Full text of "The Life of the Hon. Henry Cavendish: Including Abstracts of His More ..."

See other formats

Google

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project

to make the world's books discoverable online.

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover.

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the

publisher to a library and finally to you.

Usage guidelines

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to
prevent abuse by commercial parties, including placing technical restrictions on automated querying.
We also ask that you:

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for
personal, non-commercial purposes.

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the
use of public domain materials for these purposes and may be able to help.

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find
additional materials through Google Book Search. Please do not remove it.

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner
anywhere in the world. Copyright infringement liabili^ can be quite severe.

About Google Book Search

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web

at |http: //books .google .com/I

tfr-'J

"if-

li-;.

•5fl|!

iilMillMililETY.

S urn s.w-w. s •••• .*.." S .

^^ilil

».

Ill

THE '

LIFE

THE HON^?^ HENRY CAVENDISH,

INCLUDIirO

ABSTRACTS OP HIS MORE IMPORTANT

SOIBNTIPIO PAPERS,

AffB A

CRITICAL INQUIBT INTO THE CLAIMS OF ALL THE ALLEGED
DISCOVERERS OF THE COMPOSITION OF WATER.

GEORGE WILSON, M.D., F.R.S.E.

LECTUBE& ON CHEinSTRY, EDINBURGH.

LONDON:
PRINTED FOR THE CAVENDISH SOCIETY.

MDCCCLI.

(.ON DON :

PRINTBD BY HARRISON AND SOU,

ST. martin's LANE.

PREFACE.

When the Council of the Cavendish Society did me the
honour to ask me to write the life of the great philosopher with
whom the Society had associated itself by its name> I willingly
undertook the task. During the enforced leisure of a long
illness^ I commenced, in 1842, to collect materials for a pro-
jected work on the lives of the Chemists of Great Britain, in
which Cavendish should occupy a prominent place ; and I had
made some progress in my task when the Cavendish Society
was founded. Although an original member of that association,
I had no share in determining the selection of the name by
which it is distinguished, nor was Cavrendish an object of greater
interest to me than the other great philosophers of our country,
whose lives I proposed to write. When, however, at the call
of the Society, I laid aside the more general undertaking in
which I was engaged, and turned my attention solely to the
works and character of the Honourable Henry Cavendish, cir-
cumstances had occurred which gave him an importance in the
eyes of the lettered public, such as no other chemist at the time
possessed. He prosecuted zealously and successfully so many
branches of knowledge, that the students of nearly all the
physical sciences may consider him as an illustrious brother;
nor have I any wish to assert that Chemistry is entitled to
claim him aa peculiarly hers* It so happens, however, that his
memory haa been specially honoured by chemists, among whom
Sir Humphry Davy, Faraday, and Thomas Thomson have been

139548

VI PREFACE.

foremost. And it is also the case, that his chemical essays have
furnished to some the occasion for a denial of his intellectual
capacity and of his moral worthy which has to a great extent
thrown his defence into the hands of the chemists. I have written
thisYolumeas a student of chemistry; and having only a limited
space at my disposal^ I have dwelt less upon Cavendish's purely
physical researches^ than I should have done had I been free to
expatiate upon his merits as a natural philosopher. His physical
researches^ however, especially those on electricity and on the J
density of the earth, have not been overlooked in the suc-
ceeding pages; and the value of these memoirs is so fully appre-
ciated by men of science, that they do not demand special criti-
cism. I have given prominence, accordingly, to his discoveries
in chemistry, and in the science of heat, but especiaUy to the
former. It has been impossible to do otherwise. Within a
very recent period, Cavendish has been the occasion of the
keenest controversy that has interested chemists for a long time,
and much of this volume is occupied with its discussion. The
controversy turns upon the question. Who discovered the com*
position of water, — Cavendish, Watt, or Lavoisier? and it has
been prosecuted at greatest length in reference to the rival
claims of the English philosophers. The points in debate are
not merely questions of priority, and of relative intellectual
merit, but also of moraUty ; for charges of plagiariflm, and of
unfair dealing towards each other, have been brought against
the rivals, nor have their friends and acquaintances escaped
reproach, including the entire Boyal Society at one period of
its existence. Cavendish, in truth, has during the last ten
years been the object of attack or of defence to a much larger
number of writers of great eminence, belonging to different
professions, than any one could have anticipated would interest
themselves in the reputed author of a solitary discovery made
eighty years ago. I have undertaken, accordingly, a delicate
and difficult task, in writing a work which compels me to pass
under review the judgments of men of such note in science and
letters as Arago, Dumas, Brougham, Brewster, Jeffrey, Har-

PREFACB. Vll

court, Whewell, and Peacock, at whose feet I have been accus-
tomed to sit as a humble disciple. I may be allowed, there-
fore, to explain briefly in what spirit I have undertaken my

The volume consists essentially of three distinct portions.
The first, a biographical narrative ; the second, abstracts of
0(nentific papers ; the .third, a criticism of the asserted merits
of all the claimants of the discovery of the composition of
water* This critical inquiry has, for convenience of reference,
been printed immediately after the abstracts of the chemical
pi^rs, but those upon heat throw light upon it also* In the
abstracts there is nothing polemicaL It is otherwise with
portions of the biography, and with the critical inquiry.

It was open to me to write as a partizan, as an advocate, or
as an historian* I have chosen the last character as the only
befitting one. I do not pretend to bear witness to my own
impartiality, of which others must be the judges, but I can at
least testify to the spirit in which I have sought to write; and
candid readers, I think, will acquit me of partizanship. The
conclusions to which I have come in reference to Cavendish's
priority and merits as a discoverer, and his integrity as a man,
are such that I can rank myself amongst his most hearty
admirers and defenders. Had I written, however, ordy as his
advocate, I should have left much unnoticed which I have
recorded* Thus I have been at pains to point out the defects
of his theories, as well as their excellences, and to indicate the
merits of his rivals, as well as their faults. The reputation of
Lavoisier, and of Watt, is as sacred a thing in my eyes as
that of Cavendish ; and I should be the first to regret if the
tone of this work should seem at variance with the catholic
spirit of esteem for all great philosophers, which is an essential
elemenl of vitality in associations like the Cavendish Society.
Whilst thus, however, I have endeavoured to be impartial,
and to make the biography a faithful sketch, not a eulogy, I
have deemed it an essential part of my duty as a biographer to
vindicate the moral character of Cavendish from even the

Vm PREFACE.

shadow of suspicion. It has been impossible to do this, ^viritli-
out censuring those who have called his good name in question.
If in uttering censure I have forgotten what is due to great
authorities in literature and in science, even when they are in
error, I shall deserve and bow to reproof; but if I have only
reluctantly fulfilled an imperative though invidious duty^ and
have justified my censures by showing that they are deserved^
I shall hope to be vindicated at the hands of my readers.

I count it a great advantage that I had studied all the ]l
earlier portion of the literature of what may be succinctly styled \
the Water Controversy, before I had any temptation to take a 1
side in the dispute. I also congratulate myself on having been
compelled to look at Cavendish's discoveries and character
from two exactly opposite points of view. In 1845, Mr. i
Muirhead, the able editor of " the Correspondence of the late if
James Watt on his discovery of the theory of the Composition J
of Water," was introduced to me, and by that gentleman, who '
is the most zealous of Watt's defenders, and the most unhesi-
tating of Cavendish's assailants, I had everything that could
be said in favour of Watt urged upon me in the strongest
terms. The publication, also, of the Watt Correspondence in
1846, led to my obtaining the friendship of the late lamented
Lord Jeflfrey. He had known and esteemed Watt, and he
welcomed the publication of the Watt Correspondence, as
furnishing a becoming occasion for exalting the honour of
his old friend. Before his Lordship published his judgment
on the rival claims of Cavendish and Watt in the Edinburgh
Remew for 1848, I had many conversations with him on the
subject. Chemistry was a science in which he had always
taken great interest, and it continued to the last to engage
his attention. With his estimate of the relative merits of
Cavendish and Watt I could not concur, and he listened to my
earnest defence of the former with all the frank courtesy and
love of fair dealing which so eminently characterized him.
Against Cavendish he entertained no animosity or prejudice,
and he was most willing to praise him ; but he thought that

PREFACE. IX

liV^att had been wronged, and he was solicitous to see him
righted, so that he pressed me with all the arguments which
he perceived might be urged in favour of his great client,
whose case he has so skilftiUy and eloquently pleaded. He
did not even refuse to discuss (I may say to debate) contested
points with me, and I defended Cavendish in the strongest
terms which courtesy sanctioned.

My zeal in Cavendish's cause made no difference in Lord
Jeffi-ey's kindly dealings towards me, and he was the first in
I whose hands I purposed to place this volume, in which many of
his conclusions are called in question.

Having thus had the claims of Cavendish's English rival
brought before me in the amplest way, I have been secured
against under-estimating what may be said in favour of Watt.
Lord Jeffrey's article, indeed, is by much the ablest defence of
Watt that has appeared.

After Lord Jefirey^s decease, the Rev. William Vernon
Harcourt, the ablest of Cavendish's defenders, most kindly
put himself in communication with me, and furnished me with
his estimate of the position in which Cavendish's claims were
placed by the publications in favour of Watt, which had
appeared since 1846. I cannot concur in all Mr. Harcourt's
conclusions, but I am indebted to him for many valuable
suggestions, and for much assistance. In particular I owe to
him an introduction to the Earl of Burlington, who placed at
my disposal the whole of Cavendish'^s papers in his possession,
and obtained for me much information concerning his illustrious
ancestor's personal history. The papers on Electricity which
Cavendish left behind him, are at present in the hands of that
accomplished Electrician, Sir W. Snow Harris, who, in the
kindest manner, drew up for me an abstract of them, accom-
panied by a commentary. It is matter of great regret to me,
that I have not been able to print either the abstract or the
commentary in this volume ; but I trust that they will yet be
made public^.

I have thus had access to many unpublished documents,
which are fitted to throw light on Cavendish's merits and his

PREFACE.

personality, and I have largely availed myself of them. For
the opinions expressed in this work, I alone am responsible.
I have accepted and solicited information and assistance from
every party known to me, willing or likely to furnish aid ; but
as it was manifestly impossible for a single writer to represent
the diversified opinions of all the members of a large societjr
upon a contested question, I requested the Council of the
Cavendish Society to allow me to write in my own name. No
one accordingly but myself is committed to the conclusions
contained in this volume.

It remains for me to express my obligation to the many
scientific men who have assisted me in this work. To Robert
Brown, Esq., of the British Museum, I have been indebted for
interesting particulars concerning Cavendish and Blagden, and
I am under similar obligations to Dr. Thomas Thomson, of
Glasgow. M. Fran9ois Delessert, of Paris, also, has made me
his debtor for much valuable information. From B. H. Blagden
Hale, Esq., of Cottles, I have received various interesting
papers, which have been of very great service to me. To 4
Dr. Percy, and Dr. James Russell, of Birmingham, I am under
great obligations for their good offices in procuring for me the
loan of a number of important unpublished letters of Dr.
Priestley's, which have been confided to my care by his grand-
daughter. Miss Finch. Francis Wedgwood, Esq., of Barlaston,
has allowed me access to the papers of his celebrated ancestor,
Josiah Wedgwood; and through Professor Graham, I have
obtained from Mr. Hudson, the Secretary of the Royal Agri- |
cultural Society of England, several of the very few extant
letters of Cavendish. To Mr. Redwood, also I am much ,
indebted.

Dr. Davy has largely contributed to the materials from
which the biography has been written, and so have Professor
Brande, Mr. Konig, of the British Museum, W. H. Pepys,
Esq., J. G. Children, Esq., and Henry Lawson, Esq., of Bath.
I have to thank the Rev. Joseph Romilly, and Frederick
Fuller, Esq., of Cambridge, the Rev. C. J. Heathcote, of
Upper Clapton, and C. R. Weld, Esq., of the Royal Society,

I «

■I
.1

PREFACE. XI

for their ready communication of all the information which
their official positions enabled them to fiimish in answer to my
queries. I have also to acknowledge the liberality with which
my fiiend^ the Bey. Dr. Yaughan, of Manchester^ has permitted
me to make any use I pleased of papers contributed to the
JSriHsh Quarterly Review , in which I published, in 1845, a short
biographical sketch of Cavendish. To others also I am indebted,
but it might savour of parade to enumerate the names of each.
Once for all I would say, that from every one to whom I
have applied for information, I have received it, and that much
that is most valuable in this volume has been unsolicitedly sent
to me. I have reserved for special thanks Charles Tomlinson,
Esq., of London, to whose untiring zeal, skilful investigation,
and cordial unflagging co-operation, I am indebted for the
larger part of the materials from which the last chapter of the
Biography has been compiled. To him I am also indebted for
the striking and hitherto unknown portrait which graces this
volume, and he has done me the favour to read the proofs of
this work, besides assisting me in every other way m which
either he or I perceived that he could be of service.

I have lastly to notice that since the publication of the fVati
Corresponderice, in 1846, the only lengthened notices which have
appeared in reference to Cavendish, have been Sir David
Brewster's Article in the North British Review for 1847, and
Lord Jef&ey's Paper in the Edinburgh Review for 1848. Both
of these writers pronounce against Cavendish, and refer to the
Watt Correspondence as decisive of the merits of Watt ; but I
tlunk it will appear from the following pages, that the admirers
of Cavendish have every reason to congratulate themselves on
the publication of the "Correspondence;" and for my own
part I believe that it furnishes the most decisive evidence in
favour of Cavendish, and as such I have constantly quoted
from it.

24, Browx Square, Edinburgh,
March i 1851.

CONTENTS.

Life of Oavbhdish.

Pagb

Chap. I. Genealogy of the Cavendishes. Early history of the Hon.

Henry Cavendish 1

II. Qeneral Sketch of Cavendish's scientific researches and

discoveries 19

III. Controversy between Cavendish, Watt, and Lavoisier, con-

cerning the discovery of the Composition of Water 54

IV . Concluding events of Cavendish's life. — ^Estimate of his moral

and inteUectaal character 158

Cavsvdish as a Chemist 1.91

Three Papers containing Experiments on Factitious Air .... 195

Experiments on Rathbone Place Water 209

An Account of ti New Eudiometer 215

Experiments on Air 231

Experiments on Air. Second Series 255

A Critical Ivqvibt ihto the Claims or all the alleged Authors of
THE Discovery of the Compositioit of Water. The Water
Controversy.

1. Preliminary Discussion 265

2. Bibliography of the Water Controversy 269

3. Questions in dispute between the Principals in the Water

Controversy 277

4. Researches which led to the discovery of the Composition of

Water 279

Question of Reality. Nature of the Discovery claimed by Ca-
vendish, Watt, and Lavoisier, and imputed to Mongb.

5. Cavendish's experiments and conclusions concerning the

Composition of Water 282

6. Priestley's experiments, and Watt's conclusions from them,

concerning the Composition of Water 265

7. On the signification of the term inflammable air, as used

by Watt to denote the combustible element of Water 297

XIV CONTENTS.

8. On the fall signification of the term phlogiston, as employed

by CaYendisha nd Watt 319^

9. Experiments and conclusions of Lavoisier concerning the

production of Water from its elements 33*

10. Experiments and conclusions of Monge concerning the result

of the inflammation of hydrogen and oxygen in dose
vessels 347 «

QUBSTIOV OF PbIOBITT. '

11. Who first discovered and taught that Water is a compound /

of hydrogen and oxygen ? Date of Cavendish's experi-
ments and conclusions 353

12. Date of Priestley's experiments, and of Watt's conclusions

from them concerning the Composition of Water 404

13. Date of Lavoisier's conclusions concerning the composition of

Water 406

Question of Plagiarism.

14. Allied Plagiarism of Cavendish 407

15. Interpolations in Cavendish's and Watt's papers of date

1784 413

16. Erroueous dates in Cavendish's and Watt's papers of 1784.... 419

17. Alleged plagiarism of Lavoisier 426

Genebal SumABT.

18. Relative merits of Cavendish, Watt, and Lavoiaer 432

Cavendish as a Natural Philosopheb.

Papers on Heat 446

An attempt to explain some of the principal phenomena of
Electricity by means of an elastic fluid 466

An account of some attempts to imitate the effects of the
Torpedo by Electricity 466

Experiments to determine the density of the earth 470

Cavendish's Appabatus.

Metallic Eudiometer 476

•

LIFE OF CAVENDISH.

CHAPTER I.

GENEALOGY OP THE CAVENDISHES. EARLY HISTORY OF THE

HON. HENRY CAVENDISH.

The great majority of the distinguished Chemists of Great
Britain have sprang from the middle or lower ranks of the
people^ but two of the most famous of them^ the Honourable
Robert Boyle^ and the Honourable Henry Cavendish, were men
of illustrious lineage, and Cavendish was much the more high
born of the two.*

No one could well be more indifferent, than Henry
Cavendish was, to the external advantages which birth and
fortune gave him, yet few of those who set the greatest value
on these, could boast of a descent such as his.

His family traced their, pedigree, by unbroken and unques-
tionable linLs, to Sir John Cavendish, Lord Chief Justice of the
King's Bench in the reign of Edward III ; and, according to
learned genealogists, they could go much further back, and
derive their descent from a Norman family, famous in the
days of the Conquest. Cavendish could thus look back, across
eight centuries, to the founder of his family, and through the long
interval which elapsed between the first appearance of his
ancestors in England, and his own days, could point to his prede-

* The blood of these fiimiliea has mingled within a recent period. Lady
Cbarlotte^ the daughter and heir of Richard Bojle, Earl of Burlington and Cork,
Was married to William, fourth Duke of Deronahire, and was mother of William,
the fifth duke, as weU as of Lord George Henrj Cavendish and others. — (Collins'
Petroffe, 4th edition, vol. i., p. 333.) The Earl of Burlington is thus by descent
both a Cavendish and a Boyle, and has a pedigree on which a lover of science must
look back with peculiar pleasure.

2 LIFE OF CAVENDISH.

cessors as famous in the history of their country^ and as connected
by intermarriages with the most illustrious houses of the king-%
dom^ not excepting the royal families of England and Scotland. ^t

Some doubts have been expressed by Sir Egerton Brydges J
in his edition of Collins^ Peerage^ as to the descent of Lordl,
Chief Justice Cavendish^ from the Norman Gemons,t but even/
he expresses only a doubt, and that by no means strongly ; and/
those learned genealogists, Dugdale| and Collins,§ both unhesi-'
tatingly derive the Cavendishes " from Robert de Gernon, a-
famous Norman who assisted William the Conqueror in his
invasion of this realm, A.d. 1066 :" so that I shall take for
granted that he was their forefather.

It would be out of place in a sketch like the present^ to
enter into any minute particulars concerning a family history
so well known as that of the Cavendishes ; I shall confine myself^
therefor^ to the names of a few of its more distinguished mem-
bers, so as to connect together the Norman warrior of the 11th
century, with the philosopher of the 18th.

Robert de Gernon received laige grants of land after the
Conquest, and his immediate descendants were long famous in
Norfolk and Essex. A younger son of the family, Roger
Gernon, seated at Grimstone Hall, in Suffolk, died in 1318,
having had to wife the daughter and heir of John Potton, Lord
of Cavendish, in Suffolk, by whom he left issue four sons, who
all took the surname of Cavendish, as was customary in those
days. II

This surname is variously spelled by its older possessors,

* Lady Arabella Stoart^ who was grand-daaghter of Sir William Cavendish, a
lineal ancestor of the philosopher, was great-grand-daoghter of Margaret Tudor,
daughter of Henry VII. and Elizabeth of York. She was thus the niece of Mary,
Queen of Scotland, and cousin of her son James I.— Miss Strickland's lAvu qf tht;
Queena qf Scotland, toL i., preface, p. ix.

t Vol. i., p. 302.

X Baranoffe ofEngUmd, vol. ii.> 1676, p. 420.

§ Peerage of Englandt 4th edition, vol. i., pp. 279 — 283.

11 Ihidt p. 281 : " Our surnames are chiefly derived fh>m this origin [a
territorial possession] or from personal peculiarities, — from trades and employments,
or from the Christian name of the father or mother. Of these, the first is the most
aristocratic, denoting a descent from an ancient baron, or at least, the lord of a
manor." — (Lord Campbell's Livea qf the Chancellors, vol. v. page 174.) The
Btimame Cavendish comes within the last category, being derived from the manor
of which Potton had the lordship.

LORD pHIEF JUSTICE CAVENDISH. S

lavendisbe^ Cavendysh, Cavendjrsich^ but \b now uniyeraally
itten Cavendish. It was also frequently written Caundish
ind Candisb, and even, when spelled Cavendish^ was pronounced
a dissyllable. That accomplished critic and scholar^ Thomas
^e Quincey, has pointed out to me, tiiat up to a recent period,
id perhaps even at present, Cavendish has been pronounced
if it were written Candish. In illustration of this he adduced
Sonnet of Wordsworth^ which cannot be read rhythmically^
less Cavendish be made a dissyllable. The passage is as
roUows :—

AT PURMBflS ABBIT.

« * * * • i»

" Even as I speak tbo rising Sim's first smile
Gleams on the grass-crowned top of yon tall Tower,
Whose cawing occupants with joy proclaim
PrescriptiTe title to the shattered pile
Where, Cavendish^ thine seems nothing hot a name." !*

The first bearer of the name of Cavendish was the Chief
Justice already referred to. His father's property was small,
and his prospects indifferent. He devoted himself, however,
to the study of the law, and soon acquired great reputation as
an advocate. *' Such was his reputation,'' Lord Campbell tells
us, ''that in the year 1366, Edward III, after the peace of
Bretigni, being desirous of making himself popular by good
judicial appointments, raised John de Cavendish to the office of
Chief Justice of the King's Bench, although he had not filled
the office of Attorney or Solicitor^General, or even reached the
dignity of the coif.*'t

The fate of this first Cavendish was a sad one. He was
continued in his high office by Richard II with an increased
salary ; but after acting as Judge for sixteen years, he ^as
cruelly murdered in one of the insurrections which marked that
reign. As one of those to whom the suppression of Wat Tyler's
insurrection was entrusted, he had become an object of venge-
ance to the rebels, and their feelings of revenge towards him
appear to have been greatly deepened, by the fact that his son

* Wordsworth's Poetical WorkM^ royal Svo. 1845, p. 217. I am indebted fo^
this reference to my friend, Alan Steyenson, Esq.^ the engineer of the Skerryrore
Ughthoose^ who is profoundly acquainted with the writings of the poet.

t Lives qf the Chitf Juetieee qf England^ vol. 1., p. 94.

4 LIFE OF CAVBNBISH.

and namesake^ an Esquire of the body of Richard 11^ was th
party who slew Wat Tyler, after he was wounded by Sir Willia
Walworth, mayor of London. For this act, the younge
Cavendish was knighted on the spot by King Richard, an
granted a pension, which, however, cost his father his life
When the insurrection revived under the ferocious Jack Straw, [
he plundered and burned the house at Cavendish, and beheaded
the Lord Chief Justice, after cruelly insulting him.*

The immediate descendants of the Chief Justice may b
passed over very briefly. His son, the second Sir John, was
succeeded by his eldest son William, who had an only son^
Thomas, who in turn was succeeded by an only son of the same
name, who was the father of Sir William Cavendish, the founder
of the political greatness of the Cavendishes. Thomas, the
father of Sir William, died in the fifteenth year of Henry VIII,
1524. He had two sons, George and William, both of whom
deserve a place in the history of the Cavendishes : William was
the second son ; he began life with fewer advantages than his
brother, but rose to much greater worldly distinction, and
the merits which belong to George have been added by many
writers to his own. This William Cavendish is the founder of
the modem distinction of his family, or, at least, shares it only
with his celebrated wife, of whom more will be mentioned pre-
sently. His father was a clerk in the Exchequer in the reiga
of Henry YIII, and appears to have trained his son to exact I
business habits, of which he afterwards reaped the reward. He
must have been a man of talent and capacity, for he early
attracted the attention of Henry, and this at a period most ,
fortunate for him. <

' In 1530, he was appointed one of the Commissioners for
visiting, and securing the revenues of the religious houses which
the King was then busily confiscating. He continued in this
and similar offices for several years, and, as might be expected,
received a large share of the confiscated church property. In
1546 he was knighted, and made Treasurer to the King, a place
of great responsibility and honour. He was soon aftenvards
made a Privy Councillor, and received from Edward YI grants

* lAvet qf the Chief JusHce» qf Enfflmd, p. 95. Collins' Peerage, 4th
edition, toI. i., p. 284.

ELIZABETH HARDWICKB. 5

^f lands in seven different counties. He died in 1547, leaving
.large estate to his heirs.*

The descendants of Sir William Cavendish rose with almost
unexampled rapidity to the highest distinctions^ but before this
can be understood^ reference must be made to his wife, a very
remarkable woman, who, as Mr. Hunter has justly remarked,
was more than an equal sharer with him in establishing the
fortunes of the family. Elizabeth Hardwicke occupies a con-
siderable space in the private and public history of the peerage
of our country, and was in all respects an extraordinary person.
She was a younger daughter of a country gentleman in Derby-
shire, who could afford her only a very trifling portion. She
rapidly, however, provided for herself. At the age of fourteen
she was married to Robert Barley or Barlow, of Derbyshire, a
youth little older than herself. He was at the time a confirmed
invalid, and died very soon after his marriage, which appears to
have been merely a form, but he left all his estates settled upon
his bride and her heirs. After fourteen years of maiden widow-
hood, she became the third wife of Sir William Cavendish, who
had no male issue surviving by his former marriages. Three
sons and three daughters, who were bom of this marriage^
survived to establish the greatness of the family. Sir William
Cavendish married this lady in 1547, and died ten years
after. Some four years later she became the wife of Sir William
St. Loe, captain of the guard to Queen Elizabeth, and possessor
of several estates in Gloucestershire. He was a widower, and
had children by his former marriage, but such was his devotion
to his second wife, that, in compliance with her demand, he
settled all his property upon her and her heirs, and as he died
without issue by her, ^^ she lived to enjoy his whole estate,
excluding his former daughters and brothers.^'f

The period of St. Loe's death is uncertain. Before long,
however, Elizabeth Hardwicke was again a wife. Her fourth
and last husband was George Talbot, sixth Earl of Shrews-
bury. He had a family by a former wife, and would not
settle his property away from them. But Mistress St. Loe,

♦ Collins' Peeragct toI. i., p. 289. Rev. Joseph Hunter, in Singer's Caven*
dUh, voL U. p. xxxiY.

+ Craik's Rtnnance qf the Peerage, vol. iii., p. 149.

6 • LIFE OF CAVENDISH.

would not accept him till he had arranged to give two cs€
his children in marriage to two of hers. Her eldest son, accor-
dingly, Henry Cavendish, was married to the Earl's daughter, the
Lady Grace Talbot, and her daughter Mary to his second son and
ultimate successor Gilbert. Elizabeth Hardwicke had no children
by the Earl of Shrewsbury, so that all the wealth and influence
she had gained by her four marriages^ were brought to bear upon
the advancement of her children by Sir William Cavendish.
The Earl of Shrewsbury^ at the period of his second marriagCi
was the most powerful nobleman of the realm, and his Countess
shared with him in the responsible office of warden of the
unfortunate Queen Mary of Scotland. This office brought the
Countess into frequent communication with Queen Elizabeth,
with whom she was a favourite. The Earl died twenty-
three years after his marriage to Mistress St. Loe^ and she
survived him seventeen years, dying in 1608, on the threshold of
her ninetieth year; seventy-five years after the death of her
'first husband. Her whole history concerns the Cavendishes,
for they were both directly and indirectly gainers by all her
marriages. She is said to have been beautiful, but she was
fifty before she married George Talbot, and her whole career
points to other gifts than that of mere beauty as the source of
her prosperity. She must, however, have been a fasci-
nating person, for all her husbands were devotedly attached
to her^ although in the end, her last spouse, who had the longest
trial of her, exchanged his early devotion for something like
positive hatred, fairly tired out of his patience by her insatiable
ambition and rapacity. The Countess, nevertheless, was far more
than a match for the well-intentioned, but stolid Earl Talbot.
She has been hardly judged by most of her biographers. Sir
Egerton Brydges accuses her of rapacity, from which she cannot
easily be defended ; Mr. Hunter accuses her of ** reserve, perfidy,
and even tyranny towards her last husband,^' and quotes with
approbation Mr. Lodgers description of her, that she was
'^ a woman of masculine understanding and conduct ; proud,
furious, selfish, and unfeeling.^' Her latest biographer. Professor
Craik, is less severe; and the better points of her character appear
to have been overlooked by her harsher judges. The affectionate
terms in which she is referred to by her last three husbands,

SIR Wi;.LIAM cavendish's FAMILY. 7

-nrfaose letters still remain in attestation of their regard for bet,
«re sufficient proofs that, with all her shortcomings, there was
much that was loveable about her. And even from the least
favourable accounts of her that have come down to us, it seem«
plain that there was nothing mean, malignant, or cruel in her
character. On the other hand, she was bold, resolute, and
straightforward in her dealings, and spared no pains to advance
the wel&re of those to whom she was attached. Her worst
faults were her insatiable ambition, and love of pomp and mag-
nificence, which could not be gratified without the command of
an amoimt of wealth which it required skilful management to
procure. She had a passion for building, which her descendants
inherited, and spent enormous sums in the erection of splendid
palaces, which should perpetuate her name. One of these (Hard-
wicke Hall) still remains^ and ranks among the noblest man-
eions in Derbyshire. A second, named Oldcotes, has disappeared;
and the splendid structure which she erected at Chatsworth was
taken down by her descendant, the first Duke of Devonshire,
and replaced by the still more magnificent building which now
occupies its site.*

Whatever were the Countess of Shrewsbury's faults, she
was indefatigable in promoting the interests of her children ;
and not more indefatigable than successful. This may be judged by
the fact, that whereas she was originally, the nearly portionless
daughter of a country Squire, and Sir WiUiam Cavendish, though
of good descent, but a younger son of a clerk in the Exchequer^
the whole of their children, who survived infancy, either obtained
titles, or were married to those who had them. One son
was Knighted, another was made Baron and Earl, two of the
•daughters became Countesses, a grandson was made a Duke,
and a grand-daughter. Lady Arabella Stuart^ was for some years
heiress presumptive to the throne of England. Another duke*-

* The acoonnt of Eliiabetfa Hardwicke in the text, I have taken partly from
Collins' Peerage, vol. i., article Cavendish, Duke of Devonshire; and from Dr.
White Kennet's Memoir* of the Family qf Caoendiah, p. 65. I have been chiefly
indebted, however, to the sketch of her life contained in the Rev. Joseph Hunter's
Enayt Who wrote C(nendish*9 WolaeyT published in, Singer's Cavendish, vol. ii., p.
ilvi.; and likewise to the very full and lively account of this lady given by Professor
Craik in his paper entitled <* Bess of Hardwick and the Talbots," in his Romance
qf the Peerage, vol. iil, p. 145.

8 LIFE OF CAVENDISH.

dom was added to the family in a later generation, and their
fortunes have ever since been in the ascendant. For this unu*
sually rapid elevation the Cavendishes were mainly indebted, at
kast in the first generation, to their mother ; for their &thef,
it will be remembered, did not survive his third marriage more
than ten years, and he could do little for the advancement of
his family. It was chiefly, indeed, through the wealth and in-
fluence which her subsequent marriages brought his widow, and
especially through her connexion with the Court, which her
position as Lady St. Loe and Countess of Shrewsbury gave her,
as well as through the interest of her grand-daughter. Lady
Arabella Stuart, James the First's cousin, that she was able to
advance her family so rapidly in social rank. She owed most, how-
ever, to her own energy and boldness, for the connexion by which
she was the greatest gainer, viz., that with the royal family, was
brought about solely by her address, and not without hazard to
herself. The marriage, indeed, which she effected between her
daughter Elizabeth and the Earl of Lennox, James the First's
uncle, excited the wrath of Clueen Elizabeth so much, that she
committed the Countess for some time to the Tower.'*'

The Countess of Shrewsbury wedded Henry, her eldest
son, to the Lady Grace Talbot, a younger daughter of her last
husband. Henry was not, however, her favourite, and in her
own imperial way, she set aside the claims of primogeniture, and
preferred her second son, William, who was raised to the Peer-
age by James the First, and is the progenitor of the Earls and
Dukes of Devonshire, and of Henry Cavendish, who is the
central object in this sketch.

Before, however, tracing his descent, two of his collateral ances*
tors must be referred to, who have a special claim to be remem«
bered in connexion with one whose celebrity depends upon his
intellectual, not upon his social or political greatness. The one of
these is George, the elder brother of Sir William Cavendish,
who may be regarded as the most distinguished student of
literature among Henry's ancestors, and has earned a high place
in the literary annals of our country. He owes this to his
admirable Life of Cardinal JVoheyy which all critics are agreed
in considering *' as one of the very best specimens of English

* Cmik's RomancB of the Peerage^ toI. ii., p. 353.

GEORGE CAVENDISH. 9

biography/' George Cavendish^ nevertheless^ has narrowly missed
bemg defrauded of this honour, which has been transferred by
many writers to his more fortunate younger brother. Sir William,
Later authors were led astray on this matter by Bishop Kennet^
who in his ill-judged zeal to heap honours upon the progenitor
of the first Duke of Devonshire, attributed to him the author-
ship of the Xrf/e ofWoUey^ without, however, giving any express
authority for his statement.*

* Kennet's work is entitled^ A Sermon preached at the funeral qfthe Might
Noble William, Duke qf Dewmehire, 8fe,, with some Memoire qf the family qf
Cmemdieh. Bj Wliite Kennet, D.D., Archdeacon of Huntingdon/' In my copy
of this work, a page has been interleaved by some former possessor, on which the
following corions criticism is written^ " It was by the interest of Bishop Burnet^
that Kennet was appointed the preacher on this occasion, and the sermon gave great
offence, and made some say that ' the preacher had built a bridge to Heaven for
men of wit and parts, but excluded the duller part of mankind from any chance of
passing it.' This charge was grounded on the 34th and 35th pages, but by the
interest of the second Duke, to whom it is dedicated, he immediately obtained the
deanery of Peterborough/' The passage referred to as having given offence in the
sermon, is one in which Dr. Kennet discusses the efficacy of a death-bed repentance,
" This," says he, ''rarely happens but in men of distinguished sence and judgment.
Ordinary abilities may be altogether sunk by a long vitious course of life. The
dnUer flame is easily extinguished. The meaner sinful wretches are commonly
ffiven up to a reprobate mind, and die as stupidly as they lived ; while the nobler
and brighter parts have an advantage of understanding the worth of their soul
before they resign it. If they are allowed the benefit of sickness, they commonly
awake out of their dream of sin, and reflect, and look upward. They acknowledge
an infinite Being; they feel their own immortal part ; they recollect and relish the
Holy Scriptures ; they call for the elders of the church ; they think what to
answer at a judgment seat." The italics are the author's own. The clergyman
who could write thus must have forgotten the most memorable, and most certainly
efficacious case of a dying repentance on record, and what was the character of him
to whom it was said, " This day thou shalt be with me in Paradise."

It may seem a digression to criticise Kennet's work. The memoirs attached to
the sermon, however, are frequently referred to as of special authority in reference to
the history of the Cavendishes. They are unworthy of this praise. The author
had access to no peculiar sources of information, and has committed more than one
error ; but the main &ult of the work consists in its tone, which is that of fulsome
and extravagant adulation of all the Cavendishes, to an extent as unworthy of the
author as it was unnecessary in the case of a family whose intrinsic merits were so
great. Some allowance must be made for the fashion of the time, which permitted
a style of compliment which the taste of the present generation disowns ; and some
palliation, though no apology for the author's fault, maiy be found in the fact tliat he
had an eye to preferment in what he was writing. His compliments, however, are
sometimes even ludicrous, as when he praises the Countess of Shrewsbury for her
dutiful behaviour to her last husband, and represents her as having had her temper
and virtue exercised by the rumours at one time afloat concerning an undue intimacy
between him and Queen Mary of Scotland (p. 70). This was a most unlucky

to LIFE OF CAVBNDISH.

From him the error was transferred to the later editions of
Collins' Peerage, including that by Sir Egerton Brydges,
and likewise into the Biographia Britanmca, and other works*
Kennet's mistake was fully exposed in 1814, but this has not
prevented its re-appearance in Lord Campbell's recent work,
the Lives of the Chief Justices of England.* No one, how-
ever, who has read the Rev. Joseph Hunter's very interesting
tract, entitled WTio wrote Cavendish's lAfe of fFolsey ? t can
doubt for a moment that the author was George Cavendish^
Sir William's elder brother.

The honours of the Cavendishes, it may be noticed, gain in
a two-fold way by the transference of the authorship of the Life
of Wolseyy from William Cavendish to George, for it not only
places two distinguished men instead of one, in the list of their
ancestors, but it saves Sir William Cavendish's character from
a reproach which would have gone far to lessen the honour
which the authorship of the work in question would have con-
ferred upon him. The writer of that work professes to have
been a zealous Roman Catholic, and laments bitterly the ruin
of the religious houses, which the Reformation had occasioned. %
Sir William Cavendish, however, as we have seen, owed hia
advancement in life to the zeal with which he acted as the agent
of Henry VIII in suppressing the monasteries, and to the share
of the spoil which he received. It would be impossible, accord-
ingly, had he been the author of the Life of Wolsey^ to defend
him from the charge of gross hypocrisy and double-dealings
inasmuch as he professed to be a Roman Catholic, and a very
earnest and zealous one, whilst the Cardinal's usher, but never-

reference, for both the Queen aod the Earl of Shrewsbary accused the Countess of
being the very party who spread the scandalous report, and her husband declared to
her, " there cannot be any wife more forgetful of her duty, and less careful to
please her husband, than you have been.'' — (Romance of the Peerage^ vol. iii., p.
221.) The last years, in truth, of the Earl of Shrewsbury's life were embittered by
the quarrels between himself and his Countess, which were matters of public
notoriety. A full and yery curious account of the whole matter will be found in
Mr. Craik's paper, "Bess of Hardwick and the Talbots," already referred to,
Kennet's worst error, however, is that concerning the authorship of the lifo of
Wolsey, and plainly arose from his inconsiderate desire to praise a direct ancestor of
the Duke of Devonshire.

♦ Vol. i., p. 95.

t Reprinted in Singer's Cavendith, vol. ii.

X Singer's Cavendi»h, vol. ii., p. xxxrii.

JHOaUS CAVENDISH. 11

thelesB did not liesitate, when lie became the 'King's servant, to
display the greatest activity in suppressing the religious houses
whose overthrow he had affected so much to deplore. As the
case now stands, however, he is quite free from this stigma on
his character. He was from the first period, at least, of his
public apf>earance, a Protestant, and coidd not be charged with
inconsistency in abetting King Henry. His brother, on the
o&er hand, lived and died a devoted adherent of the old faith,
which condemned him to poverty and obscurity.*

Of the literary merits of the Life of WoUey^ I need not
speak, as they are so universally acknowledged. No later
bearer of the author's name has eclipsed him in literary repu-?
tation, although many of his collateral descendants have been
more accomplished scholars.f

Another ancestor of the Honourable Henry Cavendish was
referred to as deserving a place, in this record. This is Thomas
Cavendish, who, as one of those who first rounded the globe,
appears peculiarly entitled, although only a collateral ancestor,
to claim kin with him who first weighed it. He was a descen-
dant of Roger, the second brother of Lord Chief Justice Caven-
tiish, and the son, therefore, of Roger Gemon. He was, as the *
old writers record, '^ the third man, and the second Englishman
which sailed round the globe.'' He set sail from Plymouth on
July 21st, 1586, and after losing two of his ships, landed safely
at the same port on September 9th, 1588. His second voyage
was less successful. He could not succeed in passing the Straits
of Magellan, and being driven back to the Coasts of Brazil,
he was deserted by many of his associates, and died of grief
and chagrin, at the unfortunate issue of his voyage.

I now return to Sir William Cavendish, with a view to trace
very briefly, the descent of the special subject of my sketch from
him. His. elder brother Henry died without issue, and William,

* The forttme of the brothers was strikingly different. George Cavendish's
son bad to sell the manor from which his family took its name, and his grandson
became a tradesman in London ; whilst William's son became a large landed pro-
prietor, and an earl, and his grandson was' raised to a dnkedom. — Hunter, in Singer's
GsMHifwA, Tol. ii., p. Wiii.

t The best edition of the life of Wolsey is that entitled ** Tk% Lift qf Cardinal
Woi$ey. By George Cavendish, his gentleman-nsher ; with notes by Samuel Wdler
Smger, 1825." A part of Cavendish's narrative has been recently reprinted in the
third edition of Gait's Li/e qf'Wol»ey.

12 LIFE OF CAVSNDISH«

the second son, who was his mother's favourite, inherited a large
estate. His early elevation to the Peerage appears to have been
mainly owing to the influence of his niece. Lady Arabella Stuart^
who was then in favour with James I, and obtained from him a
promise that one of her uncles should be made a Baron,
William was selected for this honour, and became Lord Caven-
dish, accordingly, at the christening of a Princess who died in
infancy.* He was some years later created Earl of Devonshire^

Thus far, it appears that the Cavendishes were singularly
favoured by fortune; and the first Earl seems to have been
more indebted to his claims as the King's cousin, and to the
zealous solicitations of his mother, than to his own merits. He
was a man, however, of good capacity,t and the talent which
displayed itself in his direct and collateral descendants soon
showed that the Cavendishes knew well how to take at its
flood the tide in their affairs, and make it lead them on to
fortune.

The first Earl died in 1625, and was succeeded by his
son. Meanwhile the younger brother of the former, a man of
great ability, was knighted by James I, and took an active
part in the public proceedings of the period. His still abler
son, William, played a conspicuous part in the Civil War, and
was distinguished by his zeal in the Royal cause. He bore in
succession nearly every title of rank, and rose from being a
Knight of the Bath, to the dignities of Earl, Marquis, and Duke
of Newcastle. He is identified with the history of our country;
and his second wife, Margaret, the literary Duchess of New-
castle, has secured for him additional remembrance, by the
pleasing memoir which she has written of him. His only sur«
viving son died without male issue, and the dukedom of New-
castle became extinct. It has since been repeatedly revived.

The second Earl of Devonshire, who was a man of talent,

* Cndk's Romance of the Peerage^ vol. ii., p. 368.

t Collins, in his Peerage^ tells us that he was one of the first adyentarers who
settled a colony and plantation in Virginia. He had also a larg^e grant of land in the
Bermudas, and called his estate there Cavendish, a name which it still retains, or at
least recently did. I presume that from this plantation is derived the name of a
well-known variety of tobacco, which — so strange a thing is fame — ^has spread the
name of Cavendish more widely than all the patriotic deeds, or scientific and literary
achievements of its most illustrious bearers have done.

HENRY CAVENDISH. 13

and possessed of many accomplishments^ sumved his father
only three years, and was succeeded by his son, who, like all
the other Earls as well as Dukes of Devonshire down to the
present day, as well as the founder of the family, bore the name
of WiUiam.*

The third Earl came to the title in his eleventh year, and
bore it for fifty-six years. His son, the fourth Earl, and first
Duke of Devonshire, is the most distinguished member of the
family, unless, perhaps, we except the first Duke of Newcastle.
His forefathers were zealous Royalists, and stood by the Stuarts
in all their vicissitudes of fortune. The fourth Earl, however,
firom his early youth sympathised strongly with those who
opposed the encroachments which James II sought to make
upon the liberties of the English people, and his quick temper
and high sense of honour brought him into direct collision with
the King himself. In the end, accordingly, he took a most
active and prominent part in furthering the accession of William
III, for his services to whom, he was created Duke of Devon-
shire, besides receiving many other honours. He died in 1707^
and was succeeded by his eldest son, William, who died in
1729. Lord Charles Cavendish, the third son of this second
Duke, was the father of the Honourable Henry Cavendish, the
subject of our memoir.f

Lord Charles Cavendish married Lady Anne Grey, fourth
daughter j: of Henry, Duke of Kent, and by her had two sons,
Henry and Frederick.§

• Romance of the Peeraffe, vol. iii., p. 276.

t William, second Duke of Devonshirei died in 1 729, and was succeeded by
his eldest son, Williamy who died in 1755. His successor, the foarth Dnke, died in
1764. William, the fifth duke, died in 1811, and was succeeded by the present
bearer of the title, who is the sixth duke, and ninth earl of DeTonshire.

X In the article "Cavendish," in Collins* Peerage^ she is stated to have been
the Duke of Kent's tkird daughter (4th edition, vol. i., p. 330). In the account,
however, of Henry de Grey (vol. ii.. p. 521), she is called the fourth daughter;
and the names of three elder sisters are given. I presume, therefore, that she was
in reality the fourth in female descent. Her father was the thirteenth Earl, and first
Duke of Kent, but died without male issue in 1740, when the dukedom became
extiDCt. It was revived in 1799 in favour of Prince Edward, foarth son of George
ni,thefkther of Qneen Victoria. — Und.^ p. 519.

§ In Collins* Peerage (4th ed. vol. i., p. 330), Henry appears as the younger
brother. But this is a mistake; Henry is styled " the eldest son" on his funeral
tablet in the church of All Saints, Derby. In the books, also, of St. Peter's CoU

14 LIFE OT CAVENDISS;

I close the account of diis long genealogy vnth one remai4r.
Could we trace the character of any family, whatever its rank,
tihrough as long a period as we can, that of the Cavendishes, we.
should probably find one line differ very little, at least moralfyy
from another. A very few men and women would appear o£
great genius and great virtue ; and a very few would also be
found remarkable only for the magnitude of their crimes. A'
larger number would occur, characterised by the possession of
fiair talents, and displaying in their conduct an average morality ;
and a considerable crowd would present itself of persons who
were only the creatures of circumstances, and were guided and
controlled by the greater abilities and more fixed principles of
their neighbours.

Families, no doubt, like races of men, present specific cha-
racteristics, which are transmitted through many generations,
unaltered. These, however, occur more in reference to intellec-
tual than to moral peculiarities, in which preeminently our
neighbour is our brother ; and the infallible effect of varied inter-
marriage, without which a stock soon becomes extinct, is to lessen,
or obliterate, or reverse extreme peculiarities, whether physical,
intellectual, or moral. There can never, accordingly, be a very
great difference between the characteristics of two families of
the same nation, provided we study their history through a suffi-
ciently long period, and set aside a few exceptive cases, which
by their peculiarity and rarity afford the best proofs of the
existence of the law which they trangress.

A history, like that of the Cavendishes, should thus have
an interest for every one; and this is my apology for dis-
cussing it with some fulness. If there is anything exceptive in
their annals, it is the high moral character which for so many
centuries the family has maintained. Other high-born English
families of old descent have given to their country as many noble
women in every sense of the word, as many patriots, statesmen^
and men of science and letters, but there are not many of them

lege, Cambridge, lie is called ** Filiiis natu maximas." An interesting sketch of bis
younger brother, Frederick, who was somewhat eccentric, but a man of excellent
parts, and remarkable for his benevolence, will be found in the Gentlemcm^t
Magazine, 1812, p. 289. He was very unlike Henry, both in temper and tastes,
and the brothers seldom met, but they are said to hsve been sincerely attached to
each other.

SCIBNTinC BIOGBAPHIBS. 1 5

who have not some page in their history blotted by the record
of the misdeeds of their ancestors. The Cavendishes may well
be forgiven^ if they look back with complacency on a family
history which displays so few shortcomings as theirs does; and
we may all feel pride in numbering among our great philosophers
the descendant of such a stock. To him I now turn.

Sir James Mackintosh has accused Ireland of being incuriosa
suorwn ; but the charge may be preferred against all the divi-
sions of the empire* so far at least as the men of science are
concerned. No other European nation has so imperfect a series
of biographies of her philosophers^ as Britain possesses, and it
is little creditable to us, that we have often to turn, as in the
case of Cavendish, to the Memoirs of a foreign society, for the
best record of the personal history of even our most famous
students of physics* Cavendish's position is not peculiar in
this respect, for we still Ipok for a more complete life of Newton
than has yet appeared,* and among philosophers who have
recently departed, we are yet without biographies of Toimg,
WoUaston, and Dalton, So careless has his own country been
of the memory of Cavendish, that although he was for some
fifty years a well-known and very distinguished Fellow of the
Royal Society, a member for a lengthened period of the French
Institute, and an object of European interest to men of science,
yet scarcely anything can be learned concerning his early his-
tory. This no doubt is owing, in great part, to his own dislike
of publicity, and to the reserve and love of retirement which
strongly characterised him. Long before his death, however, he
was so conspicuous a person in the scientific circles of London,
that the incidents of his early life might readily have been
ascertained. They were not, it should seem, enquired into by
any biographer. Had he been a poor man of obscure birth,
this might not have surprised us, but we have seen that he came
of one of the oldest families in England, nor was he a far-off
branch of it, for he was the grandson of a Duke by both parents,
and the nephew and the cousin of one, besides counting kin on

* Since thh was written, Mr. Edleston, of Trinity College, Cambridge, has
mnnonnced his Corretpondence of Sir Isaac Newton, Sfc, which will doubtless
prove a most Talaable addition to our scientific biography. That snch a work,
however, should not appear till 1851, is the best jastificaticm of the statement in the
text.

16 LIFE OF CAVENDISH.

all sides with the aristocracy of Great Britain. He was further^
immensely wealthy ; so wealthy indeed^ that^ as M. Biot epigram-
matically puts it^ he was ^^le plus riche de tous les savans,
et probablement aussi, le plus savant de tous les riches.''^
Nevertheless^ his biographers cannot so much as agree upon
the country of his birth, although his death occurred so late as
1810.

Cuvier^t Thomson, t and Kopp,§ tell us that he was bom in
England, whilst the contemporary notices of his death represent
him as bom in Italy. The latter is the true account. He
was bom on the 10th October, 17^1, j| at Nice, whither his
mother. Lady Anne Cavendish, had gone for the sake of her
health.

His mother died when he was some two years old, and I
have been able to learn nothing concerning his earliest years.
The first notice I find concerning him goes back to iT^^j when
he became a pupil at Dr. Newcomers school at Hackney, an
institution celebrated in its day for its excellent management,
and largely attended by the children of the upper classes.^

He remained at Hackney school till 1749, but no means
now exist of ascertaining the precise nature of his studies, or
what progress he made in them.'*'*

* Bioffraphie Universelle, 1813, tomeyiL^p. 456.

■h Bloget Historiquet, tome ii., p. 79.

% History qf Chemistry, vol. i., p. 338.

§ Geschichte der Chemie, toI. i., p. 230.

II I state this on the authority of Lord Burlington. A similar account is given
in the sketch of his brother Frederick, published immediately after the latter's
death. ** Lady Anne CaTcndish was in bad health on her marriage, and went shortly
after to Nice, for the benefit of the waters there, attended by her husband. Henry
was bom at Nice, but his mother returning to England, Frederick drew his first
breath in the country of his ancestors." — Gent. Mag, 1812, p. 291.

^ Lord Campbell refers to it in his Lives of the Chancellors, as " a most ex-
cellent school at Hackney, kept by the Rev. Dr. Newcombe [Newcome], a sound
classical scholar, and a strict disciplinarian." — Vol. v., p. 367.

♦• I state this on the authority of the Rev. C. J. Heathcote, of Upper
Clapton, who possesses the only papers of the Hackney seminary that remain, and was
good enough, at my request, to examine them for any records of Henry Cavendish.
They consist, however, only of a list of plays acted by the boys of the school, in
which Henry's name does not appear, and of a catalogue of the dates of admission to
the school, and of departure from it, of its different members. Four other Caven-
dishes appear on this list, besides Henry and his brother, bat no further particukrs
are supplied.

RESIDENCE AT CAMBRIDGE. 17

From school he went directly to Cambridge, where he matri-
culated in the first rank, on 18th December, 1749.*

He was now eighteen, and remained at Cambridge till 1753,
but did not graduate. Frederick Fuller, Esq., Fellow and Tutor
of St. Peter's College, Cambridge, has done me the favour to
make enquiry concerning Cavendish's residence at Cambridge,
and has furnished me with the following particulars, which are
believed to embody all that is known concerning his occupations
there. He entered at St. Peter's College, as the following
extract from the books of that college shows: ^^ Nov. 24, 1749,
Honorabilis Henricus Cavendish, viri Honoratissimi Domini
Caroli Cavendish Filius natu maximus, h schola publica de
Hackney, annos habens octodedm, more solito examinatus et
approbatus, admittitur ad mensam sociorum sub Tutoribus et
Fidejussoribus M"* Stuart et Cox." Mr. Fuller adds, ^^ Caven-
dish (as appears from other old books in the College treasury)
commenced residence on the 24th of November, 1749, and
resided very regularly and constantly until the 23rd of February,
1753, when he left without taking his degree. As he had
then resided the full time — or nearly so, within a few days —
required for the degree, and we can hardly suppose that Caven-
dish feared the examination, there must have been some parti-
cular reason for his neglecting to take the usual course.
Perhaps he may have objected to the tests, which were then
very stringent.

'^Two more of the Cavendish family were studying here at
the same time with the Honourable Henry Cavendish, viz.,
Henry's younger brother, the Honourable Frederick Cavendish^
who was entered April 10, 1751^ and, like his brother, left
without taking any degree ; and his cousin. Lord John Caven-
dish, the fourth son of the Duke of Devonshire, who was entered
February 21, 1750, and took the degree of M.A. in the year
l7o3. All three of them were educated at the same school at
Hackney. Among Cavendish's cotemporaries at St. Peter's
College, were, the Earl of Euston, afterwards the Duke of
Grafton, celebrated by Junius ; Gray, the poet ; and Jeremiah

* Extract from Matriculatioii Book^ kindly ^ilnUhed by the Rev. Joseph
RomiUyj senior Felloff of Trinity College, Cambridge, and Regbtrar of thd
Unirenity.

18 LIFE OF CAVENDISH.

Markland, the Greek critic, who was senior fellow of the college
at the time, and always in residence/'

After leaving Cambridge, Cavendish probably went to
London 'j^ but his personal history is a blank for the next ten
years, although it cannot be doubted, from his subsequent
writings, that it was mainly spent in mathematical and physical
studies. He joined the Royal Society in 1760, but did not
contribute anything to its transactions till 1766, when he pub-
lished his first paper On Factitious Airs. Here, accordingly, I
shall suspend all reference to biographical details, and reserve
the rest of his personal history until I have discussed his labours
as a philosopher.

* Mr. Tomlinsofi has dnim my attention to the fiict that Carendkh Tinted
along with his brother Frederick. Una visit was probably paid when theylwere
yoimg men (for they had little, if any, intercourse in after-life), and before Henry
became fSunous ; otherwise some account of the journey would in all likelihood have
been made public. I have no information, however, as to the date of the joumey>
ita objeo^ or the amount of time which it occnined.

CHAPTER II.

GENERAL SKETCH OF CAVENDISH'S SCIENTIFIC RESEARCHES

AND DISCOVERIES.

Few of our men of science have been so catholic in their
tastes as Cavendish^ so far at least as physics are concerned.
He was an excellent mathematician^ electrician^ astronomer^
meteorologist^ and geologist^ and a chemist equally learned and
originaL In the fullest sense of the term^ indeed^ he was a
natural philosopher^ and had he published during his lifetime
all the researches which he completed, his reputation would
have been much wider and more varied even than it was. He
was exactly the opposite of a certain class of thinkers, whose
fertility of invention, and skill or success in research, are far
below their desire of distinction, and who are diligent in
coining every thought, though it be but a farthing's worth,
so as to put it into immediate circulation. Such men have
nothing to reveal in private; the public are already in pos-
session of all they know. Cavendish, on the other hand,
dealt with his discoveries as with his great wealth, and allowed
the larger part of them to lie unused in his repositories. His
published papers, accordingly, give but an imperfect notion of
the great extent of ground over which he travelled in the course
of his investigations, and of the success with which he explored
it. I shall endeavour, in the following pages, to give some idea
of his unpublished ^s well as of his published papers, although
so Sear as chemistry is concerned, Mr. Harcourt has leffc little
to be done in this matter by his analyses of the Cavendish
MSS.*

The obscurity which hangs over Cavendish's private history,
especially in his early days, makes it impossible to determine
what induced him to devote himself at particular periods to one

* Brititk AmeiatUmBep^rt, 1839, p. 45.

20 LIFE OP CAVENDISH.

branch of science rather than to another. He appeared first
before the public as an author on chemistry^ although from his
early devotion to mathematical and strictly physical studies, it
might have been expected that he would first have appeared as
a writer on subjects connected with them. It is very probable,
indeed, that some of the mathematical and physical essays which
remain among his unpublished papers, are of earlier date than his
first published chemical researches of 1766, but the absence of
dates from the majority of his MSS. prevents any conclusion
being drawn on this point. It will presently appear, also, that
he could, with the greatest ease, change his subject of study, and
that he was in the constant practice of carrying on together,
widely dissimilar enquiries.

In the sequel of this work, I shall discuss his researches in
a classified order, so as to bring all referring to one branch of
science under ODe head ; in the personal narrative they must be
taken in their chronological order, at least to a considerable
extent. I do not propose, however, to adhere quite strictly to
this, for Cavendishes life is so barren of incident, that with the
solitary exception of the Controversy concerning the discovery
of the composition of Water, almost no connexion can be traced
between the events of his history and the researches which he
prosecuted. It will sometimes, accordingly, be preferable to
depart from a strictly chronological order, which would carry us
forward and backward, from one topic to another, and in so
doing diminish our means of doing our author justice.

The account, moreover, of his merits and labours as a
philosopher, contained in the sketch of his Life, will be more
dogmatic than at first sight might seem desirable ; but as full
abstracts of his most important published papers are given in
the sequel, the authorities for all the statements contained in
the earlier part of this work will be found given at length there,
and although not always minutely referred to, will be discovered
without difiiculty, from the many sections into which the book
is divided. This chapter, in truth, is intended to inform the
general reader, who may not care to study minutely the more
elaborate exposition of Cavendish's merits, with which the body
of the work is occupied, what he actually did achieve as a dis-

EXPEBIMENTS ON ARSENIC. 21

coverer in physical science ; and to prepare the way for the
intelligent appreciation of the controversy in which certain of
his discoveries involved him^ as well as for a just estimate of his
intellectaal capacity. I shall chiefly refer to his published
essays^ as his actual reputation rests upon them ; but an excep-
tion will be made in the case of the Chemical MSS., and some
slight reference will occur to all his unpublished papers.

