory, 499

, equations of, 499

, pressures of, 501

, observation and theory compared,

507

, heat liberated, 501

, volume of gases, 501

, total combustion of, 400

, saltpetre and charcoal, 491

, saltpetre and sulphur, 491

, saltpetre, sulphur, and carbon, 491

Powders with a lead nitrate base, 517

with a nitrate base, 477-517

with a barium nitrate base, 517

with a chlorate base, 518-526

with a sodium nitrate base, 515

, chlorated, properly so called,

523-526

, combustion lists, 525

Pressure, calculation of, 26, 140

, gaseous, 9, 18, 20

- , measurements of, 20

, specific, 28

, characteristic product of, 32

PBIESTLEY on the electric spark, 191,

192

Primers, importance of, 54
Priming, manufacture of, 297
Products, formation of gaseous, 11 ^
Proposition on chemical combination,

11

Prussian blue, 328
Pyroxyl, 433

, wood-pulp, 445

, wood-meal, 459

RAMMELSBEBG on sulphites, 484
Eandanite, 75
Eapidity, molecular, 38, 531
Reactions, 35

, heat in gaseous, 10, 11

, measure of, 14

, pressure in explosive, 20

— , duration of explosive, 35

, general ideas of, 35

, origin of, 36

, molecular rapidity of, 38, 531

, analysis of, 42

, rapidity of propagation of, 47

2 o

562

INDEX.

Keactions, transmission of, 53

, intermediate mode of, 57

, difference in, 63

. See under the substances

between sulphur, carbon, their

oxide, and salts, 479
RECHENBERG on phenol, 277

on nitromannite, 283

REFFYE, DE, apparatus of, 20
Be'gime of detonation, 112

of ordinary combustion, 112

REGNAULT on pressure, 10

on specific heat, 19

and the water calorimeter, 145

, eudiometer of, 151

on vaporisation of water, 158

on absorption of ammonia, 224

RICQ, Captain, the recorder of, 21
RODMAN punch, the, 20
ROSCOE on perchloric acids, 351, 352
Roux on explosions, 52

on charges for bombs, 58

on heat liberation, 509, 514

RUHMKORFF, machine of, '/20-229
RUMFORD, apparatus of, 20

S

SAINT-ROBERT, DE, on ballistics, 16

on velocity of powder- combustion,

49

Saline compounds, 482

Salts, formation of, 117

, solid, 126, 127

in solution, 135

and acids, 119

SARRATJ on explosion of nitroglycerin,
16

, memoirs of, 16

, researches of, 23, 29

, new method of, 25

on dynamite, 31

on gun-cotton, 52, 288, 449, 450,

451, 458

on explosions, 55

on charges for a bomb-shell, 58

on powder-combustion, 205

on picrates, 277

• on nitroglycerin, 283, 424, 425

on nitromannite, 428, 429

— on picric acid, 462

on potassium picrate, 463

on calibration of crushers, 472

on an air-bomb, 509

SCHEURER-KESTNER on coal-combus-
tion, 489

SCHISCHKOFF on potassium nitrate,
160

on powders, 477

on heat liberation of powders, 509

on volume of gases of powders,

510

SCHLOESING on the regime of combus-
tion, 113

SCHLOESING on ozone, 208

on oxidation of ammonia, 214

— , cold process of, 311, 447
SCHONBEIN on ozone, 208, 218

, discovery of gun-cotton by, 444

SCHONE on sulphur and carbon, 483
Schulze chronograph, 21, 22

powders, 445, 517

, analysis of, 459

SEBERT, memoirs of, 16

, experiments of, 17

, apparatus of, 2i

on velocities of propagation, 55

449

— on gun-cotton, 445
Selenide, nitrogen, 253

, VERNEUIL'S formula of, 26S

, an endothermal compound, 263

Shock, phenomena of, 50
SILBERMANN, mercury calorimeter of,

145

. See FAVRE

Silent discharge, action of the, 222

Silica, amorphous, 436

Silver cyanide, from acid and base, 233

fulminate, nature of, 470

hyponitrite, 185

nitrate, 166

oxalate, 366, 475

Sodium acetate, 416

nitrate, 515, 516

nitrite, formation of, 167

Solfatara, phenomenon of the, 393
SOUBEYRAN on a chlorine compound.