Cavendish did not give to the world his earliest researches.
He probably kept many back. Two lengthened investigations^
at least, the one chemical^ the other physical, were completed
and laid aside, in a condition ready for publication^ before he
commenced contributing to the Tramactuma of the Bayal
Society of London, in which all his papers were published.

The first of these investigations was entitled^ Experiments on
ArsemCy and remains among his papers^ both in a rudimentary
and completed shape. It appears to have been written out for
the instruction of some friend^ to whom allusion is made more
than once in the course of the Essay^ and the experiments referred
to in it are as early^ at least, as 1764, as appears from a date in
the Note-Book. The Rev. W.V. Harcourt has published large
and judiciously selected extracts from the MSS., which will enable
the reader to understand the scope of this early investigation of
. Cavendishes.* It may suffice here to state that the paper in
question contains an elaborate enquiry into the differences
between regulus of Arsenic (Metallic Arsenic), white Arsenic
(Arsenious Acid, As Os), and Arsenical Acid (Arsenic Acid, AsOg).
The properties of Arsenic Acid, and of various of its salts, the
arseniates^ are described with no little accuracy. The true
nature of the difference between arsenic and its two acids.
Cavendish did not know ; but he held arsenic acid to be '^ more
thoroughly deprived of its phlogiston^ than arsenious acid ; and
the latter to bear a similar relation to metallic arsenic. These
phrases are equivalent to the statement, that arsenic acid con-
tains more oxygen than arsenious acid^ and the latter more than
metallic arsenic^ which we know to be the case. The paper^
is otherwise remarkable for its speculations on the nature of the
*^ red fumes/ (nitrous acid, produced by the action of the air on

* BritUh AuoHaiion Report, 1839, pp. 50—58.

22 LIFE OF CAVENDISH.

nitric oxide) which attended the action of nitric acid on arseni-
ous acid^ and for its discussion of the theory of the solution of
metals in acids^ and the reduction of the former by heat and
inflammable matter.

Contemporaneously with this investigation^ or a little later,
Cavendish engaged in an extensive series of Experiments on
Heat. The date February S, 17^5, occurs in the record of an
observation on the 89th page of his Notebook of Eaperiments.
There can be little, if any doubt, accordingly, that we must go
back many weeks into 1764, for the commencement of the
researches in question. Mr. Harcourt has given extracts firom
the record of these,* and I shall give an abstract of them with
a commentary, in the section of the work devoted to the con-
sideration of Cavendish's observations on Heat. They were
written out for a friend whose name is not given, but were not
publicly referred to till some nineteen years after their com-
pletion, when certain of the results were quoted in a paper
published in 1783, on the Congelation of QmckMver.

They are very remarkable researches, and had they been
made public in 1764 or 1765, they would have given Cavendish
chronological precedence to Black in some of his discoveries,
and equality of merit in others. They would have entitled
him also to rank above Black's pupils and imitators ; such as
Irvine, Crawford, and Wilcke. This at least is certain, that
Cavendish discovered for himself, and announced with admirable
clearness, the fundamental laws of specific heat ; and collected,
probably before any one else, tables of the specific heats of various
bodies. With scarcely any knowledge also of what Black had done
towards the exposition of the laws of Latent Heat, and guiding
himself by a totally different theory, as to its relation to solidity
and liquidity. Cavendish investigated for himself the evolution
of heat which attends the solidification of Liquids, and the con-
densation of Oases or Vapours, and the converse ** generation of
Cold/' as he styled it, which accompanies the liquefaction of
Solids and the Vaporisation of Liquids. I shall explain his
views at length in the commencement of the section of the
abstracts devoted to the discussion of his papers on heat.

* BritUh Atioeiaiim Report, 1839, pp. 45—50.

ESSAY ON FACTITIOUS AIRS. 23

Cavendishes earliest pubUc contribution to science was^ as
has been mentioned^ his paper on Factiticms Ain^ published in
the TVansactums qf the Bxryal Society for 1766. It consisted
of three parts : a fourth^ which was not published, remains in a
state of perfect completion^ ready for the press, among his papers.
It was evidently intended to be read to the Royal Society, for it
contains a reference to the ^^ Former Experiments read to this
Society/' For some reason, however, it was withheld, but
has since been published by the Rev. W. V. Harcourt. *

Those four papers, for the several parts are equivalent to
separate essays, are occupied with the discussion of the proper-
ties of Hydrogen, Carbonic Acid, and the gases evolved during
the fermentation, putrefaction, and destructive distillation of
vegetable and animal matters. They contain the first distinct
exposition of the properties of hydrogen, and the first fiill
account of those of carbonic acid, besides investigations into the
combining proportion of the latter, and the properties of car-
bonates. They recount also the first tfticcessfiil attempt to
determine the differences in density which characterise the gases,
and suggest the probability of there being more kinds than one of
inflammable air. A paper whi(9i was published by Cavendish
in the PkUoMophicai Tramactiomlor l767>may be considered as
an extension of the research into the properties of carbonic
acid. It is occupied with an account of the Analysis of one of
the London pump waters (that, namely, of Rathbone Place),
which was remarkable for the quantity of calcareous earth which
it deposited when boiled. Cavendish showed that the earth was
originally retained in solution by carbonic acid, which the
boiling dissipated, so as to allow the earth to precipitate. The
other constituents of the water were determined also, and the
whole research is curious as one of the earliest tolerably success-
ful attempts to analyse a natural water. Abstracts of these
papers are given in the sequel, which those who wish to see
their exact contents will consult. I refer to them here, only
as showing the prominent position which Cavendish took from
the first as a discoverer in chemistry. He may be counted
the third in order of time among the four great English pneu<
matic chemists of the eighteenth century, the other three

* Britiih AMoeiaiian lUpwi, 1839.

24 LIFE OF CAVENDISH.

being Hales^ Black, and Priestley. Hales was the earliest
enquirer into the properties of elastic fluids, and, without injus-
tice to his illustrious predecessors, the immediate disciples of
Bacon, and early contemporaries of Newton, who had made some
progress in investigating the properties of the gases, he may
be called the father of pneumatic chemistry in England. His
great merit was to point out that elastic fluids may be obtained
from an immense variety of organic and inorganic substances, of
which they are as important constituents as the solids or liquids
which may be separated from them. Hales did not recognise,
unless very imperfectly, that those elastic fluids were chemically
unlike, and specifically distinct, so that he spoke of them as if
essentially identical with each other and with the atmosphere ;
and had no other name for them than simply air. His writings
belong to the first third of the preceding century. The second
of the pneumatic chemists, Black, appeared a little after the
middle of the century, and by his celebrated essay on Magnesia
Alba, demonstrated that there existed at least one gas totally
distinct from the atmosphere, and able by its addition to
bodies, or its removal from them, to alter immensely their
physical and chemical properties. Black thus rose to a higher
discovery than that reached by Hales. The latter had shown
that the most solid stone might owe half or more of its
weight to the presence of an imprisoned or solidified air;
but he had paid little or no attention to the efiect which the
removal of this air had in altering the chemical properties of the
substance from which it had been extracted. Black demonstrated
that the fixed or solidified air did not merely increase the bulk
and weight of the solid, but determined in a most striking manner
its chemical properties, so that a substance which, when saturated
with a peculiar air, was a bland, innocuous insoluble powder,
or crystalline solid, became by the expulsion of this air soluble,
caustic, and corrosive; and the diflerence between marble or
chalk on the one hand, and quicklime on the other, was shown
to be entirely dependent on the presence or absence of the gas,
which Black nhmed fixed air, and we name Carbonic Acid.

Twelve years after the publication of Black's paper, namely
in 1766, Cavendish published the first of the essays we have
been considering. He took up the investigation of fixed air

INVESTIGATION OP HYDROGEN. 25

where Black and his pupils had left it^ and examined in parti-
cular its properties when free, on which Black had published
scarcely anything.

Thus far Cavendish appears rather as the follower of Black
than as an independent observer^ although by a reference to his
paper it will be seen that his investigation of the properties of
free Carbonic Acid was equally original and accurate. He
struck out^ however^ in addition^ a new path for himself^ and
added to the solitary fixed air, a second gas equally distinct from
it and from atmospheric air in properties. This was Hydrogen^
of which Cavendish cannot be called the discoverer, for many
of his predecessors, Boyle among others, had encountered it ;
but no chemist had carefully examined its properties, or at least
had described them. His predecessors, indeed, knew only as
much about the gas as a navigator who merely touches at a strange
island, knows of its geography and various products, to whom
we cannot deny the merit of being its discoverer, although we
often assign much more credit to some later visitor who surveys
and describes the new territory. The mere discovery of
hydrogen was no great feat ; for the most random experimenter,
who, with or without purpose, handled the more powerful re-
agents, was likely to encounter a phenomenon of which the
conditions are so simple as the evolution of hydrogen from
the contact of iron and an acid ; and the ready and explosive
combustion of the gas when it meets flame could not fail to
attract the attention of the most heedless observer. There can
be little doubt, accordingly, that among Cavendish's predeces-
sors backwards through several centuries, there were many who
could assert equally good claims to be called the discoverers of
hydrogen, of which, nevertheless, they knew exceedingly little.
Cavendish did not claim to be one of them, but he could claim
a merit which was much greater. Boyle, MayoWj and Brown-
rigg had preceded him in showing how gases may be collected,
but no one had given an example of the mode of examining
them. Cavendish's examination, accordingly, of the properties
of carbonic acid, and hydrogen, has all the interest that
attaches to the first demonstration of a method of pursuing a
novel investigation. It is easy to look back from our thoroughly
appointed laboratories, filled with the apparatus which some

26 LIFE OF CAVENDISH.

ninety years have added to the chemist's instrumentSj since the
date of the investigation we are discussing, and to criticise and
depreciate the methods and results it records ; and this has been
done laigely and unreasonably. Yet if we consider how much
more genius is requisite for the devising of an apparatus or
method of research which is quite new than is needed for its
indefinite extension and improvement^ and if we further judge
the experimenter of 1766, not by his successors of 1840 or 1850,
but by his contemporaries, we shall not hesitate to assign a very
high rank to Cavendish, as one of the earliest investigators of
the chemical properties of the gases. We find him, for example,
collecting the elastic fluids on which he experimented, with
various precautions to secure their purity, observing carefully
from how many different sources they could be procured with
identical properties, and determining with numerical precision the
relative volumes yielded by diflferent processes. The questions
of their permanent elasticity, their solubility in different liquids,
their combustibility or power to support combustion, their
specific gravity, and likewise their combining equivalent, were
all carefully enquired into. The apparatus employed, though
deficient in delicacy according to modem standards, was unex<«
ceptionable in principle, and wherever that was possible, was
made to yield quantitative results, so that this earliest analyst
of the gases introduced the principle of rendering all descrip*
tions of phenomena as precise as possible, and endeavoured
from the first to attach a numerical value to each. We shall
find, in truth, that we have done little more in later times
than extend, improve, and as we complacently say, perfect
Cavendish's processes for the analysis of gases, and that we
differ from him more in our mode of interpreting certain of the
phenomena he witnessed, than we do in our methods of inves*
tigating elastic fluids. Thus, he mistook altogether the source of
the hydrogen which he procured so abundantly from the solution
of iron, zinc, and tin, in sulphuric and muriatic acids, and referred
it to the metals in which he supposed it to exist in a peculiar
state of combination. Water at this time, it will be remem-
bered, was supposed to be an element, and the composition of
all the acids was unknown. No gas had been certainly traced

DIFFERENT INFLAMMABLE AIRS. 27

to a liquid as its source, otherwise than as dissolved in it like
carbonic acid in a mineral water ; whilst Hales and Black had
shown that the most fixed and solid bodies might yield from
their very substance large volumes of elastic fluid. It has
occurred to me as possible, that this may have been one reason
which induced Cavendish to suppose that the hydrogen came
out of the metal rather than out of the liquid. The considera-
tion, however, which mainly weighed with him, and the only
one to which he himself refers, was the belief common to him
with the majority of his contemporaries, that the metals con-
tained a peculiar combustible principle named phlogiston.
This Cavendish supposed to abandon tiie metal, and, assuming
the form of an elastic fluid, to show itself as the inflammable
air. To his opinions on this point I shall have occasion to refer
frequently in the sequel.

Hydrogen was thus the first of the combustible gases
examined, and for many years, as we shall affcerwards find, great
confusion exbted in the mind of chemists as to the number and
nature of the different inflammable elastic fluids ; nor did this
begin to cease till the composition of water and of carbonic acid
was ascertained. Cavendish, however, had clearer views on this
very important point, than most of his fellow chemists. He
ascertained that vegetable and animal matters, by putrefaction
and destructive distillation, yielded inflammable air. He was
not aware of its exact nature, but he satisfied himself by the
test of specific gravity, and the volume of common air required
for its combustion, that it was not identical with Hydrogen,
which accordingly he distinguished as the ^^inflammable air
from metals.'^ He further observed that *'the nature of the
inflammable air was not quite the same'^ from animal as from
vegetable substances. We shall afterwards find that he turned
these observations to excellent account in the researches which
led him to the discovery of the composition of water.*

Between the years 1767 and 1783, Cavendish did not appear
before the public as an author on any subject directly connected

* CtTendJih's obsenrations on the peculiarity of the inflammable air from
oiganic bodies, will be fonnd partly referred to in the third lection ^of the ezperi-
menti on factitiooa air; but chiefly in Part IV, publiflhed from the MS. by Mr.
Harconrt, BritUh AM90ciation Sepwt, 1839, pp. 58-62.

28 LIFE OP CAVENDISH.

with chemistry, but it appears from his MSS. , that he continued
to prosecute chemical enquiries. Among his papers is one on
which he himself has written, '^ communicated to Dr. Priest-
ley/^ the contents of which are referred to by the latter in his
account of Experiments and Observations made in and before the
year 177^9 so that Cavendish's communications to him cannot
have been later than that date. The paper in question has been
printed by Mr. Harcourt,* and is remarkable as containing one
of the earliest distinct accounts of nitrogen. Cavendish pre-
pared it by passing atmospheric air repeatedly through red hot
charcoal, and removing the carbonic acid produced, by caustic
potash. He gives the following description of it: ^The specific
gravity of this air was found to differ very little from that of
common air; of the two, it seemed rather lighter. It extin-
guished flame, and rendered common air unfit for making l>odies
bum in the same manner as fixed air, but in a leas degree, as a
candle which burnt about 80'^ in pure common air, and which
went out immediately in common air mixed with ^ of fixed air,
burnt about 26"^ in common air mixed with the same portion of
this burnt air.^'t Cavendish gave no special name to nitrogen,
which he referred to generally as mepkitic air. It was after-
wards minutely described by Lavoisier and Scheele, and was
distinguished by Priestley and his contemporaries, by the name
phlogisticated air. The quotation adduced above, shows incon-
testably that Cavendish discovered nitrogen for himself, and
had ascertained with great precision its chief properties ; but in
the absence of precise dates, I hesitate to adopt Mr. Harcourf s
conclusions, that the paper from which I have quoted contains
^^^Q first clear description of nitrogen as a distinct gas.'' Dr.
Rutherford, of Edinburgh, the reputed discoverer of nitrogen,
published his Thesis De Aere Mephiiico, in 177^- His process
for procuring the gas, for which he had the same general term
as Cavendish, viz., mephitic air, resembled that of the latter
chemist, except that he employed atmospheric air vitiated by
respiration, not by combustion. This he passed through
caustic potash, and tested by lime-water, which it did not
precipitate, whilst it possessed the power of extinguishing life

* Sriiisk Auoeiation Report, 1839> p. 64.
t Op. Cit. p. 65.

BLECTSICITY OF THE TORPEDO. 29

aad flame** The dates of publication^ or announcement, of
Cavendish and Rutherford^s observations; are thus the same,
whilst the dates of their experiments are uncertain. We cannot
in these circumstances give precedence to the former, but it
is certain that he was an independent discoverer of nitrogen.

For a season after making the researches referred to. Caven-
dish appears to have laid aside chemistry for other departments
of physics. In 1771 he published the elaborate paper on the
theory of the principle phenomena of electricity, which appears
in the PkUoscphical Transactums for that year. In 177^
appeared the curious and interesting accounts of his Attempts to
imitate the effects of the Torpedo* In the abstract of this paper,
the exact significance of which has been a good deal mistaken,
I shall point out the novel and important conclusions which
it brought to light. It may suflBce, therefore, here to say, that
the singular power which the torpedo possesses, of benumbing
those that touch it, had been referred with great ingenuity and
force of argument, by Walsh and others, to its possessing the
means of discharging electricity at will. In many respects,
however, the action of the torpedo differed from that of any
known electrical apparatus, so that some refused to believe that
the benumbing sensation occasioned by the fish was an electrical
phenomenon ; whilst others went to the opposite extreme, and
regarded animal electricity as quite different in kind from that
procured from other natural or artificial sources. To decide
between those views, or supplant them by a more accurate one,
was no easy matter, for the torpedo could only be procured in
the Mediterranean, and other warm seas, nor did the resources
of electrical science, in the year 1776, when only one of its
departments had been studied, suggest the means of questioning
the torpedo as to its singular powers, even if it had been more
accessible than it was. Cavendish, accordingly, attacked the
problem in another way. He tried whether he could not suc-
cessfully imitate the effects of the living fish, by a piece of
apparatus constructed in imitation of it, and placed in connec-

* Thwuon** Sfyttem qf ChemUtry, toI. i. p. 203. Dr. ThomiOD adds, by
way of comment to his account of Rutherford's obserrations, " When Hawksbee
passed air through red-hot metallic tubes, be must have obtained this gas, but at
that time the difference between gases was ascribed to ftimes held in solution." — See
Pkil, TWfW. AOr. r. p. 613.

30 LIFE OF CAVENDISH.

tion with a jEriction electrical machine and a Leyden battery.
He succeeded so weU in his imitation^ that all doubts as to
the identity of the torpedinal benumbing power^ with common
electridty^ were removed. The demonstration of diis identity,
however, was far from a simple matter, for the apparent differ-
ence between the force displayed by the torpedo^ and that
exhibited by the electrical machine, or Leyden jar, was v^y
great. Cayendish, notwithstanding, not only showed that the
two forces were in essence identical, but deduced from his study
of their apparent differences, conclusions concerning the extent
to which electricity, not of animal origin, may vary in its modes
of manifestation, which are now r^arded as having marked an
era in the progress of the science. The most illustrious of his
successors, Faraday, among others, have borne testimony to the
light which was thrown upon every department of electrical
enquiry, by Cavendish's demonstration, that the most opposite
effects might be obtained from electricity developed in the same
way, by causing its intensity to vary as compared with its quan-
tity. Thus, though voltaic electricity was not discovered till
some quarter of a century after the publication of the paper on
the torpedo, Faraday found the theory which Cavendish sug-
gested, sufficient to explain the curious and apparently contra-
dictory voltaic phenomena which he observed so late as 1833.
^' The beautifdl explication,'* observes he, " of these variations
afforded by Cavendish's theory of quantity and intensity,
requires no support at present, as it is not supposed to be
doubted."* When it is remembered, that in 1776, friction and
atmospheric electricity were the only forms of that force which
had been studied, and that they were but imperfectly known, we
cannot but admire the sagacity with which Cavendish found, in
the very perplexities of animal electricity, the means of explain-
ing it, and rose at once to the recognition of a law so wide in its
bearing, that whilst it interpreted the difficulties of the existing
science, it also furnished a key to the problems which were to
vex the students of the voltaic, thermic, and magnetic electri-
cities of a later century. In none of his essays does Cavendish
appear to greater advantage than in this.

He had now been before the public, so far as one of the most

* Faraday's JSsperimental Beiearchei on Mkctricitpf seriea iii., p. 81.

DESCRIPTION OF MBtEOROLOGICAL APPARATUS. 31

reserved of men can be said to have been so^ for ten years, and
as all his published papers had been communicated to the Royal
Society, it is not surprising that he should hare been selected by
that body in 1776 to describe the meteorological instruments
which were made use of in their apartments. The Society had
commenced in 1773 recording their observations with the ther-
mometer, barometer, rain-guage, hygrometer, variation-compass,
and dipping needle, and Cavendish was applied to, to give an
account of these. His father. Lord Charles Cavendish, had
devoted himself to meteorology, and had paid special attention
to the improvement of the thermometer and barometer, so that
probably our philosopher was early trained to the points essential
to the accurate construction and employment of these instru-
ments. What familiarity he had with the others mentioned above
does not appear, but the fact of his being entrusted by the Society
with the description of all their meteorological apparatus, shows
how entire was their confidence in the extent of his acquire-
ments and the accuracy of his observations. Their confidence
in him is the more remarkable, that he had published no paper
before 1776 referring to any of the instruments he was called
upon to describe. That he had paid great attention, however,
to the thermometer before this period, is certain from his
unpublished papers on heat of 1764 and 1765 ; and as he made
no concealment of his researches, although he did not publish
them, there can be no doubt that many of the members of the
Society were well aware of his familiarity with meteorological
instruments. The most important part of this paper is his
description of the best method of accurately graduating ther-
mometers, which will be found specially referred to in the
abstract of his papers on Heat.

The succeeding year 1777> or perhaps rather 1778, marks
the period when he commenced his most important chemical
researches; he styled them Experiments on Air. They were
carried on with frequent, and sometimes long interruptions till
1788, and no part of them was published till 1783. They led
to the discovery of the constant quantitative composition of the
atmosphere, the compound nature of water, and the composi-
tion of nitric acid. Their discussion occupies the greater part
of this volume.

32 LIFE OF CAVENDISH.

In 1783^ alsO; Cavendish published his first paper on heat^
embodying some of the results he obtained in 1764 in reference
to the freezing or solidiiying point of liquids. He returned to
the subject again at intervals onwards till 1788, so that his
papers on heat and his experiments on air, so far as published,
extend over the same period. It will be more convenient, how-
ever, to take those on each subject continuously, than to discuss
alternately the researches on heat and the experiments on air, and
I shall take the papers on heat first. They are three in number,
and belong to the years 1783, 1786, and 1788. All of them
refer to congelation ; the first to that of quicksilver, the second
and third to that of the mineral acids and of alcohol. Their
contents are of more interest as containing a statement of Caven-
dish's views regarding the laws of liquefaction and congelation,
than as reporting mere phenomena. They are all, indeed, com-
mentaries upon observations made in North America by officers
of the Hudson Bay Company on the effect of great natural
cold, assisted by powerful freezing mixtures in congealing mer-
cury, nitric acid, oil of vitriol, and spirits of wine. These
observations were made under Cavendish's directions, and at his
cost, and abstracts of his commentaries on them will be found
in the sequel. The most important of these papers was that on
the freezing of quicksilver. This metal, which had long been
imagined by the older chemists to owe its apparent permanent
fluidity to some anomalous peculiarity, was frozen in a thermo-
meter in 1759, O.S., by Professor Braun, of Petersburgh, who
observed that its congelation was accompanied by a descent of
the mercury, through many hundred degrees, and came to the
conclusion that the freezing point of the metal was some 300^
or 400^ below Farhenheit's zero, but was unable to determine
the exact point of congelation. In drawing this startling conclu-
sion, which implied, if true, that the most enormous differences in
temperature might occur between unlike regions of the globe, and
that in northern latitudes the thermometer might fall through
two, three, or four hundred degrees in a few hours, Braun con-
founded two phenomena. The one of these was the contraction
which accompanies the cooling of liquid mercury ; the other the
further contraction which attends its solidification. The con-
traction due to both these causes, exaggerated by the peculiarities

FREEZING OF MERCURY. 33

which attend the freezing of mercury in capillary tubes^ was
referred by Braun solely to the first. To his conclusion the
majority of the natural philosophers of Europe assented^ but
there were two who, from the first, discredited it. These were
Cavendish and Black, who, unaware of each other's intentions,
and conducted to their results by independent researches, sug-
gested tbe same way of ascertaining the true freezing point of
mercury, which they felt satisfied was much higher than Braun
had imagined. This method, which was put in practice by
Governor Hutchins at Albany Fort, Hudson Bay, consisted in
freezing the mercury, not in the tube of the thermometer, but
in a separate vessel, in which the instrument was plunged,
and guarding carefully against the whole of the external quick-
silver becoming congealed. Cavendish, like Black, knew
well from his experiments of 1764 and 1765, that when a
liquid begins to congeal, it acquires a temperature which
remains constant till the whole of it is frozen. He did not
doubt that mercury would resemble, in this respect, the liquids
upon which he had experimented, and he furnished Mr.
Hutchins accordingly with an apparatus founded on this prin-
ciple, which was employed with complete success. The result
was, that the freezing point of mercury is not more than 39°
or 4QP below Fahrenheit's zero, a determination which aU sub-
sequent observation has confirmed. To Cavendish and Black
is thus owing the merit of overturning the extravagant conclu-
sions regarding the lowest natural or attainable temperature
which Braun had promulgated. The merit due to either may
seem small, for the observation of the exact freezing point of a
particular liquid would not, at the present day, secure much
honour for its observer. The determination of the congealing
point of quicksilver, however, is a point of interest in Caven-
dish's history, not in itself, but as an evidence of the very clear
and broad view which he had taken of the relation of heat to
liquefaction and congelation, and his exact acquaintance with
their essential phenomena at a time when, with the exception of
his great contemporary Black, there was not, probably, a philo-
sopher in Europe who understood the true nature of congela-
tion. Mr. Hutchins^s observations were not made till 178^^

34 LIFE OF CAVENDISH.

but the directions by which he was guided had been laid down
byCayendish in 1764 and 1765.

The other two papers on heat which discuss the congelation
of the mineral acids and of spirits of wine are of less importance.
They supply additional proofs of Cavendish's familiarity with
the laws of heat ; but as they are abstracted and commented
on in the sequel^ I need not refer to them more fully here»
One of these papers contains a very curious example of an
acquaintance on Cavendish's part with what we should now call
the law of reciprocal combining proportion by weight, which is
generally supposed to have been first discovered by Wenzel and
Bichter, and to have been fully announced by Dalton.

The experiments on air to which I now turn, supplied mate-
rials for four papers, besides leading to the observation of many
phenomena which were never made public. Of these un-
published observations Mr. Harcourt has given an account,
including a lithograph fac-simile of a portion of Cavendi^'s
laboratory note-book.* I have also referred to those private
records in the sections of the Water Controversy devoted to the
dates of Cavendish's experiments and conclusions, so that I
shall only allude to them here very generally.

In the interval which elapsed between the publication of
Cavendish's first chemical papers and those we are now discuss-
ing, Priestley, the fourth of the great English pneumatic che-
mists, had appeared on the field ; while Scheele, in Sweden, and
Lavoisier, in France, besides other less distinguished observers
in different parts of Europe, had effected the discovery of
nearly all the gases known to us at the present day, and their
study engrossed the attention of every chemist. In particular,
the relation of the atmosphere to combustion demanded ex-
planation, and the nature of the change which the air underwent
when inflammables, burned within confined portions of it,
deprived it of the power of further supporting combustion.
At this problem all the active chemists of Europe were now
working, but with very unequal success, owing to the false
theory of combustion which the majority espoused, and the
erroneous opinions which were current concerning the con-

* BritUh AstociatioH Report^ 1839> pp. 45-68.

STAHL^S THEORY OF COMBUSTION. 35

stitution of atmospheric air. The nature of Cavendish's most
important chemical discoveries cannot be understood without
a reference to these points, which, however, shall be very brief.

Boyle, Hooke and Mayow in England, and Rey in France,
besides other early discijdes of the school of Bacon, understood
the true nature of combustion in air much better than the im-
mediate predecessors of Lavoisier. The former held as we do,
that a burning body is literally fed by the air, and they appre-
hended with considerable clearness, that burning combustibles
add sometiung to themselves from the atmosphere. Some of
these observers were also well aware that combustibles are con-
verted by combustion into substances possessing greater weight
than the original inflammable. In an evil day, however, Beccher
and Stahl, two men of unquestionable genius, devised a theory of
combustion which led all chemistry astray for half a century.
According to their view combustion consisted in the emission
from the combustible of a peculiar fiery principle, to which the
name phloffision was given. It was present in all inflammables,
however different their appearance and properties. When they
burned, it passed out of them into the air which surrounded
them, and by its loss they became changed in character and quite
incombustible ; but if phlogiston was restored to them, they
recovered their original appearance and properties, among the
rest, their combustibility.

Much has been said by the historians of chemistry in praise
of this theory as having served, in spite of its inaccuracy, to
guide chemistry to great results, at a time when the science was
not ripe for a juster theory. From this statement I must totally
dissent. Its devisers assuredly were men of rare gifts, and their
theory, welcomed by their fellows and immediate succes-
sors as a great boon to the science, exerted for some forty or
fifty years a strange fascination over all the chemists of Europe*
These forty years, however, were like those spent by the
Israelites in the wilderness, after their glimpse of the Promised
Land. Had Stahl and Beccher carried out the conclusions
which the early disciples of Bacon had imperfectly annoimced,
we should not have waited till the close of the eighteenth
century, and the advent of Lavoisier, for the true interpretation
of the nature of combustion. A Joshua would have been found

36 LIFE OF CAVENDISH.

some half a century sooner^ and the goodly land which the
chemists cultivate^ would exhibit a much wider extent of fertile
territory than it does at the present day.

It is a vain thing assuredly to speculate on what would
have been, if what has been had not been^ or had been other-
wise. The progress of science, we will not doubt, is deter-
mined by great laws, which we may some day be able to trace ;
and as it is, we perceive clearly enough that human progression
is not a continuous upward flight, but an alternation of risings
and fallings, which greatly retards the degree of additional ele-
vation attained within a given time, so that we need not wonder
that chemistry should exhibit in one of its epochs a retro-
gression for some half a century; but at least we should refrain
from calling it progress. We may believe that such relapses
will become fewer and fewer, as the human race grows older,
but in the meanwhile no service can be rendered to the cause of
truth by affecting to deny that, especially in the early history of
the sciences, we find long periods of total stagnation, and the
tide even ebbing, when by our calculations it should have
overflowed.

So large a portion of this work is unavoidably occupied,
with discussions concerning phlogiston, that even the most
general reader must have some conception of its meaning.
StahPs theory of phlogiston was not a refined speculation. It
scarcely deserves to be called a scientific hypothesis. It
really amounted to nothing more than the assertion, that a
tbody was combustible because it contained something com-
bustible; which was equivalent to the identical proposition
that a body burned because it burned. This declaration
instead of being a refinement of philosophy, to which only
a man of science could reach, was but the reduction to terms,
of a vulgar belief. It was a poetical, rather than a scientific
thought ; for the natural tendency of every untrained imagi-
native mind, as we see in children, and in the early history
of all nations, is to impute every manifestation of power, to the
presence in the body manifesting it, of some inner principle
more or less self-sustaining, and resembling a living or vital
agent.

The same spirit, which made the Greeks people the winds

PROPERTIES OF PHLOGISTON. 37

and the wayes^ the rivers and the trees^ with gods ; which makes
the savage regard the compass needle as animated ; and the
child demand to see in some visible shape^ the motive principle
of a watch or moving toy ; led the Chemists of the seventeenth
and eighteenth centuries to declare that a candle burned because
it contained a burning principle. I have sometimes thought
that this theory was in part occasioned by the spectacle of the
sun and other heavenly bodies unceasingly emitting heat and
light. I have fotnd, however, no reference to this striking
phenomenon in the writings of the phlogistians ; and however
much the unbroken radiance of the sun might justify a popular
belief in the power of combustibles simply to emit light, it could
never justify the assertion of this even as a probable truth, for
this would have been to explain one mystery by another. It
seems to me, on the whole, that whilst poetry might have wel-
comed the doctrine that a blazing body throws off light and
heat, as a bell utters a sound, or a flower exhales an odour, that
science could only accept it as an hypothesis of no great likeli-
hood or high value, and which at all events required at once to
be tested, as to its utility as an interpreter of known phenomena,
and a guide to the discovery of new ones.

The doctrine of phlogiston, however, was not dealt with
thus. Instead of being treated as a doubtful hypothesis, it was
employed as a perfect theory ; and phenomena at variance with
it were either wilfully overlooked, or compelled to adjust them-
selves to its Procrustean bed. A true hypothesis, or one in the
main, true, is always found capable of explaining more than it
professed or expected to explain. But the phlogiston hypo-
thesis transgressed its own self-imposed conditions, and failed to
explain the most simple and essential phenomena of combustion.
Thus its presence in bodies was held to confer upon them com-
bustibility, yet when transferred from a blazing combustible to
air, instead of rendering the latter inflammable, and changing it
into a gas which could be kindled, it changed it into one which
was totally incombustible and at once extinguished flame ; for
phlogisticated air in its simplest form was our nitrogen. Again,
phlogiston was held to be a material and therefore ponderable
substance, so that its escape from a combustible should have
caused the latter to diminish in weight ; yet the metals and

38 LIFE OP CAVENDISH.

phosphorus were known to increase in weight by combtistion.
Thus the lameness of the phlogiston hypothesis was betrayed at its
first step^ and it had to be furnished with a crutch^ in the shape
of an assumption that it was a principle of levity, so that a body
containing it weighed less than if it were absent, before it could
move a step further. Many of the Phlogistians, indeed, did not
adopt this assumption, which confounded so strangely absolute
weight with specific gravity; but they ignored the phenomenon
of increased weight, which they could not explain, and stood in
the anomalous position of professors of a Quantitative Science,
who should weigh and measure at every step, and yet had put
aside the balance as a useless thing. This, however, was not
all. That a burning body changed the quality of the air around
it, whilst itself undergoing a complete change of properties, had
not escaped the attention of the phlogistians. Beccher and
Stahl, although they made no investigation into the nature of the
change which air underwent when it supported combustion, were
aware that a limited quantity of air in which a combustible had
burned till it was extinguished, could not a second time support
combustion, a fact indeed which was matter of universal belief
from the earliest times.

Such, then, was the crude and clumsy h3^olhesis which
was recognised as a fundamental law of all chemistry, at the
period when Cavendish commenced his Experiments on Air.
Their object was to ascertain what Beccher and Stahl should
have ascertained before they promulgated their hypothesis, viz*^
what change does combustion effect upon air. The discovery
of oxygen, of nitric oxide, and of other gases, and the experi-
ments which Priestley, Scheele, and Lavoisier had been assidu-
ously making for some years, had directed the attention of
chemists to the fact, that air not only became irrespirable and
unable to support combustion when exposed to the action of
burning inflammables, but at the same time underwent a dimi-
nution in volume, so that a portion of it was to appearance
lost* To discover what became of the lost air was a question

* Barbarous as was the nomenclature introduced by the phlogistians, and com-
plete as has been its abandonmenti they employed one phrase for which it would be
well if we possessed an equiralent. This was phlogisticated air^ a term applied in
its widest acceptation to air so vitiated as to be unable to support respiration or

COMBUSTION IN AIR. 39

y which; in 1777 y greatly interested the active chemists of Europe,

and Cavendish's attention was specially directed to the problem,
by the researches of Scheele on this point, as appears from a
statement in the MSS. note-book of the former. Priestley and
Lavoisier had, contemporaneously with Scheele, investigated
the same subject; and all three had made some progress,
especially Lavoisier, in explaining the problem. When those
researches commenced, air was universally reputed to be a
simple or elementary body. It was liable, according to the
phlogistians, to vitiation, by the addition to it of phlogiston, so
that it was referred to as being more or less phlogisticated,
according to the degree of its power to support respiration and
combustion. When oxygen was discovered by Priestley and
Scheele, it was regarded by them as air altogether respirable,
and exhibiting a maximum power of supporting combustion,
because it was quite free from phlogiston. It was named
accordingly de-phlogisticated air, and for a season the atmos-

^ ' phere was referred to as consisting of two parts, a *^ dephlo-
gisticated part^^ and a " phlogisticated part,*' which differed
from each other only in degree. By-and-by those parts were re-

! garded as differing in kind, not merely in degree ; the dephlogis-

ticated part, or dephlogisticated air, being our oxygen, and the
phlogisticated part or air, our nitrogen. Cavendish's enquiry
began before this later view became general. He had proceeded
but a short way in his attempt to discover what became of the
air apparently lost during combustion, when he was arrested in
his researches by the necessity which their successful prosecu-
tion laid him under of ascertaining the quantitative composition
of atmospheric ain The problem which originally interested
him presented itself in this shape. If any combustible, such as
hydrogen, phosphorus, or a candle, was allowed to bum till it
went out, in a portion of air confined over water, the volume of
the air was observed to diminish as the combustion proceeded,
and at its close the water was found to have risen through
about a fifth of the space originally occupied by the air.
Cavendish, with the precision which characterised all his

oombastion. We have an equivalent phrase, so far as respiration is concerned, in
the term irrespirablef but we are compelled to employ a tedious circumlocution
when we wish to refer generally to air incapable of supporting combustion.

40 LIFE OF CAVENDISH.

researches, sought to ascertain the extent to which air could
thus be diminished in volume, before he decided why it was
diminished. His note-book shows that he investigated the two
questions simultaneously, but he first announced the result of
his enquiry into the composition of the air. This he published
in 1783, and, like all his other papers, the modest title which it
bore, An Account of a New Eudiometer^ conveyed a very im-
perfect idea of its contents. Referring the reader for a full
analysis of these, to the abstract of the paper in the sequel,
I may notice here, that it is ostensibly devoted to the expla-
nation of an instrument for determining the proportion of
oxygen in air, by observing the contraction which followed it»
mixture with a given volume of nitric oxide. Priestley, the first
investigator of the properties of nitric oxide, had devised this
process, but was too inaccurate a manipulator to make good use
of it. A more careful observer, indeed, than the ingenious
Priestley might have been led astray, by the employment of his
nitrous gas, for it can combine with oxygen in various propor-
tions, according to the mode in which it is mixed with air, so
that the amount of contraction may be very diflferent, although
the volume of oxygen is the same. Priestley, however, and the
great majority of his contemporaries were either ignorant or
heedless of this fact, and conceived that the purity of air might
be accurately measured, by observing the amount of contraction
which attended the mixture of it with nitric oxide over water ;
the air being the purer the greater the extent of contraction.
Experimenting in this way, they could not miss arriving at the
conclusion, that the atmosphere varied excessively in purity ;
and we find them accordingly travelling from place to place,
analysing what they called the good air and the bad air of
different localities, and coming to the most extravagant conclu-
sions as to the relative purity of specimens of it, in which all the
refinements of our modern analysis would fail to detect any
difference. The instruments which they employed, they character-
istically termediBttrftometo**, or measurers of the goodness of the
air; the object of the analyst being to determine the freedom of
the air from phlogiston, which rendered it bad in proportion to
the amount of it present.* By the performance of an immense

* The word eudiometer, which remains in our nomenclature as the solitary

DISCOVERY OF THE COMPOSITION OP AIR. 41

number of elaborate experiments^ Cavendish succeeded in per-
fecting a process, by means of which he could employ nitric
oxide so as to occasion a constant amount of contraction, when
mixed with different portions of the same specimen of air.
Having certified this, he applied his method to the determina-
tion of the two important questions : Is the atmosphere constant
in composition ? And if so, what is its composition ? To solve
these problems, he experimented for some sixty successive days
in 1781, making many hundred analyses of air. He compared,
also, air collected at one period of the day with that collected at
another, and that of the town with that of the country. He
came to the conclusion which all subsequent observations have
confirmed, that no sensible difference can be detected by
Eudiometrical analysis between the purity of different specimens
of atmospheric air. It was universally such ^^ that the quantity
of pure air in common air is H"* or as we should now word
it, the per centage by volume of oxygen in air is 20.83. This
number approaches very closely to that obtained in our most
recent analyses, and is remarkable for its accuracy, when we
consider how totally the great majority, not only of Cavendish's
contemporaries, but also his successors, even among living
philosophers, failed to obtain any constant results with nitric
oxide eudiometers. Cavendish is entitled to be called the
discoverer of the constant composition of the atmosphere, and
its first accurate analyst. It may be noticed here, to prevent
subsequent confusion, that the atmosphere had long occupied
his attention. So far back as 1766 he had imperfectly analysed
it, by observing the loudness of the report which it gave when
detonated with hydrogen. This device might be called an
Acoustic Eudiometer. Whilst engaged also in the enquiry which
we have been discussing, he checked the results obtained with
nitric oxide, by observing the diminution which air underwent
when exposed to liver of sulphur dissolved in water, and when
exploded with hydrogen in a shut vessel by means of the

relic of the phlogiston vocabularyy is now used in the sense of a measurer of the
amount of oxygen present in air. By its introducers, however, it was intended to
refer to a measurer, not of the presence of oxygen, but of the absence of phlogiston.
I hare referred to this point more fully in the abstract of the paper under discussion.
* MS. Laboratory Note-Book, p. 109.

42 LIFE OF CAVENDISH.

electric spark. The apparatus last referred to is the one
generally named at the present day Cavendish's Eudiometer^
and is the instrument which^ from its interest in connexion
with the discovery of the composition of water^ has been
selected by the Cavendish Society as their emblem^ and placed
on the title-page of their publications. Cavendish, however,
never named this instrument a Eudiometer, nor was it his
device, but Volta's, The Society^s emblem represents the
instrument as it is constructed at the present day, not as it was
used by Cavendish."^ He concludes this paper with an estimate
of the nature of the information which the eudiometer supplies,
which he shows to be very much smaller than the majority of
his contemporaries imagined. His views in this respect accord
with those universally entertained at the present day, and are
another monument to the caution and sagacity with which he
kept himself free from the prejudices of his time, and anticipated
conclusions which were not generally accepted till a recent
period.

The protracted eudiometrical enquiry we have been consider*
ing, taught Cavendish the importanttruth, that when air was dimi-
nished in volume, or a portion of it to appearancelost,by the effect
of burning combustibles or so-called phlogisticating agents upon
it, the maximum amount of diminution which could be produced
was equivalent in round numbers to one-fifth of the original
volume of the air. He knew, also, that it was the dephlogisti-
cated part, or pure air (oxygen), of the atmosphere, which
disappeared during combustion, so that he was now fully
prepared to enquire what had become of the lost oxygen. His
account of this enquiry forms the first series of his Experiments
on Air, which was read to the Royal Society in January 1784,

* The inBtrument, as Cayendish describes it, was a glass globe, with a brass
stopcock, wires for passing the electric spark, and a hook, or other arrangement for
hanging it to the beam of a balance. As constructed at the present day, and repre-
sented in the title-page of this and the other publications of the Cavendish Sodety,
it is pear-shaped, and provided with a glass stopper, which can be held firmly in its
place by screws, besides a glass stopcock in addition to a brass one, and a moveable
stand to which it can be fixed, after it has been exhausted at the air-pump, and filled
with a mixture of hydrogen and oxygen. Were Cavendish alive among us, he would
not recognise the modem instrument as resembling any part of his apparatus, and it
would startle him to hear it called his eudiometer, by which term he would under-
stand the nitric oxide apparatus which he described in 1783.

cavendish's eudiometer, 43

exactly a year after the paper on the New Eudiometer* When
he commenced those researches, he found an opinion prevailing,
that die production of fixed air, or carbonic acid, is the invari-
able result of what he called the phlogistication, and we should
call the deoxidation, of atmospheric air. He readily disproved
the truth of this view, and also of another notion, that nitric, or
sulphuric acid was produced in those circumstances ; and having
disposed of these erroneous opinions, he proceeded to observe
with great care, what was the product of the combustion of
hydrogen in air and in oxygen, Priestley, and a friend of his,
Mr. Warltire, had already experimented on this subject, with a
detonating globe of the same kind as that referred to previously,
as called at the present day. Cavendish's Eudiometer.* Their
experiments were made partly in metallic, partly in glass vessels,
and when employing the latter, they observed a deposition of
moisture follow each explosion, but Priestley paid no attention
to this phenomenon, and Warltire referred it to the condensation
of water which had been diffused in the state of vapour through
the gases. It at once, however, attracted the attention of
Cavendish, who from the first appears to have anticipated that
in the deposited water would be found the oxygen, which dis-
appeared during the combustion of hydrogen in air, and the
explanation of the diminution in volume which attended the
vitiation of air. It will be remembered that in his paper
on hydrogen, of 1766, he had represented this gas as itself
phlogiston. He now experimented accordingly upon it, not as
an individual combustible which would yield a certain product,
but as the phlogiston which was present in all combustibles, and
the product of whose combustion would represent the universal
product of combustion. He first employed hydrogen and air,

* The instrument which Volta introduced for firing ezplosiTe mixtures of gas,
by means of the electric spark, and which still bears the name of Volta*8 Eudiometer,
was a tube or cylinder open at one end. I do not know whether Volta erer em-
ployefl a shut globe« but Priestley and Warltire certainly did before Cayendish, as he
freely acknowledged, and they in their turn, as well as Watt, referred the device to
Volta, so that it must be regretted that this apparatus has been called Cavendish's
Evdiometer, especially as Monge used an exactly similar apparatus, which he also
refers to Volta. It is the admirable use which Cavendish made of the detonating
globe, not the devising of it, which justifies its employment as the Cavendish
Society's symbol. The instrument itself might, perhaps, best be called, without
reference to any one's name, the Spark Eudiometer.

44 LIFE OF CAVENDISH.

varying their relative proportion, till he ascertained that ratio
in which, after their explosion in a shut vessel, the air was found
diminished one-fifth, whilst the residual air was free from oxygen^
and possessed the properties of nitrogen. In place of the oxygen
which had thus disappeared, and a volume of hydrogen twice
as great which had burned. along with it, there was found a
certain amount of liquid. The globe, moreover, had remained
shut during the experiment, so that nothing had been allowed
to escape, and nothing ponderable had been lost, for the vessel
was found to weigh the same after the electric spark had passed^
as before the explosion. In short, there was exactly the same
weight of matter in the globe after the explosion as before, but
the oxygen originally present in the air, and twice its volume of
the hydrogen which had been mixed with it, had disappeared
as gases, and were replaced by a volume of liquid, which, of
necessity, exactly equalled them in weight. Cavendish, accor-
dingly, unhesitatingly concluded, that in the circumstances de-
scribed, " almost all the inflammable air, and about one-fifth part
of the common air, lose their elasticity and are condensed into
the dew which lines the glass.^^ Having demonstrated in this
way that the lost gas was accounted for, and remained in the pro-
duced liquid, he proceeded to investigate the nature of the latter.
The globe explosions yielded too small a quantity of liquid for
a full analysis. He burned together, accordingly, by direct
combustion, a large volume of hydrogen with 2i times that
quantity of common air within a glass cylinder, and collected
the liquid produced. This he found to be without taste, or smell,
or action on colouring matter, and to leave no sediment on eva-
poration ; in short, he observes, " it seemed pure water,'^ and
his full conclusion was, " that this dew is plain water, and conse-
quently, that almost all the inflammable air, and about one-fifth
of the common air, are turned into pure water .^'

The proceeding quotation contains the account of the first
conclusion that was drawn concerning the compound nature of
water, and the possibility of producing it out of hydrogen, and
the oxygen contained in air. Cavendish proceeded to try
whether free oxygen, if detonated with hydrogen, would in like
manner yield water. The method of procedure here was simple,
for it was only necessary to fill the globe with a mixture of one

PRESENCE OF NITRIC ACID IN WATER. 45

volume of oxygen and two of hydrogen, and to explode it by
the electric spark, to secure the entire conversion of the contents
of the globe into water. Cavendish came as near this result,
as a slight mistake in the adjustment of the combining volumes
of hydrogen and oxygen, and the limits of error in such an
experiment, at the period when it was made, on the whole per-
mitted. In the course of these trials, however, an unexpected
and perplexing phenomenon showed itself. The liquid instead
of being pure water, was found in certain cases to consist in
addition of an acid, which analysis proved to be the nitric, and
a long and difficult investigation had to be prosecuted into the
source of this acid, the composition of which it must be
remembered was totally unknown in 1784. This startling phe-
nomenon, on which the chemistry of the period could throw no
light, would have stumbled the great majority of Cavendish's
contemporaries, as we may assert very safely, for it led not only
Priestley, but even La Place astray; and it was probably ignorance
of the phenomenon on the part of Watt and Lavoisier, which saved
them from being entangled in difficulties in their investigation into
the nature of water. Cavendish solved the problem without one
false step, and whilst he avoided the confusion in which it
involved others, he built upon it an additional great discovery.
After ascertaining that the appearance of nitric acid was not
dependent on the source from which the oxygen wasT prepared,
and that the acid did not show itself unless more than a combi-
ning measure of oxygen was detonated with the hydrogen, he
traced its production to the presence in the globe of a little
nitrogen, derived from the atmospheric air which had originally
filled it, or had become mingled with the hydrogen and oxygen
during their preparation or collection. He amply verified this
conclusion, by showing that the artificial addition of nitrogen to
hydrogen, mixed with more than one-half its volume of oxygen,
increased the amount of nitric acid produced at each detonation,
and on the other hand, that if the hydrogen instead of the oxy-
gen was in excess, no nitric acid appeared, although nitrogen
was present. In this way he demonstrated that the only product
of the combustion of pure hydrogen and oxygen is pure water ;
but he was further led to a view of the composition of nitric acid,
which he carried out in the second series of his experiments on

46 LIP£ OF CAVENDISH.

air, and which secures to him the honour of being the discoverer
of the composition of nitric acid, as well as of that of water.

Avoiding here any minute discussion of Cavendish's opinions
on the nature of water, or of the controversy in which the pub-
lication of the paper under discussion involved him with Watt
and Lavoisier, both of which questions are fully discussed in
different portions of this volume, I would draw attention to the
following considerations* The general conclusion to which
Cavendish came concerning the nature of water, was in his own
words, ^^ that water consists of dephlogisticated air united with
phlogiston ;'' and as dephlogisticated air was his term for oxygen,
and phlogiston his term for hydrogen, this statement closely
corresponds to the modem view of the nature of water intro-
duced by Lavoisier. The two views cannot be considered iden-
tical, yet this is certain, that Cavendish was the first who con«
sciously converted hydrogen and oxygen into water, and taught
that it consisted of them.

His identification, however, of hydrogen and phlogiston, and
his inheritance of the prejudices of the early phlogiston school,
led him to the erroneous conclusion that every combustible
contains hydrogen, and that the deoxidation of air and the
oxidation of combustibles, are invariably accompanied by the
production of water. In this respect he erred, but we may for-
give the discoverer of so great a truth as that of the composition
of water, for over-estimating its importance. To this, and to
the other points glanced at in this sketch of the first series of
experiments on air, I have referred fully in the abstract of the
paper, and in the chapters devoted to the discussion of the
Water Controversy.

The second series of experiments on air was read to the Royal
Society in June 1785, about eighteen months after the reading
of the first. The paper was occupied with the discussion of a
point which had only been imperfectly examined in the earlier
experiments, viz., the cause of the diminution in volume which
attends the passage of the electric spark through air. Caven-
dish had imagined that this was owing to the combustion of
inflammable matter in the apparatus. He now demonstrated,
that though this might be an occasional and slight cause of the
diminution of air exposed to electric sparks, it certainly was not

NATUBS OP NITRIC ACIB. 47

the chief cause of its diminution, which, as it appeared, was
nudnly owing to the combination of the nitrogen and oxygen of
the air^ to produce nitric acid. In his first series of experi-
ments, he had shown that if nitrogen mixed in small quantity
with hydrogen, was exploded by the electric spark, along with
excess of oxygen, nitric add was produced. He now showed
that the hydrogen might be omitted, and that if a mixture of
pure nitrogen and pure oxygen be exposed to electric discharges,
it will yield nitric acid ; and farther, that if those gases be
mixed in a proper proportion, over caustic potash, they may be
entirely condensed, or converted by the electric spark into nitric
acid, which combines with the alkali to form nitre. On the other
hand, he pointed out that pure nitrogen, or pure oxygen, if taken
singly, was not sensibly a£Fected by the spark. The negative
result with the single gases, as contrasted with the positive result,
when both were taken, demonstrated, as Cavendish showed,
that the acid resulted from a combination of the two to produce
it. He did not, however, although he was aware of the possi-
bility of such an explanation, affirm that nitric acid was a direct
compound of nitrogen and oxygen, as we do at the present day.
True to the doctrine which he had announced in his earlier
pnper, he contended that nitrogen was a compound of hydrogen
and nitric acid, and that the effect of the passage of the electric
spark was to occasion the combination of this hydrogen with the
oxygen, whilst the nitric acid separated. He Was thus not the
direct asserter of the modern doctrine of the composition of
nitric acid, which he deliberately set aside, although aware of
the terms in which Lavoisier would have announced it ; and in
this he perhaps showed a little wilfulness, as well as neglect of
that quantitative method of procedure which he in general
practised more than any of his brethren, and which if it had
led him on this occasion to employ the balance, would have
compelled him to change his theory. His views, notwithstand-
ing, concerning the combination of the two gases to produce
the acid, were most explicit and unhesitating, and saved him
from the error of Priestley and La Place, who inferred from the
earlier observations on the productions of nitric acid, that it, as
well as water, was a compound of hydrogen (or at least inflam-

48 LIFE OF CAVENDISH,

mable air) and oxygen. Cavendish, accordingly, has always
been considered the discoverer of the composition of nitric acid,
although in strictness of speech this is not the exact merit which
an historian at the present day can assign him, seeing that he
manifestly regarded nitric acid as a simple, or at least undecom-
pounded body, whilst nitrogen, according to him, was a com-
pound ; his view 1)eing exactly that of a consistent phlogistian,
who did not employ the balance, and in whose eyes nitrogen
was a body constituted like a metal, of a calx (the nitric acid)
and phlogiston. This opinion appears at the present day
extremely different from that we entertain, but in 17^5, when
the most enlightened chemists were only in a transition. state
from the phlogiston to the antiphlogiston doctrines, the differ-
ence went for much less than it does now-a-days : nevertheless,
it should be freely acknowledged that Lavoisier put the true
interpretation upon Cavendish's views. Yet the merit which
can be assigned to the former is in reality small, for he discre-
dited the reality of Cavendish's observations, and could not
succeed in verifying his results, which would scarcely have been
the case had he tried the necessary experiments with the same
strong conviction of their reality which prompted his obser-
vations on the production of water. Cavendish was induced, by
the failure of Lavoisier and others, to return to the subject;
and in 1788 he published the last part of his Experiments on
Air, in the shape of a record of the successful repetition, by a
Committee of the Royal Society, of his observations on the
conversion of a mixture of nitrogen and oxygen into nitric acid
by the electric spark. The foreign philosophers, in truth, had
in one case at least been quite successful, but did not perceive
that they were so.

In the latter part of the original record of his trials on this
subject. Cavendish prosecutes the enquiry into the nature of
nitrogen, which he had in part pursued in 1772. He draws
attention to the fact, that it was uncertain whether the phlogis-
ticated (nonK)xygenous) part of the atmosphere consisted
entirely of a single gas ; and he proceeded to test this by trying
whether a given volume of the phlogisticated part was entirely
convertible into nitric acid by explosion with oxygen. He found

LATER ESSAYS. 49

that it was, and thus supplied a demonstration of the homo-
geneous nature of nitrogen, such as none of his contemporaries
could have given.

The paper of 1788 was the last upon chemistry which Caven-
dish published. The opinions of Lavoisier, which were then
rapidly becoming general, diverted chemistry into channels
foreign to those into which Cavendish had guided it; and
although he regarded the triumphs of the great French chemist
with characteristic composure, and even interest, he appears to
have been repelled by the new aspect which chemistry was
assuming, from engaging in its further prosecution. The catho-
licity of his scientific tastes and the serenity of his nature made
a change of studies, or rather the abandonment of one among
the many he always prosecuted, an easy matter.

His remaining published papers refer to meteorology and
astronomy. In 1790 he published an essay on the height of a
remarkable aurora seen in 1784. In 1792 an elaborate paper
appeared, on Tfie Civil Year of the Hindoos^ of which it may
suffice to give here the title. In 1797 a letter from him to Mr.
De Mendoza Y Rios was published, containing A method for
Reducing Lunar Distances. In the succeeding year appeared
the paper which, next to those on the discovery of the com-
position of water and of nitric acid, and that on the torpedo,
has made him most famous. This is the celebrated enquiry
into the density of the earth, which was communicated to
the Royal Society in 1798. The apparatus he employed was
contrived by his friend the Rev. John Michell, who died
before he had an opportunity of experimenting with it.
It came into Cavendish's hands, who adopted the principle
it embodied, but had the chief part constructed afresh, so
as to ensure greater accuracy in the results it was expected
to yield. Without attempting here to describe minutely
the nature of the experiments made with this apparatus, it
may suffice to say that they consisted in observing, with many
precautions, the movements of a long lever delicately suspended
by the centre, so as to hang horizontally, and furnished at either
extremity with small leaden balls. When two much larger and
heavier balls of the same metal were brought near the smaller
ones, the latter were attracted towards them with a certain force.

50 LIFE OF CAVENDISH.

the measurement of which supplied one essential datum for the
determination of the mean density of the eardi. It was deter-
mined by Cavendish to be 5*4; in other words, our globe,
according to him, is nearly five times and a half heavier than the
same bulk of water would be. The experiments referred to are
so difficult of performance, and the density of the globe is a
point so important in reference to a multitude of astronomical
and geological problems, that the ^^ Cavendish experiment'' (as
the researches taken as a whole are generally called) has always
been regarded with the greatest interest by men of science. In
1837^ Professor Reich, of Freiberg, published the account of a
careful repetition of the Cavendish experiment. He made the
density of the earth, as Cavendish had done, 5*4. In a later
repetition, however, made by the late Francis Baily, the astro-
nomer, with extraordinary precautions to secure accuracy, a dif-
ferent result was obtained. After four years of protracted, though
interrupted trial, during which several thousand observations
were made, the final result was, that the density of the earth was
higher than Cavendish had represented it, namely 5*6, or more
than five and a half times that of water. No greater compliment
to Cavendish's accuracy can be desired than that afforded by the
bict that so accomplished a natural philosopher as the late Mr.
Baily, provided with all the improvements which forty fertile
years had added to mechanical contrivances, and aided by the
counsel of many distinguished natural philosophers, with a
committee of the Astronomical Society taking part in the pro-
ceedings, and Government defraying the cost of the experiments,
did not, after nearly three years of actual experimenting, find a
greater difference than that stated above, as distinguishing his
result from Cavendish's.

The last paper which the latter published, was on an im-
provement in the manner of dividing astronomical instruments*
It appeared in 1809, the year before his death.

Such then, is a brief sketch of all Cavendish's published
essays as well as of his unpublished papers referring to Che-
mistry.

I have made an exception in the case of the chemical essays
to the general rule which the conditions of this volume necessi-
tate, viz. that attention should be limited to published papers only.

UNPUBLISHED PAPERS. 51

Cavendishes unpublished essays, however^ upon chemical subjects
(including heat), with the exception of that on arsenic, to which
I have referred very briefly, have so direct a bearing upon his
claims as a discoverer^ the discussion of which occupies the
largest part of this volume, that I have thought it necessary to
notice them, especially as I profess chiefly to deal with his
merits as a chemist, and write at the request of a society
instituted for the promotion of chemistry and the sciences
nearly allied to it. It would be doing Cavendish great
injustice, if no reference were made to his other unpublished
papers^ for although his actual reputation is based upon what he
made known to the world in his lifetime, and is all that the
historian of chemistry is concerned with, it is otherwise with his
biographer, who is not only entitled, but is required to form his
estimate of his capacity from all the materials accessible to him.
Unpublished papers, of the genuineness and authenticity
of which there is no question, may at any time be produced
as posthumous works, and alter the accepted estimate of an
author^s merits* The limits of this volume, nevertheless, do
not permit the publication of abstracts, however brief, of all
Cavendishes MSS. ; nor am I competent to the task, if there
were room for its being undertaken. I shall be satisfied with
drawing attention to these papers, and with concurring in the
hope expressed by Mr. Harcourt, that at least the more important
among them will yet be published in full.

The references to them which follow, are founded upon a
personal inspection of the original MSS., which, with the excep-
tion of those referring to electricity (at present in the hands of
that most competent critic of their merits, Sir William Snow
Harris), were placed at my disposal through the good offices of
the Rev. W. V. Harcourt, and the courtesy of Lord Burlington,
to whom they belong. Thus much may be said of them
here. The portion most interesting in reference to the personal
narrative, is contained in a parcel entitled " Journeys.^'
It contains the account of a series of tours made through Eng-
land during the summer and autumn of 17B5, 1736, 17^7 f
and 1793 ; besides the record (in Cavendishes handwriting) of a
journey undertaken by Dr. Blagden through Belgium in 1789,
and an untitled essay apparently occupied with a summary of

£ 2

52 LIVE OF CAVENDISH*

the geological observations made in those journeys, and at all
events, containing a general sketch of the upper formations pre-
valent throughout England. The records of those tours are in
the shape of journals, drawn out at leisure, and for the most
part, in Cavendish's handwriting. Some of them have been
transcribed by a clerk, but they have been revised by Cavendish,
who has corrected the blunders committed by the transcriber in
writing unfamiliar scientific terms, and has supplied the gaps
which occur at intervals, where the copyist apparently has been
unable to decipher certain words in the text from which he
transcribed. Cavendish appears to have been accompanied in
these journeys by Dr. Blagden, who is frequently referred to.
The Rev. J. Michell is also referred to, but I am not certain
that he was a party in any of the excursions. They were so
extensive as to include the larger part of England, but especially
the Southern, Midland, and Western counties. Cavendish was
as far as Whitby on the north*east coast, and Truro on the
south-west, and one journey was devoted in greater part to a
tour through the southern counties of Wales. He does not
appear, however, to have visited the south-eastern counties of
the island.

The object of these journeys was entirely scientific, and was
mainly the investigation of the geology of the districts passed
through, but much attention was also devoted to the mining
operations and metallurgical processes which could be witnessed
in the same places. Of the value of the geological observations
I do not pretend to form an estimate, but they were evidently
made with the same caution and precision which marked
Cavendish's experimental enquiries. Heights were determined
by the barometer ; the temperature of springs carefully ascer-
tained; the thickness, inclination, relative succession, and
physical appearance of the rocks minutely noticed, and specimens
of the characteristic minerals collected for analysis. The journals
recording those observations, including the general sketch of the
arrangement of the superficial strata of the island, deserve the
attention of the historians of English geology. They embody
the results of three very able observers. Cavendish, Michell, and
Blagden, who, so far back as 1785, commenced a patient and
extensive exploration of the geology of the country, the record of

SCIENTIFIC JOURNEYS. S3

which can scarcely fail to contain much that is interesting to
students of science at the present day. These journals also
contain^ as already stated^ rainute accounts of various manufac-
turing processes related to chemistry. In Wales, CornwaU^
Derbyshire, Warwickshire, Staffordshire^ Yorkshire, and else*
where^ Cavendish witnessed the mining, reduction, and working
of tin, copper, lead, iron, steel, alum-schist, and the like, which
he describes at great length, and with his usual admirable clear-
ness. All interesting pieces of machinery also, which he
encountered on his way, such as Watt's improvements on the
steam-engine, which were explained to him at Birmingham by
the great engineer himself, are referred to, and occasionally
figured ; and the recipes for producing peculiar effects, whichhe
obtained from those engaged in the chemical operations he
witnessed, are carefully recorded. No allusions occur to merely
personal incidents or adventures ; to the scenery, except as a
geological character ; or to individuals, however famous, unless as
authorities on some fact thought worthy of being mentioned.
The journal is limited to recording, in the fewest possible words,
the strietly scientific observations made during a series of tours>
which frere prosecuted on Sundays as well as week-days, with
one undeviating purpose, which nothing was allowed to disturb.
The remaining MSS. consist of papers on subjects included
under mathematics, mechanics, optics, physical geography, me-
teorology, and astronomy, besides miscellaneous observations.

54 LTFE OF CAVENDISH.

CHAPTER III.

CONTROVERSY BETWEEN CAVENDISH, WATT. AND LAVOISIER,
CONCERNING THE DISCOVERY OF THE COMPOSITION OF WATER.

In the preceding chapter I have endeavoured to give some
account of what Cavendish effected^ alike by his published and
unpublished labours^ towards the extension of physical science.
A detail of his labours seemed the best preparation for a just
estimate of his merits as a scientific enquirer and his integrity
as a man. In truths unless to those who are already familiar
with what he has done to extend our knowledge of physics^ an
exposition of his deserts would be little else than an appeal to
their prejudices or ignorance. To those^ however, who have
interested themselves in the recent literature of chemistry, and
who have made its progress in our own country a matter of
special study, the position which Cavendish at present occupies
in the eyes of no inconsiderable section of the learned public,
will prepare them for the discussion which follows. Those who
for the first time concern themselves with his labours, will be
surprised to learn that the modest, retiring, and almost inordi-
nately cautious man, whose early history and scientific researches
have been detailed, has been accused within the last few
years, both of incapacity and dishonesty, and by no obscure
assailants, for among them rank more than one of our distin-
guished men of letters and science, whose connexion with the
vexed question I have to consider, gives it an interest apart from
that which it intrinsically possesses.

The ancient Egyptians counted no man's reputation certain,
till death had set its seal upon him and his survivors had
solemnly pronounced judgment upon his life and character.
Their judgment, however, was a swift one, and was passed
before the grave closed over its object.

It could have been wished that the posthumous ordeal to

CAVENDISHES RIVALS, 55

which Cavendish has been subjected^ had followed his decease
as speedily as it would have done had he prosecuted chemistry
some thousand years ago^ in the land where tradition asserts it
to have had its origin. In that case he would have been tried
by those who could speak from personal knowledge as to the
charges preferred against him^ and he would have been quickly
acquitted. As it was^ he went down to the grave with the
gathered honours of more than threescore years and ten upon
bis head^ and all who could have witnessed to his integrity had
followed him to the tomb^ before he was summarily denounced
before a great public body as having unfairly claimed the honour
of making a famous discovery, and of having done a brother
philosopher a grievous wrong. These accusations have reference
to the claim which he so modestly asserted^ to be entitled the
discoverer of the composition of water, and were first made
extensively public some ten years ago.

The experiments on the composition of water were made in
the summer of 1781. Owing, however, to the additional obser-
vations which were necessary in order to ascertain the nature of
tiie acid which showed itself, and the long series of experiments
which Cavendish, with the extraordinary patience and caution
which characterised him, thought it necessary to perform before
publishing his entire results, as well as in consequence of other
researches, which were prosecuted simultaneously; his paper
entitled Experiments on Air, was not read to the Royal Society
till January 1784. The delay which thus happened has caused
his claim to be entitled the discoverer of the composition of
water to be contested; the rivals put forth being no other than the
celebrated James Watt and the great French chemist, Lavoisier.

When Cuvier, in 1812, as Secretary of the French Academy,
read an floge on Cavendish, who was elected in his old age a
member of the Institute, he said, in reference to the subject of
his notice, that " his demeanour, and the modest tone of his
writings, procured him the uncommon distinction of never having
his repose disturbed either by jealousy or by criticism.***

It was reserved for Cuvier*s distinguished successor, Arago,
the present perpetual Secretary of the French Academy, to
become Cavendish's accuser. When it fell to Arago's lot to

* BlogH HUt, torn* ii.« page 104.

56 LIFE OF CAVENDISH.

write the floge of Watt, which was published in 1839, he came
to this country to collect materials for the purpose ; and in the
course of his researches among Watt's published and private
papers, he arrived at the conclusion that Watt, and not Caven-
dish, was the discoverer of the composition of water. He
published this, accordingly, to the French Academy, accom-
panied by what amounted to a direct charge against Cavendish
of deceit and plagiarism, inasmuch as he was said to have
learned the composition of water, not by experiments of his own,
but by obtaining sight of a letter from Watt to Priestley.
Throughout the whole ^loge, moreover. Cavendish is made to
appear as a mean, jealous, vain, and dishonest person, who by
a cunning trick appropriated to himself the discovery of another,
to whom he did not make even a show of restitution till he was
detected in the fraud.

The scientific men of Great Britain were startled at the
charge brought against Cavendish. Of all her illustrious
philosophers, he was, without exception, the very last in refer-
ence to whom it was possible to believe that the accusation
could be true. A man to whom applause had ever been hateful,
and who had systematically avoided and declined the honours
which his countrymen would willingly have conferred upon him,
was not likely suddenly, and on a single occasion, to grow
covetous of distinction and to seek to gain it by fraud. More-
over, it soon appeared that Arago had studied the papers of Watt
(as was natural in his Eulogist) much more fully than those
of Cavendish, and that his views were in consequence, to a great
extent, one-sided. No time was lost in calling in question his
accuracy. The French Academy had heard the one side argued,
the British Association was to hear the other.

In August 1839, soon after Watt's Eloge was published, the
British Association for the Advancement of Science met at Bir-
mingham, and the President for the year, the Rev. W. Vernon
Harcourt, in one of the most eloquentof the many able addresses
which have been delivered at its opening meetings, took the oppor-
tunity of vindicating Cavendish, and of pointing out the mistakes
which he regarded Arago as having committed. At a meeting of
the French Academy subsequent to this, Arago affirmed that Bar-
court's account was insufficient to establish Cavendish's claim as

CHARGES AGAINST CAVENDISH. 57

superior to that of Watt^ and brought forward the distinguished
French chemist Dumas as concurring in his opinion. When the
report of the British Association for 1 83 9 was published^ Harcourt
replied to these observations in a postscript to his address, which
contains a most able analysis of the documents bearingupon the
subject^ and a thorough discussion of the whole question. He was
at the trouble even to publish a lithograph fac-simile of Caven-
dish's Laboratory Note'Book, that no room might be left for com-
plaint of incompleteness in his statement* Since 1839 various
additional writers have taken part in the discussion, to whom I
do not make any reference here, as their writings are described
and criticised in the earlier sections of the portion of the body of
the work entitled The Water Controvert. This controversy, in
its original shape, was not carried on in public, so far as the
English rivals. Cavendish and Watt, were concerned, but the
publication of the Watt Correspondence has admitted us behind
the scenes, or rather has converted the complaints which Watt
made in private letters to his friends, into public impeachments
of Cavendish's capacity and fair-dealing. Lavoisier was from the
first accused by the English chemists, by Cavendish publicly, by
Watt privately, of having acted unfairly towards them. Watt,
however, was much more concerned to defend himself against
his English than his French rival, whilst Cavendish took no
formal notice of Watt's implied claim to priority in the disputed
discovery, and limited himself to repudiating the demands of
Lavoisier. No reply was made by any one of the three illus-
trious rivals, to the implied or asserted superior claims of his
opponents, a matter greatly to be regretted, for hopeless obscurity
now darkens some of the most important questions on which the
controversy turns, and, as might be expected, these are the very
points on which the partisans of the several claimants pronounce
most confidently.

As an incident in Cavendish's pergonal history, I have at
present only occasion to refer to the origin of the Controversy;
but as his character has been much more seriously assailed in
the publications that have appeared in consequence of its revival,
and as this has enabled us to appreciate the merits of the
original dispute much better than we should otherwise have
done« both aspects of the question will be considered, and to
some extent together.

58 UFB OF CAVENDISH.

Omitting all reference to the precursory expenments and
speculations which shadowed forth the diacorery of the com-
pound nature of water long before that was effected, I take up
the question in the spring of 1781. Some time before the 18tfa
April (for the exact date is unknown) of that year, Dr. Priestley,
availing himself of Volta's ingenious electric eudiometer, made
what he calls a ^' random experiment," with a view ^' to enter-
tain a few philosophical friends." It consisted in exploding a
mixture of common air or oxygen with hydrogen, in a shut glass
vessel, by sending an electric spark through it. When the spark
had passed, and the explosion was over, the sidesof the glass vessel
were found to be bedewed with moisture, but to this pheno-
menon Priestley paid no attention. One of the philosophical
friends who witnessed this experiment, was Mr. Warltire, a
lecturer on Natural Philosophy in Birmingham, whose name,
otherwise unknown in the history of science, emerges for a
moment from obscurity in connexion with Priestley's random
explosion, and immediately fades again into oblivion. With
great ingenuity he proposed to test by a trial similar to Priest-
ley's, whether *^ heat is heavy or not." To avoid the risk of
injury from the explosion, he employed a copper flask, which
he filled with a mixture of air and hydrogen,* and then
weighed the vessel and its contents. When an electric spark
was passed through it, and the mixture exploded, great heat was
evolved, and after the flask had cooled, it was weighed again to
ascertain if it had become lighter by the loss of the heat which
had been given off. In several trials of which Priestley'and
Withering were witnesses, the flask appeared on the second
weighing to have lost weight ; from which Warltire seems to
have concluded that heat is a ponderable body.f

* Priestley and Warltire both call the gas they used, ''inflammable air f' it
WM probably hydrogen, and certainly at least contained it, for it yielded water on
combofltion. The meaning of the term "inflammable air/' as employed by the
chemists of last oentnry, is one of the most important questions in dispute in the
Water Controversy, and will be found Ailly discussed ia one of the sections devoted
to that subject.

t These cesearches are detailed in the Appendix to Priestley's second toIum
of BapeHmmtt and Obiervatitm* on Air, 1781. The appendix is wanting m some
copies of this Tolume. Those which contain it have the pages marked by an
asterisk. The fint ii • 895.

PRODUCTION OP WATBR FROM ITS ELEMENTS. 58

WhUst tbese experiments were in course of trial by Priestley
and Warltirot Cavendish was engaged in the experiments re-
ferred to in the preceding chapter, as undertaken with a view to
ascertain what change air underwent, when bodies were made to
bum in confined portions of it till they were extinguished.
Among other combustibles whose e£Eect upon air he was trying
in this way to discover, was hydrogen, and when he heard of
Waritire's experiment, he repeated it.

He employed, however, like Priestley, a strong glass vessel,
instead of a copper flask, performing the experiment otherwise,
generally as Warltire had made it. The result as to loss of
weight he could not verify. He found occasionally a slight dif-
ference between the first and second weighings, but commonly
none at all, and he rejected, in consequence, the conclusion at
which Warltire seems to have arrived as to the ponderability of
heat. The deposit of dew, however, on the sides of the vessel,
which Priestley disregarded, and the cause of which Warltire
totally misapprehended, he looked upon as a phenomenon
'Mikely to throw great light*^ upon the subject he was pur-
suing^ '^ and well worth examining more closely.'^ The details
of the experiments he made in the course of this enquiry are
given elsewhere. It will be sufficient, therefore, to mention
here, that hydrogen and air were exploded in various propor-
tions, till that one was discovered (namely, 2 measures of
hydrogen and 5 measures of air) which secures the entire with-
drawal of the oxygen of the air and its conversion into water.
Pure hydrogen and oxygen were then taken in the proportion
warranted by the results obtained with air, t.e., one measure, or
nearly so, of the latter gas, to two measures of the former.
When this mixture was fired, no gas remained in the globe after
the explosion, but instead of the hydrogen and oxygen which
had lost their gaseous form, a certain weight of pure water was
found. And as the vessel and its contents had undergone no
change in weight, or parted with anything ponderable during
the explosion, whilst a certain volume of gas had been replaced
by a certain volume of water, the conclusion was unavoidable,
that the ponderable matter of the gas was in the liquid, and
therefore that it consisted, weight for weight, of the hydrogen
and oxygen, which had lost the elastic form in producing it.

60 HFE OF CAVENDISH,

Such^ accordingly) was the inference which Cavendisli drew^ not
certainly with all the precision with which we apprehend that
truth at the present day, but with as much clearness as any
predecessor of Lavoisier has done.

Cavendish) however^ did not immediately publish this or any
other conclusion y as warranted by his experiments. He was at
no time in haste to publish ; he was prosecuting other researches
in which he was interested^ and he contemplated an extensive
enquiry, which he afterwards carried out, into the nature of com-
bustion, the causes of the vitiation of the air, the properties of
the atmospheric gases, and certain other topics, before the com-
pletion of which he was not anxious to make his views public.
There was a special reason, moreover, for delay. The liquid
which resulted from the detonations was very carefully analysed,
and proved in all of the experiments with hydrogen and air, and
in some of those with hydrogen and oxygen, to be pure water ;
but in certain of the latter it contained a sensible quantity of
nitric acid. Till the source of this was ascertained, it would
have been premature to conclude that hydrogen and oxygen
could be burned into pure water. It would have been well for
Cavendish, however, if he bad published at once his results-
such as they were, or failing that, had preserved entire silence
regarding them, till his enquiry was completed. In either case
there would have been no Water Controversy. As it was, he
made them known himself to Priestley, and by his friend
Blagden to Lavoisier, and the effect was, that through the first
Watt came, in a way to be presently mentioned, to enter the
lists as his rival in England, and through Blagden, Lavoisier was
led to the observations on which he founded his claim to be
called the discoverer of the composition of water.

The researches of Cavendish which have been referred to,
were made in the summer of 1781, soon after the publication
of Warltire's Experiments on the ponderability of heat,* and
were communicated to Priestley before 26th March, l783,t and
to Lavoisier before 24th June of the same year.]: They were
not read to the Royal Society till January 1 5 th, 1784 ; and thus

* PMlotophical TratuaeHotu, 1784, p. 134.

t Watt Corretpondenee, p. 17.

% Mimoira dt VAeadimie det Seiencei pour 1781 (printed in 1784), p. 472.

PRIESTLEY'S REPETITION OF CAVENDISH'S EXPERIMENTS. 61

it happened, that Watt and Lavoisier, although their researches,
whether original or not, were later than Cavendish's in point of
time, nevertheless, in consequence of having announced their
views with a certain publicity in 17B3, appear with a primd
facie character of priority to him, as claimants of the disputed
discovery. Much of the complication of the Water Controversy
has resulted from the fact, implied in the preceding statement,
that whilst Cavendish can establish beyond question, priority of
observation, Priestley, Watt, and Lavoisier come before him as
having giyen more or less overt publication in a wriilen form to
views concerning the nature of water similar to his.* It will
conduce to perspicaity to separate the French from the English
claims, and in order of time the latter fall to be considered
first.

In his published paper (Experiments on Air) of 1734,
Cavendish states that the experiments on the production of
water from its elements, which he made in 17B1, were mentioned
by hinti to Priestley, '^who in consequence of it made some
experiments of the same kind as he relates in a paper printed in
the preceding volume of the [Philosophical] Transactions.^^
The volume referred to is that for 17B3, in which Priestley
with his wonted candour has acknowledged his obligation to
Cavendish, 1>efore circumstances compelled the latter publicly to
claim it.'t' The significance of these experiments, which Priestley
did not repeat without committing a serious blunder, of which,
however, he was not aware, was imperfectly appreciated by
Priestley, but he gave an account of them (without ap-
parently referring to Cavendish, as their originator) to his friend
Watt, who at once perceived their value, and wrote to Priestley
demonstrating what conclusion his experiments warranted.
Tliis letter, which was a commentary on all the researches of
Priestley, referred to in his paper of 1783,^ Watt designed to be
publicly read to the Royal Society along with the paper on
which it commented, and had this been done, the Water Con-

* LaToiAer'i comnranication to the French Academy in June 1783, was per-
haps an oral one, bat aa it is registered in the records of the Institute, I refer to
it as having been written.

t PMIowpMcal Tran»aeH<m$t 1783, p. 426.

X Ibid., 1783, p. 398.

62 Un Oy CAVENDISH.

troversy might never have arisen. At all events it would
have assumed a different aspect, and Lavoisier could not have
preferred any claim to be the discoverer of the composition of
water by synthesis. Fortunately, or unfortunately, however,
for all parties, Priestley discovered, after receiving Watfs letter,
that a series of researches which had seemed to both to demon-
strate the transmutability of water into atmospheric ur, was
totally fallacious, and as this had interested Watt quite as much
as the transmutability of inflammable air and oxygen into water,
he requested that the letter might not be read, and his request
was complied with.* The letter, however, was shown to various
members of the Royal Society, and remained under the nominal
custody of the president. Sir Joseph Banks, for about a year, when
Watt, roused by the news that Cavendish had laid his views before
the Royal Society (January 15th, 17B4), claimed to have his
letter read, and implicitly asserted the priority of his conclu-
sions, which were thus dated from April 26ihy 1783. This is all
that we learn from the documents published in the lifetime of
Watt and of Cavendish, and the question between them, as it
appears in these, is entirely one of relative priority, which might
be disposed of without much difficulty. But from the Watt
Correspondence, which was published in 1846t, we discover that
Watt was induced to believe that Cavendish had borrowed from
his letter, the views which he published as his own; and in
private Cavendish was deliberately accused of shameful plagiarism.
It is this charge which has so embittered the Water Controversy,
especially since its revival, and the circumstances which led to
its being preferred are now to be considered.

James Watt was not a man disposed by nature or circum-
stances rashly to accuse a brother philosopher of unfidr dealing.
** In his temper and dispositions,^' as L<Mpd Jeffirey, who knew
him well, tells us, '^ he was not only kind and affectionate, but
generous and considerate of the feelings of all around him j . . .
and such,** he adds, in another part of his eloquent eulogium,
"was the influence of his mild, character and perfect fairness
and liberality, even upon the pretenders to these accomplish-

* Philoaophieal TraMaciiwu, 1784, p. 330.

t Correspondence of (he late Jamee WaU on hie dise<wery qf the Thewy of
the Composition of Water, ^c. Edited by J. P. Muirhead, Esq., P.II.S.E.

RIVALl^Y BETWUN CAVBNDISH AND WATT. 63

mcnto^ that he lired to duarm even envy itself, and died, we
▼erily believe, without a single enemy /^ *

Hie philosopher who has been aoeused of borrowing from
Watt without acknowledgment, although a much more reserved
and less demonstrative person than his rival, bore as high a cha-
racter among the majority of his contemporaries and successors.
Without prejudging the question, accordingly, which is now
to be considered, and seeking only to warn the reader of
the twofold perplexity of the problem before us, I may
remind him of the high opinion of Cavendish, expressed
by Cuvier in the passage already quoted from the Ehge.'f
Sir Humphry Davy also, who was aware of the feelings
entertained by Watt towards Cavendish, and cherished the most
friendly regard for the fc»rmer, oflSnred this unsolicited tribute to
the character of his rival, '^ unambitious, unassuming, it was
with difficulty that he was persuaded to bring forward his
important discoveries. He disliked notoriety, and he was, as it
were, fearful of the voice of tame.*'t

At the very threshold of the Water Controversy we thus
encounter a perplexing dilemma. Two unusually modest and
unambitious men, universally respected for their integrity, famous
for their discoveries and inventions, and possessed of rare intel-
lectual gifts and accomplishments, are suddenly found standing
in a hostile position towards each other, and although declining
to publish their own unquestioned achievements, are seen conten-
ding for a single discovery, which the one believes the other to
have learned at second-hand from revelations made to a common
Mend, and which that other accuses bis rival of having gathered
from a letter that he was allowed to peruse. A misunder-
standing such as this woidd never have occurred had Watt and
Cavendish been intimate in 1783. As yet, however, the friendly
intercourse which afterwards subsisted between them had not
commenced. The one was resident in London, the other in
Birmingham, and each was informed of the other's doings by

* The eulogium will he found appended to the EnglUh tranilations of Ango's
Eloge of Watt, the eiact titlea of which are given in the section of the Water Con*
if^Mi^ entilled « BibUography." It w aleo contained in tthe Encyclop^ia BriU
toMCM, article ITa/^, and in the ooUeded BaMya of Lerd Jelfrtj.

t Siogm Hist, torn, tt., p> 104.

X CMIeeted Wwrks, vol. vii., p. 128;

64 LIFE OF CAVENDISH.

third parties^ upon whom mainly, though unequally, rests the
blame of having occasioned the Water Controversy ; nor does it
materially lessen our regret to find that those who made mischief
between the great philosophers, did so with the best intentions.
The parties in question were Dr. Priestley, J. A. De Luc, and Dr.
afterwards Sir Charles Blagden, all men eminent in science and
of unblemished character. Through the first, a knowledge of
Cavendish's experiments passed to Watt, and a knowledge of
Watt's conclusions to Cavendish; by the second, Watt was
informed that Cavendish had deliberately pilfered his theory ;
and the third, who was Cavendish's assistant, reported the
tatter's conclusions, as well as those of Watt, to Lavoisier, whom
he accused in the name of both th€English philosophers, of having
appropriated their ideas. Blagden, also, made certain alterations
in the MS. of Cavendish's first Ea^periinents on Air, and whilst
superintending, in his capacity of Secretary of the Royal Society,
the printing of that paper, and of Watt's rival essay, suffered
certain typographical errors to occur, which involved himself
and his principal in accusations of unfairness. The critics of
the Water Controversy w^ho have advocated the claims of Watt,
have been unsparing in their denunciations of Blagden, and in
their praises of De Luc, whilst Priestley has been alternately
praised and blamed, his evidence being eagerly claimed by their
respective advocates, when it went to favour Watt or Cavendish,
and as summarily refused when it militated against the views
they sought to defend. These hval claims solely affect Priest-
ley's intellectual reputation, which has sufiered, and must suffer,
by his share in the discovery of the composition of water. His
moral character is unsullied, perhaps even exalted, by the part
which he took in this curious controversy. No one of Caven-
dish's advocates, so far as I am aware, has come to the defence
of Blagden, although it is unquestionable, that if he be convicted
of the chaiges brought against him, Cavendish's honour must
suffer also; nor has any admirer of Cavendish disputed the justice
of the eulogium which the friends of Watt have one and all
passed upon De Luc. I hope, however, to be able to show^
that, with the exception of a little excusable carelessness in
correcting printers' proofs, Blagden was guiltless of any wrong
towards Watt or unfairness towards Lavoisier, and that to

PROCEEDINGS OF LAVOISIER. 65

De Luc belongs the unenviable distinction of having been the
deliberate mischief-maker who provoked the Water Controversy.
In what way Priestley, De Luc, and Blagden, came to be
involved in the dispute between Cavendish, Watt, and Lavoi-
sier, will appear from the following account*

The experiments which Cavendish made on the formation of
water, in 1781, he communicated, as we have already seen, to
Priestley, not later than the spring of 17B3, and to Lavoisier in
the summer of the same year, Priestley announced the result of
his repetition of them to the Royal Society, and Lavoisier the
result of his similar trials to the French Academy in June 1783,
whilst Watf s letter had been in Priestley^s hands, but withheld
at its author's request from public reading, since April 26th of
that year. All parties went on with their researches, and towards
the end of the succeeding November, Watt took courage to
have his views published, and proceeded to throw them in an
amended and fuller form, into the shape of a letter to his friend
De Luc, which bore date 26th November, 1783, and was
intended for public reading at a meeting of the Royal Society.
That letter, however, was scarcely despatched, before news
came from Paris, that Lavoisier had stolen a march upon his
cautious and too tardy English rivals, and we find Watt writing
to Kirwan, on the 1st December, announcing that ^^ M. Lavoi*
sier has read a memoir, opening a theory very similar to mine,
on the composition of water ; indeed, so similar, that I cannot
help suspecting he has heard of the theory I ventured to form
on that subject^ as I know that some notice of it was sent to
France.^'*

This suspicion was confirmed by Kirwan, who informed
Watt, on the authority of Blagden, that he had explained to
Lavoisier the views of both the English observers concern-
ing the composition of water.t Watt felt very indignant
on learning this, and all the more, that De Luc sought to
defend Lavoisier from the charge of plagiarism, only, however,
with the efiect of increasing Watt's indignation at him.:t
Meanwhile, the letter from Watt to De Luc was not pre-
sented to the Royal Society, and Cavendish, who had completed
one great section of his protracted Experiments on -4ir, com-

♦ Watt Corr., p. 37. f Qp- Cit,, p. 39. % Op, Cit., pp. 40-42.

66 LIFE OF CAVENDISH.

monicated his results to that body on January ISth^ 1784. In
this paper he formally and publicly announced^ besides much
else, the views which he had in the preceding year made known
to Priestley and Lavoisier. De Luc, who was in Paris at the
time, heard some account of them towards the end of February
after his return to England, and requested to be allowed to read
Cavendish's manuscript, which was at once granted^ and, on the
1st March, he wrote to Watt, giving him a sketch of its contents.
The spirit in which he did this, cannot be sufficiently regretted,
for it led him to prepossess Watt with the darkest suspicions
against Cavendish, and this at a time when the indignation of
the former at the proceedings of Lavoisier, had rendered him
peculiarly sensitive to his claims as a discoverer, and jealous of
any interference with them. Had De Luc shown but a small
portion of the charity towards Cavendish, in estimating the ori-
ginality and independence of his views, which he so vuinly and
extravagantly extended to Lavoisier, no misunderstanding would
have occurred between Watt and Cavendish. De Luc, how-
ever, plays so important a part in the Water Controversy, that
it is necessary the reader should have some acquaintance with
his character. It will bear the closest inspection.

Jean Andr^ De Luc M^as born at Geneva in 1727> where, up
to his forty-sixth year, he divided his time between commercial
tasks, scientific studies, and a prominent share in the religious
and political disputes which agitated his native republic in the
latter years of the past century. He was in high esteem among
his fellow-citizens, but a reverse of fortune, and an increasing
wish for scientific occupation, induced him some time after 1770
to abandon Geneva and repair to England, where he became
reader to Queen Charlotte, George I IPs consort. This office,
however, did not necessitate constant residence at the British
Court, so that he was able to gratify his interest in social as well
as scientific progress, by lengthened journeys oh the Continent.
In 1798 he was appointed Professor of Geology at Goettingen,
and he spent several years thereafter in Germany. After the
battle of Jena he returned to England and resumed his duties as
reader to the Clueen. He resided chiefly at Windsor, where he
was highly esteemed by ail the members of the royal circle. He
died on November 7th, 18l7t in his 91st year.

CHARACTER OF DE LUC. 67

De Luc was an honourable, earnest, and accomplished person ;
not a man of genius, but certainly one of great talent. He pub-
lished works on yarious departments of science, and especially
several treatises on geology, which excited much interest at the
period of their publication, but are now forgotten. As a me-
teorologist, however, he will long be remembered. We are
indebted to him for the first accurate hygrometer, and for im-
portant improvements on the thermometer and barometer. On
the two latter instruments, indeed, he is still an acknowledged
and esteemed authority ; and from the records of his life that
remain, it would appear that he was an upright, intelligent
observer, a most indefatigable worker, and a sincerely religious
man. The space at my disposal does not allow me to refer to
De Luc at greater length, but I wish to guard against seeming
to depreciate either his intellectual capacity or his moral worth.
On the other hand I would pointedly refer to his possession of
these, as investing his interference in the Water Controversy
with an authority which it would not have possessed, had he
been a less accomplished and worthy person.*

The 1st of March, 1784, marks the commencement of the
Water Controversy, so* far as the English rivals are concerned.
On that day De Luc wrote to Watt, informing him that
Cavendish expounded and proved his system word for word,
without saying anything regarding him.f On the fourth of the
month X De Luc writes again, zealously defending Lavoisier
and La Place from the charge of plagiarism, but urging it
against Cavendish, who is declared to have used the very
words (vo8 propres termes) which Watt employed in his
letter to Priestley, of the contents of which neither he nor
his assistant could be ignorant. De Luc is charitable enough
to suppose it possible that Cavendish was an unconscious
pilferer, and acknowledges that the readiness with which he
and Blagden complied with his request to see the MS. of the
Experiments on Air, although aware of his intimacy with Watt,

* For the particalara of De Luc's Life 1 am indebted to the BiograpMe

UmoerttUe and the Penny Cyeiopadiot the last of which contains an excellent sketch

of the philosopher, with a list of his writings.

f ** On expose et prouve votre systeme, mot pour mot, et on ne dit rien de
vous,"—Waa Corr., p. 43.

t Ibid., pp. 44-47.

P 2

68 LIFE OF CAVENDISH.

as well as the unblemished characters of those he suspected,
were at variance with the notion that Cavendish was a deliberate
plagiarist. That he was one however, whether aware of it or
not, was a point which from the first De Luc considered as
beyond dispute, and explained on the more charitable hypothesis
of Cavendish'^s proceedings, by the suggestion that be had
unconsciously derived the idea of the compound nature of water
from Watt's letter to Priestley, which had excited little attention
when read by Cavendish, but had nevertheless planted a germ in
his mind, which expanded after many months into the theory
which he ultimately published. At the same time De Luc, as if
desirous to provoke Watt to the highest pitch of indignation,
counselled him, if anxious to increase his wealth, to avoid exciting
the ill-will of others ; in other words, not to enter the lists
against Cavendish as the claimant of a discovery, if he wished
to sell his steam-engines. A more certain method of inducing a
highly honourable, courageous, and unsordid man like Watt, to
defend his claim to the last, could not be conceived. The most
wilful mischief-maker, in truth, could hardly have devised a better
means of provoking jealousy towards Cavendish than to give
Watt such advice as De Luc did. When Watt accordingly replied
to De Luc he resented this advice in strong terms, and referred
«cornfully to '•' the illustrious house of Cavendish,** as one he
could despise so far as his pecuniary fortunes were concerned.*
He declines in the same letter to make an illiberal attack on
Cavendish ; nevertheless, he counts it barely possible that his
rival had heard nothing of his theory, and in apparent forgetful-
ness of his concession of at least a bare possibility the other way,
lie speaks of '* the plagiarism of Mr. C.*' ; so eflFectually had De
Luc infected him with his own ungenerous and unjust suspicions.
That the prejudgment of Cavendish's fair-dealing, of which
Watt and De Luc were guilty, was unjust, however excusable on
the part of one goaded into indignation as the former was, does
not admit of a moment's denial. As yet, neither of Cavendish's
summary condemners was entitled to become so much as an
accuser. The only proofs which De Luc professed to give of
plagiarism, were, that Cavendish used the same words in an-
nouncing his views that Watt did ; and that he must have read
Watt's letter to Priestley. Granting, however, for the sake of

♦ Watt Corr., p. 48.

SUSPICIONS OF DE LUC. 69

argument^ both those points^ it did not follow that Cavendish
borrowed from Watt; for two chemists of the phlogiston school,
who arrived, independently, at the same conclusion, could not
but use similar terms in stating it, since the nomenclature of
that school was an extremely restricted one, admitting of very
little variation in the terms which its disciples employed. • Watt
in truth wrote to De Luc ^* On the slight glance I have been able
to give your extract of the paper, I think his [Cavendish's]
theory very different from mine ;''* so that the supposed suspi-
cious identity of language, on which De Luc dwelt, does not
seem to have been recognised by Watt. And even if Cavendish's
words had been absolutely identical with those of his rival, which
they are not, as will afterwards appear, the identity could only
have justified the apprehension that they might have been
borrowed, not certainly the summary conclusion, that they must
have been. So also, before the perusal of Watfs letter to
Priestley of April, 1783, was set down as the certain source of
Cavendish's conclusions, published in 1784, it should have been
ascertained whether he had held any such views before he read
the letter, a thing most i)robable when it is remembered that he
had been investigating the subject since 1781. Yet neither Watt
nor De Luc made any enquiries of this kind, but at once decided
that it was in the highest degree unlikely that Cavendish had
acquired his views concerning the composition of water in any
other way than by pillage; and the slender charity which
acknowledged it as barely possible that it might be otherwise
speedily gave way.

De Luc thus did Cavendish and also Watt a great wrong, by
hastily deciding the case against the former, and filling his rival's
mind with suspicions against him. He was not to blame for
zealously espousing Watt's cause. He had been made a party
to it, to some extent, by the letter which Watt addressed to him
for public reading in November, and he was under obligations to
his friend at Birmingham, for assistance towards procuring mate-
rials for a projected work on heat and elastic fluids, circumstances
which would materially increase his desire to serve h5m.t But
I do blame him unhesitatingly and severely, for impeaching
Cavendish as he did, when a little reflection must have shown

• Wait Corr.t p. 48. t Ibid., p. 5.

70 . LIFE OF CAVENDISH.

him that he was in many respects disqualified from being umpire
between the English rivals.

De Luc was not resident in London^ but as reader to the
Queen he followed the motions of the Courts and spent much of
the year at Windsor,* so that he was entirely out of the way of
hearing what researches Cavendish might be prosecuting. We
have it on his own authority also, that he rarely attended the
meetings of the Royal Society, so that he placed another bar in
the way of knowing what Watt's rival had been doing.f His
acquaintance with English, moreover, though creditable for a
foreigner, was limited. Miss Burney (afterwards Madame
D'Arblay), who was attached to the Court at the same time as
De Luc, refers to him in her diary, under December 1st, 1785,
in the following terms : '^ Upon Mrs. Delany's coming to
Windsor, the Gtueen had Cecilia read to her again, and by
M. De Luc, who can hardly speak four words of English !"{
Some allowance must no doubt be made for the wounded vanity
of the authoress of Cecilia, who would set down De Luc's broken
English as a personal wrong; but after we strip the statement of
the exaggeration that plainly pervades it, we cannot credit De
Luc with a great command of our language a year before
he provoked Miss Burney by the style in which he read her
Cecilia to the Queen. He had thus another obstacle in the way
of learning what researches were engaging the attention of the
English philosophers. Moreover, he was not a chemist. When
Watt sent him the letter expounding his views on water, DeLuc
excused himself from an immediate reply, ^^Le language chimique
ne m'etant pas bien familier,"§ a fact which he totally forgot
when not long after he summarily pronounced that Cavendish's
words were suspiciously identical with those of Watt, of which
one so ignorant of chemical nomenclature was by his own show-
ing a very imperfect judge.

Lastly, De Luc went to Paris in December 1783, and passed
there the month of January and a portion of February, || so
that he was absent from England when Cavendish's paper, Eape^

* Watt Corr.t p. 249. Note to Lord Brougham's Historical Note, by Mr.
James Watt (jon.).

+ Watt Corr,, p. 43.

X Diary and Letters of Madame D*Arhlay, vol. ii,'p. 364.

$ Watt Corr. p. 3. || Jbid,, p. Iviii.

DE LUC THB MISCHIEF-MAKER. 71

riments on Air, was read^ and thus had not the advantage of
hearing the comments upon it which other philosophers made.

When all these disqualifications are considered, it will appear
that there was scarcely one among Cavendish's scientific con*
temporaries less entitled to judge him than De Luc, and that he
most culpably hastened to condemn him, before he was justified
in so much as accusing him. The haste, in truth, was extreme.
On the very day on which De Luc received Cavendish's MS., he
wrote to Watt impeaching his rival ; and commenced an analysis
of his paper, which he completed and despatched in four days,*
without, so far as appears, making enquiries at any of the mem-
bers of the Royal Society, or others, as to the proceedings of
Cavendish. De Luc has justly earned the title of the mischief-
maker, par excellence^ in the Water Controversy. Nor can
Watt be acquitted of the charge of having judged Cavendish
hastily and uncharitably. He would not publicly accuse him^
but he did not spare him in private, and he acted in public on
the principle that his rival had wronged him. This appears
from the JVatt Correspondence, which does not exalt our esti-
mate of the generosity of feeling of the great engineer, although
much allowance must be made for the sinister influence of
De Luc's well-intentioned but unhappy interference. The
immediate effect of this was to induce Watt, as an injured man,
to seek redress. Having occasion, accordingly, to visit London,
he had an interview with Sir Joseph Banks, the President of
the Royal Society, and it was arranged that both the letter to
Priestley, of 26th April, 1783, and that to De Luc of 26th
November, 1783, should be successively read. The former,
accordingly, was read on the 22nd, and the latter on the 29th
April, 1784.t

From Sir Joseph Banks and the other officials of the Royal
Society, Watt received every courtesy, but his jealousy of
Cavendish was not to be appeased, the more so that De Luc
fomented it. What transpired at the interview with the Presi-
dent does not appear, but it ended in his addressing a note to
Watt, asking him to have his Letters on Air read to the Royal
Society.! Watt, who regarded the president's request as made

* Watt Corr., pp. 43-47. t iAW., p. Ixu. $ Ibid,, p. 49.

72 LIFE OF CAVENDISH.

in a *^very civil manner,^' desired De Luc to call on him, and
settle about the reading, which accordingly he did ; not, how-
ever, without seeking to disabuse Watt's mind of the notion
that Sir Joseph Banks had any peculiar desire to have the
letters read, since he had informed De Luc that they should
certainly be read if Watt wished it, but not otherwise. *

It may reasonably be suspected that De Luc was not the
most suitable person to arrange matters with Sir Joseph. At
all events it was a most ungracious act to complain as De Luc
did, as if it were alike the duty of the president of a society
to go about beseeching aggrieved authors to trouble its peace by
claims of priority, and to authorize them to say that they did
so at his request.

Watt had too much good sense to expect such entreaties
from Sir Joseph Banks, or to refuse to request the reading of a
letter which had remained unread only in compliance with its
writer's explicit desire, and in the end he declared that ** Sir
Joseph Banks has behaved with great civility and kindness in
the affair of the letters.'' t The tone, however, of his imme*
diate reply to De Luc, is that of an injured man, distrustful of
the whole Royal Society, and regarding his rival with undi'-
minished suspicion. ^^In relation to Sir Joseph Banks," be
writes his friend, ^^ he wants the paper to be read, not, as you
observe, because he is attached to me, but because he feels as a
slight put upon the Society, the withdrawing it ; and perhaps
thinks his own honour a little called in question, which I do not
wish him to think, as he has always behaved in a friendly man-
ner towards us." J That Watt held some one's honour to be
compromised by the treatment of his letter, appears probable
from the preceding reference ; and a little further down he
betrays more strongly the feelings of jealousy which he obsti-
nately entertained towards his rival. ** After the reading of this
paper of Mr. Cavendish's, and being civilly requested to publish
in the same channel, I think it would savour a little of resent^
ment and coioardice to decline it any farther." § And in the
close of the letter, his general distrust of the Royal Society
breaks out again ; — *^ I shall certainly send the letter to yourself

• JVaft Corr,, p. 50. t Tbid., p. 60.

X Wait Corr,y p. 51. § Ibid,, p. 51.

PAIR-DEALING OF THE ROYAL SOCIETY. 73

through your own hands^ and I assure you I should have been
much better pleased that you had been the president and
members of the society who should publish it; but circum-
stances compel me to give it to the other^ and I hope it will
answer your end as well^ after they have had their will of it/^ *
It is painful to read these passages. In April 1783^ Watt sent a
letter to the Royal Society, which its fellows were most willing to
hear read^ but its author changed his mind and withdrew it. As
soon as he indicated a wish^ in the succeeding springs to have
this letter publicly read^ he was met more than half-way by Sir
Joseph Banks^ the friend of Cavendish^ and encouraged to pre-
sent it to the Royal Society. Though Cavendish, moreover^
had been guilty of all he was accused of^ the Society had not
implicated itself in his guilty real or imagined^ by receiving his
paper^ which it had not the shadow of a plea for refusing ; and
it left it open to Watt, as it did to every one else, to reclaim
against the demands of any of its members. More it could not
possibly have done^ for even if it had been satisfied that Caven-
dish had wronged Watt, and robbed him of the credit of his
discovery, he had put it out of the society's power to assist
him in vindicating his rights, by declining, in 1783, to have
his letter published. It was very unreasonable, therefore, of
Watt, to suspect the Royal Society as he did, but his pre-
judiced feeling towards it did not abate. On the same day on
which he wrote toDe Luc (12th April, 1784), he wrote to Sir
Joseph Banks^ courteously thanking him for his kindness ;
nevertheless, with a pettishness and an exaggerated modesty,
unworthy of so truly great and modest a man as Watt certainly
was^ he requests that the Royal Society '^ will also excuse the
defects of my style, which must naturally be concluded to savour
more of the mechanic than of the philosopher.'^ t

Sir Joseph Banks did his best (April 15th) to appease Watt's
irritation by thanking him for communicating his letters to the
Society, and spared him the necessity of requesting their publi-
cation, by informing him that he wished them to appear in the
next volume of the Philosophical Transactions, although, in
strict rule, a committee of the Council, and not the President,

• WaU Corr.f p. 51. t Ibid., p. 53.

74 LIFE OF CAVBNDISH.

were the judges of tiliu. Two dajrs later (April l7th,) Watt,
writes to De Luc, mentioning Sir Joseph's good offices, and
among other things, says : '^ Do as you think proper ; I am
sure you have my reputation in the matter more at heart
than I have myself/'* A fact too true, and not to be forgotten,
as showing how important an agent De Luc was, according to
Watt's own testimony, in determining the temper in which he
regarded Cavendish.

Of the same date is a letter from Watt to Sir Joseph Banks,
in which he refers to his letter to De Luc of November
26th, 1783, and points out certain alterations made in it,
'^ lest it should be said by anybody that the letter was fabri-
cated at a later date than it bears. If anything of that kind
should be started, M. De Luc can produce the original in my own
hand writing which can be compared with this present copy."t
From this allusion it appears that Watt was anxious to esta*
blish his claims as dating from November 1783, and that he
suspected certain parties of having an interest in denying
this. That one of these was Lavoisier, is certain, from an
allusion in the letter, and we cannot doubt that another was
Cavendish.

Four other letters passed between London and Birmingham,
referring merely to alterations in style and other little matters,
from 23rd April to 5th May inclusive, in which I do not find
anything calling for notice.^ On the llth, Sir Joseph Banks
writes Watt, informing him that both his letters ^' appeared to
meet with great approbation from large meetings of Fellows ;"§
and on the 12th De Luc writes, delighted with the steps Watt
had taken to authenticate his letters and dates, and adds, " le
Chevalier Banks s'y est prSte volontiers/'j] Watt acknowledges
in reply (May 14th), his obligations both to Sir Joseph and to
De Luc ;T and surely he might now be content. He had perilled
his claims to priority over Cavendish (considered for the present
as dating only from January 1784), by the voluntary and unso-
licited withdrawal of his letter to Priestley of April 1783, and
the delay which he permitted in making public his letter to De
Luc of November 1783. These letters, hitherto private, were

* Wait Corr,,^, 55. f I^id., p. 56. % Uid., pp. 57.59.

§ Ibid., p. 59. II Jbid., p. 60. If Ibid,, p. 60.

UNREASOJIABLE SUSPICIONS OF WATT. 75

SOW accepted by the Society as public documents, and whatever
they entitled their writer to claim, he could now claim from the
period at which they were written. The satisfaction, however,
with which Watt learned this, was not very abiding, nor did it
in any degree lessen his indignation at his rivals. On the 15th
May, he writes to his friend Mr. Fry, of Bristol, ^^ I have had the
honour, like other great men, to have had my ideas pirated..
Soon after I wrote my first paper on the subject. Dr. Blagden
explained my theory to M. Lavoisier at Paris, and soon after
that, M. Lavoisier invented it himself, and read a paper on the
subject to the Royal Academy of Sciences. Since that, Mr.
Cavendish has read a paper to the Royal Society on the same
idea, without making the least mention of me. The one is a
French financier ; and the other a member of the illustrious
house of Cavendish, worth above 100,000/., and does not spend
1000/. per year. Rich men may do mean actions. May you
and I always persevere in our integrity, and despise such
doings.'^* I make this quotation with great pain. It is mourn-
ful to think that such a man as Watt should have written thus.
The repetition of the sneer at the ^^ illustrious house of Caven-
dish,'' and the singular endeavour to connect the wealth of his
rivals with the wrong they had done one who was in the way
himself to become a wealthy man, show a bitterness of feeling,
and an unreasonableness, greatly at variance with the prevailing
temper of a naturally generous man. The blame, however, be-
longs more to those who published a private letter containing
such unhappy passages, than to him, who, in the confidence of
friendship, gave utterance to his unsparing indignation.

From the preceding account it will appear that Watt and
De Luc made no enquiry into the proceedings of Cavendish,
but proceeded from the first on the assumption that he had
committed theft, and that the final conclusion to which they
came, after taking the steps which they deemed requisite to
establish Watt's claims, rested on exactly the same suspicions
as their earlier inference. It will further be observed, that no
charge was brought against Cavendish, or complaint of any kind
made by Watt in either of the letters which had been read to the
Royal Society. He writes to Blagden (May 27, 1 784), " My only

* Watt Corr,, p. 61.

76 L1F£ OF CAYENDISIL

reason for wishing my letter to Dr. Priestley to be read before the
Royal Society was^ to shew them what my ideas on the subject
were, at the time it was written."* Nor does anything at variance
with this appear in the later letter to De Luc. It is of importance
that this should be noticed, for many critics of the Water
Controversy write as if Watt had publicly accused Cavendish
of plagiarism, implying of necessity the priority of Watt, and
accordingly they draw unfavourable conclusions from the absence
of a reply. But Watt made no charge of plagiarism, and did
not even claim priority. He simply asserted that his views
concerning water went back to 17B3^ and he left it to his rivals
to show that theirs were of earlier date. Had a conference
occurred between Watt and Cavendish^ the former could not
have failed to learn how long the latter had preceded him in
experimenting on the production of water from its elements,
and would not haye refused to believe that his rival had enter-
tained views concerning its true nature, before him. He after-
wards, indeed, acknowledged that Cavendish had preceded him
to some extent,t and he would have done it at the time, and
more fully, had he been better aware of his rival's proceedings,
^luch, however, as it must be regretted that Watt made no
enquiry into Cavendish's researches, in justice to the former, it
should be acknowledged, that it is a delicate matter to contest
priority, especially where no formal charge has been preferred ;
andthatsofar as his published papers represent him, Watt appears
to advantage as standing only on the defensive. The Corres-
pondence, however, which the friends of Watt have made a public
document, shows him claiming priority, accusing Cavendish of
plagiarism, and of unfairly making no reference to him ; and
when these charges are considered, we cannot acquit Watt of
blame in never seeking a conference, directly, or through a
common friend, with Cavendish, before he reproached him.
A conference might have been declined, but at least it should
have been invited ; and some positive evidence of unfair beha-
viour tow^ards him should have been in Wattes hands, before he
circulated, even in private, reports to the discredit of Cavendish.
We are now to see what steps Cavendish took to right
himself and to defend his good name. They were, as might be

♦ Wcii Corr., p. 63. f ^^«'- Trant, 1784, p. 332.

cavbndish's defence. 77

expected from his character, yery few. No overt charge had
been preferred against him, nor is there the least reason to
imagine that the contents of Watt's letter to his friends would
be divulged to him. The reclamation, however, of priority by
Watt, which followed so closely on De Luc's request to be
allowed to read Cavendish's Eivperimenis on Air (MS.), and in
which De Luc took so active a part, made sufficiently manifest
what Watt's feelings were. The inaccurate account, moreover,
of Cavendishes researches, even in the amended version of the
letter to De Luc, showed that its writer questioned, if he did
not deny, that any one had theorised before himself on the com-
position of water. It is not unlikely, also, that Sir Joseph
Banks called Cavendish's attention to the interviews he had had
with Watt and De Luc, and apprised him of the claims set up
for the former. At all events the public reading of Watt's
letters, with their inaccurate reference to the researches of
1781, showed plainly enough what demands their writer made ;
and Cavendish took advantage of the opportunity which elapsed
between the public reading (January 1784) and the printing
(July 1784) of his paper, to make three additions to it, which
introduced new elements of strife into the controversy, and
brought newnames conspicuously forward as concerned in it.
Two of these additions occur in the body of the paper, in the
handwriting of Blagden, who was Cavendish's assistant; the
other is a postscript written by Cavendish himself, in whose
writing the text of the paper is also.* In the first of these
insertions, or, as the friends of Watt love to call them, ^^ inter-
polations," Cavendish states that *^ All the foregoing experi-
ments on the explosion of inflammable air, with common and
dephlogisticated airs, except those which relate to the cause of
tlie acid found in the water, were made in the summer of the
year 1781, and were mentioned by me to Dr. Priestley, who, in
consequence of it, made some experiments of the same kind, as
he relates in a paper printed in the preceding volume of the
Transactions [I783.]"t The remainder of this insertion reports
a similar but more extended account of Cavendish's researches,

* The MS. remains in the archires of the Royal Society. A full account of it
trill be found in the sequel.

t Phil. 2Vfln#. 1784, p. 134.

78 UFB OP CAYBNOISH.

as given by Bb^den to Lavoisier in 1783i^ &nd will be referred
to again.

The paper by Priestley in the Philasophieal TVansactions of
1783, to which Cavendish refers, is the one on which Watt's first
letter was a commentary. It is entitled Experiments relating to
Phlogiston^ and the seeming Conversion of Water into AirJ^ The
second part (p. 4 14) is entirely devoted to the latter subject, and
recounts an extensive series of experiments on the transmutability
of water into atmospheric air, in the possibility and reality of
which, Priestley as well as Watt had, for a season, entire faith.
The former, however, even at the period when he was most satis-
fied with his experiments, which he after%vards found to be quite
delusive, found many persons sceptical as to their significance in
demonstrating the alleged transmutation. With a view, accor-
dingly, to strengthen his doctrine, that water might become air,
by showing that air might conversely become water, he " gave
particular attention to an experiment of Mr. Cavendishes, con-
cerning the reconversion of air into water, decomposing it in
conjunction with inflammable air'^ (p. 426). He then describes
the experiment he made, in which he employed oxygen and
inflammable air (from charcoal), and states, at the close of
his description, ^^ the result was such as to afibrd a strong
presumption that the air was re-converted into water, and there-
fore, that the origin of it had been water^^ (p. 427)* In the
passages quoted, Priestley, it will be observed, uses the word
airy in a vague sense, in conformity with the fashion, or rather
ignorance, of the time, for as yet the specific characters of the
gases were imperfectly known ; and the prevailing idea was,
that they were only modifications of atmospheric air. Moreover^
Priestley was mainly anxious to employ Cavendishes experiments
to show that gas might become liquid, and, therefore, that liquid
might become gas ; whilst he fell back on his own researches to
prove that the particular liquid water, might become the par-
ticular gas, atmospheric air. This, however, and certain other
peculiarities of Priestley's statement, do not concern us at
present. It is of importance, as being the record of the very expe-
riments on which Watt founded his views concerning the nature

• Phil Tram. 1783, p. 398.

cavendish's REVBLATIOKS to PRIESTLEY. 79

of water, so that we may learn from it what Watt stated in
neither of his letters (and seems unaccountably enough^ not to
have known) ^ namely^ that he was indebted to an avowed
repetition of Cavendish's eicperiments^ for the grounds of those
conclusions which he so summarily decided his rival could only
have borrowed from him.*

How much Cavendish told Priestley, does not certainly
appear from any existing document ; and the abettors of Watt's
claims, accordingly, represent Priestley as having been informed
solely of experiments of Cavendish's, without having been
enlightened as to what his conclusions were. This view,
however, is quite untenable, for whatever difficulties may attend
the determination of the exact nature and amount of the infor*
mation Cavendish gave Priestley, the passage just quoted is
sufficient to prove that conclusions as well as experiments were
communicated by the former. For the full proof of this, the
reader is referred to the detailed analysis of the question given
in the sequel. It will be sufficient here to say, that Cavendish
did not make random trials ; neither did he, as some have
asserted, interest himself only, or chiefly, in ascertaining
whether Warltire was right in thinking that by the detonation
of hydrogen with air or oxygen, it could be shown that heat was
ponderable. What interested him, as he tells us himself, was
the deposition of liquid, which succeeded each explosion, and
which promised to throw light on the important problem, what
becomes of the air which disappears or loses its elasticity during
combustion. He prosecuted the enquiry till he ascertained how
much of atmospheric air, namely, one-fifth part of it by measure,
and what portion of it,, namely, its oxygen, disappeared during its
maintenance of combustion ; and till he discovered that if hydro-
gen were the combustible burned in it, or exploded with it, a
measure of that gas equal to about twice the bulk of the oxygen,
lost its elasticity simultaneously ; whilst in place of both gases-
their conjoint weight of liquid was found. He verified this result

* The friends of Watt aisidaottsly, but, as I believe, qoite Tainly, deny that
Watt's oonclnsions had reference to the particular experiments recorded by Priest-
ley in the passage qnoted in the text ; but, with the exception of Arago, they
seknowledge that Watt founded his theory upon an avowed repetition of Caveudish's
axperiments. The subject wiU be found fully discusied in the scqueL

80 LIFE OF CAVENDISH.

l)y substituting for atmospheric air, oxygen, which he mixed with
twice its volume of hydrogen ; and showed that (with the
exception of a trace of impurity) a globeful of this mixture
could be completely deprived of elasticity ; and that the weight
of gas which disappeared was replaced by an exactly equal
weight of liquid. Finally, he ascertained that this liquid was
pure water. Thus much, Cavendish affirms, he told Priestley.
Thus much, in effect, Priestley acknowledges that Cavendish told
him, and Priestley's acknowledgment was made before Caven-
dish's reference to it, and details exactly such experiments
as have been recorded, only the repetition was exceedingly
imperfect and inaccurate compared with the original trials. The
more impartial, at least, of Watt's advocates, substantially
acknowledge all that has been stated above ; and the question
so keenly contested between them and Cavendishes friends, is,
did Cavendish's revelation to Priestley include only such details
as those recorded, or did it also embrace conclusions ? I think
there are few impartial readers who will feel much difficulty in
acknowledging that conclusions were communicated. In truth.
Watt's friends have one and all carefully avoided giving their
definition of an " experiment." They would, apparently, have
us believe, that it is only the handling of certain pieces of
apparatus, and the unreflecting observation of certain pheno-
mena, without any hypothesis as to their cause, or any conclusion
as to their significance.