240

Stability, tests of, 381
STAHL on nitric acid, 207
Stirrer, the, 147
Succinates, 126
Sulphates, 126
Sulphide, nitrogen, 261
Sulphur and potassium carbonate, 483
Sulphurous acid gas and carbon, 479

and carbonic oxide, 481

gas, decomposition of, 479

Summary — Book L, 527
, Book II. 535

— , Book III., 536
Synchronous vibration, 80, 533

Table of authorities, 124

of compounds, 125-144

of weights of gases, 542

TAIT on nitrogen monoxide, 191
Tamping, 41, 56
Tartrates, 127

Temperatures, theoretical, 64
THENARD on acetylene, 228
Thermo-chemistry, 114, 527

and molecular work, 114, 115

, calorific equivalence, etc., 115-

121

INDEX.

563

Thermochemistry, theorems on re-
actions, 115-117

, salts, 117-119

, organic compounds, 119, 120

, heat of combination, 120-122

, maximum work, 122-124

, principles of, first, 114

, , second, 115

f 1 third, 122

THOMSEN and the water calorimeter, 145

on potassium nitrate, 160

on nitric oxide, 161

on nitrogen trioxide, 167

on nitric oxide and oxygen, 168

on nitric peroxide, 169

on ammonia, 237, 238, 242, 302

on hydrocyanic acid, 303

on chloric acid and chlorates,

344-347

on bromic acid, 357

on iodic acid, 361

THOTJVENEL on nitrification, 210
TRATJZL on priming, 80

, " La Dynamite " of, 432

on the pyroxyl base of dynamite,

433

on nitroglycerin and dynamite,

441

Trimethylamine, 255
Trinitrophenol, 277
TROMENENO on powders, 509
TROOST, 124

on nitrated derivatives, 264

• on sulphur, 402

TRTJCHOT on soils, 235

TURPIN on nitric peroxide, 397, 398

U

UCHATITJS, eprouvette of, 20
UPPMANN on powder, 20 et passim

Vacuum, explosives in a, 48
Valerian ic acil, 187
Vapours, acid, 341
Vegetation, 217, 233, 244

Veloci meter, 21

Velocity of powder-combustion, 48

et passim

VERNEUIL on nitrogen selenide, 263
Vibration, sonorous, 82, 87

, synchronous, 80, 533

VIEIJJ.E on nitrogen sulphide, 261

on diazobenzene nitrate, 291

on detonating mixtures, 389,

391

on nitrocelluloses, 446, 447, 449

. See SARRAU

VIGNOTTI on powder, 510
VINCENT on trimethylamine hydro-
chloride, 257

VIOLETTE on inflammation of powder,
493

on mixture of powder, 513

VOQT on nitroglycerin, 51, 439
Volatilisation, table of, 140
Voltaic arc, action of the, 221
Volume, gaseous, 8, 9, 18, and under
the substances

W

Water, oxygenated, 86
WEBER on nitrogen pentoxide, 181
WETZLER on sodium nitrate, 515
WIEDEMANN on gases, 11
WOOD on potassium nitrate, 160

Xanthale, 517
Xanthine, 517
Xyloidin,264,286

Zinc acetate, 486

oxide, 130, 486

, specific heat of, 141

, density of, 144

and cyanogen, 318

ZORN on hyponitrite, 185
on hyponitrous acid, 186

CONDON : PRINTED BT WILLIAM CLOWES AND SONS, LIMITED,
STAMFORD STREET AMD CHARING CROSS.

RETURN TO the circulation desk of any
University of California Library
or to the

NORTHERN REGIONAL LIBRARY FACILITY
Bldg. 400, Richmond Field Station
University of California
Richmond, CA 94804-4698

ALL BOOKS MAY BE RECALLED AFTER 7 DAYS
2-month loans may be renewed by calling

(415) 642-6753
1-year loans may be recharged by bringing books

to NRLF
Renewals and recharges may be made 4 days

prior to due date

DUE AS STAMPED BELOW

JUL 0 1 1991

21A-50m-8,'61
(Cl795slO)476B

General Library

University of California

Berkeley

YC

LIBRARY USE

RETURN TO DESK FROM WHICH BORROWED

LOAN DEPT.

THIS BOOK IS DUE BEFORE CLOSING TIME
ON LAST DATE STAMPED BELOW

LD 62A-50m-2,'64
(E3494slO)9412A

General Library

University of California

Berkeley