That the experiments of many men are no better than this
may at once be acknowledged. But the less extreme advocates
of Watt confess that Cavendish had a purpose in his trials, and
that he was watchfully observant of the appearance of moisture as
a product of combustion ; and yet they will have us believe that
he — ^whose death Sir Humphry Davy mourned as the greatest
loss English science had sustained since the decease of Newton,
and whose skill as an interpreter of nature the scientific world
universally acknowledged — was, on this solitary occasion, no
better tihan a child amusing itself with an electrical machine and
some glass vessels. Even this demand upon our faith might
with some modification be honoured, had we no further infor-
mation than that Cavendish went through certain operations
with certain pieces of apparatus. But it is not denied by hia

CONVERSION OF HYDROGEN AND OXYGEN INTO WATER. 81

opponents that he learned so much from these alleged aimless
and uninterpreted trials^ that although one of the most reserved
and uncommunicative of men, he broke through the silence he
so much loved^ and reported his experiments to Priestley. On
the latter^ moreover^ his narration made such an impression, that
he proceeded to repeat the experiments reported to him, and
to make them known to the world. The notion that any man
should minutely report to another observations which taught the
reporter nothing, is surely absurd, and the character of Cavendish
makes the absurdity more glaring. But it is deepened, when we
consider to whom he divulged his researches. His con6dant
was Priestley, who, least of all men, except his colleague Warl-
tire, could be expected to take an interest in Cavendish's trials,
unless they brought to light something important. They were
an avowed repetition of the Birmingham experiments, which
Priestley made to entertain his friends, and Warltire tried to
determine the ponderability of heat. On the latter point.
Cavendish could only say that he found no proof that heat was
heavy in the results of his explosions. He could, however, confirm
the statement that each detonation was followed by the appearance
of a liquid,and could tell Priestley that this was the most important
phenomenon his random trial had brought to light, and that it
did not result as Warltire imagined, from the mere deposition of
water previously diffused as vapour through the gases ; for if
these were taken in a certain proportion, they could be entirely
condensed, and converted into water. This was a perfectly new
conception to Priestley. Neither he nor Warltire had made the
most distant approximation to it, nor could they have done so,
for it was only by a series of quantUaiive observations that such
a fact could be discovered, and theirs were solely qualitative.

Here, then, was a truth which Cavendish had learned from
his accurate, quantitative repetition of Priestley and Warltire's
inaccurate qualitative trials, and which no one in the world but
himself knew till he made it known to Priestley. In what pre-
cise terms the latter was informed of the truth we do not know,
but it is unquestionable that he was made acquainted with con-
clusions as well as facts. The facty Cavendish ascertained (to
take, for brevity's sake, the simpler case of hydiogen and oxygen,
omitting that of hydrogen and air), was this : — When two mea-

82 LIFE OP CAVENDISH.

sures of hydrogen and one of oxygen are exploded together in
a shut globe, they lose their elasticity, or cease to exist as gases,
and in their place is found an equal weight of water. This fact
might not have been regarded by Cavendish as justifying any
conclusion, or it might have been seen to warrant various
conclusions, among which he hesitated to select, and therefore
published none. But he did reveal a conclusion to Priestley,
and it was this : that hydrogen and oxygen were turned or con^
verted into water ; which was equivalent to saying that water
consists of hydrogen and oxygen. That he told Priestley thus
much, appears from the terms in which Priestley refers to
Cavendish in his paper of 1783. The language is very
remarkable. Priestley does not make the slightest allusion
to Cavendish's experiments as having been a repetition of
his and Warltire's. They had been so modified by their
repeater, and had brought to light so unexpected a truth, that
Priestley preferred no claim to them, but spoke of them con-
sidered as a whole, as '^ an experiment of Mr. Cavendishes
concerning the re-conversion of air into water.'' He called it
r^-conversion, because he believed that water was convertible
into air (gas or gases), and regarded Cavendish's experiments aa
complementary to his own. So convinced was he of their
importance, that he repeated them with gases carefully prepared,
so as to exclude from them moisture ; compared the weight of
the gases burned with that of the water produced, and found it,
as nearly as he ^^ could judge, equal," and, in consequence, he
came to the conclusion that there was ^^a strong presumption
that the air [inflammable air firom charcoal, and oxygen] was
re-converted into water, and therefore that the origin of it had
been water." His repetition was very inaccurate, and he com-
mitted the grievous blunder of substituting for Cavendish's
hydrogen the mixture of combustible gases obtained by heating
charcoal, which he conceived to be an anhydrous* gas. These
matters are referred to elsewhere, but they do not concern us
here ; neither does Priestley's want of entire confidence in the
significance of the experiments as establishing the conversion of
the gases into water. The wonder is, considering how inaccurate
his experiments were, that he got so far as to entertain '^ a strong
presumption" concerning the lesson they taught.

PRIESTLEY AN ACKNOWLEDGED IMITATOR OF CAVENDISH. 83

What is pre-eminently important is^ that long before Watt
had written his first letter, or had been supplied by Priestley
with the account of his repetition of Cavendish's experiments,
which was the basis of the conclusions concerning the compo-
sition of water announced by Watt, Cavendish had taught Priest-
ley the truth, which Watt, after learning it from the pupil,
declared the master had borrowed from him. Much anxious
endeavour has been made by the advocates of Watt's claims, to
deny this, to the extent, at least, of asserting that Cavendish
told Priestley only facts, and that it does not appear when he
communicated them. As for the latter point, which may be
noticed first, it is quite unnecessary to enter into any minute
enquiry into dates. It is not denied that it was between the
summer of 1781 and the spring of 1783, that the commu-
nication was made by Cavendish to Priestley, nor can it be
denied that Priestley's experiments were later than this com-
munication, and Watt's conclusions later than Priestley's
experiments. And as for the communication of Cavendish
having had reference only to ^^ facts,'^ if> among these, the
friends of Watt include the conversion of hydrogen and
oxygen into water, there need be no dispute between them
and the defenders of Cavendish ; for the conversion in
question was a conclusion from the phenomenal data sup-
plied by the globe-detonations, as every logician must acknow-
ledge ; so that his assailants must cease to affirm that Caven-
dish taught no conclusion. Priestley does not say in so many
words, that he was informed that inflammable air and oxygen
could be burned into water, but he implies this most plainly.
He tries Cavendish's experiment not as an impartial repeater
of it : he anticipates that it will establish the convertibility
of gases into water ; he hopes that it will, and so enable him
to strengthen his own converse doctrine, and he thinks, after
repeating it, that if it does not infallibly demonstrate, it> at
least, strongly warrants the belief, that a conversion of inflam-
mable air and oxygen into water can be effected. Had he
faithfully imitated his teacher's method of experimenting, he
would have been entirely satisfied of this, and I believe
he only hesitated, because his erroneous metihod of pro-
cedure rendered it impossible that he should have made

G 2

84 LIFE OF CAVENDISH.

more than a distant approximation to ascertaining the cardinal
truth, that there was equality of weight between the gases burned
and the water produced. But though he had completely failed
to verify the alleged conversion, his reference to Cavendish
would not be the less important. The experiment is '^Mr.
Cavendishes,'' as Priestley declares in two different parts of bis
paper. It is not, moreover, an experiment on the ponderability
of heat, or on the possibility of hydrogen and oxygen being
substituted for gunpowder (on which Priestley speculated), or
intended to amuse philosophical friends, but ^^ on the (re) con-
version of air into water." The first announcement to the
world, that its ancient faith in the elementary nature of water
had become a gray superstition, which, decaying and waxing old,
was ready to vanish away, was made by Priestley in the name
of Cavendish. With the latter the revolutionary doctrine
originated. He had since 1781 chronicled in his note-books,
that water which had been supposed to be " one and indivisible,"
a substance having no unlike parts, or unresembling ingredients,
could be manufactured like soap or glass, like any dye, or drug,
or pigment. He gave Priestley the recipe for its manufacture,
and Priestley, after trying it, published it to the world. It was
re-issued as a recipe of their own by Watt and Lavoisier, with
modifications which I do not at present consider. I am content
to affirm that every impartial reader must acknowledge that
Priestley's experiments from which Watt drew his conclusions,
were confessed repetitions of earlier researches of Cavendish's,
which had led him to a conclusion which he communicated to
Priestley, the substance of which was, that water could be com-
pounded out of inflammable air (according to Cavendish, hydro-
gen) and oxygen. This was as true and as full a doctrine of the
composite nature of water, as that which Cavendish or Watt
published afterwards ; and the /?rtm^y*ad« probability of Caven-
dish's originality and integrity assumes a very different aspect,
when regarded from this point of view, than it does from that of
De Luc's hasty, one-sided suspicions. Before looking at the
matter, however, in this light, it may be well to make, once for
all, a general reference to the part which Priestley took in the
Water Controversy.

His position was a remarkable one. Of all those who from

ARBITRATION OF PRIESTLEY BETWEEN CAVENDISH AND WATF. 85

first to last, have taken part in the Water Controversy, Priestley
alone was the friend of both the English rivals, and in circum-
stances to learn what each had discovered for himself, and had
borrowed from his brother philosopher. This was plainly the
belief of both Cavendish and Watt, each of whom implicitly
appealed to Priestley to attest his originality and priority, Watt
in his letter of 1783, Cavendish in the addition made to his
paper in 1784. But the umpire thus selected, never decided, so
far as any existing document shows, in favour of either claimant,
although he long survived the period of the original controversy,
and again and again commented on Cavendish and Watt's
theories of the nature of water, with the authors of which he
continued from first to last to be on the most friendly terms.
Yet Priestley, whatever were his faults, was certainly a very frank
and candid person, by no means reluctant or afraid to utter his
opinions on any subject or on any occasion. Here too, he was
invited by both parties to arbitrate between them ; and it appears
at first sight not a little perplexing, that he did not come
forward^ either to assign the whole merit to one or other of
the claimants, or to apportion it between them, according to
his conviction of their relative deserts. From his decision
there could have been no appeal, and had he assumed the
office of mediator between Cavendish and Watt, there would
have been no Water Controversy. The advocates of Watt's
claims have eagerly caught at anything that seemed to show
that Priestley favoured their client's cause, but Priestley, as
even they tacitly acknowledge, utters on the subject, but an
^ uncertain sound.'^ The following passage was written by him
in 1785, the year after the commencement of the Water Con-
troversy. It contains the only deliberate comparison of Caven-
dish and Watt's merits, which, so far as I am aware, he ever
published. Watfs advocates quote the latter part of it, as
serving their cause ; Cavendish's friends have not appealed to
it. I give it entire, that the reader may judge of its import for
himself:

^^ In the experiments of which I shall now give an accounts
I was principally guided by a view to the opinions which have
lately been advanced by Mr. Cavendish, Mr. Watt, and M.
Lavoisier. Mr. Cavendish was of opinion that when air is

86 ' LIFE OF CAVENDISH.

decomposed, water only is produced ; and Mr. Watt concluded
from some experiments, of which I gave an account to the
Society, and also from some observations of his own, that water
consists of dephlogisticated and inflammable air, in which Mr,
Cavendish and M. Lavoisier concur with him/'*

The friends of Watt quote this passage, from the words
^^Mr. Watt concluded,'' and so make the impression that
Cavendish only concurred in a conclusion which had first been
drawn by Watt. The entire paragraph, however, has a very
different import. It may, I believe, be justly referred to, as
showing that Watfs letter of 1783 contained the first statement
which Priestley had seen, that '^ water consisted of dephlogisti-
cated and inflammable air," and, therefore, that Cavendish had
not used those words in explaining his experiments of 1781- The
difference, however, between the language imputed to Watt, and
that imputed to Cavendish, is so slight, that it applies only to a
single word, and the advantage is on the side of Cavendish.
The fViends of Watt are, I believe, not aware of this, and those
of Cavendish have not referred to it, for Priestley's language is
obscure, and none of the critics of the Water Controversy have
offered an interpretation of it. I cannot here enter into the proofs
of the justice of the following statement. These will be found
at great length in the sequel,in sections 5, 6, 7^ and 8 of the Water
Controversy. It must suffice here to state dogmatically, that
the decomposition of air signified, in the language of Priestley,its
deoxidation, or rather the combination of its oxygen with some
combustible or oxidisable body, such as carbon, sulphur, or a
metal; and that the doctrine imputed to Cavendish by Priestley
was, that the universal product of oxidation is water. He did
Cavendish no wrong in this imputation. It has been so
entirely overlooked, that such was his doctrine, that his enemies
represent him as indifferent to his discovery of the composition
of water, whereas his fault was to exaggerate its importance, for
it is undeniable that Cavendish held that every oxidisable body
contained hydrogen, and that every oxidation yielded water as
one of its products. He asserts this himself, and allows him-
self to be taxed with holding the doctrine by Kirwan, who
animadverted on his paper of 1784, and called forth a reply

* Phil. Trana. 1785. Reprinted in Ejeperimentt on Air, 1786, p. 71.

PRIESTLEY'S ACKNOWLEDGMENT OF CAVENDISH'S ORIGINALITY. 87

from Cavendish,* in which, though he commented on other
matters, he suffered to pass mmoticed the affirmation of his
critic, that water is a product of every oxidation. It was so,
however, only because, ex hypothesis every combustible and oxi-
disable body contained hydrogen, and yielded water during its
oxidation, to the extent that it contained hydrogen. A/oriioriy
therefore, pure hydrogen could yield nothing but water by its
combustion, and the quantity of that which should be pro-
duced, would be determined solely by the weight of it which
entered into union with oxygen. Priestley, therefore, in 1785,
confirms the appeal which Cavendish made to him in 1784, and
gives additional meaning to the unsolicited reference to the
latter's experiments which be published in 1783. In both state-
ments he substantially asserts that his first instructor in the
true doctrine of the composition of water was Cavendish, and that
he taught that the sole product of the combustion of hydrogen
and oxygen was water. Priestley, accordingly, gives Cavendish
the priority, both in order of naming the rivals, and by re-
asserting his statement of 1783, that Cavendish then held the
doctrine, that water consisted of inflammable air and oxygen.
It thus appears that Priestley does not for a moment sanc-
tion the notion that Cavendish had arrived at no conclusion
concerning water, till after Watt's letter was written, but, on
the contrary, imputes to Cavendish an original, independent,
and prior view, which, moreover, he announced in language of
his own. To Watt, again, he imputes the first embodiment of
the theory of the composition of water, in terms which Caven-
dish, as well as he, afterwards used. But he says nothing
concerning either of the rivals being under obligation to the
other for his facts or his theory, nor does he profess adhesion
to their views. PriesUey thus leaves the question very much
where he found it, and can be claimed by the advocates of
neither of the English rivals as having decided in favotur of their
client. He did not do so, I believe, for the very sufficient
reason that he thought both were in error, and considered it,
accordingly, a very unimportant matter to decide whether
Cavendish or Watt first committed a blunder. Strange,
indeed, as it may seem, it is quite certain that Priestley lost

♦ Phil. TrwM. 1784, p. 170.

88 LIFE OF CAVENDISH.

faith alike in Cavendish's experiments and in his own repetition
of them, and came, in 17B5, to affirm that it was impossible to
bum hydrogen and oxygen into an equal weight of water, and
that the product of their combustion was nitric acid, as well as
water. He regarded the language of Cavendish and of Watt,
accordingly, as at variance with the facts of the experiment they
professed to expound, and he naturally thought it needless to
enquire which first went astray. Nevertheless, it is not a little
difficult to understand how it happened that Watt was so pro-
foundly ignorant, as he appears to have been, of Cavendish's
experiments which preceded and were the occasion of Priestley's.
The latter told all the world publicly and explicitly, that he
followed Cavendish, but he did not apparently make this known
to Watt, for although the latter was minutely acquainted with
the contents (though not, perhaps, with the very words) of
Priestley's Essay of 1 783, and wrote his first letter as a Com-
mentary upon that Essay, he does not once allude to the
experiments upon which he based his theory of the composition
of water, as primarily Cavendish's, but speaks of them as if
they had been original researches of Priestley's. An imperfect
and inaccurate reference to Cavendish's priority Was eventually
made, but not till 1784, immediately before the letters to Priestley
and De Luc were publicly read to the Royal Society.* That
Watt made an ungenerous concealment of his knowledge of
Cavendish's researches, I do not think, but it is important to
notice that he can be acquitted of this, only on the supposition
that he had either been left in ignorance of Priestley's obli-
gation to Cavendish, or was not alive to the importance of the
information on this point, which Priestley gave him. To this
question I shall return, so far as it affects the feelings which
Cavendish may be supposed to have entertained towards Watt.
Meanwhile, I only notice, that as the reference made by Watt
to his English rival's labours is much less accurate and
ample than it should have been, by one who, apart from his
priva*^e sources of information, had access, in 1784, to the pub-
lished declaration of Priestley in his paper of 1783, which it is
difficult to believe Watt would not read in the Philosophical
Tyansactions ; it seems certain that he undervalued the import-

* Phih Trant, 1784, p. 332.

SINGULAR IGNORANCE OF PRIESTLEY AND WATT. 89

ance of Priestley's statements concerning Cavendish's previous
researches^ for he wrongs them both in his reference to the
latter. He does Priestley injustice by attributing to Cavendish
the first observation, that moisture is produced during the com-
bustion of oxygen and inflammable air, an observation which
Cavendish never claimed to have been the first to make, but
expressly assigned to Warltire and Priestley, to whose joint
experiments he referred as the source of his own knowledge of
the fact.* At the same time. Watt wronged Cavendish by not
attributing to him the earliest observation of the truth, that the
weight of water produced during the combustion of inflammable
air (hydrogen) and oxygen, equals the weight of the gases which
lose their elasticity during its production, although Priestley
gave the whole credit of the discovery to Cavendish. We have
no means now of determining in what proportions the blame of
this misstatement is to be divided between the two philosophers
of Birmingham, but that it did not originate in any wilful con-
cealment on the part of either, I entirely believe. The most
ungenerous of critics must find it impossible to impute to
Priestley any jealousy of Cavendish, and only those who are as
unjust towards Watt, as some of his partizans are towards his
rivals will accuse the great engineer of intentional injustice.
Priestley plainly must have been singularly unfortunate and
sparing in his reports to Watt, of the connexion between his
researches and those of Warltire and of Cavendish, or Watt
would not have failed to do justice to his informant, if to no one
else. Watt, however, must have been very indifferent to the
authorship of the facts on which he based his theory, or he
would have sought out information for himself. In Priestley^s
volume On Air, published so far back as 1781, he would have
found the records of experiments made by two of his own
townsmen (Priestley and Warltire) on the combustion of inflam-
mable air and oxygen, even if his friend and neighbour had not
privately communicated them to him, which he is so likely to
have done; yet in 1784 he gave an account of matters totally
at variance with the statements in that volume. Nor can he
have conferred with Priestley on this subject, frequently as they
conversed together on the convertibility of water into gas, or he

• Phil. Trans. 1784, p. 126.

90 LIFE OF CAVENDISH.

would have been referred to the account of 1781^ and set right
as to the nature of the observations which Warltire, Priestley^
and Cavendish had severally made on the appearance of water
when inflammable air is burned.

Priestley thus occupies a very singular position in the Water
Controversy. He was the confidant alike of Cavendish and
of Watt) as to their theories of the nature of Water ; he was
implicitly appealed to by both^ to decide between them as
claimants of the disputed discovery ; and he was apparently in
possession of information which should have enabled him to
dispose of the Appeal. In no document published^ however, has
he done so, nor is there any prospect of unpublished papers
throwing much additional light on his views. The late Mr.
James Watt (junior) states that ^'Inquiry was made of Dr.
Priestley^s son (since dead), as to his father's papers in 1783-4.
He supposed them to have been burned at the time of the Bir-
mingham riots in 179I> which was confirmed by a search he
caused to be made in America.'^*

That this supposition was a just one can scarcely be doubted^
but an important series of letters from Priestley to others, during
the period when he was engaged in his researches into the nature
of water, was not exposed to the destruction which befel the papers
in his own possession in 1791> &nd this series has since returned
to the possession of his descendants. Through the kindness of
Miss Finch, of Birmingham, a grand-daughter of Dr. Priestley,
I have been favoured with the loan of thirty-eight letters, written
by him to his celebrated friend Josiah Wedgwood, and his son
Thomas. The greater nimiber of the letters are to the elder
Wedgwood, especially the earlier ones, and they extend through
1781, 1782,1783, 1784, 1787, 1788, 1789, 1790,1791, and 1792,
i. e., exactly over the period of most interest in reference to the
Water Controversy.f The nearest surviving relative also of
James Keir, Esq. (whose accomplishments and skill as a
chemist were highly honoured by Priestley), has granted me
the perusal of twelve letters addressed to him by the latter.

* Letter to Mr. Mnirhead. Watt Corr.f p. zIt.

t These letters were sent to Miss Finch by F. Wedgwood, Esq., of Barlaston,
Stone, Stsffordshire^ and, as I learn from him, constitute the entire extant corres-
pondence between Priestley his grandfather Josiah, and hisunde Thomas Wedgwood.

CORRESPONDENCE BETWEEN PRIESTLEY AND WEDGWOOD. 91

Their dates (so far as they are dated) are 1782, 1787? and
17B8. Many more passed between the fellow-chemists, but the
greater number have perished.* Prom the collateral relatives of
Priestley, likewise, resident in Leeds, I have obtained three
letters written by him, of dates 1786, 1791^ and 1792. These
were obtained for me by Mr. William O. Priestley, a student of
great promise, who attended the University of Edinburgh last
winter (1849-50), and in whose father's possession they now
are. They were entirely on personal matters, and did not throw
any light on the Water Controversy.

The letters to the Wedgwoods^ and those to Keir, but
especially the former, are full of references to Priestley^s views
on the composition of water; and in them, if anywhere,
might we expect to find some expression of opinion concerning
the good faith of Cavendish and Lavoisier ; but although their
names^ and those of Watt and Blagden, are referred to in con-
nexion with the theory of the composition of water, not the
slightest allusion occurs to any jealousy between the rivals.t
Some extracts from them, including all the direct references to
Cavendish, Watt, and Lavoisier, will best show what light they
throw on the Water Controversy. The correspondence may be
said to be almost entirely chemical. Mr. Wedgwood, with the
liberality which characterized him, supplied Priestley with as
many clay and porcelain retorts, tubes, and other pieces of appa-
ratus, as he chose to ask for, and made of the shapes and mate-
rials which he prescribed. The burden, accordingly, of most of
Priestley's letters is such as the first sentence of the first epistle
may illustrate : " 26th May, 1781. To Josiah Wedgwood, Esq.
Dear Sir, — I must take the liberty to give you this trouble

* For the knowledge that Bnch letters existed, and the introductionB which
eaebled me to proeare them, I am indebted to my fiiend Dr. Percy, of Birmingham,
who, besides his own aedre co-operations, secured me the good senrices of Dr. James
RuaseU, of the same town, who spared no trouble in farthering mj wishes.

t A. warm and lasting iHendship subsisted between Keir and Priestley, and the
loss of any part of their correspondence is greatly to be lamented, especially that
referring to water. My informant in reference to the Keir Correspondence, states
that '* unfortunately, part of these letters, perhapst he most interesting in the col-
lection, for they told of his experiments previous to the discovery of the decomposition
of water, have, 1 fear, been consumed at the fire at Abberley Hall, having searched
for them in vain. These letters showed how very nearly Dr. Priestley touched upon
that important discovery."

.S.!feJ*^^.i9:••jSl•l,~•s^;*J• .. ., p„„i„ „,_„,,

a their receipt is

e experiments

:d, and the results

k much interest

if earthenware,

r his peTsonal

_^jirally anxious to satisfy

48^ of the apparatus sup -

J^]Ql9tfii|}ting his friend's pecu-

^ -^ ^^ — feiEllSiequently made in the

Hm R^^t^ft#fw^^hUn]inutene8s of detail in

m^ VntH R Hi€*UM|r|S|Qyif^derab1e justice test his

* f -JJE^J«lE9tE87l'60>l9t>yj by the legal maxim,

• @t»^^!5Jtf8*^^J^x:^tentibus eadem ratio."

" 'a*'*^'^ii^'^^^^ nothing to say.*

' • '" ^r*^^*^^^^^'^ earlier letters are the

'^''S- ^^^^^^^^'^ ^«^l Correspondence,

a£E&£E igt^^lconvertibility of water

"" ""'^"'"^""g^l^.stley so much, and led

■, of 1783.

pJCSA^I^nything specially refer-

"■"j^l&^ihiefly occupied with the

j^rt^sjl^ous earthy substances,

-Jt,Wl5;'-'*«'*3i»|^'ti4ii^»§S^«^»^»terraTieous fires " (May

.W^-"."*j|j4 a^j-I^^^Mg^ reports an experiment

ir*^ ^*^ ^-s^bP^kUfi^ulgil^^^which he reduced oxide

f"* -^ '■a'ftilV'!' ^^ lf*tCw«idai^5iS-'«re to JoiUh Wedgwood. u>d
■• sf--¥.^-«,*'*^''«^iM|w^>@D^^^ or greed in Meeptiiig

f^f'3D{pSnrS^S£3M£S«>'3M^TlSp«<«''' "?>■<> genenllj taking
^>&:^!^UI^M^Sl^'if/f^is£t, KB pliinlr, on the ODo hmd.

- ^!i*^)^7^'^*!i|^i^f(^^^^^^(^*(:l''><>*>Bde;me>i(> of the kind.
r^L^«rif;!^i:^.?&aIfi^iiS:''bligrtion, noi UQConsdoui of
I^a-'^'f: *^°'^*^*€l*^*^^'^*tl'' ^ >"itln°g> trmtlal Uboura

-ss^

SUPPOSED TRANSMUTATION OF WATER INTO AIR. 93

of lead to the metallic state^ by heating it by the sun's rays con-
centrated by a lens in an atmosphere of inflammable air, which he
thinks is thus proved to he phlogiston. It was the prosecution of
this idea which led him to theorise on the nature of water, and
to interest himself in the speculations of Cavendish and Watt.
On 16th September (1782) he announces his supposed dis-
covery, that charcoal is entirely convertible by heat into inflam-
mable air, a mistake which afterwards led himself and Watt
seriously astray, and which figures conspicuously in the Water Con-
troversy. On 8th December he announces that he can convert
water into permanent air, by distilling it with lime at a strong
heat; and on 8th January, 1783, that the lime is unnecessary,
and that the mere heating of water in one of his friend's earthen
retorts, was sufficient to convert it into air. Great demands are
now made on the stock of retorts at Etruria ; and Wedgwood,
probably not a little surprised to find that his vessels possessed
so wonderful a property as that of transmuting water into
atmospheric air, sends two sagacious queries along with a fresh
supply (23rd January, 1783). Priestley had explained to him
that though the retorts were air-tight, a portion of the water
placed within them came through their walls, and vaporised
from their outer surfaces. Wedgwood in reply asks, "If
water passes through the retort outwards, may not air pass
inwards r^'* This query Priestley disposes of in the negative,
very ingeniously, but he afterwards found, as Watt did also,
that Wedgwood had detected the true state of matters, and
that the supposed transmutation of water into air was only an
exchange of steam, and the atmospheric gases through the pores
of the retorts, which became permeable at a high temperature.
For the present, however, Priestley proceeded with increased
alacrity in his transmutations, and discovered, as he imagined,
that hydrogen and oxygen could be burned into carbonic acid
(7th March 1783) ; a mistake which introduced confusion into
his own notions, and those of Watt, concerning water.f

* This has been copied by Josiah Wedgwood's secretary, Mr. Chisholm, in
Priestley's letter (Jan. 23rd, 1783), as an interlineation after it had been received.
Francis Wedgwood, Esq., of Barlaston, Stone, Staffordshire, by whom Priestley's
letters to his grandfather (Josiah) were sent to Miss Finch, frpm whom I had them
in loan, kindly enabled me to verify this point.

t Watt Corr,, p. 17.

94 LIFE OF CAVENDISH.

On the 23rd of the same month, Wedgwood is informed of
an experiment tried by his friend, which is the most important of
all that he made^ so far as the Water Controversy is concerned.
It is thus announced : '^ I have lately made such experiments
relating to the conversion of water into air, as must, I think^
satisfy even Mr. Kirwan. By the electric explosion I decom-
pose dephlogisticated and inflammable air, and I find the weight
of the latter in the water I get from it.'' * The wording of this
statement is ambiguous, and might seem to imply that only the
weight of the inflammable air burned was found in the water
produced by its combustion. There can be no doubt, however,
that Priestley signified that the weight of both the gases con-
cerned in the combustion, was equalled by that of the water they
jrielded. This at least is the proposition which he undertakes to
prove in his paper of l7B3,t and there also he informs us, that
his experiment was a repetition of one previously tried by
Cavendish. To Wedgwood, however, he was as silent on this
point, as he seems to have been to Watt, and apparently for the
same reason, that the result interested him much less than his
own fancied discovery, that water could be transmuted by porous
retorts, into atmospheric air.

An important interval here elapses, during which he ad-
dresses no letters to his friend, and which we know was occu-
pied by himself and Watt, in drawing up for the Royal Society
the conclusions they had drawn from the experiments in question.
On the 6th of May, however, Priestley writes from London to
Etruria, explaining that he has discovered the delusive nature of
his supposed transmutations of water into air, and verifying
Wedgwood's sagacious conjecture as to the true nature of the
phenomena witnessed with the porous retorts. This discovery
led Priestley to alter the concluding part of his paper and its
title, and induced Watt to withhold his letter from pub*
lication.:^

Priestley's paper, meanwhile, was read to the Royal Society,
and he proceeded with his researches into the evolution of gases

* The same ezperiment is referred to in the VTati Chrretpondenee, p. 1 7, under
date March 26, 1783.

t Phil. Trmu, 1783, p. 427.
t Watt Corr., p. 25-30.

PRIESTLEY'S RECANTATION. 95

from substances heated in Mr. Wedgwood's excellent retorts.
Among other things, he heated mtre, and obtained from two
ounces of it, 812 ounce measures of oxygen, and he tells
Wedgwood, that his own idea is, that the nitrous acid of the
nitre is charged into the air, but that Mr. Watt '^ still thinks
that it is water that furnishes the air'' (24th July, 1783.)*

His next letter to Wedgwood is of date 16th January, I784,t
and contains some very remarkable passages. The italics in the
quotations mark the words underlined by Priestley. ^' The
great problem with us aerial philosophers (not navigators), of
late, has been to find what becomes of dephlogisticated and
inflammable air when they are made to unite, as by explosion, &c.;
some saying that they make water ^ others fixed airy ^c. The foU
lowing experiments show that, in different circumstances, they
make both, and also, that dephlogisticated air incorporates with
iron in a great proportion."

This letter was written the day after Cavendish's paper on
Water was communicated to the Royal Society. It is not pro-
bable, however, that a knowledge of its contents can have
reached Birmingham before the letter was written. It cannot be
doubted, nevertheless, that Priestley referred to Cavendish and
Lavoisier's views. From the Watt Correspondence it appears
that in November 17B3>t Priestley was aware of Lavoisier's
opinions, and he had known Cavendish's since at least the preoed-

* Watt pabliflhed thU opinion in 1784. " Nitre/' says he, <' besides its water
of crystaltisation, contains a quantity of water <u oneqfiU eiementary parit, which
water adheres to the other parts of the nitre, with a force sufficient to enable it to
SQStain a red heat. When the nitre is melted, or made red hot, the acid acts upon
the water, and dephlogisticates it ; and the fire supplies the humor with the due
quantity of heat to constitute it air, under which form it immediately issues." PkU.
Tratu. 1 784, p. 336. I need scarcely say that Watt was mistaken in imagining that
nitre contains water, and that the oxygen which the salt yields when heated is derived
from that liquid. His views on this subject, which he illustrates at length in his
paper, will be referred to frequently again.

t This letter is endorsed by Josiah Wedgwood, June 1784. Priestley's own
date, however, exactly corresponds to those of his letters of 1783, in three of which
January is written as in the letter of 1784. Francis Wedgwood, Esq., of Barlaston,
tells me that he suspects that his grandftither did not always docket his letters when
he received them, and has given cogent reasons for this opinion. PriesUey's indis-
tinct writing of Janr- might easily be mistaken for June, at a hurried glance. I
have asked the opinion of three unbiassed parties, including a great grandson of Josiah
Wedgwood's, and they all read the word January*

X Watt Corr.f p. 35.

96 LIFE OF CAVENDISH.

ing March. They were the only parties who contended that
inflammable air (hydrogen) and oxygen make water^ and on/y
water^ when they combine. Kirwan was the great advocate of
the production of fixed air, or carbonic acid, by the combustion
of the gases referred to ; and Priestley and Watt held, on the
authority of the experiments of the former, that both water and
carbonic acid might be produced by the union of inflammable
air and oxygen.* Priestley, however, mentions no names, and
does not hint that Watt was aggrieved, although he knew that he
complained of Lavoisier's behaviour to him. His silence in
this respect is in keeping with the whole tenor of his proceed-
ings, and shows how utterly heedless he was of the personal
differences which attended the discovery of the composition of
water. He was one of the most zealous of polemics on all sub-
jects, and never more serene and complacent than when, like
Ishmael, his hand was against every man, and every man's
hand against him. Yet living, as, like the fabled Salamander,
he delighted to do, in an atmosphere of fire, he refused to take
a side in a controversy which might have been expected to rouse
all his combativeness ; and on perhaps the solitary occasion in
his life, when he was solicited to become a combatant, he quietly
answered the contending rivals by ignoring the fact that they
were at issue, and acting as if there were either no dispute, or at
least nothing worth disputing about.

The indifference which Priestley thus displayed, in assisting
to verify the views of his friends, he showed as conspicuously
in reference to the truth and consistency of his own views. In
the letter from which I have been quoting, a statement occurs
of the greatest importance in reference to more than one vexed
question in the Water Controversy. It is as follows : — '^Ano-
ther experiment shows a remarkable difference between inflam-
mable air from metals and that from charcoal. Having mixed a
quantity of each of them with half as much dephlogisticated air,
I exploded that from iron, and found neither water nor fixed
air; but exploding the mixture that contained the inflammable
air from charcoal^ 3^ ounce measures of the mixture yielded an

* See In illustration of this the papers of Cavendish, Watt, and Kirwan, in Phil.
TVans, 1784. Priestley's letter speaks for itself.

PRIESTLEY'S MISTAKES CONCERNING INFLAMMABLE AIR. 97

evident quantity of water and f ths of an ounce measure of pure
fixed air/*

This singular passage^ it will be observed^ contains an implicit
contradiction of what Priestley had published to theworld in 1783.
In announcing in his paper of that date^ his repetition of Caven-
dish's experiments on the conversion of gases into water^ he had
affirmed that he could burn the inflammable air from charcoal
{a mixture of hydrogen with different compounds of carbon)
-along with oxygen, into their conjoined weight of water. He now
states^ in contradiction of that incredible statement, that car-
bonic acid as well as water results from the combustion in ques-
tion; but as if to atone for correcting one error, he commits
another quite as great, and makes the extraordinary assertion,
that inflammable air from iron, t. e> hydrogen^ may be exploded
with oxygen, and yet no water be produced.

It is further evident that he regarded the combining measure
of the charcoal gas as identical with that of hydrogen^ and
double that of oxygen, a fact^ as it will afterwards appear^ of
some importance, in reference to the exact doctrine taught by
Watt in opposition to Cavendish, concerning water. Before,
therefore, either Cavendish or Watt had given his views to the
world, Priestley had lost faith in the conclusions of both. He
held it possible that hydrogen might be burned and yet water
not be produced, and he believed that carbonic acid was as
common a product of the combustion of inflammable air and
oxygen as water.

On the 23rd of January, 1784, he returns to the question of
the source of the oxygen yielded by melted nitre, and announces
that he has changed his opinion concerning the presence of
nitrous acid in oxygen, as he had found that the acid was
partly volatilized, partly dissolved by the water during the coU
lection of the gas from nitre. ^^This,*^ he continues, "greatly
favours Mr. Watfs hypothesis, that air [oxygen] is dephlogis-
ticated water. In this business I am little more than the
jbellows-bhwerJ^'^ This is the most important reference to the
opinions of Watt concerning water, and to Priestley's share in
producing them, which I have found in the Wedgwood Corres-

* tliat '' air" bere signifies oxygen, is erident from the nature of the experi«
meni^ and from Wiitt's exposition of his views, referred to in the preceding note.

* . -•- •- »,

' ' ' - • • «-

98 LIFE OF CAVENDISH.

pondence. With characteristic frankness^ the latter disavows all
but a slender share in leading to their formation^ and empha*
tically underlines the title he gives himself. In truths as he
now dissented from them, he could not with any propriety claim a
share in them. The hypothesis which he attributes to Watt has not
reference merely to the evolution of oxygen from nitre, but to the
general doctrine announced in his letter to Priestley, of April 1 783,
which as yet, at its author^s request, remained unread at a public
meeting of the Royal Society. That doctrine was, that water
consisted of phlogiston, or inflammable air, and dephlogisticated
air. By depriving water of the former, oxygen was obtained,
and by such a loss of the phlogiston of the water hypothetically
present in it, nitre was supposed to yield oxygen. I do not
comment upon this passage because it is less full and explicit
than one which Priestley published m 1785,* which has already
been quoted, and will be referred to again.

A blank occurs here in the correspondence^ for several
months^ at the critical period when De Luc and Watt were
accusing Cavendish of plagiarism, and when, possibly, had
letters remained, we should have learned Priestley's opinion con-
cerning the claims of the rivals. The letter next in chronological
order bears date 8th November, 1784, and was therefore written
after Cavendish and Watt's views on the compound nature of
water had been given to the world in the Philosophical Transact
turns for that year. Nor can Priestley have been without some
private information concerning the feelings entertained at Bir-
mingham towards Cavendish. His letter, however, betrays no
consciousness of these, and is chiefly occupied with an account of
the repetition of Lavoisier's experiments on the decomposition
of water, but his name is not mentioned. After referring to the
passage of water and spirits of wine, in the state of vapour,
through a red-hot copper tube, he continues: ^'Iron, I find,
gained one-third in weight in this process, and gives one-half
more inflammable air than it does when dissolved in acids, the
reason of which I believe to be, that much of the phlogiston is
always retained in the solution of metals in acids. On com-
paring the experiments, I now think that the inflammable air is
furnished by the iron, and that there is no decomposition of the

* PAt/. Tran9. 1785, reprinted in ExperimenU on Air, 1786, p. 71.

PRIESTLEY'S THEORY OF THE ORIGIN OF HYDROGEN. 9&

water. Mr. Watt thinks so too/' According to this statement,
Priestley and Watt imagined that when steam is passed over
red-hot iron, it displaces the phlogiston hypothetically present
in it, and unites with the calx of the metal. This appears more
distinctly from what follows : — ^^ Iron that is thus increased in
weight, and has yielded so much air [hydrogen] (which, by the
way, has not the least offensive smelly which has been so much
complained of in filling balloons), is reduced to its former state
by heating in charcoal. In this process, instead of yielding
water J as we all imagined it would, it yielded a prodigious quan-
tity of inflammable air, but of a peculiar kind, for it is about as
heavy as common air, owing, as I found, to its containing a great
quantity of fixed air combined with it, so as not to be separated
by lime-water, but only by decomposition with pure air by the
electric spark.'* The expectation of Priestley and his friends
evidently was, that the phlogiston of the charcoal would displace
the water from the calx of iron, and so reproduce the latter in
the metallic state ; and the whole passage shows, as other state-
ments do also, how incredulous Watt was, concerning Lavoi-
sier's great discovery that water may be analysed into hydrogen
and oxygen.*

The remaining letters of Priestley to Wedgwood are of
much later date than those already referred to. They con-
tain, however, several important allusions to the nature of
water, which may best be introduced here as enabling us fully
to understand the point of view from which Priestley looked at
the rival claims of Cavendish, Watt, and Lavoisier.

In a letter not dated by its writer, but endorsed 1787,t
Priestley renews the correspondence with Etruria, which bad
been interrupted for a considerable interval, for the very cogent
reason naively acknowledged at the commencement : ^^ Having
been engaged in courses of experiments that did not require the
use of earthen retorts, &c., I have not troubled you of a long

♦ The heavy mflammable air referred to by Priestley, was evidently (in greater
part at least) carbonic oxide, obtained by heating the oxide of iron (produced by the
passage of steam over the red-hot metal) with charcoal ; and the fixed air he sup-
posed to be contained in it, was produced by the union of the carbonic oxide with the
oxygen with which he exploded it.

t The endorsing is in the handwriting of Josiah Wedgwood, as his grandson,
Francis Wedgwood, Esq., has enabled me to verify.

H 2

100 WFE OF CAVENDISH.

time. But now, haying many things in view which I cannot do
without your assistance, I am obliged to have recourse to it.'^
The letter, however, does not allude to water, and is the only
one of that year. On 8th January, 1788, he explicitly refers to
the different views entertained concerning the composition of
water, and seems at length to accept the office of umpire
between the disputants, though not exactly in the way they
desired : —

^^As the experiments in which I am now engaged, pro-
mise to be of some consequence with respect to what has of
late been the subject of philosophical discussion, I give yoa the
earliest account of the probable issue of them.

^^ They completely refute the hypothesis of dephlogisticated
and inflammable air composing only water. The decomposition
of them always produces acid, and Dr. Withering finds it to be
as yet in all cases the nitrous. They give reason to think that the
great quantity of water that has been found in this case is nothing
more than was either diffused through the airs, or was necessary
to their aerial form. I almost conclude that water is the basis
of all kinds of air. One of my experiments (on terra ponr
derosa) proves that it [[water] is a considerable part affixed air,
not less than one-third of its weight; though it has been
thought to consist of nothing but dephlogisticated and inflam^
mable air.

*^ My experiments seem to render doubtful the conclusion
that Mr. Cavendish draws from his, as I get nitrous acid from
dephlogisticated air, without any that is phlogisticated. This is
the case whether the dephlogisticated air be got from manga-
nese, red precipitate, or red lead/'

This important passage contains the first reference to an
opinion from which Priestley never afterwards receded. Water
he believed to be present in all gases, and hydrogen and oxygen
he held might produce by their combustion, nitrous (nitric) acid.
Cavendish had encountered the same apparent phenomenon,
but had shown that it resulted from the presence of a little
nitrogen in the gases burned together. Priestley, however,
thought he had secured perfect purity of the hydrogen and
oxygen, and discarded this explanation. He thus dissented
alike from Cavendish, Watt, and Lavoisier, and totally aban-

*. •> k
» » »

PRIESTLEY'S ERROR AS TO THE NATURE OP NITRIC ACID. 101

doned his original doctrine^ that inflammable and dephlogisticated
air were entirely convertible into water.

He returns to the subject in his next letter (endorsed by
Mr. Chisholm^ Mr. Wedgwood's secretary, March 18th, 1788),
in which> referring to his observations, reported in the preceding
epistle^ he says, ^' These experiments, I cannot help thinking,
prove the decomporition of water to be a fallacy, and establish
the doctrine ot phlogistonJ^'

On August 18th, (1788), he writes, "I see clearly the cause
of the fallacy in M. Lavoisier's experiments and my own, in
which we found ^rc watery when I now always find some acidJ^
He then refers anew to experiments, which prove, as he thinks,
that inflammable air from iron (hydrogen), in certain circum-
stances, unites with oxygen to produce carbonic acid ; so that
he had now come to the conclusion, that water, carbonic acid,
and nitrous (nitric) acid, might all be produced by the union of
hydrogen and oxygen. " The objection," he continues, '' that
Mr. Cavendish and Dr. Blagden made to my experiments, was,
that the acid I procured was from phlogisticated air [nitrogen],
but this I have abundantly obviated, for in all the processes the
more there is of this air (or of any other kind that cannot be
decomposed by it), the less acid I find.'' *

Assured as he thus thought himself, of his accuracy, which,
however, he never asserts dogmatically, Priestley proceeded to
set his contemporaries right, on a point on which he alone was
all in the wrong, and in his next letter to his friend (Oct. 9,

* Had Priestley studied Cayendish's paper of 1784 he would not have made this
mistake. Cavendish showed that the conditions for the production of nitric acid
were, excess of oxygen, a moderate proportion of hydrogen, and a very small one of
nitrogen. If the last were greatly increased, no nitric acid appeared. This negative
result was procured when atmospheric air (containing 4-5ths of its volume of nitro-
gen) was detonated with hydrogen. Cayendlsh and Blagden appear to have pointed
out thia to Priestiey when he communicated hia second paper on phlogiston to the
Royal Society (Phil. Trant. 1788), for in a letter to Wedgwood, not dated, but
evidently written before that quoted in the text, he says, " the only objection that was
made to my conclusions was, that the add I got was from the phlogisticated air,
which I could not exclude." Both letters are important, as showing that Cavendish
and PriesUey discussed a second time the question, what is the product of the com-
bttstion of hydrogen and oxygen ? which had occupied them in 1781 and 1783 ; and
that the latter, in 1788, in spite of all that had been written on the subject, in England
and France, fell into the mistake against which Cavendish had guarded himself and
others in 1784.

102 UFE OF CAVENDISH.

1788), hs announces that he has "drawn up a third paper to
send to the Royal Society,** to show what he has done, '^ in a
business so much agitated as the doctrine oi phlogiston.^^ ^ No
further reference, however, to this question, which in Priestley's
estimation was inseparable from that of the true nature of water,
occurs earlier than I790.t In an undated letter (endorsed by
Josiah Wedgwood's secretary, Mr. Chisholm, October 1790) of
that year, he writes, *^ My chemical pursuits have been directed
to the great question now depending on the decomposition of
water, S^c. But still, whether I decompose the two kinds of air
by an explosion in a copper tube, or by a slow burning, or in the
manner of the French, I never fail to produce acid, though
they now say they find none at all, and even have made ounces
of water perfectly pure/*

In the next letter (February 16, 1791), he enforces the same
erroneous doctrines still more emphatically. ^^It was ob-

* Among some letters which Mr. Frands Wedgwood, of Barlastoiii has allowed
me to peruse^ is one in the handwriting of Mr. Chisholm, Josiah Wedgwood's secre*
tary, addressed to Priestley, hut unsigned. It is evidently, however, from Wedgwood,
and in answer to Priestley's letter referred to in the text. This appears from the
writer's acknowledging ''your good letter of the 9th" (9th October, 1788)> and
declining his friend's offer to send him a separate copy of his paper. The following
extracts from it are of interest in reference to the Water Controversy : —

"I must, therefore, once for all, beg your acceptance of my best thanks for the
early communications, from time to time, of your truly valuable discoveries, wfaidi
now become more and more interesting ; and I most sincerely vrish you health with
every convenience for the prosecution of them.

'' I cannot forbear expressing my particular satisfaction to find that my old
favourite, phlogiston, is likely to be restored to its former rank in the chemical
worid. ••♦•••

Mr. Watt's conjecture of nitrous acid being contained in inflammable air, as the
vitriolic is in sulphur, pleases me much, though I confess there is one circumstance
which appears rather unfavourable to it ; for I understand it to be by combusHon that
the acid is detached from the phlogiston, and one would expect the niinms acid to
be rather decomposed than developed by that process."

The allasion here, evidently, is to the appearance of nitric acid as a product of
the combustion of apparently pure inflammable air (hydrogen ?) and oxygen, which
all chemists now agree with Cavendbh in referring to the presence of nitrogen as an
impurity in the gases, which when burned together yield the add, but which Watt,
like Priestley and La Place, referred to an erroneous source.

This letter is the only one from Wedgwood to Priestley of which I have any
knowledge. The MS. from which I have quoted was probably a copy of the original.

t Tlie letter quoted in the text is the last belonging to 1788. There is only one
Tielonging to 1789 ; it is addressed to Frands Wedgwood (the son of Josiah), and
refers to private matters.

CONCLUSION OF PRIESTLEY^S EXPERIMENTS ON WATER. 103

jected^ though on insufficient grounds, to my former experiments^
that the acid I 'produced came from the phlogisiicated air that
was necessarily mixed witti the dephlogisticated that I made
use of. But I now with great certainty make air so pure, that I
am confident it contains no mixture [of] phlogisticated air what-
ever^ and yet the explosion of this air^ with a due proportion of
inflammable air^ produces more acid than when the air I used
was less pure. I also use no air pump, filling my copper vessel
with water, and displacing it by the mixture of air to be
exploded.

^ Admitting, therefore, what I am not disposed to dispute,
that the slow combustion of the two kinds of air by the French
philosophers, produces nothing but the purest water ^ it must be
admitted that a different mode of combining the same elements
in my process makes nitrous €udd. ; • . . The French experiment
makes nothing against the doctrine oi phlogistouy as it only
proves that it enters into the composition of water/'*

On February 26th, (17^1)^ be reiterates this declaration more
emphatically, ^ I an at pleasure make either nitrous acid or
pure water from the same materials, viz., dephlogisticated and
inflammable air. If there be a surplus of the dephlogisticated
air, the result is always acid, if of the inflammable air, it is mere
water. Extraordinary as this is, it is uniform, so that both M.
Lavoisier and myself have been right. The doctrine oiphlo^
ffiston, however, stands firm, and it only appears it is one element
in the composition of water.

^' I shall send a paper on this subject to the Royal Society in
the beginning of the next week. It will decide this long contest J^

The italics in the last sentence are mine, and the words are
very significant. The chosen umpire of the English rivals, it
will be seen, put them out of court, made himself a party to the
dispute, compromised matters with the French rival, and de*

* Priestley's disuse of the air-pninp, in the hope, evidently, that he would avoid
contaminating with nitrogen the hydrogen and oxygen which he exploded together,
was of no avail in securing purity of the gases. The water with which he filled his
copper vessel was certain to contain air (and therefore nitrogen) dissolved in it, which
would be displaced by the hydrogen and oxygen^ and become mixed with them. The
conditions for the production of nitric acid as laid down by Cavendish were therefore
Mcnred, provided only exceat of oxygen was present, and that there was, appears
from the next letter (Febmary 26th, 1791) referred to in the text.

104 LIfE OF CAYENDISHr

clared the contest at an end. His judgment^ we may be certain^
satisfied no one but himself. It was ignored by his contempo*
raries, and has been reversed by his successors.

The remaining letters to Wedgwood contain no further
reference to water. The '^invasion of the Goths and VandaLs,"
as Priestley styles the Birmingham riots of July 14th^ 1791, is
referred to in his next epistle^ which is dated from London
(July 26th, 1791), and a fresh supply of retorts is requested to
replace those which had been destroyed by the rioters. Two
other letters of 1791, one to Josiah, the other to Thomas
Wedgwood, refer to the same subject. The last two of the series
addressed to Thomas Wedgwood (Feb. 25, and March 17^
1792), allude to the same topic, and to some private matters.

Priestley's letters to Keir are of less importance. He
explains to him in two- letters of date 178^^ his reduction of
metallic oxides by hydrogen, when heated by the sun's rays
concentrated by a lens, and his conclusion that phlogiston is
*^ the same thing with inflammable air in a combined state.''
The next letter is of date, Dec. 15, 17^7 y and refers to the acid
produced by the combustion of apparently pure hydrogen and
oxygen. From it we learn the fact, not mentioned elsewhere,
that Priestley at first supposed the acid to be sulphuric* ^^ I
have procured a solution of copper in some acid, which to all
appearance is the same, and I think the vitriolic." In the trials
referred to, the explosions were made in a copper vessel, a por-
tion of which was dissolved by the acid produced. Had Priestley,
however, called to mind. Cavendish's Experiments on Air of
1784 and 1785, he would have tested at once for nitric acid,
which he afterwards found the acid to be. Other allusions to
the appearance of an acid occur of the same nature as those in
the letters to Wedgwood, but less numerous, and eight of the
letters to Keir are not dated, except that some of them have the
day of the week in which they were written, marked on them.
The following is the only additional passage which it seems
necessary to quote. It is from an undated letter : ** That water
is essential to every kind of air, I am now strongly inclined to
believe, having found it to be so in respect to inflammable and
fixed air, and so great a quantity being found on the decompo-
sition of dephlogisticated air. It is probably that which

SINGULAR CHANGES IN PRIESTLEY'S VIEWS. 105

^Tes them their aerial form, and may be called their common
baiisJ^

I have quoted fully from the Wedgwood Correspondence,

because one important influence in occasioning the protraction

of the Water Controversy, has been neglect or misapprehension

of Priestley's views, especially on the part of the advocates of

Watt- The letters extend over a period of ten years, during

which no chemical enquiry interested Priestley more than the

nature of water, and the downfall of the doctrine of phlogiston,

which the alleged compound character of water threatened.

He knew the views of Watt, Cavendish, and Lavoisier ; he was

jealous of none of them, and he never wronged them, except

unintentionally, when he misunderstood them. Watt, moreover,

was his friend. Cavendish only his acquaintance, and Lavoisier

in some respects his rival. Wedgwood also, was a friend of

Watt's, as several references in the correspondence, to pieces

of apparatus sent from Etruria to him, show.* There were

few, therefore, to whom Priestley was more likely to have

expressed his opinion concerning Cavendish's good faith than

to Wedgwood. The absence, accordingly, of the slightest

reference to Watf s English rival as having wronged him, shows

how little importance Priestley attached to the accusation. It

is singular, however, that no reference should occur to the

rivalry between Cavendish, Watt, and Lavoisier, in the writings

of so zealous a friend, and so frank and outspoken a person as

Wedgwood's correspondent; yet his silence is perhaps not

difficult to account for. It was impossible, changing his views

so rapidly and entirely as he did, that he should feel much

interest in defending opinions which he disbelieved. A glance

at the Wedgwood Correspondence will illustrate this. Before it

commenced in 1781^ Priestley had exploded hydrogen and

oxygen, as a random experiment, and attached no importance

to the appearance of water which resulted from the explosion.

He reported, however, hisfriend Warltire to believe that the water

was simply deposited from the oxygen, in which it had preexisted

* This will probably not be dispnted by any one. I may state^ howerer, that
Mr.Francis Wedgwood informs me that his grandfather Josioh was, ''according to his
belief, on Tery friendly terms with James Watt/' and that his nncle Thomas and
his father were on the same footing with Mr. James Watt, junior*

106 LIFE OF CAVENDISH.

ready formed. In 178^9 he Lad demonstrated^ as he thought^ that
water was transmutable into atmospheric air; and that hydrogen
and oxygen can be burned into carbonic acid. In 1783, he
abandons the former view, and annomices on the authority of an
inaccurate repetition of Cavendish's experiments^ that inflam-
mable air and oxygen can be burned entirely into water. In
1784 9 he asserts that these gases may be burned into carbonic
acid and water, but that it is possible to explode hydrogen and
oxygen together^ without obtaining either ; whilst the inflam-
mable gas from heated charcoal yields both. In 1787? Le affirms,
that inflammable air and oxygen always produce nitrous acid
by their combustion ; and that water is essential to the exist-
ence of every gas. In 1788, he urges, in strong terms, what he
had formerly stated less decidedly, that the decomposition of
water is an entire fedlacy ; and finally, he declares in 1791; that
if oxygen be burned with excess of hydrogen, it yields nothing
but pure water ; but if burned with a deficiency of that gas it
affords nitrous acid as well as water. He thus ended in I791j
with the doctrine he had learned from Waltire in 1781, that
water is simply deposited from the gases which yield it when
they bum. He returned to the belief which, perhaps, he had
never abandoned) that water was an element, and that it was
present in every gas, and that nitrous acid was as certain a
product of the combustion of hydrogen as water; nor does it
seem beyond question that he ceased to believe that carbonic
acid may be produced out of the elements of water.

It thus appears, that when Watt's jealousy was most strongly
roused against Cavendish and Lavoisier in the end of 1 783 and the
beginning of 1 784, Priestley had totally lost faith in the experi-
ments of Cavendish, and in his own repetition of them (on which
Watt had founded his theory of the nature of water), as well aa in
Lavoisier's similar researches. Nor did Priestley return to a
modified faith in them till 1790, when, moreover, he affirmed
that inflammable air and oxygen could produce nitric acid, which
Cavendish, and Lavoisier totally disbelieved, and from the first
had denied.

The tenor of the Wedgwood Correspondence is thus entirely
in keeping with the published statements of Priestley ; and it
seems to me in no degree probable that any material evidence

PRIESTLEY'S DEPORTMENT AS UMPIRE. 107

towards the adjustment of the claims of Watt and Cavendish^
has been lost to us in the missing papers of Priestley. He
showed no reluctance to return to the subject in his published
papers, where, had he held a decided view as to the reality and
greatness of the asserted discovery, he would certainly have told
us with his accustomed frankness, what share he thought Caven'*>
dish and Watt had in making it. But if he did not do this in
his familiar letters, it is not surprising that he should write in
his published papers, as if there were no rivalry concerning the
discovery ; and any one who reads the passage I have previously
quoted from his Ewperimerds and Observations relatinff to Air
and Water {IJSS'^), without any other knowledge than it
supplies, would infer, as he would also from the Wedgwood
letters of the same date, that the most cordial harmony reigned
between Cavendish, Watt, and Lavoisier, and that they were
exactly at one in their conclusions. Yet Priestley certainly
knew that Lavoisier was accused by the English diemists of
unfair appropriation of their views ;t nor is it probable that he
can have been ignorant of the feelings entertained towards
Cavendish by Watt. The probability is all the other way. We
may, therefore, safely accept Priestley's published statement X
as all he cared to utter respecting the disputed claims, and it
amounts only to an ascription to the rivals, of identity of belief
as to the cardinal facts, with a difference of expression as to
their significance. Cavendish, Watt, and Lavoisier are thus
treated as independent asserters of the same truth. The hm*
guage in which Watt expresses it is preferred, and the others
are said to concur in it, but priority is implicitly ascribed to
Cavendish, and the later date of Watt's conclusions is marked
by their being stated to have been drawn " from some experi-
ments of which I (Priestley) gave an account to the Society,*'
which are unquestionably the repetition of Cavendish's experi-
ments, contained in the paper of ] 783. Priestley thus declined
to take a side in the Water Controversy, and nevertheless
retained the esteem of the rivals. Nor would it be fair for us
to blame him, because we may imagine that had he boldly said,

* Ante, p. 85.

t Watt Corr., p. 35, and PhiL Trtnu. 1784, p. 134.

t BxperimenlB and Obiervatiom on Air, 1786, p. 71 1 or Phil, TroM, 1785^
p. 279.

108 LIFE OF CAVENDISH*

Cavendish knew thus much in 178I9 Watt thus much in 1783^
we should have been enabled to do entire justice to both. It
should not be forgotten^ in judging Priestley^ that he did not
claim his own share in the discovery o{ the Composition of
Water, nor take any steps to correct the inaccurate account
which Watt gave,* or the defective one sanctioned by Caven-
dish,t of his priority as the observer of the very important
initial fact, that drops of a liquid inferred to be water bedew the
sides of a vessel in which inflammable air and oxygen are
burned. He who was so utterly indiflferent to his own share in
a discovery, could not be expected to concern himself much
about the amount of merit which should be parcelled out among
the three great claimants of it, who had forgotten him, and were
abundantly able, without any help, to defend the right, which
each asserted, to have the lion's share. It was no resentment
of their indifference to him which kept Priestley silent, but the
far more powerful motive, that he thought the lions were fight-
ing for a shadow. It was a delusion, according to him, at the
period when the rivalry was keenest, that water was a compound,
and a mistake that it equalled in weight the burned gases which
produced it. Hydrogen and oxygen always yielded nitric acid,
sometimes carbonic acid, when burned together. Priestley thus,
for his part, desired to cry peccavi^ and no doubt thought that
the great rivals would soon be among the penitents also. He,
at all events, set them the example, and avoided fomenting
what he must have considered a most needless and idle dispute.
Priestley, then, must be acquitted of everything like evasion,
cowardice, or partiality, in his dealings towards the original
disputants in the Water Controversy. His bias must have been
in favour of Watt, but so far was he from giving way to this,
that the latter evidently distrusted him, as the entire trans-
ference of his defence to the hands of De Luc, and the total
absence, in the Water Controversy, of any appeals to Priestley,

♦ Phil, Trans. 1784, p. 332.

t PAit. Tram, 1784, p. 126. Cayendish'B reference if taken along with PriesU
ley's own account of his experiments and of Warltire's, which is adduced as the
anthority for the reference, cannot mislead ; but if read alone, it would certainly oon-
Tey the impression that Warltire, not Priestley, first observed the depo&ition of
moisture to follow the detonation of inflammable air (hydrogen, or one of its com-
pounds) with oxygen.

WATTES DISTRUST OF PRIESTLEY. 109

except as a witness to the authenticity of the letter of 17835
seem to show. An enquiry into the nature and source of this
distrust will materially assist us in deciding how far^ and in what
way, Priestley was to blame for the origin of the Water Contro-
Tersy. Wattes distrust, I believe, had reference solely to the
accuracy of Priestley^s experiments and conclusions, not to his
friendly feelings, which he knew remained unabated. After
learning that he had been led astray by his friend, as to the
transmutability of water into air. Watt withdrew his miire paper,
and as Priestley, before its publication, discovered his error as
to the charcoal-gas yielding on combustion only water, and
declared that hydrogen yielded none, it was vain to quote his
early experiments as justifying Watt's conclusion. When Watt
published his views accordingly, which he did not do till
Cavendish and Lavoisier had put it beyond question, that water,
weight for weight, is the only product of the combustion of hydro*
gen and oxygen, he took care to fortify Priestley's original state-
ments, which the latter was now in private retracting, by adducing
the Parisian repetition of Cavendish's experiments as ^^ clearly
proving " the *^ essential point," that " the deflagration or union
of dephlogisticated and inflammable air, by means of ignition,
produces a quantity of water, equal in weight to the airs ; and
that the water thus produced appeared, by every test, to be pure
water/^ *

Watt was plainly justified in preferring the advocacy of
De Liuc to that of Priestley in 1 784, and the last must to some
slight extent be reprehended for being the innocent cause of
some part of the jealousy of Watt towards Cavendish, and for
having abridged our means of doing justice to either. Had he
told Watt all that Cavendish told him, and told it as Caven-
dish's, the great engineer would never have listened to De
Luc's insinuations against his rival's honour. Had he accu-
rately repeated Cavendish's experiments, moreover, or accu-
rately observed what his own researches brought to light, the
question between the English claimants of the disputed disco-
very would have been greatly simplified, and Lavoisier would
have been altogether excluded from a claim to the initial disco-
very .f It cannot but be regretted, therefore, that the medium

• Phil. IVau. 1784, p. 333. t ^'oiL Corr., p. 34.

110 LIFE OF CAVENDISH.

of commnnication between Cavendish and Watt should have
been Priestley, whose great and peculiar gifts as a scientific
observer, were unsuited for a delicate quantitative investigation,
such as an enquiry into the product of the union^of hydrogen and
oxygen preeminentiy was. He cannot be acquitted, also, of
hasty observation, which led him into strange inaccuracies, nor
of inattention to the views of others, which led him uninten-
tionally to misrepresent them; and the contradictory statements
on the same subject which he printed side by side, witiiout
comment or attempt at reconcilement, imply a peculiarity of
intellectual organisation singularly disqualifying him for the
office of umpire to which he was, without seeking it, preferred.*
I cannot say thus much in depreciation of Priestiey,
whom I honour for capacity, courage, honesty, earnestness,
ingenuity, energy, and almost imequalled industry, without seek-
ing to guard against doing him injustice. It has been one of
the most unfortunate results of the Water Controversy that
it has led to accusations against many besides the principals
involved in it, and I count it no digression to save Priestiey^s
character from being misapprehended in connexion with it. It
should be remembered, then, that Priestiey had seen so much
of the evil of obstinate adherence to opinions, which time had
rendered decrepit, not venerable, and had been so richly rewarded
in his capacity of natural philosopher, by his adventurous explo-
rations of new territories in science, that he unavoidably and un-
consciously over-estimated the value of what was novel, and held
himself free to change his opinions, to an extent not easily sym-
pathised with by minds of a different order. Some men love to
rest in truth, or at least in settled opinions, and are uneasy till they
find repose. Tbey alter their beliefs with great reluctance, and
dread the charge of inconsistency, even in reference to trifling
matters. Priestley, on the other hand, was a * Follower after
Truth,' who delighted in the chase, and was all his life long
pursuing, not resting in it. On all subjects which interested

* Thxis, in the yolume of Erperiments and Observaiiotu on Air for 1786^ he
reprints the statement of 1783, that the inflammable air from charcoal and oxygen
can be bamed into their united weight of pure water, although he had announced to
Wedgwood in 1784, that he had discovered that this was a mistake (as it oertainly
was), and that carbonic add as well as water was a product of the combustion. His
different views, indeed, on water, are irreconcileable.

PRIESTLEY'S PURSUIT OP TRUTH. Ill

him, he held by certain cardinal doctrines, but he left the out-
lines of his systems to be filled up as he gained experience, and
to an extent very few men have done, disavowed any attempt to
reconcile his changing views with each other, or to deprecate
the charge of inconsistency* It is impossible not to admire his
moral courage and candour in frankly acknowledging this, and
his faith in Wisdom justifying herself without his entering on a
defence of his alterations in belief. Nevertheless, I think it
must be acknowledged by all who have studied his writings,
that in his scientific researches, at least, he carried this feeling
too far, and that often, when he had reached a truth, in which
he might and should have rested, his dread of anything like a
too hasty stereotyping of a supposed discovery, induced him to
welcome whatever seemed to justify him in renewing the pursuit
of truth, and thus led him often completely astray. Priestley,
indeed, missed many a discovery, the clue to which was in his
hands, and in his alone, by not knowing where and when to
stop. This peculiarity, however, is not more manifest in his
researches into the true nature of water, than in his other sci-
entific investigations, and no special blame belongs to him for
his errors in connexion with the former.

Further, from the most praiseworthy motives, Priestley never
concealed a discovery till he had completely realized its cer-
tainty, or hoarded a truth, content like Cavendish to have learned
it for himself. As one who knew him well* has justly said,
he was characterised by ^' extreme ingenuousness of character
and honesty of purpose. Whenever he discovered a new fact
in science, he instantly proclaimed it to the world, in order that
other minds might be employed upon it besides his own/^
Priestley greatly furthered the progress of chemistry by this
speedy publication of his discoveries, but it unavoidably led
him to make known unfinished researches, and involved him in
contradictions which a less liberal and more cautious mode of
procedure would have prevented. We must set off theimmensje
service he rendered science by the free* and rapid publication
of his ingenious observations, against the harm he occasionally
did by premature announcement of supposed discoveries.

* The late T. L. Hawkes, of Birmingham > Dr. RoBBeU baf famished me with
Mr. Hawket* judgment of Priestley's character.

112 LIFE OP CAVENDISH.

I must also plead in defence of Priestley^ a very curious
mental peculiarity of his, to which he has drawn attention in
his autobiography, as it will assist us in explaining his behaviour
in the Water Controversy. Priestley^s words are as follows :—
'^ As I have not failed to attend to the phenomena of my own
mind, as well as those of other parts of nature, I have not been
insensible to some great defects, as well as some advantages
attending its constitution; having from an early period been
subject to a most humbling failure of recollection, so that I
have sometimes lost all ideas of both persons and things that I
had been conversant with. I have so completely forgotten
what I have myself published, that in reading my own writings^
what I find in them often appears perfectly new to me, and
I have more than once made experiments, the results of which
had been published by me/'*

To this psychological characteristic of Priestley I refer the
more confidently that my attention was drawn to it by his grand-
daughters, as a well-known peculiarity of their ancestor, without
any reference to its bearing upon the Water Controversy. Dr.
Russell, of Birmingham, also informs me that Priestley^s in-
timate friend and associate, Mr. Thomas Lakin Hawkes, bore
testimony to his " lack of memory, which he ascribed to his great
mental activity .'' This appears a true theory of what was a source
of strength, as well as of weakness to Priestley. To be able utterly
to forget what had shortly before entirely occupied his thoughts
was in one respect the best possible preparation for a new
research. And those who, like myself, have read with astonish-
ment, the mere list of Priestley's varied publications on nearly
every branch of human enquiry, will find in his '^ most humbling
failure of recollection,^' one key to the success with which he
prosecuted so many and so diverse enquiries.*!*

It was also, however, a source of weakness, and probably
furnishes the explanation of some of the transactions in the
Water Controversy, which appear at first sight so inexplicable.

* Life qf Priestley, Centenary edition, p. 74, where a remarkable example of
total failure of recollection is given.

t Priestley's case was not singular. I have heard it remarked by a competent
judge, that many of our most successful barristers owed their success to their power
of completely dismissing from thdr thoughts a case in which they had been engaged
as soon as it had been decided*

PRIESTLEY'S TREACHEROUS MEMORY. 113

I have no special point to prove by referring to ,' it ; but it
seems to me, that the possession of a memory so treacherous as
Priestley's was^ may perhaps explain how it happened that
Watt was unaware, as he seems in 1783 to have been, that the
experiments on which he based his theory of the composition
of water, were repetitions of those of Cavendish, and, therefore,
why he omitted all reference to the latter's priority ; and further,
it may throw some light on the existence in Priestley's paper
of 1783, of the incredible description given there of the results
of the combustion of charcoal gas and oxygen, and in general,
may help to account for the many opposite theories of the
nature of water, which Priestley published and republished
without hinting at their contradictory character, much less
seeking to remove it. One who could totally forget his own
experiments, must have been at least equally liable to forget
those of others, and when he substituted charcoal gas for
hydrogen, in repeating Cavendish's experiments, it may have
been, not as a supposed improvement, but in entire forgetfulness
that hydrogen was the gas he was desired to employ. At all
events, Priestley's singular obliviousness must be taken into
consideration by all who are aware of the difficulty of recon-
ciling his dealings towards Cavendish and Watt, with the only
tenable hypothesis of his character, viz., that he was a very
honourable, ingenuous, and truthful person, the possessor of a
most active, versatile intellect, which education and unceasing
exercise had trained to a rare degree of acutencss.

Having thus disposed of the share which Priestley took in
the Water Controversy, after he was appealed to by Cavendish
in the first of the three passages which be added to his paper of
1784, before it was printed, I proceed to the consideration of
the second, so-called interpolation which has led to much dis-
cussion, both in reference to its contents and to its writer. Like
the first interpolation, it is in the handwriting of Sir Charles
Blagden, but the important part which he took in the controversy
is best reserved for later discussion, and as the added passage
went forth to the world as an authorised expression of Caven-
dish's sentiments, and formed an integral part of his published
paper, it is of much more importance to determine what its
significance is, than to settle why it was penned by Cavendish's

114 LinS OF CAYSNDISH.

amanuensis^ Blagden^ and not by himself. It forms one amon<;
the personal incidents of the Water Controversy^ inasmuch as it
contains the only direct reference iirhich Cavendish made to
Watt throughout the discussion. It is as follows : — ^^ As Mr.
Watt, in a paper lately read before this Society, supposes water
to consist of dephlogisticated air and phlogiston deprived of
part of their latent heat^ whereas I take no notice of the latter
circumstance, it may be proper to mention, in a few words, the
reason of this apparent difference between us* If there be any
such thing as elementary heat^ it must be allowed that what
Mr. Watt says is true.*^' Cavendish then proceeds to explain
(as he had already announced to the world in 1783) that he is
not a believer in elementary heat; and that even if he were, he
should not think it necessary to insist upon its evolution accom-
panying the union of hydrogen and oxygen, as of more import-
ance than its development during other chemical combinations.
The passage is elsewhere commented on, so far as the doctrine
it teaches is concerned, I consider it here only in its character
of a personal reference by the one rival to the other.

The advocates of Watt read this passage by the light of the
suspicions which they inherit from De Luc^ and some (Sir
David Brewster and Mr. Muirhead, amongst others) have
gone the length of affirming that it contains an implicit acknow-
ledgment of Watt's priority, and a confession of the wrong his
rival had done him in not mentioning his name in the original
draft of the Eaperimenta on Air, communicated to the Royal
Society on January 15th, 17B4. Yet certainly his words do
not convey this meaning, although those who contend for it
rest their argument solely on the words. These make no allu-
sion to obligation ; they make no acknowledgment of priority;
they do not even assert identity of opinion^ but are chiefly occu-
pied with the assertion and defence of a difference in the doctrine
which Cavendish taught. Watt, a pupil of the school of Black,
was a firm believer in the materiality of heat, and attached
great importance to its evolution when water was produced from
its elements. Cavendish was of the school of Newton, and
regarded heat as immaterial, neither did he think that there

* PUl, Tram, 1784> p. 140. The passage is enclosed between square brackete
in Mr. Mvirhead's reprint. FFa//. Cbrr., p. 135,

caybndish'^ liberal treatment of watt. 115

ir&8 anything specnally remarkable in its evolution when hydrogen
and oxygen united, in which opinion the chemists of the present
day are universally at one with him. But when he found Watt
insisting on its special importance in reference to the produc*
tion of water, he added a passage to his paper to show that he
had deUberately, not through ignorance, avoided enlarging
on this.*

In this commentary Cavendish treats Watf s doctrine as being
generally the same as his own, a fact on which the advocates of
the latter build much, but we shall presently find, that though
the statement by the rivals of their doctrines was substantially
identical, there was an important difference indicated by Caven-
dish in his first interpolation, in their use of the termpUogUton
or inflammable air, to denote the combustible element of water.
This does not, however, concern us at present. What is
important is the information which the passage quoted gives us
as to the anitmts of Cavendish towards Watt. It is creditable
to him in the highest degree. The advocates of Watt, when
they reproach Cavendish for so tardy a reference, as they count
it, to their client's views, contrive most singularly to forget two
things. Firstly, that Watt had made it impossible for any one
(except Priestley, to whom it was addressed) to quote from his
letter of 1783. as containing his views on the composition of
water, by withdrawing it from public notice, so that it was
rendered b private letter. If any reason were given at the time
tar this withdrawal, it must have been the same as that announced
more recently, namely, its author's loss of faith in certain of the
conclusions contained in it, whilst Cavendish was entirely igno«
lant what the conclusions were which Watt had abandoned, so
that for anything he knew, or could know, they were the very
opinions which, in 1784, were reclaimed as never having been
distrusted. Secondly* — The advocates of Watt have still more
strangely forgotten, that in neither of his letters did he make
the slightest acknowledgment of his obligation through Priestley
to Cavendish, for the experimental data on which his conclu-
sions were founded. He set the example (perhaps unconsciously)

* In the preceding chapter, and in the abstracts of his papers on heat, it is
shown that Cavendiah held this doctrine before the Water Controversy arose. See
his paper on the freesing of mercnrj in Phil, Draw. 1783.

I 2

116 LIFE OF CAVENDISH.

of injustice to his rival (if injustice there were on either side).
Of deliberate injustice I have already acquitted Watt^ and I
am willing to believe that he had forgotten, or was ignorant of
Cavendishes experiments having been made in 1781. I shall
further concede, for argument's sake, that Priestley, whose share
in the matter it is so difficult to determine, was alone responsible
for Watt's ignorance of Cavendish's researches ; and further^
that the latter were, as nearly as possible, mere observations of
phenomena, from which no conclusions were drawn. Neverthe-
less, Cavendish had first observed that equality of weight
between the gases burned and the water produced, without the
certainty of which Watt's conclusions were, as he acknowledged,
baseless. The former, therefore, on the lowest estimate of his
deserts, might well feel astonished that his name was utterly
passed over in Watt's letter of 1783, and that the truth which he
had taught Priestley was referred to, as if discovered by his
f)upil, whilst that pupil freely acknowledged that he had learned
it from Cavendish. This omission by Watt might, with great
show of justice, have been set down by Cavendish as a wilful
and imfair disregard of his claims. De Luc, we have seen^
without the slightest enquiry, gave vent to gratuitous suspicions
against his friend's rival. Had Cavendish or Blagden copied
his evil example, he might have pleaded not surmises, but the fact
confessed by Watt, that he was indebted to Priestley for his
experimental data, and could have confidently asked if it were
very credible that Priestley's intimate friend and fellow-towns-
.man could be ignorant of the truth, so willingly published by
the former, that he had only imitated Cavendish. A primd facie
mew of matters would certainly render it probable that Watt
had deliberately concealed what Priestley told him.

When Watt, therefore, in April 1783, addressed to the
Royal Society a document expounding his views on the nature
of water, and totally omitting Cavendish's name, he showed a
disregard of the claims of the latter to at least the experiments
which might have been pleaded in justification of a corresponding
omission of all reference to him. At the present day, with our
perfect assurance of the unimpeacheable integrity of Watt, and
our knowledge of the contradictory and perplexing dealings of
Priestley, we may find nothing strange in the silence of the

CAVENDlSirS LIBERAL TREATMENT OF WATT. 117

former* But Cavendish might well wonder^ that when every
one else knew that Priestley had only copied his experiments^
Watt alone^ who took such an interest in the copy, should refer
to it as an original ; and had Cavendish or his friends been
found looking suspiciously at Watt^ we must have conceded that
they had grounds for their suspicion. In saying this I pur-
posely take the lowest possible ground for Cavendish, and assert
that even had he borrowed his conclusion from Watt without
acknowledgment, he could plead that the latter had borrowed
his facts without acknowledgment from him, so that Cavendish
had only followed a bad example, and been guilty of reprisals.

But such low ground I gladly leave. Cavendish needs no
such defence. We have hitherto looked at his dealings towards
his rival, through the distorting medium of De Luc's suspicions,
which the advocates of Watt would have us believe, supply a
fiftultless magnifying-glass for discerning his virtues and Caven-
dish's transgressions* The position, however, which Cavendish
himself assumes in the interpolation under notice, is that of one
who had done Watt no wrong, and owed him nothing, and from
this point of view I now look at his dealings. And plainly, if it
were the case, that Cavendish had interpreted his experiments
and had come to the conclusion which he ultimately connected
with them, before he reported them to Priestley, he owed Watt
less than nothing. Because Watt hastened, in 1 783, to publish his
interpretation of phenomena which were known to both, whilst
Cavendish preferred to prosecute his researches before he made
any part of them public, there certainly lay no claim on Cavendish
to refer to Watt, when he formally made known to the
world what he had discovered for himself* No one will
assert that two independent discoverers of the same truth are
required by justice (although they may be by courtesy and
good feeling) to expatiate on each other's merits. And in the
particular case of rivalry before us. Cavendish knew that he had
long preceded Watt (however ignorant Watt might be of the
&ct), and could appeal to Priestley as a witness to the truth,
that he had demonstrated to him that hydrogen and oxygen could
be converted or compounded into water, before the possibility
of such an occurrence had become known to Watt. The latter
only asserted this truth in other words and later in time. Will

118 LIFE OF CAVENDISH.

the most exacting critic say that Cavendish was even in cour-
tesy^ required to point out^ that another had later than himself
drawn a conclusion identical with his, from similar experiments ?
But this is not all. It was Cavendish's very experiments on
which Watt was really, however unconsciously, founding his
theory ; for all that seemed true in Priestley's avowed repetition,
was untrue as regarded it, and true only as connected with the
accurate original trials.

However ignorant Watt might be of this, Cavendish was
well aware of it : he might have asked at the hands of those who
were so sensitively jealous of Watt's rights, whether he had
none, and with what show of justice they reprehended him for
not acknowledging that Watt, who omitted all reference to
him, should have it punctiliously conceded that he had drawn
from experiments performed by his rival (which had reached him
half-interpreted, at least, through Priestley), conclusions which
Cavendish, their original performer, had from the first taught as
inevitable consequences of the truths he had discovered. If the
advocates of Watt should feel tempted to dispute the relevancy
of this argument, their own client would shut their mouths.
Watt had taken his case out of court, and forbidden any one to
offer an opinion upon it. His letter of 1783 he had suddenly
withdrawn (he did not at the time say why), and he certainly
-would have reproached any one who, after its withdrawal, had
imputed to him the belief that water was transmutable into
atmospheric air ; nor Was there any clue by which it could have
been discovered by those not on terms of intimacy with him, that
he still held faith in the convertibility of inflammable air and
oxygen into water. The letter of 1783, indeed, which Watt's
advocates speak of somewhat magniloquently, as having been
deposited "in the archives of the Royal Society," was at one
time in the hands of Priestley, at another in those of Sir Joseph
Banks, and had actually to be reclaimed firom those of De Luc,
in 1784, before it could be read publicly ; nor would Sir Joseph
read it, as we have seen, till he was a second time formally
requested to do so by Watt. It is most idle in these drcum-
stances to affirm that Cavendish or any one else was bound by
a document, which was half-public, half-private, which its writer
partly retracted, partly avowed, without informing any but his

cavendish's liberal treatment of watt. 119

intimate firiends, to what extent^ and for what reason^ he dis-
believed some portions of it, and hesitated to publish others ;
which when ultimately reclaimed was neither in the hands of
Dr. Priestley^ to whom it was addressed, nor in those of the
president, or secretaries, or council of the Royal Society, who
were its official guardians if it were consigned to the keeping of
the society, but in those of De Luc, who had no apparent right
to it, and from whom the president had to demand it.

It thus appears that Watt had put it out of the power of
Cavendish, however much he had wished to do so, to refer to
the opinions expressed in the private letter of 1783, and that
Cavendish is, therefore, altogether blameless for taking no notice
of Watt in the first draft of his paper of 1 784. Further, it appears
that Watfs conclusion was in statement substantially identi-
cal with Cavendish's; was drawn from his experiments; and was
formed later in time than his inference that hydrogen and oxygen
are convertible, weight for weight, into water. So that Watt only
re-announced at a later date (and not accurately) the views which
Cavendish had already taught Priestley, and this without the
slightest reference to the former. Cavendish, accordingly, so far
from being blameworthy, for not praising Watt, might with
justice have taxed the latter with wronging him, and totally
omitting his name, although (as all but the extreme partizans of
Watt acknowledge) he was at least entitled to the whole merit
of the experiments, whoever might claim the conclusions ; and
had Priestley^s imsolicited testimony to prove that the latter
were substantially his also.

Thus far, then, Cavendish's deportment to Watt was irre-
proachable ; nor was it till the latter published his views that
they could be criticised. He did this, as we have seen, in April
1784, and Cavendish commented on them in the first two addi*
lions made to his paper. The spirit in which he did so was
altogether commendable. There can be little question that one
object of the first interpolation was to counteract the impression
which Watt's letters unavoidably conveyed, that Priestley's expe-
riments were original, and to explain the doctrine which Caven*
4ish had taught him and Lavoisier.

This first interpolation, as well as the second, contains the only
expression of opinion which Cavendish ever published, and so

120 LIFE OF CAVENDISH.

far as is known^ ever uttered^ concerning Watfs claims and
opinions^ as contrasted with his own. He might have reproached
his rival for doing him injustice, or have passed over his
name in silence without wronging him. He did neither; but
accepting Watt's claim to be considered as an independent
teacher of a theory of the composition of water, he pointedly
drew attention, not only to the priority of his own views,
but to the connexion of Priestley and Watt's conclusions
with experiments made, not with hydrogen, as his were, but
with charcoal gas. He made no reference to Watt's first letter
of 1783, but commented solely on his paper of 1784, which con*
sisted of two letters and several additions and alterations of dif-
ferent dates in 1783 and 1784, and he did so solely to express
dissent from certain of its views. Cavendish's total avoidance
of any complaint of wrong done him, his temperate simplicity in
urging his own claims, and his utter indifference to personalities,
were entirely in keeping with his character. Watt appears in
unfavourable contrast with Cavendish in his treatment of his
rival. So long as he omitted all notice of him, he might be
acquitted of intentional wrong, but when he made a deliberate
reference to Cavendish's researches, he should have been care-
ful to give an accurate account of them. Yet I should be slow
to blame so generous a man as Watt, and would willingly make
large allowance for the effects of the sinister influence which the
mischief-making De Luc was exerting upon his candid spirit,
when, smarting under the sense of a supposed wrong done him
by his rival, he referred to Cavendish's researches. Watt's own
advocates, however, have selected his scanty and inaccurate
reference to his rival, as something which erred on the side
of praise, and have left me no choice but to point out how
unwise and unjustifiable this laudation is. The reference by
Watt to Cavendish, is contained in a note to his paper of
1784.* The latest date attached to the paper is November
26th, 1783, but the note was not written then, and probably
not till March 1784. Mr. Muirhead informs us ^' that the note
is not in Mr. Watt's original draft, nor in the press copy of the
letter in his own writing, sent to M. De Luc, of 26th Novem-

* P7«7. Traru. 1784, p. 332-

watt's inaccurate reference to cavendish. 121

ber, 1783; but is added at the bottom in pencil, in his own
hand/^* It seems probable, accordingly, that it was written by
Watt when he visited London in the end of March or the begin*
ning of April 1784, and had communicated withDe Luc, in whose
possession the letters, afterwards published, were. At all events,
it seems certain that the pencilled note formed part of Watt's
paper when it was read to the Royal Society, for he had no access
to his MS. after it was read and had been consigned to the
secretary of the Royal Society .f

The point is of some importance, as affecting the question
whether Cavendish was aware of the existence of this note, when
he wrote the two interpolations already considered. If it formed
part of the paper when read, as it seems unquestionably to have
done, he could not be ignorant of its contents ; and it left him no
choice but to add such passages before his paper was printed, as he
did in the two first interpolations. It was thus an interpolation
of Wattes, which led to the interpolations of Cavendish. The
note is as follows, ^^ I believe that Mr. Cavendish was the first
who discovered that the combustion of dephlogisticated and
inflammable air produced moisture on the sides of the glass in
which they were fired ;^'j: and Mr. Muirhead thus comments
upon it, *^Mr. Watt inadvertently stated, that he believed Mr.
Cavendish was the first who observed the dewy deposit ; thereby
assigning to him too much merit in place of too littleJ*'§ Watt
was certainly guilty of great inadvertence in writing thus. We
may not doubt that his ignorance was unfeigned, but we can
unhesitatingly affirm that it was wilful ; and that the error in
his notice of his rival's doings did not lie in its overpraise. In
truth. Watt's ignorance of the proceedings, not only of Caven-
dish, whom he counted his enemy, but of Priestley, whom he
knew to be his friend, adds another to the many perplexities of
the Water Controversy. Cavendish never pretended to have
been the first to observe the " dewy deposit,'^ but imputed that
observation to Warltire and Priestley; and referred to the
latter's Eafperiments and Observations on Air (1781), as con-
taining the earliest account which was known to him, of
the appearance of mobture succeeding the combustion of

• Wait Corr,, p. xxxiv. t ^^^- Cwrr.^ p, 59-61.

X PhiU Trwu. 1784, p. 332. § Watt Corr.t p. xxxir.

122 LIFE OF CAVENDISH.

inflammable air and oxygen. It is almost inexplicable that
Watt should have been worse informed on this point than
Cavendish. The latter quoted from a volume printed in 1781
at Birmingham, where Watt resided, written by his friend
Priestley, and containing an account of experiments by his
fellow-townsman Warltire; and yet it should seem that Watt
was ignorant of its contents in 17B4, and was utterly unaware
through private channels, although in constant communication
with Priestley, that he had preceded Cavendish in observing
the phenomenon in question. This is certainly strange, but our
wonder is increased when we consider that Watt had other means
of information open to him, than the volume of experiments and
observations of 1781. Cavendish had referred his readers, in
his paper of 1784, to that work, and had quoted from it the
observations of Priestley and Warltire ; and De Luc, as we have
already seen, made an analysis of the MS. of Cavendishes paper
for Watt, and sent it to him. Yet this analysis, if it were an
accurate one, must have included some reference to Warltire
and Priestley, and should have saved Watt from robbing them
of an honour which was repudiated by Cavendish, to whom he
unjustly transferred it. And further, the printing of Watt's
paper was not commenced till August 1784,* up to which
period its author was at liberty to suggest any alteration
which he thought proper. It seems most unlikely, however,
that before this period, he should not have read Priestley's
paper On Phlogiston^ and the seeming Conversion of Water
into Air, published in the Philosophical Transactions for 178S.
This was the paper on which Watt's first letter was a
commentary, and which it originally accompanied. It con-
tained many references to himself, and with the greater part
of its contents he was well acquainted. How then, did he
continue to be ignorant of the fact, that it contained an
account of Cavendish's experiments, in which it was im-
plicitly asserted, — not that be had first observed the 'Mewy
deposit," an observation which Priestley had already claimed
for himself and Warltire, — but the much more important,
and indeed, cardinal truth, that in certain proportions, a
given weight of hydrogen and oxygen may be converted into
the same weight of water ?

* Watt, Corr,, p. 68.

watt's injustice to CAVBNDISH. 123

There is something much more inexplicable in Watt's
ignorance on the points referred to, than in anything attaching,
on any hypothesis, to the dealings of Cavendish or Blagden.
Nor can the blame be transferred from Watt to any one else.
He stands self-condemned. The note appended to his paper
shows, that he had become aware that, to some extent. Caven-
dish had preceded him, and the contents of the note profess to
exhibit to what extent he had been anticipated by his rival.
Its great inaccuracy, and above all, its injustice to Priestley,
I accept as evidences of its having been the honest expression
of its writer's notions at the time when it was written, of Caven-
dish's share in the disputed discovery. But that Watt took no
pains to ascertain the truth, and chose to hazard a surmise,
rather than investigate the matter in dispute, is most manifest ;
for Cavendish's MS. or De Luc's Abstract, or Priestley's Papers,
or Priestley himself, or Blagden, or Sir Joseph Banks, or Maty
and Planta, the Secretaries of the Royal Society, or Kirwan,
besides others, could have prevented him from committing the
great mistake of which he was guilty. I dwell upon this, not in
order to show how idle is Mr. Muirhead's declaration that Watt
assigns Cavendish too much merit, but because it demonstrates
beyond all question, that Watt was incapacitated from doing
Cavendish justice by his wilful ignorance of the latter's proceed-
ings, and thus yielded at once to the insinuations of De Luc, who
was equally ill-informed concerning them.

I need not further urge that Watt's interpolation in reference
to his rival, places him in a less favourable light than Caven-
dish's interpolations in reference to him. I could, however,
say more. Cavendish might with great justice have reproached
Watt for reaping where he had not sown, and gathering an
unripe crop which was not his. whilst its true proprietor waited
only for its full ripening to celebrate his harvest* He might
have asked whether it was either just, or generous, that another,
made acquainted with his earlier researches, should step in
between him and their completion, and Watt could have de-
fended himself against the charge only by pleading his total
and extraordinary ignorance of all that his rival had done, and
all that his friend Priestley had written concerning his doings.
That Cavendish preferred no such charge is to his credit, for
had he entertained one tithe of the suspicions towards Watt, that

124 LIFE OF CAVENDISH,

Watt entertained towards him^ he could have justified his
calumniations of his rival by tenfold more numerous and more
plausible grounds for his jealousy, than De Luc or Watt did or
could show for their suspicions of him, and he might most
justly have resented in terms of the strongest indignation, the
account of his researches which Watt published to the world*
It is Cavendish, not Watt, that must be commended for a liberal
and generous interpretation of the dealings of his rival.

These illustrious men appear no more in our pages in con-
tention with each other, and the claims of the great French
rival of both will be considered latest of all. Before, however^
he can be referred to, the dealings of the third aider and abettor
in the Water Controversy, Dr« Blagden, must be considered.
He is alleged to have assisted Cavendish in wronging Watt,
and he was umpire between them and Lavoisier, against whom
he gave his judgment, so that the position he occupies is an
important one on any view of the Water Controversy, and is
rendered additionally so by the unsparing denunciations which
have been heaped upon him by the friends of Watt since the
revival of the Controversy.

Dr. Blagden was Cavendish's assistant in his scientific in-
vestigations, and acted also as his amanuensis. He penned, as
"we have already seen, two of the interpolations in Cavendish's
paper.* In May 1784, he was appointed one of the Secretaries
of the Royal Society, and in this capacity superintended the
printing of Cavendish and Watt's papers, in which certain
errors of date were permitted to occur, for which he was more
or less responsible. It was through him, as Lavoisier acknow-*
ledged, that he became aware of Cavendish's experiments on
the combustion of hydrogen and oxygen, and he wrote a letter
to Crell, in 1786, accusing Lavoisier of plagiarism, and referring
to Watt's theory of the composition of water, as similar to Caven-
dish's.f The special charges preferred against Blagden shall be
minutely considered in the detailed analysis of the Water Con-
troversy. It will suffice, therefore, here to state, that the advo-
cates of Watt have in different degrees accused Blagden of

* The third, which was a postscript in Cavendish's handwriting referring to
LaToisier, will be noticed in the sequel.

t Chemisehe Annalen, 1786, pp. 58 — 61: Translated by Mr. Mnirhead, Wait
Corr,, p. 71.

ACCUSATIONS AGAINST BLAGDEN. 125

officious intermeddling between Cavendish and Watt; ofliayiilg
done something unfair in writing the interpolations in the manu-
script of the former ; of haying been in a condition to affirm
whether Cavendish or Watt first formed a theory of the compo*
sition of water, and of having kept silence, to the disadvantage of
the latter; of having, with most culpable carelessness (if his fault
was not greater), suffered errors to occur in the papers of Caven*
dish and Watt, which gave the former a fictitious priority ; and of
having done Lavoisier some injustice^ in the report which he gave
of his experiments^ made in imitation of Cavendish's, in 1783.
These are formidable charges^ and they are urged, though
by no means in equal degree, by Arago, Sir David Brewster,
Lord Brougham, Lord Jeffirey, Mr. James Watt, junior, and Mr.
Muirhead ; and in such a shape that they more or less inculpate
Cavendish. The asserted, or suggested motive, to all these
false dealings on the part of Blagden^ was the salary which Caven*
dish paid him^ and the legacy which he bequeathed to him.
Although, accordingly, it is not incumbent on me, as the biogra*
pher of Cavendish, to enquire into the character of Blagden, it is
requisite to ascertain whether the good name of the former is tar*
nished by his connexion with the latter ; and to demonstrate, if
possible, either that Cavendish was not responsible for the
transgressions of Blagden ; or still better, that the latter was
not a transgressor, and involved neither himself nor his principal
in responsibility or blame. The latter is the proposition which
I hope to prove, and I will commence the proof by observing,
that the advocates of Watt have said either too little or too
much in reference to Blagden. If he was guilty of even a part
of the offences laid to his charge, he should have been held up
to the universal scorn of mankind, as a treacherous and venal
deceiver, who for base lucre was guilty of crimes which, had they
been amenable to a court of justice, would have been styled
falsehood, perjury, and forgery* And in that case Cavendish
should have been openly denounced also, as having connived at
crimes committed to serve him, and as standing in the position
of the wealthy receiver of stolen goods, who bribed the poor
thief to steal for him. Blagden might have pleaded with Shaks*
peare's apothecary in Romeo and Juliei, *^ My poverty, but not
my will| consents ;'' but what could the millionaire philosopher

126 UFE OP CAYENDISH.

plead in his defence ? He was the greater transgressor of die
two, and added cowardice to his other crimes. These charges
the advocates of Watt hare not in so many words preferred
against Cavendish and Blagden, and yet if their accusations were
jastifiable, they should have taken this shape.

If, on the other hand. Watt's defenders had no other evidence
to adduce, condemnatory of Blagden,than that which they have
adduced, they have said greatly too much in depreciation of his
good name, and their fault is the greater, that they generally
avoid precise charges, and leave hanging over the heads of
Blagden and Cavendish a dark ill-defined cloud of suspicion,
which from the standing-point of their detractors, looks black as
midnight, whilst it eludes the grasp of those who seek to show that
it is but a mist created by prejudice, and that the only cloud
over Cavendish's head is a halo of glory.* It is time that these
charges should be made definite, and that the questions should
be asked and answered categorically : 1* Was Blagden guilty of
the offences laid to his charge ? 2. Was Cavendish conscious
of Blagden's guilt ? If the first question be answered in the
negative, the second will not require consideration. As essen-
tial to this, I offer the following sketch of Blagden's life and
character, which I have taken some pains to render as accurate
as possible. No detailed notice of him, so far as I am aware, has
appeared in any English work, probably in consequence of his
death having occurred at an advanced age in Paris. The
esteem, however, in which he was held by the French savans,
led to a short sketch of him being published in the Moniteur for
September 22nd, 1820^ page 1296. This was written by M.
Jomard,t and through the kindness of M. F. Deleesert, of Paris,

1 A- *j ^^ ^ concluding lectioni of the Water Controwty, the opinioiis of the
fadiTiduid advocate, of Watt are referred to Mparately; snd in the action entided
Bibhography, the reader wttl find the mean, of eonsulUng their wriUngs, from which
I do not quote here any particular page., because Blagden figures .0 prominently in
SrrL'alLST r •'' ^«»»ldo»ly-rite^a«magain.tthereferen^tohi. name;
m^' ^o^r K ' T"^' "^'^ ''^^'^ •«^'*"* °f ^ '^^^^ ^ consulted. I

S:^o?^re^!S^\r *'*'.'"^' ^"^^'"^ '^ -^' i""'^ "^^ Tery^^y.
pomted «ferenrr<L^::r"l" ^^^^"^ "^'^ "^'f^ ^^^^^^ ^^ Watt, make

»ect the« with hi. .in! of ^ ' ^^"""^ ^'"'" *^ ^^"^^ "^ significantly con-
t Jomard omiMion or comminion.

lUhedbythePrenlG^^^tL^^^^ editor, of X« 2).«.n>«o« d'Effypte, pnb.

^tiremment after the return of Bonaparte'. Egyptian expedition.

MEMOIR OF fiJLAGDEN. 127

1 have obtained a copy of it, as well as of the inscription on
Blagden^s tomb in Pere la Chaise^ which is of some importance
in reference to dates. By R. H. Blagden Hale, Esq., also, of
Cottles, Melksham, Wiltshire, the nephew and executor of Sir
Charles Blagden, I have been favoured with much interesting
information. The life of Blagden was destitute of any very
remarkable incidents, and his character is of more importance
to my present purpose, than his personal career. I have applied,
accordingly, to all those who were acquainted with him, to whom
I had access, for information as to his reputation in the eyes of
bis contemporaries. By Robert Brown, Esq., of the British
Museum, and Dr. Thomas Thomson, of Glasgow, who were
friends of Blagden, I have been most kindly favoured with
their judgment regarding him ; and with Mr. W. A. Cadell, who
knew him, I have had many conversations concerning his
character. Their opinions are given in the sequel.

Charles Blagden was born at Wooton under Edge, in Glou-
cestershire, in April 1748. The day of his birth I have not
learned, but Mr. Blagden Hale informs me that he was baptised
on the 19th of the month. He had not the advantage in youth
of attending a public school or either of the universities. He
acquired, however, a considerable acquaintance with languages,
for which he must have been mainly indebted to his own exertions,
for he was early educated for the profession of medicine. He
practised this, for some time privately, but about 1776, he received
a medical appointment in the army, and went to America to do
duty with the troops.* He returned to this country, as Mr.
Blagden Hale informs me, ^^ about the latter end of 1779, or

He had aoeess to the hest sonroes of information. M. T. Delessert writes to me,
'* Sir Charles Blagden ^tait tris li^ avee ma famille, oii 11 ^tait re^a presque jonmelle-
Bent pendant ses stfjonrs ^ Paris, et en particnlier avec fen mon fr^re, M. Benjamin
Delessert* C'est mon frere qui ayait recueilli nne grande partie des renseignemens qui
ont aide ^ faire la notice r6dig^ par M. Jomard, membre de I'Institut (encore viyant
[1849])> qui avait anssi 6tk ]i6 ayec Sir C. Blagden." A memoir of B. Delessert,
who was a remarkable man in many ways, appeared in ib» Journal desDibaU, 1850.

* Jomard refers to Blagden as having been *' m^edn en chef dans les arm^,**
and Mr. R. H. Blagden Hale styles him '* Physician to the Forces." If I mistake not
thfi K^ titles are applied at the present day, only to the senior medical officers of the
army, who occnpy the highest posts. So far as I can discover, Blagden's original
position was what we should now indicate by saying that he was an army torgeon.

128 LIFE OF CAVENDISH.

beginning of 1780/' and held an appointment at of near Ply-
mouth, which he resigned after filling it a short time, and with
it his connexion with the army. Soon after this he became
associated with Cavendish, who settled an annuity upon him.
When this connexion began does not exactly appear. Mr. R.
Brown, however, informs me that Blagden was in Plymouth up
to 1782, and that a letter from him in his possession, of that
date, shows that he had not then become Cavendish's assistant*
This is a point of some importance, for Cavendishes experiments
on the production of water from its elements were made
in 1781, and cannot have been witnessed by Blagden. In
June 1783, however, he visited Paris and gave an account of
them to Lavoisier,* on the authority of Cavendish; and from a
reference in his letter to Crell, published in 1789, it appears
that he was acquainted with Cavendish's views in the spring of
1783, so that he must have become his assistant either in 1782^
or early in 1783. He continued with Cavendish for several
years, but they ultimately parted, at what period I do not know»
but probably not before 1789. Among the papers entrusted to
me by Lord Burlington, I find three letters of Blagden^s to
Cavendish, dated Dover, 1787> and couched in very friendly
terms, implying no cessation of good feeling up to that period.
These letters refer to the conferences of the English and French
commissioners in connexion with the trigonometrical surveys
of the two kingdoms. Blagden accompanied General Roy, and
Cassini and Legendre are alluded to as the most prominent
parties on the French side. From the tone of these letters I
gather that Cavendish and Blagden did not part before 1787.
In one of them, for example (September 23rd, 1787), Blagden
writes, " Be so good as to open and read, or get read, any letters
that you think may contain news.^^ The letters here referred to
must have been in German or some other language unknown to
Cavendish, and it should seem that they were addressed to Blagden
at Cavendish's residence, so that they were certain to c me under
his notice. Among the Cavendish MSS. also, is a parcel of papers
entifled « Journeys,*' referred to in the preceding chapter, one

iVftJlvl*! ^'^^i^^^'.P' ^^^' ^""'* Ch^Uche Annalen, 1786, pp. 58.61;
l£!ZTm! ^ S^^e.pour 1781, p. 472 ; Waii. O^., JJ. 39, Tl!

BLAGDEN MADE SECRETARY OF THE ROYAL SOCIETY. 129

portion of which is a journal in Caven dish's handwriting,
entitled **Dr. BL journey, 1789," and occupied solely with an
account of the geology of the districts passed through, in a six
weets* tour through Flanders and a part of Germany. Blagden
appears to have furnished a report of the geology of the territory
in question, accompanied by specimens of the characteristic
minerals, of which a list is given by Cavendish. I gather from
this, that the journey referred to, was undertaken, in part at
least, at Cavendish's desire, and that the formal parting which is
understood to have happened between him and Blagden, did not
occur till at least 1789.*

When or why Cavendish and Blagden parted, I have not
discovered, but the cause of their diflFerence, whatever it was, did
not prevent the former from leaving his assistant a large legacy
at his death. They were both reserved and undemonstrative
persons, and neither has enlightened us as to the circumstances
which brought them together or separated them. The only
answer I have been able to obtain to queries on the latter point
has been^ that the union or co-partnership was found ^^ not to
suit."

On May 5th, 1784, Blagden was elected one of the Secre-
taries of the Royal Society of London, in room of Mr. Maty,
who had resigned, f The date of his election is important, as
affecting the question when he became responsible for errors
committed during the printing of the Philosophical Transactions*
Its occasion also was peculiar. The Royal Society, in 1784,
was divided into two parties, the one assailing, the other de-
fending, the mode in which Sir Joseph Banks had fulfilled the
duties of President. When Maty, who was one of the Presi-
dent's assailants, resigned, the election of his successor became
a trial of strength between the opposing parties. ** Dr.

* Blagden accompanied Cavendisb in some if nol in all of the geological Journeys
in England referred to in the last chapter, and appears more than once to have acted
ns his geological commissioner on the continent. Under date 1788, 1 find among the
Cavendish MSS. a letter from his friend Michell,in which Blagden's geological obsenra-
tions are referred to. In reply, Cavendish writes, " Dr. Bl. has sent me the miner
[alogical] account of his journey as far as Paris," of which he proceeds to give an
abstract, conclading with a more special reference to his assistant's recovery from an
illness in the course of his journey, than he was wont to bestow on the personal
affairs of his acquaintances.

t Weld's Hiitory qf the Royal Society, vol. ii. p. 561.

130 LIFE OF CAYBNDISH.

Horsley and his party brought forward Dr. Hutton as a
candidate^ in opposition to Dr. Blagden, who was supported
by the President and Council. Before the election^ Sir Joseph
Banks circulated the subjoined card amongst the Fellows:
— * In consequence of Mr. Maty's resignation of the Secre-
taryship, at the last meeting of the Royal Society, the
President takes this method of acquainting you that, at his
desire. Dr. Blagden has declared himself a candidate for that
office. From Dr. Blagden's known abilities and habits oi
diligence, the President does not doubt but he will, if elected,
fulfil the duties of the station with advantage to the Society*
Soho Square^ March 29, 1784.'*'*

On the election-day Blagden had a hundred more TOtes than
his rival, and his appointment by so large a majority, showed
so unequivocally the strength of the President's party, that the
dissensions which had troubled the Society almost entirely
ceased. Blagden continued in office till November 30, I7d7-t
With Sir Joseph Banks he was on most intimate and friendly
terms, having access to his house at all times, and spending much
of his time there whilst in London. He had thus many opportu-
nities of meeting the distinguished men of science from foreign
countries, who received so general and cordial a welcome from Sir
Joseph ; and, till the commencement of the first French Revo-
lution, he spent a considerable portion of each year on the Con-
tinent, visiting diflFerent parts of Germany, Italy, Switzerland, but
especially France. On one of those occasions he resided for
some time at the Bavarian Court, and acquired the friendship of
Benjamin Thompson, better known as Count Rumford.

At the Peace of 1814 he resumed his visits to the Continent
with the title of Knight, which had been given to him in 1792, in
recognition of his services to science. " Every year,'^ says M.
Jomard, '^ he came to pass more than six months at Paris or
Arcueil. None of his countrymen have done more justice to the
labours and discoveries of the French, or have contributed more
than he to the happy relations which have subsisted for six years
(1814 — 1820) between the savans of the two countries.^'t

His last visit to France was paid in 1819, in the autumn of

* Weld'8 History of the Royal Society, vol. ii. p. 165.

t Op, at., p. 561. X Moniteur, Sept. 22, 1820.

CHARACTER OF BLAGDEN. 131

which he travelled to Paris, and took up his residence at Ar-
caeil, in the neighbourhood, in the house of his friend Berthollet,
where he died suddenly of apoplexy, on March 26th, 1820.*

The salient points of Sir Charles Blagden's character have
been sufficiently indicated to us by those who knew him. He
was not a man of genius, his writings display no originality,
nor has he any place jamong discoverers in science. On the
other hand, he appears to have preferred to occupy himself with
the labours of even those he might have rivalled, and was further
remarkable for the multitude of eminent men with whom he
was intimate. The large circle of British men of science who
attended Sir Joseph Banks' soirees, must have been more or less
known to him, including all his contemporaries, the Fellows of
the Royal Society. He had some acquaintance with Dr. Samuel
Johnson and Sir Joshua Reynolds, and assisted Smeaton in
writing his work on the Eddystone Lighthouse.f La Place,
Cuvier, Berthollet, and Benjamin Delessert, he could call friends,
and he had some acquaintance with Lavoisier, Denon, and
Humboldt, as well as with other contemporary philosophers,
whom he saw at the meetings of the Institute, or encountered
in the Parisian scientific circles, t

According to Jomard, indeed, he preferred France to all
other countries, and the fact of his having been on terms of so
great intimacy with the French savans, is one of importance,
for one of the accusations brought against Blagden is, that he
wronged Lavoisier, and may have been misinformed or forgetful
as to his researches. It seems well, therefore, to notice here,
how frequently he visited Paris, and how welcome he was there,
both before and after the occurrence of the controversy which
has cast a shadow over his good name.

The friend or acquaintance of so many distinguished men

* Jomard speaks of Blagden as having exhibited none of the infirmities of age at
SO, bat as he was born in 1748 and died in 1820, he can only have been 72 years
old at his death. The date of his death I take from the inscription on his tomb in
P^re la Chaise, of which M. F. Delessert has sent me a copy : " Ce monument est
•itn^ an Pire la Chaise^ au dmetiire de I'est, 19me division, snr le bord da chemin."

t Mr. W. A. Cadell is my anthority for the reference to Reynolds and Smeaton.

t This appears from the information supplied to me by Mr. Blagden Hale,
Mr. Robert Brown, and M. F. Delessert. Jomard refers generally to the intimacy
between Sir Charles BUgden and the French sayans in the passage previously quoted,
and I may add that Sir Charles left legacies to La Place, Berthollet, and the flCep-
daughter of Cuvier.

K 2

132 LIFE OF CAVENDISH.

must either have possessed some fascination of manner, or other
personal grace, which made him acceptable to the majority of
those he encountered, or his acquirements in science must have
been such as to secure their esteem ; if he did not (as one would
be tempted to imagine from the very large circle of those known
to him) possess alike the intrinsic merit and the external grace*
So at least, it should seem, thought no less severe a critic than
Dr. Samuel Johnson, whose judgment in 1780, Boswell thus
reports, ^^ Talking of Dr. Blagden^s copiousness and precision of
communication. Dr. Johnson said, ' Blagden, sir, is a delightful
fellow.' '^* Such, however, was not the general opinion. Mr.
Robert Brown, who drew my attention to this passage, dissented
from it as a description of Blagden's manner, which he thought
formal; and Mr. Cadell's judgment is, that it was ^^ stiff and
cold.'' Even Jomard says of the object of his eulogium, " Sous
des dehors calmes et quelquefois impassibles, il cachait un coeur
bien veillant et g^nereux f and in Mathias's Pursuits o/Literature,
a well known satirical poem of the last century, he figures as
"prim Blagden/'t There can thus have been no charm in
Blagden's manner ; and his reputation must have been in spite of,
not in consequence of it. It rested mainly on his acquaintance
with recorded facts in nearly all the physical sciences, and his
diligence in keeping pace with the progress of discovery. Boswell
refers to his '^ copiousness and precision,'' as being notorious. Sir
Joseph Banks confidently recommended him tp the Fellows of the
Royal Society, as a man of known abilities. Mr. R. Brown spoke
^f him as ^^ au courant du jour" in all branches of knowledge.
JTomard uses the same phrase, on which he enlarges, j: His papers
in the Philosophical Transactions demonstrate this. His essay
on the congelation of quicksilver in the volume for 1783, may be
taken as an example of his extensive acquirements, and his skill
in arranging and expounding facts.

He was a great economist of time, and very methodical and

* Croker's BoswelVs Johnton, toI. iy. p. 362.

t Seventh edition (1798), p. 72. Matfaias spares no one. The reference to
Blagden is incidental, and was manifestly made only to justify a foot note, in which
this caustic and often coarse writer attacks the Royal Society.

% ** Personne n'^tait plua an oonrant que lui des voyages nouveaux, des
recherches dans les sciences, des decouvertes indastrlelies, des productions de toute
eB^^."^M(mUeur, Sept. 22, 1820.

CHARACTER OF BLAGBEN. 133

assiduous in all his pursuits. His habits were unvarying ; so
much so^ that Jomard says he always took the same road from
his residence (at Brompton, I believe) in London^ to Sir Joseph
Banks^ house and to the Royal Society^s apartments^ although
he had his choice of several ways^ and the one he selected was
the longest.

Thus far^ there is no difficulty in ascertaining what Blagden's
character was. He appears before us as a somewhat formal
and ungenial person, more an object of respect than of love to
those who knew him. Such has been the disposition of pro-
bably the greater number of natural philosophers ; at all events,
it has characterised some of the most illustrious among them,
and I do not stop to apologize for Blagden, or to seek to lessen
in any way the impression which the faithful sketch I have
tried to give may produce.

In two respects his character has been commentedon to
his disadvantage, and these only call for special consideration.
He is accused of having been greedy of money ; and it is at
least insinuated, that this tempted him to wrong Watt and
Lavoisier, and to bestow unjust praise on his benefactor Caven-
dish. The advocates of Watt, indeed, all enter a more or less
decided protest against the impartiality of Blagden^s testimony, on
the ground of his having had a pecuniary interest in praising
Cavendish. The following passage from the life of Cavendish by
Ijord Brougham, may serve as an illustration of the mode in which
Blagden is referred to : — " Having formed a high opinion of
Dr. (afterwards Sir Charles) Blagden^s capacity for science,
he [Cavendish] settled a considerable annuity on him, upon
condition that he should give up his profession and devote him-
self to philosophy ; with the former portion of which condition the
Dr. complied, devoting himself to the hopeless pursuit of a larger
income in the person of Lavoisier's widow, who preferred marry-
ing Count Rumford. He [Cavendish] left Sir Charles a legacy
of jB 15,000 ; which was generally understood to have fallen much
short of his ample expectations.'^*

* IAve8 qfMen of Science (first series), pp. 445, 446. Blagden had reaaon to
congratulate himself on the refusal of his suit to Lavoisier's widow, if it is more than
a piece of harmless gossip, which Lord Brougham reports. Count Rumford soon
separated from Madame Lavoisier, and the marriage is understood to have been a very
unhappy one. Blagden was very kind to Rumford's daughter, Sarah.

134 LIFE OF CAVENDISH.

That Blagden amassed wealth, and did not throw it away, is
certain. In this respect he resembled Cavendish, Watt,
Lavoisier, Hooke, Newton, Wollaston, and Black, not to men-
tion others. His habits were frugal and unostentatious, and he
is understood to have speculated to profit in the French funds,
and thus to have greatly increased his wealth. According to the
judgment of his relations, however, " avarice, or a desire to grasp
or acquire money by any but fair means, he was entirely free from,**
and the verdict of less partial witnesses is in keeping with this
decision. Jomard, who had no inducement to commend Blag-
den's liberality, unless he was assured of it, and who might have
left unnoticed the question of his wealth, and the use he made
of it, enters on the subject freely, and notices it thus: ^^Sa
moderation extreme lui permettait d'augmenter de plus en plus
ses revenus. Mais autant il ^tait dans Paisance, autant il ^tait
genereux, mSme magnifique dans ses liberalit^s. Pour acquitter
un service qu*un ecu aurait paye, il donnait deux pieces d'or ;
si Ton hesitait, il en ajoutait deux autres, craignant qu*on ne fdt
pas satisfait. A Arcueil, il ouvrait g^nereusement sa bourse
aux pauvres; il coopcrait k des actes de bienfaisance, a des
etablissemens d^utilite publique. On pourrait citer de lui
encore plus de preuves de bonte de Tame que de traits de
singularity ou de bizarrerie/**

Such a statement as the one just quoted is surely sufficient
to outweigh the traditions or suspicions which represent Blag-
den as having been an avaricious man. His will also points in
the opposite direction. By it he provided liberally for his rela-
tives, but he took care also to remember his scientific friends^
whom many more wealthy philosophers have altogether for-
gotten when making the final disposition of their riches. Ber-
thollet, the daughter of Madame Cuvier, and the daughter of
Count Rumford, received each 1,000/. ; Dr. Thomas Thomson,
of Glasgow, 500/.; and La Place, 100/., ^'to purchase a

ring."t

On this point it seems needless to enlarge^ nor will I stop

* Moniteur, 22nd September, 1820.

t The probate of the will was under 50,000/. Other parties besides those men-
tioned in the text received lej^acies. BerthoUet and Ramford's daughter are under-
stood to hare received benefactions from Blagden dnring h's life-time.

INTSGRTTY OF BLAGDEN. 135

to enquire whether Blagden was dissatisfied with the sum left
bim by Cavendish, or thought to enrich himself by a wealthy
marriage. It is no such rare thing to find a legatee discontented
with the bequest left him by a very wealthy testator, or to see
even philosophers anxious to wed rich wives, that the certainty
of Blagden having done what is vaguely charged against him,
would imply that he had acted dishonourably. Against any
unfavourable interpretation which might be put upon these
transactions, even if they were unquestionable, we have the
positive testimony of those who knew Blagden, that he was a
strictly conscientious man. Such, we may feel certain, was Sir
Joseph Banks' estimate of his character, from the confidence he
reposed in him, and such also, from Jomard's account, was the
opinion entertained towards him by BerthoUet and the other
French savans who knew him best. His relatives freely bear
witness to his integrity.

Mr. Robert Brown informs me, that although reserved in
his manner, Blagden was an upright, honourable man. Dr.
Thomas Thomson, of Glasgow, writes to me thus : — ^^ Sir
Charles Blagden, you are aware, was assistant and operator to
Mr. Cavendish, in his experimental investigations. I knew him
well, and considered him as a man of perfect honour and
integrity. I have the same opinion of Mr. Watt.*' ..." Both
Watt and Cavendish acted fairly and honourably.'* .... "The
attack upon the integrity of Blagden .... I think unfair.
Blagden had no motive that I can conceive for acting the part
assigned him."* I have already quoted the letter from De
Luc to Watt, in which he acknowledges that Blagden's
character was such as to render it most unlikely that he had
acted unfairly between Watt and Cavendish. Against such
testimony as that adduced, suspicions can avail nothing. In
truth, the impeachment of Blagden belongs almost entirely to
the revival of the Water Controversy, for during his lifetime

* Dr. Thomson has also favoared me with his estimate of the share which the
English rivals had in making the disputed discovery. In conformity, however, with
the principle I have adopted thronghont this volume of discussing the Water Contro-
verfy aa a question of evidence, not of authority, I have not quoted Dr. Thomson^s
opinion on the relative merit of Cavendish and Watt, great as his title to offer a judg-
ment is. I have given, however, in full, his opinion concerning the integrity of the
disputants.

136 LIFE OF CAVENDISH.

he corresponded with Watt, De Luc, La Place, BerthoUet, and
others, who took an active share in the earlier transactions
of the dispute, without his good name being ever called in
question.

Thus much settled, I have now to consider the specific
charges that are brought against Blagden. The first accusation
is, that he was guilty of some crime, or fault, in writing the
interpolations in Cavendish's paper. The advocates of Watt do
not pointedly accuse Blagden of having acted unfairly in penning
those passages, but they speak of this transaction as if it implied
officious intermeddling on his part, and had been done to injure
Watt. There is a further charge against Blagden, that in his
official capacity as Secretary of the Royal Society, he permitted
the interpolations to appear as if they had been portions of the
original MS. The last charge is quite just, so far as the fact is
concerned, but not as implying any unfair dealing on the part
of Blagden.

It was the practice of the Royal Society in 1784, and long
before and after, to permit the contributors to its Transactions*
to alter their communications between the period of their being
read and printed. Into the proof of this I have entered at
length in the section of the Water Controversy, entitled /n-
terpolations. It will suffice, therefore, here to point out, in illus*
tration of this, the one sufficient and important fact which the
advocates of Watt have not perceived, or at all events have
not acknowledged, viz., that Watt's paper contains as many
undated and unmarked interpolations as Cavendish's, and that
Blagden, who had the superintending of the printing of Watt's
paper in 1784, no more insisted on the interpolations in it being
enclosed in brackets, or otherwise indicated, than he did on those
in Cavendish's paper. There was nothing unfair in the inter-
polations made by Cavendish, nor did Watt suffer by their being
printed continuously in the text of his rival's paper, as if origin-
ally a part of it. The importance of the added passages
depended entirely on their truth, and was not in any respect
affected by the part of the paper in which they appeared.
They were either true or false, and if true, they might have been
added as postscripts, or printed as notes, with dates attached to
them, had such been the practice of the Royal Soci3ty, without

INTERPOLATIONS IN THE PAPERS OF CAVENDISH AND WATT. 137

tlie daims of Watt and Cavendish being in the slightest degree
altered by the procedure. Mr. Muirhead^ in the partial appli-
cation of an invidious rule, has marked by brackets the inter-
polations in Cavendish's paper^ but not those in Watt's (both
of which are reprinted in the sequel to the Wait Correspondence),
and an impression is thus conveyed^ that Cavendish or Blagden,
or both^ had done something wrong in making or suffering inter-
polations. If Mr. Muirhead had supplied the brackets^ so as
to mark the added passages in Watt's paper^ it would be seen
to contain as many interpolations and anachronisms as his
rival's paper does. The title^ for example, and two at least of
the notes^ were not added till the spring or summer of 1784,
yet they are printed as part of the original text of a paper dated
November 1783. Let it be observed, however, that Watt had
not been guilty any more than Cavendish of wrong-doing in thus
altering or supplementmg his paper. The interpolations in both
cases were innocent and laudable, and were made with the entire
sanction of the officials of the Royal Society. If this be the
case, however, it must be a matter of exceedingly trifling moment,
in whose handwriting two truthful declarations were engrossed.
Only those who judge Blagden by the light cast by the unfounded
suspicions entertained by De Luc and Watt against him and
Cavendish, can regard the question of handwriting as one of
the slightest importance. Since^ however, it has been urged,
it may be noticed that it is an argument in favour of both
Cavendish and Blagden, that the interpolations were written
by the latter^ for they are very short and might have been
rapidly written by Cavendish himself, had he intended to pass
them off as parts of the original paper, whereas they appear in
the handwriting of his secretary, and on separate sheets of a
smaller size than those containing the text of the essay, and of a
different quality of paper.'^ In the second interpolation, like-
wise. Cavendish reveals by the wording, that the passage had
been added after Watt's paper had been read, and was not,
therefore, a part of his original text. That Blagden urged
Cavendish to make those interpolations, there is no proof. But
even if there were, it would imply nothing discreditable to the
former. When Watt's zealous friend, De Luc, was doing so

* Weld's History qf the Royal Society, vol. ii. p. 173.

138 LIFE OF CAVENDISH.

much for him, it was nothing less than a duty that Cavendishes
friend Blagden, should guard his reputation. So long, indeed^
as De Luc and Blagden were careful to publish only truths^ it
could not be matter of accusation against either that he did
publish them in the defence of his friend.

The second charge against Blagden is a more serious on^
and affects the errors in date in the papers of Cavendish and
of Watt, printed in 1784. The first of these was read on January
15 th of that year, as appears by its title in the Philosophical
Transactions** But in certain copies of the paper which were
printed separately, the year was marked 1783 instead of 1784.
How, or when this alteration of figure was made, and who was
responsible for it, is not quite certain* The blame, however, if
blame there be, lies as much with Cavendish as with Blagden, for
it was not part of the latter's official duty to superintend the print-
ing of private copies of the papers which appeared in the Phu-
losophical Transactions. From a reference, however, in CrelPs
Annalen (1785, part iv. p. 324), to Cavendish's research as
contained in " Experiments on Air, London, by J. Nichols,
1784, 4 Mai. p. 37,'* which had been sent to the editor by Sir
Joseph Banks, it appears in the highest degree probable, if not
certain, that the private copies were paged separately from I to
37, whilst in the Philosophical Transactions for 1784 the numbers
run from 119 to 153.f It seems, further, likely that during
the alteration of the paging the error was detected and not
allowed to appear in the Transactions; or if they were printed
first, the error may have been committed during the substitu-
tion of the one set of figures for the other. In whatever way,
and by whomsoever committed, few impartial readers will doubt
that the error was accidental, for there was absolutely nothing
to be gained by it, inasmuch as, firstly y the separate copies of a
paper reprinted from the Transections of a Society, have no
authority, unless in so far as they exactly agree with the text

^ *' Cavendish's paper is printed as having been read on January 15> 1784. It
was commenced at the meeting of that date, but the reading occupied the time of two
evenings.'' — ^Weld's History of the Royal Society, vol. ii. p. 173.

t I have tried in vain to procure access to one of the detached copies of the
Experimenta on Air. It is not a little singular that Crell should refer to his
copy as marked 1784, not 1783, as if there were no error of date on it. Should any
of my readers encounter a copy, he would do a service by printing a description of
the title-page and dates, along with the numbers on the pages.

ERROR IN DATE OVERLOOKED BY BLAGDEN. 139

of the Transactions; and secondly, the substitution of 178S for
1784 only carried back the entire question of the claims of
Watt^ Cavendish, and Lavoisier^ by a year, without, in the
slightest degree, altering their relative position as claimants of
the disputed discovery. The whole proceedings were antedated
by a year, but everything else remained as it was.

But (which is most decisive of all). Cavendish, as soon as he
became aware of the error, wrote to Mongez, the editor of the
Journal de Physique d Parts^* requesting him to correct it, and
in so doing not only acknowledged, and made reparation for the
accidental error which had occurred, but by apologizing in his
own name, and not in that of Blagden, for its existence, took
upon himself the responsibility of having allowed it to appear.

For the error in Watt's paper, Blagden certainly was re-
sponsible, and the advocates of Watt have been unsparing in
their denunciations of the secretary, whom they do not exactly
accuse of having wilfully committed the blunder, but of some-
thing as nearly approaching a wilful error as can very well be.
Watt's paper, as it was ultimately printed, assumed the shape of
a letter to De Luc, dated November 26th, 1783, but through an
error of the press, 1784 was printed instead of 1783, so that Watt
had the disadvantage of having his letter post-dated by a whole
year. That the error was accidental, the paper bore upon its face,
for the letter was stated to have been " Read April 29, 1784,"
which it manifestly could not have been, if not so much as written
till the succeeding November; yet, as it appeared in the Philoso--
phical Transactions for 1784, it was impossible that the mistake
could apply to the time of reading, which might otherwise have
been supposed to have been 1785, and the conclusion could not
have been avoided by one who observed the incongruity of the
two dates, that the error applied to the time when the letter was
written, not to that when it was read. At all events, the irre-
concileable character of the 'numbers could not fail to strike
any one who consulted the paper with a view to determine
questions connected with the chronology of the disputed dis-
covery. And if such a perplexed reader did what was impera-
tive on him, before condemning the editor of the Transactions^
but which the advocates of Watt seem never to have done, viz.,

* British Association Report, 1839, p. 65.

140 LIFE OF CAVENDISH.

consulted the errata5 whicb^ in those days of careless editings
had always a prominent place on the concluding leaf of each
volume of the Transactions, he would have found that the much
maligned Blagden had pointed out that November 1784 should
have been 1783. I will not demur to Blagden's carelessness in
permitting so unfortunate a blunder to occur, being not only
lamented^ but even reprehended. The advocates of Watt,
however, represent him as having taken the first opportunity
which his secretaryship afforded, of interfering with the printing
of the Transactions to permit, or to commit an error most un*
favourable to Watt, whereas the recency of his appointment to the
office of editor of the Philosophical Transactions should surely be
regarded as entitling him to a lenient and charitable judgment,
when he is found to have suffered an error of the press to occur
whilst discharging duties with which he was not yet familiar.
And as this has not been acknowledged as in candour it
should have been, I must point out a little more fully how
excusable Blagden^s carelessness is. Firstly y then, there are fewer
eiTors in the volume of Philosophical Transactions for 1784 than
in many that preceded it, so that Blagden cannot be accused of
having neglected his editorial duties according to the standard of
editorship of his day. Secondly, he offered to send Watt the
proofs of his paper to Birmingham if he wished it, which one
proposing to wrong Watt would certainly not have done. Thirdly,
the letters of Blagden to Watt contained in the TFatt Correspond^
ence, show a solicitude to do the great engineer justice, and
to publish his paper in the way which he preferred, altogether
at variance with the notion that from a venal or envious motive,
Blagden designed to do him wrong. Fourthly, Blagden informed
Watt that he might obtain separate copies of his paper when
he pleased, so that he exposed himself to the risk of being dis-
covered to have committed a fraud before the volume of the
Ti^ansactions was published, not to refer to the certainty of
the detection after it had appeared. Fifthly, the MS. from which
Blagden had to print Watt's paper, consisted primarily of two
documents, of one of which, parts only were to be selected for
publication by the secretary, and besides, of a title, and of certain
notes and alterations which were sent by Watt in letters to Sir
Joseph Banks and to Blagden. It is not very surprising, therefore.

INNOCENCE OF BLAGDEN. 141

that in the construction of a single methodical paper out of
several letters, an error should have occurred in copying the
date of one of them. Watt himself confounded the date of his
letter to Priestley with that of his letter to De Luc ; * and
Mn Muirhead, who puts the worst interpretation on Blagden^s
permission of a typographical error, contrived to commit one as
suspicious (i. e. as innocent) as Blagden's, in copying from the
plain print of a part of Watt's paper.f I have no fear of the
judgment which will be passed upon Blagden by those who
set aside the suspicions of De Luc and Watt, and consult the
errata of the Philosophical TVansactions for 1784.

When thus it has been shown that, according to the testi-
mony of his contemporaries and his survivors who knew him
best, Blagden was an upright, honourable man; that in the
matter of the interpolations he acted only as clerk, and was a
party to no wrong ; that the first error in date was a harmless
one, confined to a few unauthoritative copies of Cavendish's
paper, and acknowledged and corrected by Cavendish as soon
as it was discovered ; and that the second typographical error
was a self-evident one, made in very excusable circum-
stances, and corrected simultaneously with its publication to
the world : I think that I am fully entitled to repeat, that the
advocates of Watt have said greatly too much in disparage-
ment of Blagden's good name, and that they ought in justice
to retract all that they have said, or to impeach him boldly,
and produce proofs of his villany which as yet have not been
shown.

It remains, so far as this subject is concerned, that the cruel
and gratuitous suspicion that Cavendish, if he did not bribe
Blagden to lie and cheat for him, at least put his dependent in
such a position that he saw he could only earn his bread by com-
mitting fraud in behalf of his master, should for a moment be
noticed I would not willingly put in the mouths of the advo-
cates of Watt a darker suspicion than they actually entertain,
but in spite of all their cautious wording, and reservations, and
qualifications, they prefer a charge amounting in its most naked

* Watt Corr.f p. 70.

i* Watt. Corr. Reprint of Watt's paper, p. 78, where " meeting " occurs
instead of ** reading." (Note — 7th line from bottom.)

142 LIFB OF CAVENDISH.

shape to this^ that (1) Blagden was dishonest for the sake of
Cavendish ; and (2) that Cavendish paid Blagden money. That
the money was the cause of the dishonesty is further implied
where it is not affirmed; and the only point which is not
absolutely pressed^ is^ that Cavendish desired that dishonesty
should be practised, or intended to reward it. If this view of
the fair-dealing of Blagden be true, it should be shown by the
advocates of Watt, that Cavendish made Blagden's salary
dependent on his services; whereas it was secured to him in the
form of an annuity,* as was but just, when Blagden abandoned
his own profession to associate himself with Cavendish. He
had thus nothing to gain by becoming a party to any fraud
which might serve his principal. It is for the advocates of
Watt, moreover, to make out a consistent theory of Blagden's
dealings on their own hypothesis of his character, so as to
reconcile it with the fact implied in the accusation, that the
party accused, as all acknowledge, was a very shrewd person,
and, according to their view also, a very selfish one, and
nevertheless only succeeded, if the charges are true, in in*
volving himself and his principal in needless and unprofitable
suspicion by sanctioning two errors in date which no clever
rogue would have committed. I shall presently return to
the consideration of the proceedings of Cavendish and Blag-
den, as looked at from an impartial point of view; for
as yet, I have considered them solely by the light of those
suspicions of De Luc and Watt, which, according to the
friends of the latter, supply the only just criterion by which to
test the proceedings of the English rivals. Before doing this,
however, it is necessary to consider the part which Blagden
took in reference to Lavoisier's claims, as his defence of Caven-
dish's rights against them led to a reference to the demands of
Watt as late as 1786, in which the advocates of the latter find
fresh grounds of complaint against Blagden. And as this can-
not be done without special reference to Lavoisier's demands, I
shall dispose of them in the course of the discussion.

In June 17^3, Blagden during one of his visits to Paris,
gave some account of the researches of Cavendish and Watt into

• The onnuity amounted to 500/. {HUtory qf the Rovid Society, toI. ii. p.
173.) ^

IMPEACHMENT OF LAVOISIER BY BLAGDEN. 143

tbe production of water from its elements. At a later period^ as
all acknowledge^ Lavoisier^ for the first time in his experience^
observed that the combustion of a given weight of hydrogen and
oxygen yielded a nearly equal weight of water, and announced
the result to the French Academy, in June l783j whilst Watt's
letter was still in abeyance^ and before Cavendish had made his
observations known, either directly or through the medium of
Priestley's announcement of the repetition of them. When
Lavoisier ultimately published his account of these experiments,
he stated that they were first performed on 24th June, 1783, in
the presence of Blagden and others, and that the result was
reported to the Academy on the following day (June 25th). In
the first of these references he acknowledges that Cavendish
had previously burned hydrogen in close vessels, and that he had
obtained '^a very sensible quantity of water,'' but Lavoisier pro-
fessed to have been the first to observe that the weight of water
produced was equal to the weight of gases burned, and that
water was not a simple body, but consisted of hydrogen and
oxygen.*

The accuracy of Lavoisier's statement was called in question,
so soon as it became known in England, and as Blagden was
acknowledged to have been Lavoisier's informant, it came to be
a question affecting the veracity of the one or the other, includ-
ing on the side of Blagden^ Cavendish, whom he represented.
Blagden, without hesitation, impeached Lavoisier's veracity,
and implicitly accused him of plagiarism, and this in three
different ways. As early as December 1 783, when his recollec-
tion of what he had told Lavoisier must have been unimpaired,
be authorised Kirwan to tell Watt that he had explained the
English theories "minutely" to Lavoisier in the preceding
summer.t In 17B4, Cavendish stated in the concluding part
of the first interpolation in his paper, that a friend of his (whom
no one doubts was Dr. Blagden) gave an account of his experi-
ments and conclusions to Lavoisier in the summer of 1783,
and that he was with difficulty persuaded to beUeve that the
gases burned together could be converted into their joint weight

* Mimoiru de VAcadime dei Scieneet, 1781 (printed io 1784), pp. 472—475;
Watt, Corr., pp. 176—180. A detailed analysis of LaToisier's experiments is gtvea
in a section of the Water ControTerqr devoted to their consideration.

t Wait, Corr., p. 39.

144 LIFE OF CAVENDISH.

of water^ and only adopted this conclusion when he had obtained
the same result in his repetition.* It may also be noticed here,
that Cavendishes so-called third interpolation, or rather post-
script, which is in his own handwriting, was a temperate and
altogether unpolemical criticism of Lavoisier's anti-phlogiston
doctrines, which will be found fully noticed in the abstract
of the Expeinments on Air, This postscript has not given
occasion to any discussion^ and therefore is not further referred
to here.

Finally, in 1786, Blagden wrote a long letter to Crell, who
had published several inaccurate accounts of the circumstances
attending the discovery of the composition of water, which he
attributed to Lavoisier^ whose views Cavendish was stated only
to have confirmed, whilst Watt's name was not mentioned. In
this letter, Blagden, in his own name, accused Lavoisier o£
having concealed the facts that Iiis experiments were made in con-
sequence of what Blagden had told him, that the weight of water
produced equalled the weight of the gases burned, and that Caven-
dish and Watt had founded upon those researches a conclusion
identical with Lavoisier's, except that he did not employ the
word phlogiston as a name for the combustible element of water.
Without entering here into any investigation into the merits of
Lavoisier, to which I have sought to do justice in a special
section of the Water Controversy, it may be noticed that he
never replied to the accusations which Cavendish, Watt, and
Blagden preferred against him, and that none of his contem-
poraries or successors, even among his own countrymen, have
been able to defend him from the charge of plagiarism.* Nor
did Blagden's denial of Lavoisier's fair-dealing lead to any
cessation of intercourse between the former and the French
philosophers, who would certainly have resented the charges
brought against Lavoisier, had they been untrue. The part of
Blagden's letter, however, which mostly concerns us is, that in

• Phil. Trana, 1784, p. 134.

t Lavoisier's aocoant of his ezperimeDts is at rariaDce with the oontemporaaeou
account of them given by his colleague La Place, written a few days after they were
made, in a letter to De Luc. (Watt Corr,, p. 41.) La Place there asserts that up to
the 28th June, 1/83, that is three days after Lavoisier's announcement, they did not
know whether the quantity of water produced represented the quantity of gas con-
sumed. Berthollet also, in a letter to Blagden, of which I have published the greater
portion in the sequel, imputes the discovery of the composition of water to Caveadiih •

BLAQDEN'S LETTfiR TO CRELL. "145

which he states that Watt's conclusions became known to him
** about the same time" (the spring of 1783) as those which
<]!aYendish had drawn from his own experiments.

The advocates of Watt assert that Blagden must have known
"whether Cavendish or Watt first formed his theory, and that he
woxdd have stated that Cavendish was the first, if he could in
honesty have done so. Without entering into an elaborate
defence of Blagden, which will be found elsewhere in the
volume, I may notice here, that according to the interpretation
whicb Watt's advocates themselves put upon an ambiguous
letter of Kirwan's, Blagden first learned Watt's theory from
Cavendish.* Whether this be the true interpretation or not, it
is certdn that Blagden could not have been ignorant of Caven-
dbh's researches after the spring of 1783, for the first public
account of them, that, namely, in Priestley's paper, accompanied
Watt's letter to the Royal Society. On the other hand,
Blagden did not become Cavendish's assistant till 1782, and
could not, therefore, be cognizant of his experiments of 1781,
till long after they had been made, and Priestley is the only
person to whom Cavendish professes to have explained them
before the spring of 1783. It is of much more importance,
however, to notice that the object of Blagden's letter to Crell,
of 1786, was not to discuss the claims to priority between Watt
and Cavendish, but those of Cavendish against Lavoisier. The
last-named philosopher had in efifect called in question the
veracity both of Blagden and of Cavendish, and Blagden only was
in a condition to disprove Lavoisier's statement, and to remind
him of what he had told him. There was a peculiar necessity,
also, for addressing a letter to Crell, because, through imperfect
information, he had, in two different numbers of his journal,
imputed to Lavoisierwhat even he did not claim. ^ There was no
occasion, however, for Blagden to discuss Watt's claims in Crell's
journal, for they had not been noticed therein. Nor could Blagden
have spoken on his own authority in reference to them, as he, and
only he, could do in regard to the statements of Lavoisier.
An English journal was the fitting place in which to canvass

• JTatt. Corr.t p. 39.

t See in lUiutration of this the quotations from Crell's Annalen, which are pab«
lished in section 11 of the Water Controversy contained in the sequel.

146 LIFE OF CAVENDISH.

the claims of the English rivals, and Cavendish, Prieadey, and
Watt, were the only parties competent to discuss it.

I feel no difficulty, accordingly, in agreeing with the advo-
cates of Watt in their belief that Blagden's silence in reference
to the relative priority of Cavendish and Watt, was deliberate.
But I entirely differ from them as to the motive which led to
his silence. It is as unimportant as it is impossible, to deter-
mine whether Blagden knew by what period of time the one of
the English rivals had preceded the other. We are apt to forget
that the W€Ut Correspondence^ which we now see in print, put
forth by Watfs descendants as a public claim, in his name, of
priority over Cavendish, was, during the lifetime of the rivals, a
bundle of private letters. Ostensibly there had been no dispute
between Cavendish and Watt, even in 1 784, and each had, in
his own name, published all that he deemed requisite to justify
his claims. In 1786 there was nothing to revive the dispute ;
on the other band. Cavendish and Watt had shaken hands at
one of Sir Joseph Banks' soirees, and we shall find in the next
chapter, that in the year when Blagden wrote to Crell, Caven-
dish, in one of his Geological tours, found his way to Birming-
ham, and took unusual interest in studying the great engineer's
beautiful inventions. We may feel certain, therefore, that how-
ever willing Cavendish might be to defend his veracity against
Lavoisier's implied denial of it, that with his characteristic in^
difference to fame, he would, if consulted in the matter, forbid
Blagden to enter upon the question between himself and Watt.
Nor could Blagden have anything to gain by renewing the strife
between the English rivals, or by affirming, in the name of
Cavendish, in a German journal, what Cavendish could so much
better say for himself at a meeting of the Royal Society, where
Watt himself, with their respective friends, could meet face to
face. I come to the conclusion, therefore, that no reproach
attaches to Blagden for avoiding the discussion of the question
of relative priority as between Cavendish and Watt, and I will
add that' he deserves praise for his bold impeachment of so
influential a person as Lavoisier, and for hazarding the loss of
the friendship of the French philosophers, by his open vin-
dication of the priority of Cavendish.

In bringing to a close this long discussion of the parts takea

BEHAVIOUR OF THE RIVALS IN THE WATER CONTROVERSY. 147

by the rivals in the Water Controversy^ and their friends, I
would remark that, with the exception of Lavoisier, they ought
all to be acquitted from the charge of having acted, or sought to
act with deliberate injustice, in defending their own or their
friends' claims. Priestley, in particular, had it in his power to
have appropriated to himself a large part of the merit which
belongs to Cavendish or to Watt, or to both ; nor would it have
been easy for those whose ideas he borrowed to have identified
and recovered their property. Although, accordingly, no man,
deserves to be praised for being simply honest, yet it is but
just to Priestley to set off against whatever is obscure or inex-
plicable in the part which he took in the Water Contro-
versy, the certainty that his good name is if possible heightened
by his behaviour in it. On Cavendish's head no blame rests.
No explanation was asked by Watt at his hands ; no just ac-
knowledgment of even his experiments was made by his rival ;
yet he did justice to Watt, so soon as the public reading of
his paper allowed him to comment upon his views, and he
did not resent the imperfect account of his own. Blagden,
also, stands exculpated from any heavier charge than that
of excusable carelessness in correcting a printer's proof on one
occasion. Nor did the part he took in vindicating Cavendish's
claims, in any way exceed what every honourable man would
consider himself justified in doing, to defend the good name and
the reputation of a calumniated friend, as well as to assert his
own veracity. De Luc stands acquitted of any design to wrong
Cavendish, or of having acted otherwise than according to his
conscientious convictions of what was the truth. But he was
guilty of most blameable haste in coming to a conclusion con-
cerning the conduct of Cavendish, and of not less blameable
neglect in not obtaining the information without which he was not
entitled to pronounce a judgment. Watt is open to the reproach
of having listened too readily to the suspicions of De Luc, and
. of not having taken the means which were accessible to him for
ascertaining what Cavendish had done before him, although he
published a decision on Cavendish's merits. This charge has not
hitherto been preferred against Watt. He has had the good
fortune to have had a much greater number of able and earnest
advocates than Cavendish has enjoyed. These zealous defenders

I. 2

148 LIFE OF CAVENDISH.

have^ with great success^ conveyed the impression that the first
wrong was committed by Cavendish^ and that Watt only acted
on the defensive ; whilst Blagden officiously assisted^ if he did
not perhaps precede Cavendish^ in wronging Watt^ to whom
De Luc rendered only such services as friendship imperatively
demanded. Against this view of matters I protest, as a complete
inversion of the true position of the parties referred to.

The beginner of the Controversy was De Luc, whom Watt
joined, and they compelled Cavendish and Blagden to assume
the defensive, in vindication both of their intellectual and moral
reputation. Cavendish did Watt no wrong in omitting all refer-
ence to bis name in the first draft of his paper, for the private
letter of the latter had been withdrawn by himself, and was not
subject to public comment by any one. When De Luc accord-
ingly found no reference to Watt in the MS. of Cavendish's
paper, which was fireely and courteously sent to him by its
author, he should in the first place have repaid the courtesy by
communicating with Cavendish, and ascertaining what view he
entertained of Watt's opinions. For anything De Luc knew, or
was entitled to assume, he might have found that when it was
explained, that Watt had never lost faith in that part of his letter
which referred to the production of water from inflammable air
and oxygen, and intended to publish that to the world, as an
opinion he had held since 1783^ Cavendish was quite willing to
acknowledge the independence of Watt's conclusions, and their
similarity to his own. At the same time De Luc might have
learned that Cavendish had first made the experiments which
Priestley repeated in 1783, and that in repeating them charcoal
gas had been substituted for hydrogen ; and Cavendish might
further have informed him that he was the first, as Priestley
could testify, who turned or converted a given weight of hydrogen
and oxygen into the same weight of water. Had De Luc known
this, he would certainly have acted otherwise than he did, and
so, assuredly, would Watt also. Had De Luc and Watt, indeed,
only known the one fact, that Cavendish had, in 1781, dis-
covered the equality of weight between the gases burned and
the liquid produced, and that this liquid was pure water, they
must have judged him in a very different spirit, and confessed
that one-half, at least, of the merit of discovering the composition
of water belonged to him.

RETRIBUTION WHICH AWAITED DE LUC. 149

De Luc's neglect of all this prevented the possibility of an
amicable understanding between Cavendish and Watt^ so that
even to thb day^ we are in ignorance as to what their true
feelings were towards each other in 1784. Further^ De Luc's
two grounds of suspicion, viz.^ the similarity of language
between the statements of Cavendish and of Watt^ and the
certainty of Cavendish having read Watfs letter of 1783,
would have lost one half of their weight, even to a prejudiced
mind, had its ]x>ssessor been aware that the supposed plagiarist
had long preceded him whom he was alleged to have pillaged,
and had announced a theory in words which were entirely his
own. In total and wilful ignorance of all that Cavendish had
done, De Luc did not even proceed to verify his suspicions^
but acted upon them as if they had been demonstrated truths.
Equally negligent was Watt, who has written his self-con-
demnation in the pencil note attached in 1784 to the MS.
of his paper, which even the most zealous of his advocates
confesses to be strangely inaccurate.

A heavy retribution awaited De Luc for his hasty and
ill-judged interference. He was accused himself in later life of
shameful plagiarism from Black, and had some difficulty in
satisfying even Watt that the charge was unjust.* He was cer-
tainly innocent, but he may have looked back with some remorse
and with some sympathy, when he tasted in his own person the
fruits of hasty and inaccurate investigations, like those by which
he had calumniated Cavendish. His punishment certainly was
not greater than his offence, for the evil effects of his mischief-
making have increased instead of diminished by the process of
time ; and this leads me to the consideration of the revival of
the Water Controversy by Arago, on which a few words must
be said.

This Controversy will, I believe, be looked back to in later
times as exhibiting one of the most remarkable examples of a
myth unconsciously fashioned out of a legendary tradition, and
that not by poets, but by men of science. The advocates of
Watt, are not, I suspect, aware to what extent they have been
assuming as evidence points which never ranked higher than
suspicions* The revival of the Water Controversy was merely

* Edin. Rev. 1803» p. 21, and 1805> Appendixi pp. 502—515.

150 LIFE OF CAVENDISH.

the revival of the ignorant surmises of De Luc ; and it hap-
pened thus : — ^Watt retained copies of the letters which he wrote
in reference to his theory of the composition of water, and pre-
served those relating to this and kindred topics, received from
his fiiends. These letters form the Wait Correspondence^
which has been so ably edited by Mr. Muirhead. After Watt's
death, these letters passed into the hands of his son, the late
Mr. James Watt, junior, who naturally set a high value upon
them, and as naturally adopted explicitly the opinions of his
father, which he found expressed in them. With commendable
filial piety also, he showed those letters to various men of science,
^nd sought to induce them to adopt and defend the views
which they contained. In this he was not very successful till
Arago visited this country in 1834, to collect materials for his
Eloffe of Watt. He was already satisfied of Watt's right of
priority, and the perusal of these letters which were shown
to him ^^ put,'' says Mr. James Watt, ^^ the seal on his conviction,
and he requested permission to make use of them in his intended
memoir, urging that, in justice to my father's memory, and as a
matter of history, I ought not to withhold them. In conse-
quence, I arranged them in chronological order for his use,
accompanied by such brief explanations and remarks as occurred
to me.'^* The Eloge was published in 1838, and the Correaponr
dence on which so much of it was based was published, as we
have'already seen, in 1846, so that all have now the means of com-
paring Arago's conclusions with the materials which led him to
form them. I do the distinguished Secretary of the French Aca-
demy no injustice, when I say that he has committed the same fault
as De Luc did, whose unjust surmises he has revived but not
justified. This is a grave charge to bring against so great a phi-
losopher as Arago, and I feel peculiar reluctance in urging it at
■a time when all lovers of science lament that a grievous infirmity
has made him a sufferer, and abridged his means of advancing the
branches of knowledge which he has done so much to extend. But
these feelings I must set aside, content to be blamed if I can only
show the justice of my charges. I must then say that Arago has
not investigated the claims of Watt against Cavendish, with the
care which should have been bestowed by one who was about to

* Watt Corr,, pp. z» zi.

ERRORS V$ ARAGO'S ELOGB OF WATT. 1^1

i(dd the weight of so deserv^edly illustrious a name to the support
of Watt's demands, and who, before the assembled philosophers
of France, proceeded to call in question the integrity, as well as
the science, of Cavendish. I find in the Eloffe of Watt many
errors, which for ten years have contributed to spread over the
civilized world a false view of the characters of Cavendish,
Priestley, Blagden, and in truth, of the whole Royal Society.
I mention some of these, although the task is an invidious
one : but it cannot be avoided by one who would do justice
to Cavendish. I make in the first place, this general charge
sigainst Arago, that he did not study the papers of either
Cavendish or Priestley, although without a knowledge of
them, it is impossible to judge impartially between the rivals in
the Water Controversy. In consequence of this, he gives the
following singularly inaccurate account of matters : '^ Priestley
records in detail, and as his own, experiments which prove, that
the water produced by the combustion of a mixture of hydrogen
and oxygen, has a weight exactly equal to that of the two gases
which are burned. Cavendish, some time after, claims this
result for himself, and insinuates that he had communicated it
verbally to the Chemist of Birmingham.^'*

Had Arago consulted Priestley's paper, he would have found
that it did not describe a single experiment on the production
of water from hydrogen and oxygen ; and that the only experi*
ments that it does record, were made with the inflammable air
from charcoal (a mixture of various gases) and dephlogisticated
air. And if he had consulted Cavendish's paper, he would have
found that he did not insinuate, but broadly declared, that
Priestley had repeated, but with a difference in the quality of
the combustible gas, his experiments on the production of
water ; and that he referred to Priestley as having already pub-
licly acknowledged both those points.

The conclusions of his own countryman, Lavoisier, Arago
also mistakes, for he affirms that in 1784 the word hydrogen
was applied to the combustible constituent of water, although
it was plainly impossible that such a word should come into
use before the composition of water was discovered ; and it is
certain that Lavoisier does not use it in his papers on water of

* Watt C&rr.f p. 230. Mr. Mvirhead'^ tnnslation of Arago's Eloge.

152 LIFE OF CAVENDISH.

that year, but speaks of hydrogen as " air inflammable aqueux/^
or ^^principe inflammable aqa/euxP

The eulogist of Watt has also throughout his work^ taken for
granted that the word phlogiston, which was applied by Watt
as well as by Cavendish to the inflammable constituent of water,
necessarily signified hydrogen^ and in so doings has been guilty
of a most manifest petitio principiij as other advocates of
Watt, and especially Lord Jeffirey, the ablest of them all,
acknowledge. I hesitate to say more, but these are not the
only errors. Macquer is represented as having in 1776, proved
by analysis, that pure water accompanied the combustion of
hydrogen, whereas this chemist made no analysis, but only
surmised that the liquid was water. Cavendish is thus deprived
of the merit which is unquestionably his. Warltire is furtlier
stated to have first ^^ imagined that an electric spark could not
pass through certain gaseous mixtures, without causing some
change in them ;'' and yet Arago has himself shown in his
Eloffe of Volta, that he was the first to employ the electric spark
to detonate explosive gases ; and Warltire professes only ta
have imitated Priestley, who, in his turn, avowedly borrowed
from Yoltal Again, Cavendish is represented as having ob-
served no more in 1783, than that water may be obtained by
exploding a mixture of oxygen with hydrogen ; whereas Priestley
testifies to Cavendish having then discovered that the weight
of water equalled that of the gases burned. Watt's letter to
Priestley of 1783 is referred to by Arago as having been ^^ pre-
served in the Archives of the Royal Society of London,'' whereas
it had to be demanded from De Luc by Sir Joseph Banks, when
it was publicly read in 1784. Further, great importance ist
attached to Watt's letter having remained " in the Archives of
the Society ;" and Cavendish is accused of having ultimately
referred to this letter as one with which he had only become
acquainted when it was read before the Royal Society ; whereas
Cavendish says nothing whatever as to when he first became
acquainted with the letter in question, but simply concerns him-
self with certain of its contents as having been formally made
public. Arago also omits all allusion to the cause of Watt's
letter of 1783 not having been read till 1784, declaring that
circumstances, which he suppresses, « because they do not aflFect

ERRORS IN ARAGO'S ELOGE OF WATT. 15a

the present enquiry, delayed this reading for a year,*' Yet,
surely, it was a point of the greatest importance to the enquiry,
that Arago's readers should know that it was Watt himself who
withheld the letter from being read ; and this because he had lost
faith in certain of the conclusions announced in it. Cavendish^
moreover, whose case is so summarily disposed of, is styled ^^ a
pretender*' to the disputed discovery. *^ Numerous errors'' of the
press are referred to as having occurred in the printing of the Phi*
losophical Transactions for 1 784, not one of which ^^ was favourable
to Watt !'* This reference conveys a very exaggerated impression
of the wrong dates, for there was only one error in the Trans--
actions^ and that was corrected in the erratum. Arago's apology,
moreover, for Cavendish's alleged treatment of Watt, is even
more painful to read than his accusation. ^^ On the subject of
discoveries," he says, ^^the strictest justice is all that can be
expected from a rival or a competitor, however high his repu*
tation may already be." This is a moumfal announcement, if
it expresses what the secretary of the largest scientiiSc body in
the world believes to be the temper entertained by discoverers
in science towards each other ; and a woful lesson to be taught
by such an authority to youthful students, who are excused from
being generous, provided they are barely just to their rivals.

In 1840, after Mr. Harcourt had pointed out certain of those
errors, Arago sought to vindicate himself before the French
Academy ; and in reply to the fact adduced in contradiction of
certain of his assertions, that Priestley had retracted his original
statement, that the inflammable air from charcoal, when burned
with oxygen, yielded nothing but pure water, demanded whether
he was called upon to study memoirs of 1786 and 1788, in
reference to a discovery of which the latest date was 1 784. Yet
Arago had no objection to quote Blagden's letter to CreQ, of
1786, in support of the claims of Watt, and could not, there-
fore, on the plea of date, refuse to consult Priestley's papers of
that and later years, which, as we have already seen, have a most
important bearing upon all the questions in dispute.

Arago sought also to justify his substitution of the word
hydrogen, for phlogiston, in expounding Watt's views, by show-
ing that he had made a similar substitution in expounding
Cavendish's; as if this were not a petitio principii of the

154 LIFB OF CAVENDISH.

strangest kind^ since it was matter of certainty that Cavendish's
phlogiston was hydrogen, whereas it was matter of doubt and
dispute what Watt's phlogiston was, and it should have been
proved, not assumed to be hydrogen* The only proof, how-
ever, of this point, which Arago gave, was to assert that in
17B4, pUogiston, inflammable air, and hydrogen, were all names
applied to the combustible element of water. Tet the last of
these terms had not yet come into use, and the second was
applied to every combustible gas known at the time; the specific
name for hydrogen being inflammable air of or from the metals;
and phlogiston, a word which signified one thing in the mouth
of Cavendish, with whom it primarily stood for hydrogen, and
another in that of Watt, with whom it primarily stood for
inflammable air from charcoal, or perhaps for combustible gas.

The mistakes I have pointed out have remained uncorrected
by their author since 1838, and Mr. Muirhead reprints in 1846
Arago's judgment regarding the success of his own demonstration
of Watt's priority, which is as follows : ^' the settlement of a
question of priority, when it turns, as in the above instance, on
the most careful examination of printed memoirs, and the most
minute comparison of dates, assumes the character of a very
demonstration.^' •

Yet Arago's successors, not excepting Mr* Muirhead, who
also asserts (the itaUcs are his own) that the advocates of Caven-
dish ^^ can point out no inaccuracies in our statements of fact —
our dates — our references — or, we believe we might safely add,
our conclusions/'f have explicitly or implicitly denied the accu-
racy of many of Arago's statements. Few things, indeed, are
more strange in this strange controversy, than that some ten
years after Arago had pubhshed his alleged complete demon-
stration, and subsequently to the publication of the fFatt Cor-
respondence, in which Mr. Muirhead offers a demonstration still
more complete of Watt's priority, it should be conceded by Lord
Jefirey, in his review of the Correspondence, that it had not yet
been demonstrated that Watt had a claim at alL In that artide,
his Lordship, for the first time, made out a logically consistent
case for Watt, by endeavouring to show that he signified by
phlogiston, hydrogen. X
* Watt Corr., p. 233. f Watt Qfrr., p. oiii. t Sdm. J2ep., Jan. 1848.

CAUSS OF AKAOO'S ERRORS. 155

This curious result of matters has chiefly been occasioned
by nearly all the defenders of Watt looking at the question
between him and Cavendish^ through the distorting medium of
De Luc's unjust suspicions. For this^ Arago is chiefly to blame^
as having revived these in his Eloge of Watt. How seldom
foreigners^ however accomplished, can interfere to advantage in
contests between the rivals of another nation, whose works are
written in a language imperfectly known to the strangers,
is strikingly illustrated by the ill-success which has attended the
interference in the Water Controversy of two philosophers so
unusually gifted and accomplished as De Luc and Arago ; and I
will add, that the fact of neither having specially devoted him-
self to chemistry, has thrown another di£E[culty in the way of
their doing justice to the question of which they constituted
themselves judges. It has been affirmed, indeed, that the Water
Controversy is a question more of evidence than of chemistry, but
' to say this is to assert a distinction without a difference, for it is
a question chiefly of chemical evidence, and those who have laid
down this canon for their own guidance and justification, have
at once disobeyed it, and produced all the chemical evidence
they could procure to defend their views. More than half the
question, in truth, turns upon ih^ facts in dispute, not upon the
conclusions which they, if ascertained, would warrant, and the
greater number of the facts are chemical.

That men should take different sides in a vexed question
like the Water Controversy, cannot be matter of surprise.
There is room for different views being held by equally honest
and impartial observers. Mankind will never be at one in an
adjudgment of merit between Cavendish, Watt, and Lavoisier,
any more than they will be unanimous on other subjects which
concern them a great deal more ; nor shall I imitate those who
offer to furnish a demonstration, which in the end proves to be
a very one-sided decision. Where so much is doubtful, we must
be content to doubt, and I am greatly more anxious to exonerate
Cavendish from the charge of unfair dealing, than to insist upon
my estimate of his share in the disputed discovery as the just
one. I might be indifferent to Arago's imperfect acquaintance
with the necessary documents, if it affected only Cavendish's
intellectual reputation^ but I cannot pass it unoondemned when

156 LIFE OF CAVBNDISH.

it occasioned his honour to be called in question. Arago should
have studied the papers of Cavendish and Priestley much more
carefully than he did^ before he insinuated the shadow of a doubt
concerning the integrity of the former The eulogist of Watt
did not occupy the most favourable position for doing justice to
the great engineer's rival, and should have been especially cautious
how he judged him. The effect of his incaution has been most
mournful.

Who cannot but lament that for ten years some of the ablest
men of Britain have spent (nay mispent) their time, in con-
structing out of De Luc's idle suspicions a dark myth which has
obscured the good names of Cavendish, Priestley, Blagden,
and in truth of all the Fellows of the Royal Society in 1784,
from the President downwards ; besides extending its baleful
influence to the nameless printers of the Transactions, who
are involved in. suspicion because some unlucky man among
them, mistook a 3 for a 4 ? Watt, also, has been no gainer by
De Luc's ill-judged interference. Had he left matters alone,
there would probably have been no interpolations in Caven-
dish's paper, and his priority could not in some respects have
been defended in the way it now can. De Luc's successors
also, have (unconsciously as I believe) deprived themselves
in many cases, of the means of doing justice to the question
which they discuss, by their inheritance of his suspicions.
There is much reasoning in a circle, in the advocacy of Wattes
claims. De Luc's suspicions, for example, are accepted as
showing why an error of the press was permitted ; and the
error of the press is referred to as demonstrating the justice
of his suspicions. No one asks what Blagden^s character
was, but contents himself with referring to it, as if it were
what De Luc suspected it was, or rather perchance might
be. Whatever is doubtful becomes clear by a reference to
De Luc's uncharitable hypothesis, the rule being to accept it
as a true theory, and to put the worst constraction on every
thing obscure in which Cavendish and Blagden were concerned.
It is not surprising that the more extreme defenders of Watt
have done tliis so long, that they have totally forgotten that
neither De Luc nor Watt ever went beyond suspecting, and that
they both betrayed entire ignorance of the proceedings of Caven-

RELATIVE MERIT OF CAVENDISH, WATT, AND LAVOISIER. 157

dish. But it is a little strange that they should not have taken
more pains to do what De Luc and Watt left undone, namely,
prove that their suspicions were well founded. The reader
who desires to be impartial must keep this steadily in view. De
Luc and Watt are no authorities regarding the motives or the
conduct of Cavendish or of Blagden. We have all the evidence
on which they built their suspicions before us, and much more
of which they knew nothing, and we have no veil before our
eyes to hide or colour the truth, and no personal interest to
serve by taking a side. I cherish the confident expectation
that all who act impartially, will come to the conclusion that
Cavendish and Blagden deported themselves as honourable men,
and I hope also to persuade many that they had not occasion to
pilfer or connive at pilfering, inasmuch as the intellectual pro-
perty alleged to have been borrowed or stolen, was the fruit of
Cavendish's capital, and had been from the first in his posses-
sion. The discoverer of the composition of water in 1781 had
no occasion to borrow from its partial asserter in 1783.

I close this chapter by observing, that the conclusion regard-
ing intellectual merit to which I have come, is, that Watt did
not signify by phlogiston, hydrogen, and did not assert in the
equivalent terms of his own day, that water consists of hydrogen
and oxygen ; and further, that the conclusion to which he came,
such as it was, was arrived at later in time than Cavendish's
just conclusion, and was drawn from a repetition of his ex«
periments. For Cavendish I claim that he was the first
who observed and inferred that water consists of hydrogen
and oxygen; and to Lavoisier I assign the merit of having
simplified and perfected Cavendish's conclusion, and of
having been the first to prove the composition of water
by analysis. I acknowledge Watt to have been an independent
and original theorist on the composition of water, and to have
largely contributed to the dissemination of the true theory of
its nature. I must refer the reader, however, to the concluding
sections, and the summary at the close of the Water Controversy
for the fuller statement and justification of the views which I
have taken of the merits and demerits of the three great rivals.

158 LI9S or CAVENDISH.

CHAPTER IV.

CONCLUDING EVENTS OF CAVENDISH'S LIFE.— ESTIMATE OF
HIS MORAL AND INTELLECTUAL CHARACTER.

The year 1783, which has figured so largely in the preceding
diapter as the period when Cavendish announced one of his
great discoveries, was also an important epoch in his personal
history. In the spring of that year, his father died,^ and Henry
was free to pursue his own tastes uncontrolled by any one. The
incidents, however, of his later life are involved in little less
obscurity than those of his earlier days, but we have much
fuller details concerning his character in manhood than in
youth. How exactly he spent the thirty years which elapsed
between the period of his leaving Cambridge in 1753, and the
death of his father in 1783, is not at all certain. For reasons
which I have not been able to ascertain accurately. Lord Charles
Cavendish restricted his son to a snudl yearly pecuniary allow-
ance, and much importance has been attached to this fact by
certain of Cavendish's biographers, as afibrdingsome explanation
of the peculiarities of his character. The period when he
acquired possession of an ample fortune, is thus a point of
more importance, than at first sight might appear, in the history
of a man, who for a large portion of his life possessed immensely
more wealth than he knew how to spend. I am not at all
certain, however, from what quarter, or at what period. Caven-
dish inherited the riches which ultimately passed into his hands;
but from some facts to be presently mentioned, it would seem
that he had become a wealthy man before his father's death.
In the absence of any authoritative statement on this point,
I shall quote the declarations already made public regarding it,

• Collina' Peerage, addend, et corrig., vol. i. p. 565.

HIS POVERTY IN EARLY LIFE. 159

whidi are, no doubt, substantially true, although in several
respects they are incompatible with each other. Dr. Thomas
Thomson, who was acquainted with Cavendish, says in his
interesting sketch of him : — ^^ During his father's lifetime he was
kept in rather narrow circumstances, being allowed an annuity
of 500/. only, while his apartments were a set of stables, fitted
up for his accommodation. It was during this period that he
acquired those habits of economy and those singular oddities of
character which he exhibited ever after in so striking a manner.
At his father's death he was lefl a very considerable fortune ;
and an aunt, who died at a later period, bequeathed him a very
handsome addition to it, but in consequence of the habits of
economy which he had acquired, it was not in his power to
spend the greater part of his annual income/'^

Cuvier gives a fuller but somewhat different account of
Cavendish's early poverty, and the source of his wealth.f
'' Cadet d'une branche cadette, it ^tait assez pauvre dans sa
jeunesse, et ses parens le traitaient, dit-on, en homme qui
avait Fair de ne devenir jamais riche. Le hasard, ou son
m^rite reel, en decida autrement.

'^ Un de ses ondes, qui avait fait la guerre auz Indes, et qui
en rapportait une tres grande fortune, con9ut pour lui im
attachement particulier, et la lui laissa tout entiere. j:

* History of Chemittry, toL i. p. 336.

t I attach considerable importance to CuTier's statements, for although he
girei no anthoritj for them, it is probable that he derived his information from
Blagden, who must have been on terms of special intimacy with the French philo-
sopher, to whose step •daughter he left a legacy. It must be acknowledged, howeyer,
that as Blagden did not become CaTcndish's assistant till at least 1782, and probably
not till alter the death of Lord Charles Cavendish in 1783, he oonld only speak on
the authority of others as to the events of Cavendish's early history. He had peculiar
opportunities notwithstanding^ for learning these. I refer to this, because the Earl
of Burlington, a descendant of Cavendish's chief heir, cannot furnish any information
regarding the exact source of the philosopher's wealth. I am constrained, accordingly,
to quote Cuvier, unauthenticated as his statements are, because he is in some respects
the moet authoritative biographer of Cavendish. His position as secretary to the
French Academy, of which the subject of his Bloge was a member, gave him a right
which others did not possess, to demand information ; and his friend Blagden could,
directly or indirectly, afford him the most valuable assistance in collecting the mate-
rials for his Eloge, Cuvier also wrote in 1812, two years after Cavendish's death,
and therefore at a time when more, probably, could be learned concerning him than
at any later period.

* Bloget Hittoriquetf tome ii. pp. 102, 103.

160 LIFB OF CAVENDISH.

Bioty who wrote a year later than Cavier, assigns the year
1773 as the year when Cavendish became ^'le plus riche de
tons les savans*^' " Un de ses oncles/^ says he, *^ qui avait ete
General outremer, ^tant revenu de ses courses en 1773, avait
trouv^ mauvais que la famille exit neglige son neveu, et, pour
Pen dedommager, Tavait fait, en mourant, heritier de toute sa
fortune, qui se montait a plus de 300,000 liv. de rente/'*

I have not been able to discover who the General ^^ outre-
xner'^ was ; or whether it was an uncle, as Cuvier and Biot
assert; or an aunt, as Thomson states, who left Cavendish his
fortune. The question, assuredly, is one of no moment. It
would be of some importance, however, to ascertain the exact
period when Cavendish ceased to be dependent on the limited
income granted him by his father. His independence must
have been attained not later than 1783, for in this year (if not
in 1782) he settled an annuity of 500/. on Blagden. On this
view he was fifty-one years old, or, if we accept Blot's date,
forty-one, before he was the possessor of great riches. It is
certain, at all events, that for the first half, if not for a greater
part of his life, his pecuniary resources were extremely limited.
A gentleman, who had frequent interviews with him when
he visited the British Museum, understood " that he was kept
by his father on an allowance of only 120/. a-year until he was
forty years of age."* A senior member of the Royal Society
Club, also, who frequently met Cavendish at its dinners, learned
from Dr. Dryander, ^' that for some years Cavendish attended
the club regularly, and that he had only the five shillings in his
pocket to pay for the dinner, — not a penny more. His father,
it appears, allowed him to attend, and gave him the exact five
shillings to pay for the dinner .''J

Whatever obscurity or incongruity attaches to the statements
I have quoted, those who made them are manifestly at one as
to the two cardinal facts, that Cavendish was, for the- first forty
years of his life, a poor man, and for the last thirty-nine an
unusually wealthy one. No one seems able to assign a reason

t Biographie UniverseUe, tome vH. p. 456.
* Communicated by Charles Tomlinson, Esq.

X Ibid. " The expense of the dinner was limited to fire shillings, and black
pudding was always a standing dish."

CHARACTER OF LORD CHARLES CAVENDISH, 161

for his father's niggardliness towards him. There is good
reason^ however, for believing that the cause of Lord Charles
Cavendish's parsimony to his son has been misapprehended.
Dissatisfaction with Henry for not entering on public or pro-
fessional life, is the alleged ground of his father's illiberality
towards him. Cuvier, Biot, and Lord Brougham assert as much,
but give no authority for their statements, and I learn from
Robert Brown, Esq., that Lord Charles Cavendish was not a rich
man, and that he allowed his son as much as he could afford."*^
Mr. Brown also states that Henry's father, himself a scientific
man, appreciated his son's genius, and never treated him harshly
or unkindly. Our great botanist had much better and ampler
opportunities of estimating the character of Cavendish and of
his father than the French biographers of the former, or Lord
Brougham had. Mr. Brown's opinion on this point in truth
is authoritative, and it is entirely in keeping with the probabilities
of the case. It would assuredly have been very singular if Lord
Charles Cavendish had not taken peculiar interest in a son whose
scientific tastes were so congenial to his own, and who must
have displayed from an early period many marked disqualifica-
tions for success in a political or professional career.

Between 1783 and 1810, it is scarcely possible to select one
date as more entitled than another to be considered as marking
an important period in Cavendish's personal history. The
dates of his researches mark the great events of his life, but
those have been already given. His series of journeys might
at first sight seem to afford an exception to this statement, but
as the records of these were almost entirely limited to a bare

* Biot states, ** II n*eat pendant sa jeanesse que le sort r^erv^ en Angleterre
anx branches cadettes, c'est-ik-dire, une fortune tr^s mediocre. Cavendbh d^daigna
les emplois aoz quels sa naissance pouvait le porter, et ses parents, pr^nant sa
moderation pour I'apathie, s'61oignerent de lui,* '—Bioffr, Univ., tome 7, p. 455.
Blot's sketch, however, is not very accurate. He represents Cavendish as having
been bom in 1733, Instead of 1731, and as the second son of the Duke of Devon-
shire, whereas he was the grandson of the second duke by his third son. In his
estimate of Cavendish's wealth he is also in error, as Lord Brougham has shown
('* Life of Cavendish," in Lives qfMen qf Letters, p. 430). His Lordship, however,
states that Cavendish's ** family, aware of the talents which he early showed, were
anxious that he should take the part in public life which men of his rank are wont to
do, and were much displeased with his steady refusal to quit the studies which he
loved" (p. 430).

162 LIFE OP CAVENDISH.

summary of scientific phenomena, we learn exceedingly little
from them concerning their writer's personal history. The only
feet, indeed, which I have fbnnd in the diaries of those jour*
neys, which may be said to have an interest in reference to
Cavendish's personality, is a visit which, in 17^5, he paid to
his great rival. Watt, at Birmingham. The meeting was a
friendly one, and much intercourse must have occurred between
the philosophers, especially in reference to the researches and
inventions of Watt.*

No allusion is made to any difference having occurred
between the parties who met in August 17B5, for the first time
after their rivalry in 1784. Their early reconciliation, or rather
their refusal to act as if there ever had been a difference between
them, is a pleasing fact in the history of two men so estimable
as Cavendish and Watt are, and of whom it is painful to think
ill. Students of human nature tell us that we hate those to
whom we have done wrong, but there are no signs of hatred iu
Cavendish's visit to Watt. That so very shy and reserved a
person as the former should have sought out his rival and taken
the greatest interest in his inventions, is incompatible with the

* The following quotations from the MS. journal of CaTendish's journey in
1785» will iUuBtrate this :—

** At Birmingham we were informed by Dr. Withering and Mr. Watt, that the
part of the road through which we descended o£f the Lichny on the north side, when
fresh cut, appeared evidently to be a granulated quartz'' (p. 32).

" The fiBkshtonabie excellence of gilt buttons is, that they may look red^ mncb
like copper. For this purpose the gilt button, scarce polished up, is dipped into a
solution of some salts, amountiDg, as Mr .Watt said, to a kind ofaquaregia" (p. 34).

"Mr. Watt's new method of giving a circular motion by the steam-engine is
by making a small wheel fastened at the bottom of the bar suspended to the beam of
the steam-engine, pass round a larger wheel, without revolving at all on its own
axis" (p. 35;.

** Mr. Watt mentioned, that having found that some steam is condensed in the
cylinder of the steam-engine, though surrounded with steam, he made an experiment;
to discover what happened'* (p. 36).

" Mr. Watt thinks to have ascertained by experiment, that the less heat water is
converted into steam with, the more latent heat it requires to assume the elastic
form" (p. 37).

" Mr. Watt considers the heat of steam at 212 both sennble and latent, as near
1160" (p. 38).

*'The engine was of Watt's construction" (visit to Mr. Rathbone's works, Coal^
brooke Dale.) (p. 45).

*' The steam-engine at Bradley was upon Watt's construction" (p. 58).

'* Mr. Watt has contrived a furnace to bum the smoke, which he means ta
apply to the steam-engine. The draught of air is conducted backward" (p. 62).

cavendish's CIRCUIATING LIBRAHY. 163

hypotliesis, that on a solitary occasion be grudged him the
merit of a discovery, or robbed him of it. Watt, also, must
have welcomed the visit, and taken pains to explain to Caven-
dish all that was likely to interest him, for the references to the
engineer's doings are very minute, and Cavendish exhibits an
interest in his claims as an inventor, such as he rarely took in
questions of relative priority or originaUty. Thus, referring
to one of Watt's devices for regulating the motion of the steam
engine, he says, ^^ It was invented, or at least the patent for it
obtained, by a Mr. Picard, who has sold it to the present pro-
prietor ; but Mr. Watt claims the original idea.* It is not
uninteresting also to notice that Blagden accompanied Caven-
dish in this journey, and when I further add, that Watfs son
informs us that his father, ^' after becoming, in 1785, a Fellow
of the Royal Society, formed the personal acquaintance of Mr.
Cavendish, and lived upon good terms with him,''t I think I
may with confidence affirm, that any feeling of resentment
which had been entertained by Cavendish and Watt towards
each other, whilst strangers in 1784, was exchanged for mutual
respect as soon as diey met in 1785.

At this period Cavendishes reputation was widespread, in
spite of his solicitous endeavours to prevent himself becoming
fiEunous. It may be well, therefore, to refer here to his position
in London between the years 1783 and 1785, when his most
remarkable chemical researches were either made or published.
His town residence was close to the British Museum, at the
comer of Montague-place and Gower-street-J Few visitors
were admitted, but some found their way across the threshold
and have reported that books and apparatus formed its chief
furniture. For the former, however. Cavendish set apart a
separate mansion in Dean-street, Soho. Here he had collected
a large and carefully chosen library of works on science, which
he threw open to all engaged in research, and to this house
he went for his own books as one would go to a circulating
library, signing a formal receipt for such of the volumes as he
took with him.§

♦ Jtmmey in 1785, p. 33. t Watt Corr., p. iv. ,

X Infonnation supplied by Charles Tomlinson, Esq.

§ Ibid. Cavier, Eloget HUtor,, tome i. p. 104. ; Biot, Biogr, Univ. t. 7, p. 456.

M 2

164 LIFE OF CAVENDISH,

His favourite residence was a beautiful suburban villa af
Clapham^ which, as well as a street or row of houses in the
neighbourhood, now bears his name.* " The whole of the house
at Clapham was occupied as workshops and laboratory/^t " It
was stuck about with thermometers, rain-gauges, &c. A
registering thermometer of Cavendish's own construction,
served as a sort of landmark to his house. It is now in Professor
Brande's possession.^ A small portion only of the villa was
set apart for personal comfort. The upper rooms constituted
an astronomical observatory. What is now the drawing-room
was the laboratory. In an adjoining room a forge was placed.
The lawn was invaded by a wooden stage^ from which access
could be had to a large tree, to the top of which Cavendish, in
the course of his astronomical, meteorological, electrical, or
other researches occasionally ascended. §

The hospitalities of such a house are not likely to have been
overflowing. Cavendish lived comfortably, but made no display.
His few guests were treated, on all occasions, to the same fare,
and it was not very sumptuous. A Fellow of the Royal Society
reports, ^' that if any one dined with Cavendish he invariably
gave them a leg of mutton, and nothing else/'|| Another
Fellow states that Cavendish ^' seldom had company at his
house, but on one occasion three or four scientific men were to
dine with him, and when his housekeeper came to ask what was
to be got for dinner, he said, ^a leg of mutton I' ^ Sir, that will
not be enough for five.' * Well, then, get two,' was the reply.''

With this glance at Cavendish's style of housekeeping and
general social deportment, I pass on to the only two remaining
dates in reference to his personal history, which seem to require

* Cavendish Honse, Clapham Common, is a low white building, opposite the
6fth mile-stone from Cornhill^ now occupied by Mr. Herbert, whose lady has furnished
Mr. Tomlinson with some interesting anecdotes of the philosopher, given elsewhere.
He was an object of interest and perplexity to the residents in Clapham, among the
more ignorant of whom he passed for a wizard.

t Information supplied by Dr. Davy, who recdyed it from Mr. Newman, the
instrument-maker.

X Information given to Charles Tomlinson, Esq., by a Fellow of the Royal
Society.

§ Information furnished by BCr. Tomlinson, who visited Cavendish House
alld learned from its present occupant, and from Dr. Sylvester, of Clapham, the
facts which are embodied in the text.

II Information supplied by Mr. Tomlinson.

PLAYFAIR'S ESTIMATE OF CAVENDISH'S CHARACTER. 165

notice. The first is the 25th March, 1803, on which he was
elected one of the eight foreign associates of the French
Institute.* The second, and excepting the date of his birth,
the most important epoch in his histor}*, is 24th February,
1810, on which he died in his seventy-ninth year. The striking
circumstances of his death are mentioned in the sequel. He
was buried in All Hallows, or All Saints Church, Derby, where
Elizabeth Hardwicke built for herself a splendid tomb, round
which the ashes of many generations of her descendants rest in
peace beside her own.t

A more eventless life, according to the ordinary judgment of
mankind, than that of Cavendish, could scarcely be conceived.
His character, how^ever, was a very remarkable and interesting
one, and I shall try to explain what its prominent peculiarities
were. The most striking of these, at a first glance, was, a sin-
gular love for solitariness, and a reluctance to mix with his fel-
lows, which I may perhaps best denote by saying, that Caven-
dish was one of the most ungregarious of beings. The following
quotations from the writings of some of his more eminent con-
temporaries, who were well qualified to form an estimate of his
disposition, will illustrate this, as well as most of his other
characteristics.

Professor Play fair, of Edinburgh, visited London in 1782,
and was frequently present at the meetings of the Royal Society
Club, one of the very few places of comparatively public resort
which Cavendish attended. The impression made upon Play-
fair is thus recorded :— .

'^ Mr. Cavendish is a member also of this meeting. He is
of an awkward appearance, and has certainly not much the look
of a man of rank. He speaks likewise with great difficulty and
hesitation, and very seldom. But the gleams of genius break

* Biot, in Biographie Universelle, tome Til. p. 456.

f Elizabeth Hardwicke was a great proprietrix ia Derbyshire, and founded an
ahnshouse in the town of Derby. Her descendants, accordingly, were people of note
in the town and county, and were honoured after death by the citizens of the former
in a somewhat unusual way. A kind of public funeral was granted to all the Caven-
dishes, and therefore, I presume, to Henry ; for on the death, two years after him,
of his brother Frederick, he was buried " in the family vault, in All Saints, Derby,
the corpse being met, as when a Cavendish is buried has been customary, at the
entrance of the town by the mayor and thirty burgesses in mournmg.** —(Gentle*
man* 9 Magazine J 1812, p. 291). Whether this practice still continues 1 do not know.

166 LIFE OF CAVENDISH.

often through this unpromising exterior. He never speaks at
all but that it is exceedingly to the purpose, and either brings
some excellent information, or draws some important conclu*
sion. His knowledge is very extensive and very accurate ; most
of the members of the Royal Society seem to look up to him
as to one possessed of talents confessedly superior ; and, indeed,
they have reason to do so, for Mr. Cavendish, so far as I could
see, is the only one among them who joins together the know-
ledge of mathematics, chemistry, and experimental philosophy.''*

I have already quoted a reference to the same effect from
Dr. Thomas Thomson. He further states of Cavendish: —
^^ He was shy and bashful to a degree bordering on disease;
he could not bear to have any person introduced to him, or to
be pointed out in any way as a remarkable man. One Sunday
evening he was standing at Sir Joseph Banks', in a crowded
room, conversing with Mr. Hatchett, when Dr. Ingenhousz,
who had a good deal of pomposity of manner, came up with an
Austrian gentleman in his hand, and introduced him formally
to Mr. Cavendish. He mentioned the titles and qualifications
of his friend at great length, and said that he had been pecu-
liarly anxious to be introduced to a philosopher so profound
and so universally known and celebrated as Mr. Cavendish.
As soon as Dr. Ingenhousz had finished, the Austrian gentle-
man began, and assured Mr. Cavendish that his principal reason
for coming to London was to see and converse with one of the
greatest ornaments of the age, and one of the most illustrious
philosophers that ever existed. To all these high-flown speeches
Mr. Cavendish answered not a word, but stood with his eyes
cast down, quite abashed and confounded. At last, spying an
opening in the crowd, he darted through it with all the speed
of which he was master, nor did he stop till he reached his
carriage, which drove him directly home."t

Sir Humphry Davy, in addition to the eloquent eulogium
passed on Cavendish, soon after his death, left this less studied
but more graphic sketch of the philosopher amongst his papers :

* Worki of John Playfair^ toL i. appendix, LczziT.

t HUiory qf Chemittry, vol. i. p. 337. From Henry Lawson, Esq., of Lens-
down Place, Bath, who was one of the few admitted to the intimacy of Cavendiah,
I haye receiyed a similar account of his treatment of a French philosopher, which
perhaps, howeyer, is but another yersion of the incident related by Dr. Thomson.

ESTIMATE OF CAVENDISH BYDAVY, BROUGHAM, AND PEPYS. 167

«— '^ Cavendish was a great man, with extraordinary singularities.
His voice was squeaking, his manner nervous, he was afraid of
strangers, and seemed, when embarrassed, even to articulate with
difficulty. He wore the costume of our grandfathers; was enor*
mously rich, but made no use of his wealth. He gave me once
some bits of platinum, for my experiments, and came to see my
results on the decomposition of the alkalis, and seemed to take
an interest in them ; but he encouraged no intimacy with any

one He lived latterly the life of a solitary, came to the

club dinner, and to the Royal Society, but received nobody at his
own house. He was acute, sagacious, and profound, and, I think,
the most accomplished British philosopher of his time.^^*.

Lord Brougham gives the following account of his estimate
of Cavendish's character : — ^' He was of a most reserved dispo-
sition and peculiarly shy habits. This led to some singularity
of manner, which was further increased by a hesitation or diffi*
culty of speech, and a thin shrill voice. He entered diffidently
into any conversation, and seemed to dislike being spoken to.
He would often leave the place where he was addressed, and
leave it abruptly, with a kind of cry or ejaculation, as if scared

and disturbed He hardly ever went into society. The

only exceptions I am aware of are an occasional christening at
Devonshire or Burlington House, the meetings of the Royal
Society, and Sir Joseph Banks' weekly conversaziones. At both
the latter places I have met him, and recollect the shrill cry he
uttered as he shuffled quickly from room to room, seeming to
be annoyed if looked at, but sometimes approaching to hear
what was passing among others. His face was intelligent and
mild, though, from the nervous irritation which he seemed to
feel, the expression could hardly be called calm.''t

Mr. W. H. Pepys gives the following interesting description
of his interviews with Cavendish: — ^^^The first time I saw him
was at Sir Joseph Banks' house in Soho-square; it was a
general meeting of men devoted to science. I was relating to
Sir Joseph some experiments that I had been making with the
voltaic battery, when I observed an old gentleman in a com-
plete (faded violet) suit of clothes, and what was then termed a

* Daoy*9 Collected Worke, yqI. vii. p. 139.
t Livee of Men of Letters, 8fc,f pp. 444, 446.

168 LIFE OF CAVENDISH.

knocker-tailed periwig, very attentive to what I was describings
When I caught his eye he retired in great haste, but I soon
found he was again listening near me. Upon enquiry I heard
it was Mr. Cavendish, but at the same time was cautioned by
Sir Joseph to avoid speaking to him as he would be offended*
if he speaks to you, continue the conversation ; he is full of
information, particularly as to chemistry.

'*'! met him soon after at the Royal Society Club, and sitting
very near him he commenced some enquiry upon what I had
said at Sir Joseph^s, which plainly showxd he had remembered
me. His speech was hesitating and excited, but he was very
quick of comprehension/'*

Dr. Davy, who met Cavendish one or two years before his
death, gives a similar account of his appearance and manners.
^^ I well remember him, as he was in the habit occasionally^
between 1808 and 1809, of coming to the laboratory of the
Uoyal Institution, drawn there, no doubt, by the researches
then in progress. His dress was that of the gentleman of the
preceding half century. The frilled shirt-wrist, the high coat
collar, the cocked hat, in brief, almost the court dress of the
then and the present time. His appearance w*as, apart from his
dress, nowise distinguished : of fair complexion, small, and not
marked features, a feeble and somewhat hesitating voice. He
was then aged, turned, I believe, of seventy ; but though his
body seemed infirm, his conversation and queries denoted quick-
ness and acuteness, and undiminished vigour of mind, and that
I think, was the impression on my brother's mind, who always
held him, as his writings testify, in the greatest respect.''t

The following account of Cavendish is from one of our most
accompUshed chemists, who communicated it orally to Mr.
Tomlinson, from whom I received it. "When I was a very
young man — a new Fellow of the Royal Society — I always looked
upon it as a great honour to be noticed by Cavendish, and so
did the other young members of the society. We used to dine
at the Crown and Anchor, and Cavendish often dined with us.
He came slouching in, one hand behind his back, and taking off

* Letter from William Hasledine Pepys, Esq., to Lord Burlington, from
whom, with Mr. Pepys' sanction, I received it.
t Letter, April 9th, 1850.

ESTIMATE OF CAVENDISH BY HIS CONTEMPORARIES. 169

his hat (which by the bye he always hung up on one particular
peg), he sat down without taking notice of anybody. If you
attempted to draw him into conversation he always fought shy.
Dr. Wollaston's directions I found to succeed best. He said,
' the way to talk to Cavendish is never to look at him, but to
talk as it were into vacancy, and then it is not unlikely but you
may set him going !'^^

J. 6. Children, Esq., was often in the company of Cavendish,
and thus refers to his interviews with him ; '^ I am now the
Father of the Royal Society Club. I remember Cavendish well,
and have often dined at the Crown and Anchor with him. When
I first became a member of the club, I recollect seeing Caven-
dish on one occasion talking very earnestly to Marsden, Davy,
and Hatchett. I went up and joined the group, my eye caught
that of Cavendish, and he instantly became silent : he did not
say a word. The fact is he saw in me a strange face, and of a
strange face he had a perfect horror. I don't think I had been
introduced to him, but I was so afterwards, and then he behaved
to me very courteously. He was an old man when I joined the
club, and was regarded by all as a great authority •'' *

The most remarkable illustration, however, of Cavendish's
excessive shyness is, perhaps, that contained in the following
account of his reluctance to make his appearance at the soirees
of Sir Joseph Banks, which he frequently attended. It was
communicated by a senior member of the Royal Society to
Mr. Tomlinson, from whom I obtained it : "I have myself seen
him stand a long time on the landing, evidently wanting courage
to open the door and face the people assembled, nor would he open
the door until he heard some one coming up the stairs, and th^n
he was forced to go in.''

He was thus to appearance a misanthrope, and still more a
misogynist. He was reported among his contemporaries,
indeed, to have a positive dislike of women. Lord Burlington
informs me, on the authority of Mr. AUnutt, an old inhabitant
of Clapham, *^that Cavendish would never see a female servant,
and if an unfortunate maid ever showed herself she was imme-
diately dismissed." Lord Brougham tells us that Cavendish
" ordered his dinner daily by a note, which he left at a certain

* Reported by Charles Tomlinson , Esq.

170 LIFE OF CAVENDISH.

hour on the hall table^ where the housekeeper was to take it^
for he held no communication with his female domestics from
hb morbid shyness/^*

I might multiply illustrations of Cavendishes extreme repug-
nance to encounter females, but two additional instances may
suffice. At one time he was in the habit of walking in the
neighbourhood of Clapham with methodical accuracy at a
particular hour of the day. Two ladies, who watched his
movements, and had observed the punctuality with which he
reached the same spot at the same hour every day, took a
gentleman with them on one occasion to catch a sight of the
famous philosopher. '^ He was in the act of getting over a stile
when he saw to his horror that he was being watched.'^ He
never appeared in that road again, and his walks in future were
taken in the evening.f

The following incident occurred at a meeting of the Royal
Society Club, in the early part of this century, and was reported
by one of the most accomplished Fellows of the Society to Mr.
Tomlinson. ^^ One evening we observed a very pretty girl
looking out from an upper window on the opposite side of the
street, watching the philosophers at dinner. She attracted
notice, and one by one we got up and mustered round the
window to admire the fair one. Cavendish, who thought we
were looking at the moon, bustled up to us in his odd way, and
when he saw the real object of our study, turned away with
intense disgust, and grunted out Pshaw P'

In the preceding statements I have quoted largely and
verbatim from the materials at my disposal, that the reader
may have the means of forming an estimate for himself of the
difficult character of Cavendish. The portrait prefixed to this

* Livei qfMen qf Letten, Sfc, p. 446. CaTendisb, says another ant;horityj
** one daj met a maid servant on the stain with a broom and a pail, and was so
annojred that he immediately ordered a back staircase to be built"— (Inibrma*
tion commanicated to Mr. Tomlinson.)

f "His fovonrite, and, indeed, his only walk at Clapham, was down Nightingale
Lane, nearly opposite his house, from Clapham Common to Wandsworth Common,
and so round the road back to his own house. This walk he always took in the dusk
of the erening, and he always walked in the middle of the road, never on the side
path. He was never known to speak to any one, or to touch his hat to any one
who took off his. In short, his desire seemed to be alone and to be left alone."— >
(Information furnished by Dr. Sylvester to Mr. Tomlinson.)

. HIS PORTRAIT. 171

volume supplies an additional means of realising the appearance
of the philosopher^ as he was seen by those whose testimony I
have adduced.*

A striking unanimity pervades the references made to Caven-
dish by those who saw him^ and so far as the mere externalities
of his character are concerned, we cannot readily misapprehend it.
We picture him to ourselves an excessively shy, silent^ awkward,
and embarrassed person, barely enduring the looks of men, and
fleeing from the gaze of women. His reluctance to encounter
his fellows has been ascribed by some to the chilling influence
of the straitened circumstances under which his early life was
spent ; but, as I conceive, with no propriety. His poverty was
only relative, and his tastes were not expensive. Were I disposed
to impute to outward circumstances the development of an
individuality so well marked as his, I should lay most stress

* The Cavendish Society is under the greatest obligation to Charles Tomlinson«
Esq., for ^e disoorery of the portrait of Cavendish , and for the arrangements by
which he secured its being engraved at a merely nominal cost. The original is a
water-colour sketch contained in the print-room of the British Museum, of the
existence of which few were aware before Mr. Tomlinson drew attention to it. Dr.
Paris had obtained a copy, but so little was it known that a portrait was extant, that
Lord Brougham in his Xr(/« pf Cawndish (page 446) says, " It is not likely that he ever
should have been induced to sit for his picture ; the result, therefore, of any such
experiment is wanting." His Lordship's statement is true in one sense, though not
in that in which he intended it, for Cavendish certainly neither sat nor stood for his
likeness, and probably was not aware of its existence. Its history is singular, as
appears from a passage, extracted from a volume which Mr. Tomlinson found in
the library of the British Museum, entitled " Sketches qf the Royal Society
and Royal Society Club, by Sir John Barrow, Bart., F.KS, Lond. 1849.*'

Barrow, at the request of Alexander, ' the excellent draughtsman to the China
Embassy," who wished to take*a full-length portrait of Mr. Cavendish, applied to
Sir Joseph Banks, to know if the philosopher would consent to have his portrait
taken. Sir Joseph, in reply, assured Barrow that he would certainly receive a blunt
refusal, as had been the case with himself, on making the same request. He brought
about, however, a meeting between Cavendish and Barrow, who thus reports the
result: —

" I could not, however, find a favourable occasion for proposing the portrait,
and at last Alexander, who was bent upon having it, said, if I would invite him to a
club dinner, he could easily succeed, by taking his seat near the end of the table,
from whence he could sketch the peculiar greatcoat of a greyish green colour, and
the remarkable three-cornered hat, invariably worn by Cavendish ; and obtain,
unobserved, such an outline of the face as, when inserted between the hat and coat,
would make, he was quite sure, a full length portrait that no one could mistake.
It was so contrived, and every one who saw it recognised it at once. I think
Alexander told me he should leave it at the British Museum ; but whether it be
there I know not."

172 LIFE OF CAVENDISH.

upon the early loss of his mother^ whose *' affectionate kindness'*
his brother Frederick " frequently lamented that he had never
known.'** Kindly feminine care in early life, and especially that
of a mother or a sister, might have done much tx)wards infusing
human sympathies into Cavendish's heart, and rendering his
nature more affectionate and genial. He lost his mother, how-
ever, when he was two years old, and from his eleventh till his
twenty-second year was at school or college, so that there is great
reason to believe, that at the most critical period of his life, he
was not exposed to the salutary influences of a happy home,
which might have tempered the peculiarities of his character.
Whilst I say this, however, I by no means wish to exaggerate
the effect which his education among strangers may be supposed
to have produced upon him ; at best, his early orphanhood and
his comparative poverty can but have fostered singularities
over which external circumstances had, after all, little control.
Hundreds of youths have been poor, and motherless, as Caven-
dish was, and have, nevertheless, grown up warmhearted, generous,
and even enthusiastic men. Frederick Cavendish was exposed to
the same influences as his brother Henry, but became, notwith-
standing, an exceedingly cheerful, genial, and benevolent, though
somewhat eccentric man. The peculiarities, indeed, of a character
like Henry Cavendish's must be referred much more to original
conformation, than to anything else ; and whatever may have
been the restraints which his father imposed upon him, it seems
certain that one so widely connected with the aristocratic and
wealthy families of his country, as he was, might have procured
pecuniary and other assistance, towards the prosecution of any
lawful or honourable enterprise on which he wished to enter,
from some of his relatives. His brother, as well as himself,
followed no profession, and both became ultimately wealthy men.
Henry, in all likelihood, might have escaped from paternal
restraint, long before death released him from its bondage, had
he wished to have been free. All other causes, accordingly,
seem to me of slight importance, as sources of Cavendish's
peculiarities, compared with the influence which the strongly
marked original elements of his nature exerted upon him.

What these were, will appear as I proceed with the analysis

* Gentleman* 8 Magazine^ 1812, p. 289.

Ills INDIFFEHENCE TO HIS RELATIVES. 173

of his character. Whether from original or acquired indifference,
he exhibited from the first period when we have the means of
forming a judgment concerning him, a passive selfishness in all
his dealings. With his relatives he had very little intercourse.
The biographer of Frederick Cavendish says of him, that " For
his brother Henry he ever had a truly fraternal aflfection, which
seems to have been fully repaid, though they met but seldom."*
Had Henry, however, exhibited his regard for his brother largely
or openly, the terms in which it is referred to would have been
very diflFerentf His heir, Lord George^Cor^ndish, visited him
but once a-year, and remai q^j ^^jotiyliialf-an-hour at each visit.
Towards those not of his own blood, he was, if possible, still
more indiflferent. The only one whom he admitted to daily
intercourse with him, and that but for a few years, was Blagden,
and he finally became estranged from him. Sir Joseph Banks,
perhaps, knew him more intimately than most men did, yet after
all they were only acquaintances. Every one entitled to give an
opinion on the subject, to whom I have applied, gives the same
judgment on this point, which I may state once for all, in the
words of one of my informants, " Cavendish was the coldest and
most indifferent of mortals." This selfishness was entirely
passive, as I have already implied. Its strange betrayer might,
in his later years, have obtained for himself distinctions of all
kinds, but even scientific eminence, the only kind of fame for
which he cared, if indeed he cared even for that, he made no
struggle to attain, and he prevented his brother philosophers
firom placing him on the height to which they would willingly
have raised him, by keeping back many of his most remarkable
discoveries. He had thus the same law for himself as for his
brethren, and, after a fashion of his own, kept the golden rule,

* Gentleman*i Magazine, 1812, p. 291.

t The following is the only account I have received of intercourse between the
brothers. It was communicated to me by Mr. Tomlinson, who received it from a
PeUow of the Royal Society : —

"On one occasion Cavendish travelled in France with his brother Frede-
rick. On landing at Calais they stopped at an hotel, and in retiring for the night
passed a room^ the door of which was left open, and they sew in passing a dead body
laid oat for burial. Nothing was said at the time, but the next day the following
conversation took place between the brothers on their road to Paris :—

" ' Fred. C, loq. — Did you see the corpse?

«' ' Henry C, re*.— I did.' "

174 LIFE OF CAVENDISH.

and did unto others as he would have others do unto him. A
demand, accordingly, upon his sympathy, seems to have surprised
him. Sir Humphry Davy was indebted to him for ^* some bits
of platinum/' but tacitly appealed in vain for assistance in
prosecuting his electrical researches. ^' The last time I saw
Cavendish/' says Mr. W. H. Pepys (in the letter to Lord
Burlington, already referred to), ^ was at the Royal Institution,
in the apartments of Sir H. Davy (then Mr. Davy). It was
just before the subscription was entered into for the extended
voltaic battery, and upon Davy expressing regret that he feared
he should not obtain sufficient for the object, he [Cavendish]
joined most truly in deploring the want of liberality in the
patrons of science to carry it into effect. He did not seem to
think he was called upon to take any active step to forward the
desired object.'' Tet had any one asked Cavendish to sign a
cheque in Sir Humphry's name for 5002., he would probably
have done it at once. I infer as much at least from what Mr.
Pepys further tells us. — ^'At one time Mr. Cavendish had a large
library in London, which was in a bad state of arrangement. It
was proposed to him to allow a gentleman, who was not very well
off, to reside in the house, as being a clever man he would in
return arrange the books, and render the library more useful for
consultation, which Mr. Cavendish freely allowed. After this
gentleman had resided there a considerable time, and had suc-
ceeded in classing the books, he left to go to the country. Mr.
Cavendish, dining one day at the Royal Society Club, some
person present mentioned this gentleman's name, upon which
Mr. Cavendish said, ^^ Ah ! poor fellow : how does he do ?
How does he get on ?" ^^ I fear very indifferently/^ said this
person. " I am sorry for it," said Mr. C. " We had hopes
you would have done something for him, sir." " Me, me, me,
what could I do ?" *' A little annuity for his life ; he is not in
the best of health." ^^ Well, well, well, a check for ten thousand
pounds, would that do ?" '^ Oh sir, more than sufficient, more
than sufficient."

Similar acts of liberality are understood to have been per-
formed by Cavendish on other occasions. According to Cuvier,
'' il a soutenu et avanc6 plusieurs jeunes gens qiu annon^aient

HIS ANSWER TO HIS BANKERS. 175

des talens/^* Who those parties were does not appear. The
judgment^ however^ of a Fellow of the Royal Society, who had
good opportunities for coming to a conclusion on this point, is,
'^ that Cavendish did some good in a very ungracious manner ;^^t
and it could scarcely be expected, that one who took almost no
care of his own property, should concern himself much about
the prosperity of others. His famous answer to his bankers,
who were alarmed at the immense sum of money which had
accymulated in their hands, is the best proof of his indi£ference
to pecuniary afiairs. Dr. T. Thomson gave the first account
of this singular transaction.^: I give another version, from the
graphic pen of Mr. W. H. Pepys : —

'' The bankers where he kept his accounts, in looking over
their affairs, found he had a considerable sum in their hands,
some say nearly eighty thousand pounds, and one of them said,
that he did not think it right that it should lay so without
investment. He was therefore commissioned to wait upon Mr.
Cavendish, who at that time resided at Clapham. Upon his
arrival at the house he desired to speak to Mr. Cavendish.

" The servant said, * What is your business with him ^

'^ He did not choose to tell the servant.

^ The servant then said, * You must wait till my master
rings his bell, and then I will let him know.^

^^ In about a quarter of an hour the bell rang, and the banker
had the curiosity to listen to the conversation which took place.

"*^ Sir, there is a person below, who wants to speak to you.^

" * Who is he ? Who is he ? What does he want with
me?^

** * He says he is your banker, and must speak to you.'
'^ Mr. Cavendish, in great agitation, desires he may be sent
up, and« before he entered the room, cries, ^ What do you come
here for ? What do you want with me ?'

** * Sir, I thought it proper to wait upon you, as we have a
very large balance in hand of yours, and wish for your orders
respecting it.'

" ^ If it is any trouble to you, I will take it out of your hands.
Do not come here to plague me/

* Bloffe§ Hiai.f tome i. p. 104.

t Infonnation received from Mr. Tomlinton.

X Hittory ^f CkemiMtryy yoI. i. p. 336.

176 LIFE OF CAVENDISH.

" ' Not the least trouble to us, Sir, not the least ; but we
thought you might like some of it to be invested.'

« ' Well I well ! What do you want to do ?^

'^ ^ Perhaps you would like to have forty thousand pounds
invested/

" ^ Do so ! Do so, and don''t come here and trouble me, or 1
will remove it/ '**

In spite of this ungracious demeanour, and his undeviating
indifference to the affairs of his fellow-men, Cavendish awakened
an interest, almost reaching to affection, in some of those who
knew him. He had no vices. In the sight of man he was
blameless. The " good haters," whom Dr. Johnson loved,
would have been puzzled to justify themselves in hating Caven-
dish. No one could know him and not respect him. Many,
probably, longed to love him, but felt that he acted on them
like an electrified conductor on the bodies in its neighbourhood,
which it has no sooner attracted than it violently repels. Some
few, apparently, were able to make the transition from admiring
respect to loving regard. The following letter^ which had found
its way into the electrical MSS. of Cavendish, brings him more
within the circle of human sympathies than any other document
which I have encountered. It is dated 16th March, 1792. The
writer was an officer in the navy, and ultimately Hydrographer
to the Admiralty : —

*^ Dear Sib, — I was very sorry yesterday to hear that you
were prevented coming amongst us by an attack of the gravel ;
it brought to my recollection that old Balchier mentioned at the
Club one day, that nothing was more efficacious in that com-
plaint than lintseed tea. I hope, however, the complaint is
going off, as it was said you were better. That you may soon
come amongst us is the sincere wish of all your friends, and of
none more truly than of, dear Sir,

" Your most affectionate,
(Signed) "A. Dalrymplb/*

* Letter to Lord Barlington, communicated to me : " Mr. CavendUh [at the
period of his death] was the largest holder of bank -stock in England, and died worth
1,157,000/. in different public funds, the value of which was estimated at 700,0002."
He had besides a freehold property about 8,000/. a-year, and canal and other personal
property ; 50,000/. also were in the hands of liis bankers. — CC(ntleman*9 Magazine,
1810, p. 292.)

michell's letter to cavendish. 177

The Rev. J. Michell, also, who devised the apparatus with
which; in a modified form. Cavendish determined the density of
the earth, seems to have been a warm friend of the philosopher's.
Among the Cavendish MSS. I found a long letter from Michell
to him, dated 14th August, 1788. It refers chiefly to Geology,
and to a specimen of black lead which accompanied the letter.
Its conclusion runs thus : " with best respects to yourself, and
due compliments to all friends, when you see them, particularly
those of the Crown and Anchor, and Cat and Bagpipes Clubs,
I am, &c/'

There is in this passage an appeal to Cavendish's social
feelings and kindly human sympathies, which seems to show
that he possessed some warmth and geniality of character. It
Tvas with no little interest, accordingly, that I fell upon a paper
among the Cavendish MSS., evidently containing the draft of a
reply to Michell ; but I was disappointed. ^^ I am much obliged
to you *' says the writer, *^ and to Mr. B. for the plumbago, and
to you for your letter.'^ The properties of the plumbago, and
the geological questions raised by Mr. Michell, are then largely
considered, but no acknowledgment is made of his greetings,
nor does any reference occur to the Crown and Anchor or Cat
and Bagpipes Clubs.He Michell, perhaps, knew better than to
expect an answer on these points.

There are thus some faint glimpses of Cavendish occa-
sionally appearing among his fellow men in a capacity which at

* Hoping that some light might be thrown on Cavendishes character, by a
knowledge of the nature of the last of the Clubs referred to, I made enquiry con-
cerning it, and Mr. Tomlinson took a great deal of trouble in endeavouring to
discover the place of its assembling, and its object. For some time we were entirely
at fiialt ; but at length that valuable journal, Notes mid Queriet, solved the problem,
to far, probably, as it can now be solved. The Cat and Bagpipes, it appears, was
once well known : '* A public^house of considerable notoriety, with this sign, existed
long at the comer of Downing Street, next to King Street. It was also used as a
chop-house, and frequented by many of those connected with the public offices in
the neighbourhood.*' — {Notes and Queriee, Nov. 9. 1850, p. 397.) The nature of
the Cat and Bagpipes Club, of which Cavendish and Michell were probably mem*
bers, remains undetermined. One is tempted to imagine, that in the society of some
trustworthy, select few, Cavendish may have indulged in a temperate conviviality
and have unbent, for some half hour or so, from the rigid indifference which gene-
rally characterised him.

The Crown and Anchor was the tavern at which the Koyal Society Club held its
meetmgs*

116 UPE OF CAVBTa>ISH.

least redeems him from the charge of misanthropy. There
was at least one lady, also, who was ready to defend him from
the charge of being a woman-hater^ for she had to thank him
for saving her from the attacks of an infuriated cow. Unfortu-
nately she is dead^ or through her testimony we might have
learned that Cavendish did not hate women^ but was only
awkwardly shy and afraid of them.*

After all^ however, it must be acknowledged, that out of the
monVs cell, and the prisoner's dungeon, there have been very
few men who have lived for nearly four score years, and held so
little communication with their fellows, or made so few friend-
ships as Cavendish.

To the other objects of common r^ard which excite and
gratify the fiancy, the imagination, the emotions, and the higher
affections, he was equally indifferent. The Beautiful, the Sublime^
and the Spiritual seem to have lain altogether beyond his horizon.
The culture of the external senses, which the prosecution of
researches in the physical sciences^ secures to all who are suc-
cessful in their study, did nothing in Cavendish's case, to
quicken the perception of beauty, whether of form or sound or
colour. Many of our natural philosophers have had a strong
and cultivated aesthetical sense ; and have taken great delight
in one or other or all of the fine arts. For none of these does
Cavendish seem to have cared. Unlike Blacky he was indifferent
to elegance of form in his apparatus, which, provided it were
accurately constructed, might be clumsy in shape and of rude
materials. He insisted, however, on its perfect accuracy .f

* The lady in qneBtioii was Mn. Keer, foimerlj resident hi C^pfaam. She
appears to have taken mmdi interest in Cavendish's proceedings, probably firom a
feeling of gratitude, and was fond of referring to them. His interposition to save
her firom the mad cow, excited a great sensation in Qapham, where so mnch was
not expected from him, for he was considered a confirmed woman-hater. Mr.
TomUnson learned these particolars from Mrs. Herbert, the present oecapantof
CaTendish House, Clapham Common.

f Dr. Davy, who inherited a laige part of Cavendish's apparatus from Sir Hiim>-
phry, writes to me regarding it : — ** Cavendish seemed to have in view in constmc-
tion, efficiency merely, without attention to appearance. Hard woods were nevor
used> excepting when required. Fir wood (common deal) was that commonly
employed. The same disregard of mere appearances was shown in his laboratory.
A lady of rank (I believe it was the Duchess of Gordon), on paying Cavendish a visit
at Clapham, saw, I have heard it related, a long row of utensils never intended to

HIS INDIFFERENCE TO NATURAL SCENERY, 179

The grandeor of nafcural scenery — the changing aspects of
the skies — the striking differences between the inhabitants of
different parts of a region -^the historical associations insepa-
rable from certain localities^ and much else, on which Saussure^
Humboldt, Dalton,* Darwin, Forbes, and other scientific
pilgrims, expatiate so largely, find no place in Cavendish's
Journal of his Travels. Again and i^in, we read with expec-
tation; "At — I observed,^'— and look forward to something
remarkable being described, but the sentence ends — "the
barometer/' Cavendish crosses the country Uke a railway
surveyor; turning neither to the right nor to the left, or
deviating from his route only to make such announcements as,
that ^^ At Stroud we were informed that the old canal had not
yet paid any dividend to the proprietors, and that scarcely
anything but coals were brought up it, at the rate of 3^. a
ton/^f One solitary passage, however, I have found, which
infuses vitality and a human interest into these formal diaries;
and shows that their writer had deep down in his nature, the
common sympathies of humanity. He is describing the banks
of the Severn, and says, "The Terras-walk commands a remark-
able scene, from the singular appearance of these rocks all
around, but especially on the opposite side of the river Severn,,
the eastern, and from the fine view of the river underneath.
The remains of the old Castle, battered by Oliver Cromwell,
exhibit a remarkable instance of a leaning Tower, which pro-
duces a fine effect.'' j:

meet the eye, and on expressing surprise at their namber and arrangement, was
harried by them without explanation." They were employed in the crystallisation of
saline solntionsj by spontaneous eraporation.

The foUowmg anecdote, which Mr. Newman, the instrument-maker, of Regent
Street— ooommnicated to Dr. Dary and to Mr. TomUnson-— strikingly illustrates the
statement in the text : —

"My fikther for many years worked for the trade, and I remember a wind*gnage
which he made for Nairn and Blant, who had receiyed the order fix)m Cavendish. It
consisted of a train of wheels worked by a yaned fly, and it registered its lesalts in
the same manner as a common gas-meter does. When this anemometer was fimshed,
my father had to attend at Nairn and Blnnt's, where he met Cayendish, who insisted
upon his taking the whole apparatus to pieces, and then, by means of a file and a
magnifying glass, he tested the pinions to see that they were properly hardened and
polished, and of the right shape, according to fiis written directions."
* Meteorological Essays ; Appendix to second edition.

t Journey of 1785, p. 8. % n>id. p. 56.

N 2

180 LIFE OF CAVENDISH.

The character of Cavendish would be incomplete if I left
unnoticed his apparent irreligiousness. It may have been
only apparent* One so very reserved on ordinary affairs, is
likely to have been especially uncommunicative on a subject
which he might consider lay only between his Maker and
himself, I should be the last to pronounce judgment on his |

religious belief, for he gave no utterance on the subject, and
others can only surmise concerning it. It is part, however, of the I

legitimate function of a biographer, to chronicle the extent to |

which the moral and spiritual affections of our common nature ^
showed themselves in the subject of his sketch ; and even if I
wished to avoid the consideration of Cavendish's religious
character (which I do not), I should feel compelled to notice
the observations already published on this point. Biot states
that Cavendish ^^^tait d'une morale austere, religieux a la
mani^re de Newton et de Locke.*'* This reference (for which
Biot, no doubt, had some authority) is ambiguous, but I pre-
sume that it refers to the doctrinal faith of Cavendish ; for no one
would impute to him any manifestation of the religous earnest-
ness and fervour of Newton and Locke, who gave spontaneous,
public utterance to their religious convictions, and counted it a
duty and a pleasure to bear witness to what they believed to be
the truth* The significance of Biofs allusion is also rendered
doubtful, by the uncertainty that still prevails as to the exact
doctrine which Newton held concerning the Trinity.f It j

cannot be doubted, however, that the biographer intends to
signify that there was some peculiarity in Cavendish's religious
l)elief ; and that his views were, to a certain extent, what are
termed by theologians unorthodox,X and probably Arian or Unita- k

rian. Mr. Fuller, of St. Peter's College, Cambridge, writes to me
on this subject, '* I find there is a sort of hereditary belief here,
that Cavendish was not only a favourer of unitarian notions,

* Biographie UniveneUe, tome yii. p. 456.

t Into this vexed question I do not enter.. The reader will find opposite Tiewi
urged with great ability by Sir Dayid Brewster in his life of Newton (Famt/y
lAbrmy, No. 24), and by Professor A. De Morgan in his very interesting sketch of
the philosopher {Knighl't Cabinet Portrait Gallery, vol. xi. p. 109).

t I nse the word in its technical sense, as having a definite meaning attached to
it by all religions parties in this country.

HIS RELIGIOUS BELIEF UNCERTAIN. 181

but deddedly a Unitarian. I am not able to discover any
foundation for the belief except tradition.'^

Cavendish also, as already mentioned, is supposed to have
"t left Cambridge without a degree, from reluctance to submit to
the stringent religious tests applied in his day to candidates
for d^prees.*

This is all that I have been able to learn concerning his
doctrinal belief. Whatever it was, it did not lead to any open
confession of faith. Cavendish did not ally himself with any
religious body. He is understood to have ^^ never attended a
place of worship/'t The only service, indeed, in any degree
religious in which, so far as I can discover, he ever took part,
was the christening of some of his young relatives, and I am
not certain that he did more than attend the christening
dinner.!

A Fellow of the Royal Society, who had good means of
judging, states that, '* As to Cavendish's religion, he was nothing

^ * On this point Mr. Fnller lias fsroured me with the fdlowuig information : —

L "So &r as I can ascertain, there has never been any religious test at Cambridge

administered at matriculation, or at any period of a students uniyersity career pre-

^ yious to his taking a degree. In this respect we di£fer from Oxford, and we haye

I ' accordingly students of all denominations, religious and irreligious (and I eyen

I remember a Mussulman), who study here, but leaye us without taking any degree.

The first religious test is administered on taking the first degree — Bachelor in Arts,

Law, or Medicine, and it is a declaration to the effect that the candidate is hand fide

I a member of the Church of England as by law established. A candidate for a higher

degree — ^Master of Arts, or Doctor in eiUier of the foculties, must submit to a more
stringent test, yiz., he must sign the 36th Canon, the Articles, and the Liturgy of
the Church of England.

** Practically, it is found that the former test does not exclude many whose

opinions are somewhat at yariance with either the doctrine or practice of the Church,

\ but not widely so ; while the second test excludes many who do not outwardly

f dissent, but yet haye conscientious scruples on certain points.

I "When Cayendish was at Cambridge, the second and more stringent test was

administered to candidates for the Bachelor's degree as well as the Master's. The

change to the present system was introduced by grace of the Senate, dated June 23,

1772, and another grace dated March 26, 1779, prinripaUy, I belieye, through the

instrumentality of Archdeacon Paley."

t Information furnished to Lord Burlington by Mr. Allnutt, of Clapham, who
only, howeyer, professed to state his belief on the matter in question.

X Lord Burlington writes me : ** I have heard my grandmother say that he
[Cayendish] once came to a christening, and that it being at that time the custom to
make a present to the nurse, he put his hand in his pocket, and presented her with a
handful of guineas without counting them."

182 LIFE OF CAVENDISH.

at all. The only subjects in which he appeared to take any
interest, were scientific. An unaccustomed glow seemed to
come over him when some new point in mathematics was
spoken of ; but if the conversation relapsed into general topics,
or even the exciting politics of the day, he turned aside, and all
the cold indifference of his nature returned/^ *

This opinion seems confirmed by the references in Caven-
dish's Journals. The most important, perhaps, of all his
experiments on the composition of water, was made on a
Sunday ; and, in his journeys, all days of the week were alike,
so far as geological or meteorological observations were con-
cerned.

From what has been stated, it will appear that it would be
vain to assert that we know with any certainty what doctrine
Cavendish held concerning Spiritual things; but we may with
some confidence affirm, that the World to come did not engross
his thoughts; that he gave no outward demonstration of
interest in religion, and did join his fellow men in worshipping
God. What worship he offiered in private we do not know,
but the striking circumstances of his death prove that to the
last he excluded others from a knowledge of his belief or non-
belief in a future state, and in a God to whom he should be
required to answer for the deeds done in the body. He died,
and gave no sign, rejecting human sympathy, and leaving
us no means of determining whether he anticipated annihi-
lation, or looked forward to an endless life. I have reserved,
accordingly, the notice of his death till now, of which
several accounts have been given. Dr. T. Thomson writes, —

" When he found himself dying, he gave directions to his
servant to leave him alone, and not to return till a certain time
which he specified, and by which period he expected to be no
longer alive. His servant, however, who was aware of the state
of his master, and was anxious about him, opened the door of
the room before the time specified, and approached the bed to
take a look nt the dying man. Mr. Cavendish, who was still
sensible, was offended at the intrusion, and ordered him out of

* Information fumiBhed to CharlM ToaalinfOBi Eiq.

. HIS DEATH. 183

the room with a voice of displeasure^ commanding him not by
any means to return till the time specified. When he did come
back at that time^ he found his master dead.^^ *

Dr. Davy gives a slightly different account. His authority,
as he informs me, was a person of the name of Harrison, who
was at one time in the employment of Mr. Ramsden, the instru-
ment-maker, and afterwards for a period in that of Cavendish:^

^^ He died, I have been assured, in the most tranquil manner.
A person employed by him about his apparatus told me, that
the last thing Mr. Cavendish called for was a glass of water,
and then he desired to be alone ; his attendant being uneasy
respecting his state, retired to a distant part of the room. Mr.
Cavendish drank some of the water, turned on his side, and
shortly expired, without uttering a word or even a sound, much
in the mann^ of his illustrious contemporary. Dr. Black, who
died as if he had fallen asleep, with an unspilled basin of milk
on his knees, sitting in his chair.^^f

A still fuller and somewhat dissimilar description of the
closing scene of Cavendish's life, has been sent me by H.
Lawson, Esq., of Lands down Place, Bath :%

^' He went home one evening (I believe from the Royal
Society) and passed silently as usual to his study. His man-
servant observed blood upon his linen, but dared not ask the
cause. He remained ill for two or three days, and on the last
day of his life, he rang his bell somewhat earlier than usual, and
when his valet appeared, called him to the bedside, and said, —

'^ ^ Mind what I say — I am going to die. When I am dead,
but not till theny go to Lord George Cavendish, and tell him of
the event. 60 !'

** The servant obeyed.

'^ In about half an hour Cavendish rang his bell again, and
calling his servant to his bedside, desired him to repeat what he

• HUiory (ff Chemutiy, vol. i. p. 339.

t CoUeeted Work$ qf Sir Humphry Dopjf, edited hy Dr, Dmyj yoI. tIL p.

139.

X For a referenoe to this aequamtuioe of Cavendith'Sf I was indebted to Dr.
Davy, 1H10 obtained it from Mr. Newman, the instnunent^maker, of Regent Street
London.

184 LIFE OF CAVENDISH.

had been told, * When I am dead, ^c/— * Right Give me the
lavender water. Go/

" The servant obeyed^ and in about half an hour, having
received no further summons^ he went to his master^s room^
and found him a corpse/^

Dr. Elliotson, whose family, as he mentions, resided at
Clapham, so that he had good means of ascertaining the truth,
confirms the general accuracy of these accoimts of Cavendishes
death.* Nevertheless a description of the closing scene of his
life, considerably at variance with those quoted above, has been
given to the world by Sir John Barrow, on the authority
of Sir Everard Home. Sir Everard's veracity has been
called in question, but there seems little reason to doubt that
he was a faithful witness in this case ; and it is certain at leasts
that he gave a similar account to a Fellow of the Royal Society,
who reported it to Mr. Tomlinson. The substance of Sir
Everard's statement was, that Cavendish sent his servant out of
the house, '^ ordering him not to come near him till night, as he
had something particular to engage his thoughts, and did not unsh
to be disturbed by any one P' The servant, who believed his
master to be dying, summoned Sir Everard Home, who hastened
to Clapham. ^^ He found Cavendish in bed, very much exhausted^
and apparently in a dying state. Mr. C. seemed rather surprised
to see him there ; and said that Sir E. could be of no use to
him, for that he was in a dying state ; and blamed his servant
for bringing him (Sir Everard) down from town, for that at
eighty years of age he thought that any prolongation of lite
would only prolong its miseries. Sir E. insisted on remaining
with him during the night. The patient remained tranquil, and
shortly after daybreak departed this lifc^^f

After all, however, the various accounts of Cavendish's death

^ Pliynatogy, fi^ edition^ p. 1044.

t Sketches qf the Royal Society and Royal Society Club, by Sir John Barrow.
Sir E. Home, after Cavendish's death, examined his repositories, in the presence of
his servant, with the following result — according to Barrow :— " In one of the chests
of drawers they found many old-fashioned articles of old jewellery, parts of emhruidered
dresses, &c., and, among other valuahle articles, an old lady's stomacher so beset
with diamonds that when it came to be examined and valued I think Sir £. men*
tioned its worth as something like 20,000/."

HIS NEGATIVE MORALITY. 185

do not differ^ so &r as essentials are concerned ; and I would
willingly believe that the '^ something particular/^ which he told
his servant was to engage the undisturbed attention of his last^
and solemn, silent hours, was his [preparation for the unseen
world into which he knew he was about to pass.

Such, then, was Cavendish in life and in Death, as he
appeared to those who knew him best. The account I have
given of him has necessarily assumed the character of a Mosaic,
made up of fragments furnished by different hands. I have thus
supplied each reader with the means of drawing a likeness for
himself, and it only remains that I offer very briefly my own
estimate of the character of the Philosopher. Morally it was a
blank, and can be described only by a series of negations. He
did not love ; he did not hate ; he did not hope ; he did not
fear ; l^e did not worship as others do. He separated himself
from bis fellow men, and apparently from God. There was
nothing earnest, enthusiastic, heroic, or chivalrous in his nature,
and as little was there anything mean, groveUing, or ignoble.
He was almost passionless. All that needed for its apprehen-
sion more than the pure intellect, or required the exercise of
fancy^ imagination, affection, or faith, was distastefiil to Caven-
dish. An intellectual head thinking, a pair of wonderfully
acute eyes observing, and a pair of very skilful hands ex-
perimenting or recording, are all that I realise in reading his
memorials. His brain seems to have been but a calculating
engine ; his eyes inlets of vision, not fountains of tears ; his
hands instruments of manipulation which never trembled with
emotion, or were clasped together in adoration thanksgiving,
or despair; his heart only an anatomical organ, necessary
for of the circulation of the blood. Tet, if such a being,
who reversed the maxim '* nihil humani me alienum puto,''
cannot be loved, as little can he be abhorred or despised.
He was, in spite of the atrophy or non development of many
of the faculties which are found in those in whom the
'^elements are kindly mixed/^ as truly a genius as the
mere poets, painters, and musicians, with small intellects
and hearts and large imaginations, to whom the world
is so willing to bend the knee. He is more to be wondered at than
blamed. Cavendish did not stand aloof from other men in a proud

186 UFE OF CAVENDISH.

or supercilious spirit^ refusing to count them his fellows. He felt
bimself separated from them by a great gulf, which neither they
nor he could bridge over, and across which it was vain to stretch
hands or exchange greetings. A sense of isolation from
his brethren^ mode him shrink from their society and avoid
their presence, but he did so as one conscious of an infirmity,
not boasting of an excellence. He was like a deaf mute sitting
apart from a circle, whose looks and gestures show that they are
uttering and listening to music and eloquence, in producing or
welcoming which he can be no sharer. Wisely, therefore, he
dwelt apart, and bidding the world farewell, took the self-
imposed vows of a Scientific Anchorite, and, like the Monks
of old, shut himself up within his cell. It was a kingdom
sufiident for him, and from its narrow window he saw as much
of the Universe as he cared to see. It had a throne also, and
from it he dispensed royal gifts to his brethren. He was one of
the unthanked benefactors of his race, who was patiently teaching
and serving mankind, whilst they were shrinking from his cold-
ness, or mocking his peculiarities.* He could not sing for them
a sweet song, or create a ^^ thing of beauty ^' which should be
" a joy for ever,'' or touch their hearts, or fire their spirits, or
deepen their reverence or their fervour. He was not a Poet, a
Priest, or a Prophet, but only a cold, clear Intelligence, raying
down pure white light, which brightened everything on which it
fiell, but warmed nothing — a Star of at least the second, if not
of the first magnitude, in the Intellectual Firmament.

His Theory of the Universe seems to have been, that it
consisted solely of a multitude of objects which could be
weighed, numbered, and measured ; and the vocation to which
he considered himself called was, to weigh, number, and
measure as many of those objects as his allotted three-score
years and ten would permit. This conviction biassed all his
doings, alike his great scientific enterprises, and the petty details
of his daily life. Tlavra fih-pipf k^li apiOfi^y k^i araOfi^j

* CuTier reoounti this pleasing anecdote of Cayendish's austere liberality :— ^
" Un jour le gardien de ses instrumens Tint lai dire avec humear qu'un jeune homme
avait cass^ one machine tr^-prtfcieuse ; * H faui,* repondit-il, 'que lea Jetmee ffent
CMeeeni dee maehinee pour apprendre a e'en eervir ; faitu^en /aire une autre"-^
(Eloges Uistoiiqiies, t. i., p. 104.)

HIS THEORT OF THE UNIVERSE. 187

VTBB his motto ; and in the Microcosm of his own nature he
tried to reflect and repeat the subjection to inflexible rule, and
the necessitated harmony, which are the appointed conditions
of the Macrocosm of God's Universe. The little peculiarities
of his domestic affairs, which might otherwise appear trivialities,
on which only the spirit of idle gossip could dwell with relish,
have for me a much deeper interest as tokens of a strongly
developed will, which gave a singular consistency and unity to
all the proceedings of its possessor. Cavendish did all things
in the same spirit. He was a hero (to the extent of his heroism)
even to his valet-de-ohambre. Throughout his long life, he
never transgressed the laws under which he seems to have
instinctively acted. Whenever we cateh sight of him we
find him with his measuring-rod and balance, his graduated jar,
thermometer, barometer, and table of logarithms ; if not in his
grasp, at least near at hand. Many of his scientific researches
were avowedly quaniiiaiwe. He weighed the Earth ; he analysed
the Air; he discovered the compound nature of Water; he
noted with numerical precision the obscure actions of the ancient
element Fire. Each, like some visitor to a strange land, was
compelled to submit to a scrutiny, in which not only its general
features were noticed, but everything pertaining to it, to which
a quantitative value could be attached, was set down in figures,
before it went forth to the scientific world, with its passport
signed and sealed. The half-mythical calendar of the Hindoos
was submitted to the same ordeal, and made to yield consistent
numerical results. The electricity of the Torpedo; the freezing
of mercury ; the appearance of an Aurora Borealis ; the hardness
of a London pump*water; the properties of carbonic acid and
of hydrogen, and much else, were equally subjected to a canon
which knew of no limitations, and required that every pheno-
menon and physical force should be held to be governed by law,
and admit of expression in mathematical or arithmetical symbols.
It seems^ indeed, to have been impossible for Cavendish to
investigate any question otherwise than quantitetively. If he is
making hydrogen, he tells us how much zinc, or iron, or tin he
took; and what quantity of gas its solution in sulphuric or
muriatic acid yielded^ although he had no apparent purpose to
serve in measuring the volumes of elastic fluid produced. If he

188 LIFE OF CAVENDISH.

plunges a candle into a mixture of nitrogen and air^ or carbonic
acid and air^ he counts carefully the number of seconds during
which it bums, and with unwearied patience varies the propor-
tion of the gases. If he is preparing oxygen, he records in his
note-book the weight of mercury he took, the quantity of nitric
acid in which he dissolved it, and the amount of gas which the
resultant oxide of mercury yielded, although he need have
attended to nothing except that he had pure oxygen. It would^
apparently, have been painful to him to have experimented
otherwise. Nor was this all : he insisted on the trivial routine
of outward life, following a law as inflexible and imperative as
that which rules the motions of the stars. He wore the same
dress from year to year, taking no heed of the change in fashions.
He calculated the advent of his tailor to make a new suit of
clothes, as he would have done that of a comet, and consulted
the almanac to discover when the artist should appear.* He
hung up his hat invariably on the same peg, when he went to
the meetings of the Royal Society Club. His walking-stick
was always placed in one of his boots, and always in the same
one.f He dispensed charity by a singular numerical rule, not
according to the deserts of those for whom assistance was
craved, into whose wants he made no inquiry. He settled
beforehand the value of a commodity which he wished to
purchase, and referred to it as if its worth in money admitted
of as precise an arithmetical determination, as the com-

* ** Sea habUlemento ne changeaient jamais de forme, de oonleiir, ni de matiire;
oonstamment T^tn de drap gris, on saTait d'ayance, par Talmanacb, quand il fidlait
lui hire on habit neof, de quelle tftoffe et de quelle couleur il fallait le iaire ; ou ri,
par hasard, on onbliait I'^poque de oette mutation, il n^avait besoin, pour la rap-
peler, que de proferer ce aeul mot, le iaiileur,"^'CBioffraphie UniveneUe, tome vii.
p. 456.)

A similar account as to Cavendiah'spoitesiing no wardrobe, and owning but one
suit of clothes at a time, was given to Mr. Tomlinson by Mrs. Herbert^ of CaTendish
House, Clapham Common.

t " His boots were brought down and put against the dinmg-room door always
in one spot, and in one particular position, with the point of his stick standing in
one particular boot." — (Information given to Mr. Tomlinson by Mrs. Herbert.)

CuTier relates a similar fact : " Quand il montait h, cheral, il derait trouver ses
bottes toujours au m6me endroit, et le fouet dans Tune des deux, et toigours dans la
m6me !" Cuvier, however, probably mistook the whip 'for the walking stick, for it
does not appear that Cavendish was an equestrian.

HIS OBEDIENCE TO RULE IN ALL THINGS. 189

billing proportion of a chemical element or the orbit of a planet.*
When he rode out in his carriage, he measured the number
of miles which he travelled by a way-wiaer attached to the
wheels.t He would not take books out of his own library, without
giving a receipt for them, nor indeed willingly do anything
otherwise than in the most simple, uniform, and methodical
manner possible.

Such was he in life, a wonderful piece of intellectual clock-
work ; and as he lived by rule, he died by it, predicting his death
as if it had been the edipse of some great luminary (which in
truth it was), and counting the very moment when the shadow
of the unseen world should enshroud him in its darkness.

Whatever, accordingly, we may think of the ideal which Caven-
dish set before him, we must acknowledge that he acted up to it
withundeviating consistency; and that he realised it to a far greater
extent than most men realise the more lofty ideals which they set
before them. The pursuit of truth was with him a necessity,
not a passion. In all his researches he displayed the greatest
caution, not from hesitation or timidity, but from his recognition
of the difficulties which attend the investigation of nature ; from

* '* When any one called upon lum with a rahscription list for some charitable
or benerolent object, it was his custom to look do?m the list for the largegt sub*
scription. He would then puU out his cheque-book, and write a cheque for the
amount of the largest sum subscribed by any one indiyidual, neither more nor less.
This practice becoming known, some persons^ thinking, perhaps, that a small sin is
justifiable if it lead to a great good, would enter a large nominal amount in their
subscription list, and thus cheat Cavendish into a larger subscription than he would
otherwise have giTen."^-(Information supplied to Mr. Tomlinson by Dr. SyWester, of
Qapham.)

The following curious fragment of the draft of a letter I found among the
CaTendish MSS. on one side of a sheet of paper entitled ** Musical Intervals," and
ooenpiedwith figures. It is printed verbatim, but the italics are mine—
" Sir,

'' You would have heard from me sooner if it was not that I had [blank]

*' I forgot to ask you yesterday when you would have me return the plans yon
sent me. I would have told you yesterday how much I would give for the estate,
had it not been that it is so much less than what you said you had refused that I
thought it to no purpose. If, however, you have a mind, I will let you know what I
think it worth, and at the same time, at I hate kaglingt will tell you the utmost I
will give for it, but in that case you may depend upon it that I shall not offer any
more."

f The way-wiser (an antique wooden instrument) is now in the museum of
King's College, London, to which it was presented by Mr. Newman, of Regent
Street.

190 LIKE OF CAVENDISH.

his delight in reducing everything to numerical rule^ and his
hatred of error as a transgression of law. Cavendo tutus was the
motto of his fiimily^ and seems ever to have been before him.

He died as he had livedo taking no pains to perpetuate the
memory of a fame which could not be kept from proclaiming
itself^ even during his lifetime. His enormous wealth he left to
his relatives^ who would not have grudged, had he bequeathed
some small portion of his great possessions to the furtherance of
the sciences, to which bis life was devoted. But from hb
kinsmen he had received his wealth, and to them, increased a
hundredfold, he returned it.* His scientific successors, who
would have been grateful in their early struggles for some pecu-
niary help towards successfully prosecuting studies, which do not
always secure to their prosecutor even daily bread, have remem-
bered that he, who forget them in his last testament, forgot also
himself; and spent none of his wealth in prolonging his memory
upon earth. He has enriched us all, by his lessons and his
example, by his methods of research and his great discoveries;
and we have paid him only an honour which he deserved, when
we named ourselves after him, and founded a Cavendish Society.

* CaTendish left a oontiderable legacy to the Earl of Besborongh, who was not, I
beliere, a oonnexioii of his, in oonsequence, as is stated, of the pleasore he derived
from the EarFs conTersation at the Royal Society Clnb dinner. His Lordship was
not a man of soienoe. The immense bulk of Cayendish's wealth, however, went to
his brother, and to Lord Geoi^ge Cayendish and his family.

CAVENDISH AS A CHEMIST.

The published Chemical Researches of Cavendish ref^
chiefly to the gaaes^ and are contained in seven papers contributed
to the Phil. Trans. J and published at intervals from 1766 to
1788. I shall take up those papers in their chronological order^
as they form a series, naturally following each other. Their
consideration forms the best introduction to the discussion of
the controversy regarding the discovery of the composition of
water, which cannot, in truth, be understood without an ac-
quaintance with them. After its consideration, the three
remaining chemical papers, which treat of Congelation, will be
discussed.

Cavendishes pneumatic researches are remarkable for the
number of discoveries they unfolded. They contain, besides
less important announcements, the first full exposition of the
properties of hydrogen and carbonic acid; the demonstration
of the constancy in composition of aUnospheric air, and its first
tolerably accurate quantitative analysis; the record of the
&mous experiments which led to the detection of the non-
elementary nature of water, and by an extension, and slight
modification, to the discovery of the composition of nitric acid.

When Oavendish began those fruitful labours, pneumatic
chemistry had barely come into existence. More than one
chemist in different countries, and at different periods, had
noticed and described the production of permanent elastic fluids
as an accompaniment of chemical reactions. Paracelsus had
some slight acquaintance with hydrogen.* Van Helmont, the
introducer of the word gas, had distinguished more or less
explicitly carbonic acid, and certain of the combustible gaseous
compounds of carbon, and sulphur, with hydrogen.t Boyle

* Hoefer, HUU de la Ch., t. ii., p. 16.
t Qp. at,, t. u., pp. 142—144.

192 CAVENDISH AS A CHEMIST.

had encountered carbonic acid and hydrogen^* and Mayowf
was familiar with the latter*

Those chemists^ however^ had at best but a faint conception
of the individual gases^ as specifically distinct substances^ and
were too little acquainted with their unlike properties, to be
successful in convincing themselves or others, that each gas had
constant characters, by which its identity might always be recog-
nised. A belief that air might be generated^ de novo, sometimes
as it appeared, identical with atmospheric air, sometimes differ-
ent, was common probably to all the chemists of the latter
part of the eighteenth century. But beyond this they had not
got. The extension of pneumatic chemistry could result only
from a study of the differences which the several artificial airs
presented; but chemists paid little attention to those differences,
or explained them away, when they were too striking to escape
notice, and dwelt only, or chiefly, on the points of similarity, or
identity, between the gases they described, and atmospheric air.

To such a length was this exclusive consideration of the
common properties of atmospheric air ancl the gases carried, that
Stephen Hales, in his celebrated Statical Essays,]: pronounced
them substantially identical. From the details he gives, he
must have prepared in the course of his researches, oxygen,
hydrogen, nitrogen, chlorine, carbonic oxide, carbonic acid, sul-
phurous acid, and coal gas ; besides other gases. Nor did he
fail to observe the diversities in odour, colour, solubility in
water, combustibility, respirability and the like, which occurred
among those elastic fluids. Nevertheless, he looked upon them
as identical with atmospheric «air, because they agreed with it in
elasticity, and as it also seemed, from his inaccurate determina-
tions, in specific gravity. Their striking differences in sensible
characters, he regarded as resulting from the casual impregna-
tion of the one true air, with foreign matters, not as essential
and distinctive properties of specifically dissimilar elastic fluids.
After the fashion of his day, accordingly, he spoke vaguely of

* New Phy^ieO'Mechanieal Experiments, 1659; and New ExperimenU touching
the relation between Flame and Air, 1671.

t Hoefer, Hiet. de la Chim., t. ii., p. 268.

t Vegetable Statieke, collected in a vol. in 1727; and Haemattaticks, 1732.
The clieinical results are given in appendices.

DISCOVERIES OF HALES AND BLACK. 193

I the air being ^^ tainted^^ or " infected" with certain hypothetical

^'fumes/^ ^Wapours/' or "acid and sulphurous spirits."*

The Rev. W. V. Harcourt has dwelt at length on the com-
paratively clear apprehension which Boyle and his immediate
successors had of the fact, that there existed other permanently
elastic fluids than air.f But, in truth, their experiments went

^ no further than to show that a permanent gas was frequently

developed during chemical changes, and if they held that in
any case this gas was different from atmospheric air, it was but
an opinion. They entered into no proof of its justice, and,
for anything they published to the contrary, the gases they
examined might have been common air, altered in certain of its

■ properties by the intermixture with it, of other substances.

The approbation with which Hales' Essays were welcomed
over Europe, and the tacit assent which was accorded to his
general conclusion that there is but one true air, show how
slight and unabiding was the impression which Boyle's faint

I discrimination of *^ factitious airs," had made upon his suc-

* cessors. More than twenty years elapsed after the publication

of the Haemastaticks, before any one disputed the justice of

I Hales' views ; nor was it then done directly.

In 1754, however, the appearance of Black's celebrated
inaugural dissertation, demonstrated the existence of at least
one air, possessed of constant chemical properties, unlike those
of the atmosphere. I It proved this incidentally; for the chief

) * A fuller account of Hales' chemical labours wiU be found in the British

Quarterly Review for August, 1845^ pp. 229—233.

t London and Edinburgh Phil. Mag., Feb. 1846, p. 123.

X Considerable cpnfusion exists as to the date of Black's earliest publication on

fixed air, and the uncertainty which prevails on this subject has been increased by

the contradictory numbers which Prof. Robison gives in his edition of the Chemist's

I Lectures. The latter is made to say that the year in which his first account of fixed

[ air was published, was 1757 {Lecturee, vol.ii. p. 87). This, however, is certainly a

mistake, resulting from an oversight, either on the Author or Editor's part ; for a

printed copy of Black's Inaugural Dissertation is preserved in the Library of the

University of Edinburgh, and I find on its title-page the date 1754. The treatise is

styled Ditsertatio Medica Inauguralie de humore aeido a cihis orto et Magnesia

Alba, It consists of two sections; the first strictly medical, and apparently intended

to justify the presentation of the essay as a Dissertatio Medica; the second chemical,

, and containing the views of the Author on fixed fur. When the dissertation was

t published in English, the first section was omitted, and the second was entitled

'* Experiments upon magnesia alba, quick-lime, and other alkaline substances,"

and appeared along with an "Essay on Evaporation," by Dr. Cullen. The

194 CAYBNDISH AS A CHEMIST.

object of the essay was not to offer a formal denial of the pre-
vailing views^ regarding a one universal or elementary air^ but
to assign a reason for the difference in properties between the
caustic and mild^ alkalis and alkaline earths.

In his dissertation Black says very little about fixed air as
an elastic fluids and does not profess to have ascertained many
of its properties when free. He states distinctly his conviction
that it is different from common air, and gives some reasons for
his opinion; but he excuses himself from assigning to it a
specific name, and from entering on an exposition of the cha-
racters of the gas, ^^ which will probably be the subject of my
further inquiry .^^^

Nevertheless, Black made a great advance beyond all his
predecessors. Hales, confirming and immensely extending the
older views, had shown by the amplest evidence, that air or gas
was an abundant constituent of most substances. Black now
showed that in one case at least it was not less abundant than
important as an element of bodies. He entered into no minute
discussion of the secondary doctrine which his essay embodied,
and did not, in truth, except in the briefest terms, directly
enforce it, but the conclusion which his researches almost
unavoidably compelled, was, that a gas which by its absence or
presence, made all the difference between quick lime and chalky
between a mild and a caustic alkali, must be something quite
peculiar, and very unlike Common Air.f

copy from which I quote bean date 1782, and is called the fourth edition, but it ia a
simple reprint of the earlier issues. Black's experiments were first printed in
English in the second volume of ''Essays and Observations, Physical and
Literary, read before a Society in Edinburgh," p. 172. Thift volume appeared in
1770^ but Black's paper is dated June 5, 1755, the period, probably, when it was
read to the Society.

I note these dates, because soma discussion has recently occurred as to the
exact period of Black's discovery. Robison's date of 1 757 cannot be considered ^

as authenticated by Black, as it was not published till after his death; 1755, on
the other hand, was the date given during his lifetime, and is more trustworthy. |

To the world at large Black's opinions were not fully known till they were printed I

in the Edinburgh Essays in 1770; but he had publicly announced them to the Uni-
Tcrsity of Edinburgh in 1754, and taught them from the Chair of Chemistry in tiie |

University of Glasgow ftx)m 1756 downwards.

♦ Experiments upon Magneeia, &c., p. 72.

t M. Jaquin, who defended Black's general views against his foreign assailants, i

nevertheless held that fixed air was identical with common air. Hocfer, Hist, de la
Chim,, t. ii., p. 364.

INFLAMMABLE AIR. 195

Here Blacky so far as he has published his views in his
Experiments on Magnesia^ left the subject, content to show-
that carbonic acid had very marked properties whenfijced; and
Cavendish, who was to some extent anticipated by Macbride,
showed, twelve years later, that it had equally marked properties
when free. Many of their predecessors deserve most honour-
able mention in connexion with pneumatic chemistry, but
Black was the founder of the chemistry of the gases. The
word gas had no certain plural till his time, and Cavendish was
his acknowledged pupil.

His first communication on the gases is entitled ^' Three
Papers containing Experiments on Factitious Air,^' and was
published in 1766.* It begins with a definition of factitious
air^ as ^'any kind of air which is contained in other bodies in an
unelastic state, and is produced from thence by art.^^ This is
followed by a reference to Dr. Black, whom he states his inten-
tion of following, in applying the name Fixed Air to the gas
contained in the earthy and alkaline carbonates. He discusses
inflammable air, however, before carbonic acid. Of the former
he gives no definition, but he employs the name (inflammable
air) as one already in use, and familiar to his readers. Van
Helmont had pointed out that certain of the intestinal gases bum
with a peculiar flame.f Boylej: and Lemery§ recorded their
observation of the combustibility of hydrogen; a phenomenon
which it is probable many others also noticed. Hales had pre-
pared many varieties of combustible gas or inflammable air;
among others, coal gas. The fire-damp of mines had likewise
begun to attract the attention of scientific men;|| and the title
Inflammable Air^ appears to have come gradually into use, by
general consent, to distinguish all the known gases which were
combustible in air. Some such title was plainly necessary after
the recognition of fixed air as a distinct gas. It was a general
term applied to all combustible gases, but admitted of limitation
by connecting it with the source of the gas. Thus Cavendish

* Phil, Trafu., 1766, p. 141.

f Hoefer, Hist, de la Chim,, t. ii., p. 144.

X New Experiments touching the relation between Flame and Air, 1671.

§ Hoefer, Hist, de la Chim., t. ii., p. 297.

li PhU. Trans,, 1765, p. 219.

196 CAVENDISH AS A CHEMIST.

refers in his paper to inflammable mfrom the metals^ and to
inflammable air yrom putrefying animal matters.

As a necessary prelude to an account of his experiments and
their results, he gives a description of his pneumatic apparatus.
In its general arrangements it was identical with that of Hales,
and inferior to Priestley^s. Cavendish's pneumatic trough had
not the latter's simple but important addition of the Shelf, so
that, like Hales, he hung his gas-jars, or bottles, by strings,
with their mouths downwards, below the surface of the water.*
In other respects his arrangements were sufiicient, but more
effective than elegant. He made almost no advance on his im-
mediate predecessor in the invention of apparatus for collecting
and preparing gases, and left important improvements to be
suggested and introduced by his successors. Cavendish, in
truth, was not remarkable for an inventive spirit, but eminently
conspicuous for setting before him a standard of accuracy in
working, such as few of his fellow-chemists at that period cared
to acknowledge. His strong mathematical bias, induced him
to seek for quantitative results in all his researches, and
he modified apparatus to make this attainable, where the in-
struments in use were not of service. But if the apparatus
ready to his hand was sufficient for his purpose, he took it as
he found it, without spending time on its improvement. His
great caution and love of simplicity, made him averse to novel
or complicated arrangements, and he suggested very few. No
two persons, in truth, were more unlike in this respect than
he and Priestley, who was inexhaustible in contrivances, and
•unhesitating in trying them. The latter, I think, is entitled to
the first place among devisers and introducers of chemical
pneumatic apparatus; and next to him comes Hales, who
preceded him in time. Between them, they leave little merit
to be ascribed to Cavendish as a mechanical inventor, but he
made better use of his scanty apparatus as an analyst of the
gases than either of them did.

* Brownrigg, perhaps, gave the first idea of the Pneumatic Shdf in 1765. His
shelf, however, was above the level of the cistern or trough, on which it was fixed as
a perforated lid or cover, and the jars were prevented by wedges from sinking too
deeply through the holes in the board, which were wider than the jars. It was a
rack, therefore, rather than a shelf, and less convenient than Hales' and Cavendish's
method of suspension. {Phil, Trans., 1765, p. 235.)

IDENTITY OF HYDROGEN AND PHLOGISTON. 197

He divides his paper on factitious air into three parts. The
first treats of Hydrogen, the second of Carbonic Acid, and the
third of the Gases evolved during Fermentation and Putrefaction.
Some discussion has recently occurred as to whether or not
Cavendish should be regarded as the discoverer of hydrogen.
It seems needless, however, to raise the question. He does not
himself claim the discovery, which we have seen had been made
in the previous century by Boyle and others, but refers to the
gas as one already familiar to those he addresses. In one place
he prefaces his account of experiments on the explosibility of a
mixture of air and hydrogen, by the statement, ^^ it has been
observed by others.^^ Who those were, may be learned from
Dr. T. Thomson^s statement, in reference to hydrogen, that
^'its combustibility was known about the beginning of the
eighteenth century, and was often exhibited as a curiosity.*'*
He adduces two authorities in support of this statement, one, a
work {Cramer^s Elementa Docimasia) published in 1739.

The chief facts which Cavendish observed were the follow-
ing. Zinc, iron, and tin were the only metals which he found
to generate inflammable air when dissolved in acids, and that
only by solution in diluted sulphuric or muriatic acid. Zinc
dissolved in both acids with greater rapidity than iron or tin,
but yielded the same amount of gas, whichever acid was em-
ployed. Iron yielded the same quantity of inflammable air, in
specimens of dilute sulphuric acid of difierent strengths. Tin
dissolved best in warm muriatic acid. An ounce of zinc pro-
duced about 356 ounce measures of gas; the same weight of
iron 412, and of tin 202 ounce measures.

All those metals dissolved readily in nitrous (nitric) acid,
and generated air (nitric oxide), which was not inflammable.
They also dissolved with effervescence in hot oil of vitriol, and
discharged ** plenty of vapours which smell strongly of the
A'olatile sulphurous acid, and which are not at all inflammable.'^

From those observations Cavendish concluded, that when
the metals in question are dissolved in dilute sulphuric or mu-
riatic acid, '^ their phlogiston flies off, without having its nature
changed by the acid, and forms the inflammable air;'' but when
they are dissolved in nitrous acid or strong oil of vitriol, the

* Sytlem qf Chemiitryy 6th ed., yoI. i., p. 217.

198 CAVENDISH AS A CHEMIST.

phlogiston of the metals unites to the acid used for their solution^
and flies off with it in fumes^ and the phlogiston loses its inflam-
mability. The sulphurous acid which is evolved when oil of
vitriol is employed, is thus represented as being phlogisticated
sulphuric acid (as Stahl, indeed, named it) ;* a compound of the
phlogiston of the metals with the oil of vitriol of the acid. What
change the nitrous acid underwent. Cavendish was not certain;
but with that ready reference to phlogiston as the key to all
difficulties, which so strikingly characterises even the ablest
chemists of last century, he observes that the change which the
acid had undergone, *' can hardly be attributed to anything else
than its union with the phlogiston/^ The inflammable air, on
the other hand, he thought not likely to consist of any combi-
nation of phlogiston and acid, because its quality was the
same whether sulphuric or muriatic acid was used in preparing
it; and '^ also, because there is an inflammable air, seemingly
much of the same kind as this, produced from animal substances
in putrefaction ;^^ and ^' there can be no reason to suppose that
this kind of inflammable air owes its production to any acid.^'

From the preceding quotations it will appear that Caven-
dish believed the hydrogen which was evolved, to proceed, not
from the diluted acid, but from the metal as it underwent solu-
tion, and in truth to be the very phlogiston of the metal in the
gaseous form. This is the first announcement of the identity
of phlogiston with inflammable air, which ultimately became
one of the cardinal doctrines of the disciples of the later Phlo-
giston school. Cavendish afterwards changed his opinion, and
held that inflammable air was in all probability a combination
of phlogiston and water,t as will iully appear when his views
concerning the nature of water are under discussion. This
alteration of view has led to his earlier opinion being overlooked
in the course of the Water Controversy, and much unnecessary
criticism has been expended on his later, and as it is assumed,
his only conclusion concerning phlogiston.]: It seems well,
therefore, for the sake of its subsequent application, to notice
that not only did Cavendish originally hold that inflammable

* Kopp, Getchiehte der Ckemie, i. theil, p. 232.

t Phil. Trana,, 1784, p. 140.

X Watt Corresp., p. ciii., Bdinb. Re»,, January, 1848, pp. 103, 104.

DBTQNATION OF HYDROGEN AND AIR. 199

air was phlogiston, but he was the first, at least in England,
who broached this doctrine. It was afterwards taken up by
Priestley and Kirwan, and its truth apparently demonstrated by
special experiments, so that Cavendish referred to it as their
doctrine, not his own.* With him it was simply an hypothesis.
The reasons which induced him to change his view, I shall
afterwards consider, as they are important in reference to his
speculations on the composition of water. It is important,
however, to notice, that Priestley and Watt changed their opi-
nions concerning the nature of inflammable air in the same way.-f-
According to the final belief of aU three, it was what we should
now call a hydrate of phlogiston.

The properties of hydrogen which Cavendish observed
were the following. It did not lose its elasticity by keeping,
and was not sensibly absorbed by water, or by fixed or volatile
alkalis. Others had remarked its explosibility with common
air, and he proceeded to try the effect of varying the propor-
tions of air and hydrogen. A mixture of one part of inflam-
mable air and nine of common air would not burn at the mouth
of a bottle, but allowed a flame to spread through it. A mix-
ture, on the other hand, of 8 parts of inflammable air and 2 of
common air, burned, but did not explode. When about twice
or four times as much hydrogen was taken, a loud explosion
was heard. From these experiments Cavendish drew the gene-
ral conclusion, that inflammable air, like other inflammable
substances, '^cannot bum without the assistance of common
air,'' and that it must be mixed with more than its own volume
of the latter, to produce complete combustion. He seems,
however, to have over-estimated the volume of common air
required, for he mixed 2 volumes of hydrogen with more than 7 of
air, whereas 5 of the latter would have sufficed.

It is not a little curious that Cavendish should make no
reference in the record of his experiments on the inflammability
of hydrogen, to the appearance of moisture as an accompani-
ment or product of the combustion of the gas. He certainly
overlooked the phenomenon at this time, for at a later period
he referred its first observation to Warltire, who experimented
inl7Hl.t

* Phil. 7Vaii«M 1784, p. 137. t Qp. «/., p. 330. $ IHd., p. 126.

200 CAVENDISH AS A CHEMIST.

An additional series of trials was made^ with a view to
ascertain whether '^ the air produced from different metals by I

different acids/^ was equally inflammable* Five different sorts I

were used: !• A recently prepared specimen from zinc and
sulphuric acid« 2. A similar specimen which had been kept
for a fortnight* 3. Gas from zinc and hydrochloric acid. 4.
The same from iron and sulphuric acid* And 5. From tin and
hydrochloric acid. No difference could be observed in their
relative inflammability.

The specific gravity of the four last-mentioned varieties
was then tried, and the process followed is interesting as the
first successful attempt to compare the density of a gas with that
of common air. A bladder was used as the containing vessel*
It was filled with hydrogen^ emptied to get rid of traces of
atmospheric air, and filled again with the former gas* It was
then weighed, and thereafter the hydrogen was replaced by
air, and the bladder weighed a second time. A bladderful of
hydrogen was thus found to weigh about 41 grains less than
the same volume of common air. If, therefore, the density of
air be assumed to be 800 times less than that of water (which
Cavendish thought must be near the truth), then hydrogen will
be 7 times lighter than common air; but if the latter be 850
times lighter than water, as Hauksbee estimated it to be, then
hydrogen will be nearly 11 times lighter than air. Either of
these numbers, or the intermediate one which would have been
obtained had air been taken as 815 times lighter than water^
are,*! need not say, much too high for pure hydrogen. Caven-
dish^ however, was so far at least aware of this, and points
out the uncertainty that attended the settlement of the density ,

of common air, and the difficulty of preventing air from i

mixing with the hydrogen, and the diffusion through the <

latter of water-vapour. He determined the amount of water by i

forcing a known quantity of hydrogen through a glass tube
containing pearl-ashes, which he weighed before and after the
passage of the gas* In this way he found inflammable air to
contain nearly -J^th of its weight of moisture. To check the
results obtained with the bladder, he made another series of
determinations of the specific gravity of hydrogen, upon a dif-
ferent principle. His apparatus exactly resembled one of those

DENSITY OF HYDROGEN. 201

employed at the present day for ascertaining the amount of
carbonic acid in a limestone. It consisted of a glass bottle
nearly filled with dilute sulphuric acid, to the neck of which a
drying tube was luted containing pearl-ashes in coarse powder,
to arrest the water-vapour which accompanied the hydrogen.
This apparatus was carefully weighed, and also a portion of
zinc. The latter was thereafter introduced into the bottle, the
tube luted, and the whole left till the metal had entirely dis-
solved. The apparatus was then weighed a second time, and
the weight of the hydrogen which had been discharged thereby
ascertained. The volume of gas which this weight represented
was learned by a reference to former experiments, in which the
number of grain measures of gas which a given weight of zinc
would evolve^ had been ascertained. In one of the trials in this
way, of the density of hydrogen, 254 grains of zinc had been
dissolved. From the previous experiments, that weight of
metal must have set free 90,427 grain measures of hydrogen, and
the second weighing showed that the weight of this was 10|
grains. The density of hydrogen as determined in this way was
10^ times less than that of air.

Similar trials were made with hydrogen from zinc and hydro-
chloric acid, and from tin and hydrochloric acid. '' By a medium
of the experiments, inflammable air comes out 8,760 times
lighter than water, or 11 times lighter than common mP
Hydrogen, however, in reality is 14*4 times lighter than air.

Cavendish appears to have been the first who employed that
important little pneumatic instrument, the drying tube, for
depriving gases of moisture. His second method of determin-
ing the density of hydrogen was both original and ingenious;
and his first was unexceptionable in principle, and brought out
a result, such as no previous experimenter had obtained. Caven-
dish, indeed, has been referred to as *^ the first person who at-
tempted to determine the specific gravity of airs, by comparing
their weight with that of the same bulk of common air.*** Mr.
W. V. Harcourt, however, has shown that Hauksbee as well as
Greenwood preceded Cavendish in this attempt. t So also did

* Thomson's Hut, of Chemistry, yol. i., p. 343; and Lord Brougham's Lives
of Men cf Letters and Science of the time of George IlL, p. 431.
t Land, and Edinh, Phil, Mag., Feb. 1846, pp. 120, 121.

202 GAYENDISH AS A CHEMIST.

Hales, who compared the weight of 540 cubic inches of ^ air of
tartar' with that of the same volume of common air. He used
a pear-shaped glass vessel, open below^ the mouth, of which he
closed with a piece of bladder during the weighing.*

None of those observers, however, detected any difference
l>etween the density of the gases they examined and that of J
common air; partly, because* as in Hauksbee's case, the difference
was small, but chiefly, because the vessels Were too large, and
the balances they employed not sufficiently delicate. Hales,
too, had evidently prejudged the question, and did not expect
his airs to differ in specific gravity. On the other hand^ he
thought a common density next to a common elasticity, the
best proof that the various gases he prepared were ^^true air,
and not a mere flatulent vapour/'f

Cavendish, then, was not the first who investigated the
specific gravity of the gases, but he was the first who ascertained
that they have different densities.

The paper on inflammable air concludes with an account of
trials, as to whether or not it could be obtained by the action of
copper on hydrochloric acid. Cavendish found that inflamma-
ble air could not be procured in this way, but that a gas was
produced which " immediately loses its elasticity, as soon as it
comes in contact with the water.*' This elastic fluid, which
was gaseous hydrochloric acid, he did not examine. In 1772,
Priestley repeated this experiment, and speedily discovered that
neither copper nor any other metal was essential to the evolu-
tion of the condensible gas, which was yielded abundantly by
spirit of salt when it was heated, and could be collected over
mercury. He called the elastic fluid Marine Acid Air.:|:

The title of the second part of Cavendish's paper runs thus:
Eauperiments on Fixed Air, or that species of Factitious Air which
is produced from Alkaline Substances j by Solution in Adds, or by
Calcination.

The properties of carbonic acid, apart from its relation to
the mildness and causticity of alkalis, had been investigated to
some extent, as we have seen already, by Dr. Black and Mac-
bride before Cavendish studied them. As some difference of

* Vegetable Statickgy 2nd ed., p. 190. f Op. et he, eit,

% Bxperimentt on Air, 1775, Tol. L, p. 143.

BLACK ON FIXED AIR. 203

opimon has been expressed recently^ as to the extent to which
Dr. Black had anticipated Cavendish in reference to carbonic
acid^ it is desirable to notice that the former published no
detailed account of the characters of fixed air.^ From his lec^
tures it appears that^in 1754 he had discovered many of the
properties of carbonic acid^ and these, it cannot be doubted, he
exhibited to his students at the University of Glasgow, from
1756 downwards. It is certain, however, that, unless in his
lectures, he published nothing on the subject, except his
Inaugural Dissertation of 1754; and a reference to that work
will decide whether Cavendish took up new ground in 1766, or
only repeated what Black had already made known, not merely
to his clasS'pupils, but also, through the press, to students of
science at large. The only properties of free carbonic acid,
which are referred to in the ^^Experiments upon Magnesia
Alba, &c./' (which it will be remembered is the Chemical
Section of the Inaugural Dissertation in an English dress,) are
its solubility in water, and its production of a precipitate with
lime-water.f

Black thought that it performed the function of an acid,
and '^that as the calcareous earths and alkalis attract acids
strongly, and can be saturated with them, so they also attract
fixed air, and are in their ordinary state saturated with it;^^:t but
he makes no reference to its taste, or to its action on colouring
matter, as proofs of its possessing acid characters.

The only other point, except its functional acid character,
upon which he insists, is, that fixed air is quite distinct from
common air; but he reserves an investigation into the points of
difference for a future research, which he only partially com*-
pleted, and never made public through the press. The following
passage will show how much was left undone by Black: ^' Quick-
lime, therefore, does not attract air when in its most ordinary
form, but is capable of being joined to one particular species
only, which is dispersed through the atmosphere, either in the

* Lord Brougham's Livw qf Mm qf Letters and Science of George IIL*$
Reign, p. 330; and Harcourt's Letter, Lond, and Bdinb, Phil Mag,, Feb. 1846,
p. 118.

t Experiments on Magnesia Alba, p. 56.

X Op, eit., p. 50.

204 CAVENDISH AS A CHEMIST.

shape of an exceedingly subtle powder^ or more probably in
that of an elastic fluid* To this I have given the name of fixed
air J and perhaps very improperly; but I thought it better to
use a word already familiar in philosophy^ than to invent a new
name^ before we be fully acquainted with the nature and proper-
ties of this substance^ which will probably be the subject of my
further inquiry/**

Cavendish prepared carbonic acid by dissolving marble in
muriatic acid* , He found that the gas was soluble in water, and
was rapidly absorbed by the caustic alkalis^ but might be pre-
served over mercury for upwards of a year, without any loss of
elasticity or change of property. To determine the extent to
which carbonic acid is soluble in water, he made use of a mer-
curial pneumatic trough, a piece of apparatus which Priestley
has been supposed to have been the first to employ. Into a
graduated jar filled with mercury, he passed up measured
volumes of gas and water at intervals, and ascertained in this
way " that water, when the thermometer is at 55°, will absorb
rather more than an equal bulk*' of the fixed air. In the course
of these experiments, however, he found that water did not
always absorb the same amount of gas; and conceiving the
latter to be pure, he drew from this observation the conclusion,
that the ^^ fixed air contained in marble consists of substances
of different natures, part of it being more soluble in water
than the rest.** On this opinion of Cavendish*s, Black passed
the sagacious criticism: ^^I suspect, however, that this was
a deception, proceeding from the common air which water
contains, and which arises with the fixed air during the extrica-
tion of this last from the alkaline substances. '*t With this .
criticism, another celebrated chemist concurred. ^^ Dalton has J
since given,** says Dr. Thomson, "a satisfactory explanation of '
this seeming anomaly, by showing that the absorbability of
fixed air in water is proportional to its purity, and that when
mixed with a great quantity of common air, or any other gas
not soluble in water, it ceases to be sensibly absorbed.**^ It
has been overlooked, however, that Cavendish was aware of the

* Experiments on Magnesia Alba^ p. 72.

t Lectures by Robison, vol. ii., p. 91. <

X Thomson* 9 Hist, qfCAem,, vol. i., p. 342.

DENSITY OF FIXED AIR. 205

fact pointed out by Dalton. In 17B4^ the former wrote thus:
^* Though fixed air is absorbed in considerable quantity by
water, as I showed in Phil. Trans,, vol. Ivi., yet it is not easy to
deprive common air of all the fixed air contained in it, by
means of water* On shaking a mixture of ten parts of common
air to one of fixed air^ with more than an equal bulk of distilled
water^ not more than one-half of the fixed air was absorbed.*^*

In continuation of the inquiry. Cavendish ascertained that
cold water dissolves more carbonic acid than hot; and ^^that
water heated to the boiling-point is so far from absorbing air,
that it parts with what it had already absorbed.^^ Spirit of
wine, the specific gravity of which is not given, was found at
the temperature of 46^ to absorb ^' near 2^ times its bulk of the
more soluble part of this air/' Olive-oil very slowly absorbed
more than an equal bulk of fixed air, the thermometer being
between 40® and 50°.

The specific gravity of carbonic acid was determined with a
bladder in the same way as that of hydrogen had been. Caven-
dish inferred its density to be 1*57, air being 1*0. This is a
very fair determination, if allowance be made for the imperfec-
tion of the apparatus, and the presence of both water and
muriatic acid in the gas, which passed directly from the gas-
bottle into the bladder. In consequence of these impurities
and the imperfections of the method adopted, the specific gra-
vity of carbonic acid appeared to be greater than it really is.
According to the more careful trials of recent observers, it is
1*529^ not 1*570* A series of experiments was made on the
influence of carl^onic acid in arresting combustion, which led to
the observation of the curious fact, that the presence of a com-
paratively small proportion of fixed air in common air, is suffi-
cient to deprive the latter of the power of supporting flame.
Thus a small candle burned 80 seconds in a closed jar full of
common air. When the same receiver contained one part of
fixed air to 19 of common air, the candle burned 51 seconds.
When the fixed air was -^ths of the whole mixture, it burned
23 seconds, and when the fixed air was ^th of the whole, 1 1
seconds. It was extinguished immediately when the air con-
tained less than ^th of its bulk of fixed air. Cavendish draws

♦ PhiL Tram., 1784, p. 122.

206 CAVENDISH AS A CHEMIST.

attention to the circumstance that the size of the candle is an
element of importance in such trials, and that it must bear a '
certain proportion to the capacity of the gas-jar^ because ^' large
flaming bodies will bum in a fouler air than small ones/' This, \
however, did not affect the validity of his conclusion^ ^^that the |
power which common air has of keeping fire alive, is very much t
diminished by a small mixture of fixed air/' Later inquirers
have confirmed and extended this conclusion^ which still
remains, as Cavendish left it, without any theory having been
offered in explanation of it.

The last series of experiments which Cavendish undertook
in connection with this inquiry, had for its object the determi-
nation of ^* the quantity of fixed air in alkaline substances.'' As
may be anticipated, his quantitative analyses are far from accu-
rate, but they are interesting, from the period at which they
were made; and the principles on which they were conducted
are in many respects identical with those followed at the present
day. Marble was analysed by finding the loss of weight which
it underwent when dissolved in hydrochloric acid, contained in
a weighed flask provided with a drying tube, which was filled
with shreds of filter-paper to arrest moisture, instead of pearl-
ashes; '^ for," says Cavendish, ^^ pearl-ashes would have absorbed
the fixed air that passed through them." Carbonate of ammo-
nia effervesced too violently to be examined in this way. For
its analysis, three vials were taken, and weighed with their
contents in the same scale. One contained weak muriatic acid;
the second held some lumps of carbonate of ammonia, and was
corked to prevent evaporation of the salt; tlvs third, in which
the contents of the other two were to be gradually mixed, con-
tained a little water, and had a paper-cap to arrest the small
jets of liquid thrown up during the effervescence. When it had
ceased, the three vials were again weighed, tad die loss appears
to have been set down as carbonic acid, without any deduction
for the accompanying moisture. The general conelusion was,
that in proportion to the quantity of acid it can- satt!lrate, carbo-
nate of ammonia contains much more carbonid acid than marble
does; but that different specimens of the ammoniacal salt differ
considerably in composition. Cavendish then applies this
observation to explain a phenomenon which had greatly per-

ANALYSIS OF CARBONATES. 207

plexed him; namely^ the occurrence of effervescence when a
neutral solution of chloride of calcium was added to a solution
of carbonate of ammonia. This effervescence he explains by a
reference to the fact^ that as lime requires less carbonic acid to
saturate it^ than is present in the salt of ammonia^ the excess of
gas which the lime* cannot absorb flies off in an elastic form.
He refers^ in like manner, the non-precipitation of a salt of
magnesia by carbonate of ammonia, to the alkaline earth being
held in solution by the large amount of carbonic acid in the
ammoniacal salt.

Pearl-ashes were analysed in the same way as carbonate of
ammonia, with the substitution of diluted sulphuric for muriatic
acid. When the effervescence was over, the neutral liquid
was tested for free carbonic acid, as it was in the other ana-
lyses also, by the addition of lime-water. The precipitate which
was produced in the case of the pearl-ashes, was collected, dried,
and weighed; the proportion of carbonic acid in it, calculated on
the assumption that the precipitate was identical in composition
with marble, and added to that represented by the loss of weight.
The last carbonate which Cavendish analysed was bicarbon-
ate of potash. Availing himself of a suggestion of Dr. Black's,
he prepared this salt (probably for the first time), by slowly
forcing carbonic acid from a bladder communicating with a
gas-bottle, into a solution of pearl-ashes. Crystals gradually
formed, which were analysed in the same way as the carbonate
of ammonia had been. In preparing them, Cavendish was led
to suspect that the carbonic acid from marble is not homoge-
neous in composition, but consists of portions not equally
soluble in caustic alkalis, as he had formerly supposed it to
be a mixture of gases which were not equally soluble in water.
On this point, however, he speaks hesitatingly.

The large proportion of carbonic acid which he found in
bicarbonate of potash, led him to anticipate that it would
resemble carbonate of ammonia in its action on salts of lime
and magnesia. He found, accordingly, that the bicarbonate
precipitated chloride of calcium with effervescence, and that it
gave no precipitate with sulphate of magnesia in the cold, but
when heat was applied to the mixture, *^ a great deal of air was
discharged, and the magnesia was precipitated.^'

tt
ft
tt

208 CAVENDISH AS A CHEMIST.

The following table represents the composition which
Cavendish assigned to the carbonates he analysed : —

Marble 1000 grs. contained 408 of fixed air.

Garb, amm 1000 „ 533 „

Pearl-ashes 1000 „ 284 ,,

Bicarb, potash 1000 „ 423 . „

None of those numbers are accurate. It is impossible^ how-
ever, to be certain what variety of carbonate of ammonia Cavendish
anaylsed ; and it can scarcely be doubted that the pearl-ashes
were very impure. If all the salts had been pure, and quite
accurately analysed, the numbers would have been as follow : —

Marble (CaO,CO>) 1000 grs. contain 440 ofcarb.add.

Garb. amm. (2NH^0,3CO2) 1000 „ 559*32

Pearl-ashes (K0,C02) 1000 „ 318-84

Bicarb, potash (KOC02,HOCO«) 1000 „ 400

It is curious to notice, in connection with these determina-
tions of the quantities of carbonic acid necessary to saturate
different bases, how long it was before its possession of acid
characters, even when free, was detected. Black held, on
theoretical grounds, that fixed air was an acid. Brownrigg
attributed to it the peculiar taste of Spa water.* Cavendish
determined its saturating power. None of those observers,
however, made direct trial of its acid properties. Bergmann,
who called it the aerial acid, was the first who discovered that
it reddened vegetable blues.f He communicated this observa-
tion to Priestley, who mentions the fact.^

The third part of Cavendish's paper details experiments
^' On the Air produced by Fermentation and Putrefaction.*'
Macbride, following out a suggestion of Black's, had shown
that these processes yield carbonic acid, and, as he conceived,
only that elastic fluid.§ Cavendish confirmed this result, so \

far as the vinous fermentation of sugar and apple-juice was i

concerned. The gas these evolved, he found to be entirely '

absorbed by caustic potash, and to have the same solubility in
water, action on flame, and specific gravity, as the fixed air
from marble. He further showed that the common air which
had remained in contact with the fermenting liquid suffered no
change during the process, but detonated as sharply with
hydrogen as that of the atmosphere.

♦ Phil. Tram., 1765, p. 219. t ErpnimenU on Air, (1775) i., p. 31.

t Hoefcr, HtMt. de la CAim., ii., 444. § Black's Lectures, Yol. u., p. 89.

RATHBONE PLACE WATER. 209

The gaseous products of putrefaction were examined by
keeping ^^ gravy-broth'^ contained in a gas-bottle, at a tempera-
ture of about 96^ as long as it gave off an elastic fluid. This
was received in a bottle filled with solution of caustic potash^
which absorbed the carbonic acid, and left a mixture of common
air (transferred from the gas-bottle) and inflammable air (derived
from the gravy), in the proportion, as Cavendish estimated, of
one volume of the former to 4*7 of the latter. He ascertained
the specific gravity of this mixture by filling with it ^^ a piece of
ox-gut furnished with a small brass cock,'' which he found
more convenient than a bladder, for determining the density of
small quantities of gas. He afterwards filled the ox-gut with a
znixture of 4'7 volumes of hydrogen and one of common air^
and found that it weighed less than the gas from the gravy, in
the proportion of 4| to 4|. He drew the conclusion, accord-
ingly, that '^this sort of inflammable air is nearly of the same
kind as that produced from metals. It should seem, however,
either to be not exactly the same, or else to be mixed with some
air heavier than it, and which has in some degree the property
of extinguishing flame, like fixed air." Raw meat was also
found to yield inflammable air when it putrefied, but in smaller
quantity than the gravy. It was not very minutely examined,
but appeared to be of the same kind as that already described*
A fuller reference to the difference between hydrogen and other
inflammable airs, will be found in the 4th series of Experiments
on Air, printed from Cavendish's MS. BriL Asioc. Sep, 1839,
p. 60.

Experiments on Rathbone Place Water*.

This paper maybe regarded as to a great extent a continua-
tion of an inquiry into the properties of fixed air, but it is also
interesting as detailing one of the earliest tolerably accurate
analyses of a mineral water. The experiments described were
made about the same time as those detailed in the Researches
on Factitious Air, but were recorded separately, as they included
an examination of the solid as well as the gaseous contents of
the water.

* Phil. JVam., 1767, p. 92. Read to Royal Society Feb. 19, 1767.

210 CAVENDISH AS A CHEMIST.

It had long been noticed^ as Cavendish observes^ that ^^most
waters, though ever so transparent, contain some calcareous
earth, which is separated from them by boiling, and which
seems to be dissolved in them without being neutralised by any
acid, and may, therefore, not improperly be called unneutralised
earth/' The cause of the suspension of this earth was un-
known, and with a view to discover it, the Rathbone Place
water was selected for examination, ''as it contains more
unneutralised earth than most others.^'

The water in question was the produce of a large spring at
the end of Rathbone Place, and at one time was raised by an
engine to supply the neighbouring parts of London. Its fixed
ingredients were first determined. To ascertain their amount^
a measured quantity of the water was distilled, till between a
third and a fourth had been drawn off. The earth which pre-
cipitated during the distillation was collected and dried. It
was quite soluble in hydrochloric acid, and was, therefore, ac-
cording to the canons of analysis of the day, '' an absorbent
earth,'' t. e., carbonate of lime, or of magnesia, or a mixture of
both.'*' To determine its nature more minutely, a second and
larger quantity of the precipitate was saturated with oil of
vitriol, which converted it in greater part into insoluble selenite,
or sulphate of lime. The clear liquor strained from off the
selenite, yielded on evaporation a small quantity of sulphate of
magnesia, so that the precipitate contained both of the absorbent,
or, as we should now call them, alkaline earths, with which
chemists were then familiar.

The water in the still was then evaporated, first in a silver
pan, and afterwards in a glass cup, to about three oimces. In
the course of concentration, it deposited a little sulphate of
lime; and it was found, to contain another sulphate, apparently
sulphate of potash, and, in addition, chloride of sodium.

Much attention has lately been directed to the presence of
nitrates in natural watera, but Cavendish and his contempo-
raries were well aware of what has been regarded as a recent
discovery. He sought for nitric acid in the Rathbone Place
water, because ''many waters contain a good deal of neutral

* In the ordinary Isngnage of the day, day, or rather perhaps alumina, was alao
an absorbent earth ; but Cavendidi appears to limit the term to lime and magoesia.

FIXED AIR IN RATHBONE PLACE WATER. 211

salt composed of the nitrous acid, united to a calcareous earth.''
He found no nitrate present, however; and he remarks, in refer-
ence to this point, ^' as I have heard of no other London water
that has been examined with this view, but what has been found
to contain a considerable proportion of nitrous salt, it seems
very remarkable that this should be entirely destitute of it''

The water which distilled over, precipitated lime-water,
sugar of lead, and corrosive sublimate, and changed vegetable
blues to green. When mixed with a little sulphuric add, and
evaporated to dryness, it left a brownish salt, which gave off the
odour of volatile alkali when lime was added to it. Cavendish
determined approximatively the proportion of ammonia pre-
sent in the water, by adding to a measured quantity of the dis«
tilled fluid, a slight excess of sulphuric add, which was after-
wards neutralised by the addition of a known weight of carbo*
nate of ammonia. He appears, however, to have made no
allowance for the loss of carbonic acid which attends the con-
version of carbonate of ammonia into sulphate, for he deducts
the whole weight of the former salt which he added to neutra«>
lise the excess of sulphuric add, from the entire residue of sul-^
phate of ammonia, produced ; thus : '^ The volatile sal-ammoniac
(carbonate of ammonia) contained in sixty-six grains of vitriolic
ammoniacal salt (sulphate of ammonia) is 58^ grains." His
estimate, therefore, must have been far wrong. It is singular,
that after having determined the proportion of carbonic acid
which carbonate of ammonia loses when dissolved in an acid
(ante, p. 206), Cavendish should have omitted to allow for this
loss in making his calculation.

He now proceeded to investigate the nature of the gases
present in Rathbone Place water, prepared to expect that it
would yield carbonic add, from the investigations of Dr. Brown-
rigg, who had found ^' that a great deal of fixed air is contained
in Spa water," and ready to connect the solubility of the calca-
reous earth found in the former water, with the presence of the
fixed air expected to occur in it.

With this view a considerable quantity of the Rathbone
Place water was introduced into a tin pan, occupied by a dome-f
shaped funnel, with its narrow end uppermost, and wide enough
below, and laterally, to fill nearly the whole circumference oC

212 CAVENDISH AS A CHEMIST.

the pan. The water rose above the neck of the funnel, over
which a bottle filled with the water under examination was
placed, with its mouth downwards. The contents of the pan
were raised to the boiling point, and when the bottle was filled
by the air which had risen from the water, ^^ it was removed by
putting a small ladle under its mouth,'' a convenient substitute
for a cork or tray, when the necessary manipulations had to be
performed in boiling water. A second bottle full of the cold
Rathbone Place water was in the same way substituted for the
first, and that in its turn by others so long as the gas was freely
evolved* The gas was analysed by allowing it to stand for a
day over water, when ^^ much the greatest part of the air was
absorbed,'' and the water acquired the power of precipitating
lime-water, from which it was inferred that what had been
absorbed was fixed air* The unabsorbed air was then trans-
ferred to another bottle standing over a solution of ^^ sope-leys,"
or caustic alkali, which reduced it considerably in bulk, and the
residue was tested as to its identity with common air, by
detonating it with hydrogen. It must be recollected that
nitric oxide was still unknown, as well as oxygen and nitrogen^
and that no eudiometer had been devised, or any test proposed
for common air. Cavendish's instrument for analysing the
latter was quite unique, and differed from all later instruments
in being an acoustic eudiometer. He applied it, as we have
43een in reviewing his former paper, to the identification both of
inflammable and of common air. A gas supposed to be one of
these, was mixed with a certain volume of the other, and
exploded ; the loudness of the explosion was carefully noted,
and compared with the sound produced by the detonation of a
mixture in the same proportions of hydrogen and common air.
If the sound were the same, then the gas under examination, if
inflammable, was inferred to be hydrogen — if uninflammable, it
was inferred to be common air, as we have seen already in refer*
ring to the analysis of the air confined over fermented sugar
and to that of die inflammable air from putrefying meat and
gravy (ante, p. 209). In the case before us a small vial being
filled with equal quantities of the unabsorbed air from the
Rathbone Place water and inflammable air, and a piece of
lighted paper applied to its mouth, *^ it went off with as loud a

CARBONIC ACID DISSOLVES CARBONATES. 213

bounce as when a small vial was filled with equal quantities of
common air and inflammable air/^ This singular and uncertain
method of identifying atmospheric air did not satisfy Caven-
dish. He determined^ in addition, the specific gravity of the
unabsorbed gas of the water^ and found it the same as that of
common air.

It was possible that the fixed air which had arisen from the
water had been generated during the boiling. Cavendish^ how-
ever, satisfied himself that fixed air pre-existed in the water, by
adding to it lime-water, which gave an abundant precipitate.
He found that in this way he could throw down the whole
of the calcareous earth, so that the water ceased to deposit
on boiling, and was not troubled by the addition of '^ fixed
alkali."*

From the amount of lime-water needed to precipitate a
measured quantity of Rathbone Place water, and the weight of
calcareous earth which in its natural state it deposits on boiling,
Cavendish inferred that it contained ^^near 2^ times as much
fixed air as is sufficient to saturate the unneutralised earth in
it." From his whole experiments he drew the following con-
clusion as to the relation of the fixed air to the solubility of the
earthy carbonates in water. '' It seems likely from hence, that
the suspension of the earth in the Rathbone Place water is
owing merely to its being united to more than its natural pro-
portion of fixed air ; as we have shown that this earth is actually
united to more than double its natural proportion of fixed air,
and also that it is immediately precipitated, either by driving off
the superfluous fixed air by heat, or by absorbing it by the
addition of a proper quantity of lime-water."

Cavendish then proceeds to comment on the strangeness of
the fact, that the total abstraction of carbonic acid from lime,
and the addition to it of a great excess of that gas, should
equally render it soluble in water, although in its natural, inter-
mediate condition of calcareous earth it is insoluble. To lessen
the objections to his conclusions, which their strangeness in
this respect might occasion, he resolved to make a direct trial,
as to the possibility of suspending a calcareous earth in water,

* The term signifies here, carbonate of potash or soda, either or both; Cayendish
distinguishes caustic alkali by the name " sope-leys."

214 CAVBNDISH AS A CHEMIST.

((

by furnishing it with more than its natural proportion of fixed
air.'^ For this purpose he placed in a bottle^ a weighed quantity
of carbonate of potash dissolved in rain-water, and poured into
it a solution of chloride of calcium, mixed with a portion of free
hydrochloric acid, less than sufficient to neutralise the whole of
the alkaline carbonate. The bottle was quickly stopped and
well shaken. At first the mixture was turbid, but it soon
became transparent. On heating it the liquid again became
turbid, dischai^ed a good deal of air, and yielded an earthy
precipitate. In this experiment the proportions were adjusted
so as to correspond to the Rathbone Place water. The carbo-
nate of potash precipitated carbonate of lime from the chloride
of calcium^ but simultaneously supplied carbonic acid to dis-
solve the precipitate.

To demonstrate that the carbonic acid was the solvent of
the calcareous earth, Cavendish repeated the experiment, with
the same proportion of materials, but added the alkaline carbo-
nate to the hydrochloric acid, and allowed the effervescence to
be past before pouring in the solution of chloride of calcium.
The precipitate which was produced in this case '^ could not be
redissolved on shaking,'' because, as Cavendish inferred, the
carbonic acid which would have held it in solution had been
allowed to escape. I^astly, lest any should imagine that the
chloride of calcium or other salts, assisted in suspending the
calcareous earth. Cavendish saturated rain water with carbonic
acid, and added 1 1 ounces of the solution to 6^ of lime-water.
^^The mixture became turbid on first mixing, but quickly
recovered its transparency on shaking, and has remained so for
upwards for a year.'' When this experiment was repeated with
about two-thirds of the carbonic acid water, a permanent preci-
pitate was produced.

Three other London pump-waters were found to give a pre-
cipitate of calcareous earth with lime-water, and to yield a
similar residue by evaporation. From his examination of them,
along with the Rathbone Place water. Cavendish thought it
*^ reasonable to conclude that the unneutralised earth in all
waters, is suspended merely by being united to more than its
natural proportion of fixed air."

He finishes his paper with a summary of his analytical

. A NBW EUDIOMETER. 215

results. That of the Rathbone Place is subjoined^ as an example
of the quaatitative analysis of a natural water in 1766.

One pint or 7315 grains of Rathbone Place Water leaTes of ) . - .

solidrendne K''^ ^'

Carbonate of ammonia 0'9

Carbonate of lime and a little carbonate of magnesia 8*4

Free carbonic acid .... .... .... .... .... .... .... .... 4*65

Snlpnate of lime .... ..•• ..•• ...« ••.« •••• ..•• •••• 1'2

Chloride of sodium and sulphate of magnesia 7*9

In reference to the preceding analysis, it may be noticed,
that Cavendish appears to have regarded the ammonia as pre-
sent in the caustic state, which it could not be in a water con-
taining free carbonic acid. His first item is, ^' as much volatile
alkali as is equivalent to about -^ grain of volatile sal-ammo-
niac/^ The carbonic acid he gives as, ^^as much fixed air,
including that in the unneutralised earth, as is contained in
19-^ grains of calcareous earth/' Only the carbonic acid not
present in the 8*4 carbonate of lime is placed in the table as
free. It is calculated from Cavendish's datum, that carbonate
of lime contains in 1000 parts, 408 of carbonic acid

An Account op a New Eudiometer.*

The seventeen years which elapsed between the publication
of Cavendish's two first papers and the one which we are now
to consider, did more to alter and enlarge the boundaries of
pneumatic chemistry than any seventeen years have done before
or since. During that long interval, Bergmann, Scheele, Lavoi-
sier, but, above all, Priestley, besides others, had been engaged
in researches on the gases.f The great majority of these had
been discovered, and though almost nothing had been done
towards their analysis, the properties of the chief among them
had been carefully studied and were well known.

The discovery of nitrogen and oxygen had naturally directed
much attention towards the atmosphere, as a compound or mix-
ture of chemical substances, and a convenient as well as accurate

* PkU, TVmu,, 1783« p. 106. This paper was read to the Royal Society
January 16, 1783.

t Fiye ont of Priestley's six yolumes on air were published before the end of
1781.

216 CAVENDISH AS A CHEMIST.

method o analysing it was now an object of desire to every
chemist. Priestley's discovery in 1772 of nitrous gas or nitric
oxide^ which Hales had prepared, but had not recognised as a dis-
tinct elastic fluid, supplied a method of determining the amount
of oxygen present in any gaseous mixture. As nitric oxide,
when it meets oxygen, combines with it to form a compound
soluble in water, the latter rises within the vessel in which the
gases are permitted to mingle, as it dissolves the compound
which they form by their union. If we suppose a slight excess
of nitric oxide made use of in every case^ then, cateris paribus,
the water will rise higher the greater the proportion of oxygen
present, and by the degree of its elevation will measure the
relative quantities of that gas contained in different gaseous
mixtures, analysed in the same way. In actual practice, a diffi*
culty occurs in the use of this gas, which only Cavendish's
successful employment of it prevents us from calling insur-
mountable. The same volume of oxygen can combine with
very different volumes of nitric oxide, according to circum-
stances, and occasion a corresponding difference in amount of
contraction; so that equal contractions cannot be taken as
implying the presence of equal volumes of oxygen.

Two methods of employing nitric oxide were in use before
Cavendish published his paper. The theoretically simpler pro-
cess was to add nitric oxide to a measured volume of respirable
air standing over water, in small successive quantities, so long
as it occasioned diminution in the bulk of the air. The space
through which the water rose, corresponded in this case, exactly
to the volume of oxygen which the nitric oxide had withdrawn
from the air. Experimenting in this way, Priestley showed that
'• about ^th'* of common air combined with the nitric oxide^
and was absorbed by the water,*

It was found in practice, however, a very difficult matter to
adjust the proportion of the nitric oxide, so that not a bubble
too much or too little of that gas should be mixed with the air
under examination. Priestley accordingly substituted for the
method described, another, in which an excess of nitric oxide
was at once added to the air, and the diminution in bulk which
followed was noted. Two specimens of respirable air which

* Experiments and Observatiom on Air (1775), toI. L, p. 111.

PHLOGISTICATION OF AIR. 217

suffered the same contraction in these circumstances^ were
assumed to be equally pure; but the absolute amount of oxygen
contained in the airs was not ascertained by this mode of using
the nitric oxide. Priestley's standard, accordingly, was quite
arbitrary. He mixed equal volumes of nitric oxide and air in a
small wide jar, and after contraction had ceased, transferred the
residue to a narrow graduated tube, in which he measured the
diminution of bulk that had occurred. He expressed this
diminution by the number of parts remaining. Thus^ if one
measure of air^ added to one of nitric oxide, diminished from
2 measures to 1*06, Priestley called the purity of the air 1'06.*
An arbitrary standard such as this would never have been
followed, had the impression not been universal^ that no two
specimens of air had exactly the same composition. This
belief, or rather notion^ would very soon have been corrected
as methods of analysis improved, had it not been connected
with an hypothesis, which, as it was ultimately understood, was
to the effect, that as oxygen supported respiration and combus-
tion much better than common air did^ the salubrity of the
latter might be reasonably assumed to depend on the propor-
tion of oxygen present in it. This hypothesis quickly became
a theory in the hands of those who tested its truth by analyses
of air^ although Priestley, who was indirectly its originator and
its great supporter, confessed that air hypothetically bad^ was
often not to be distinguished from what was reputed the best.
The difference, for example, between the most unwholesome air
from the workshops of Birmingham, and the ^^ very best air in
this county (Wiltshire), which is esteemed to be very goody
" was very trifling.^'t Others were more successful in finding
the difference which they wished to find, and all the philoso-
phers of the day sanctioned the general belief by the name
Etuiiameter, which they gave to their instruments for gaseous
analysis. In reality, however, their view as to the salubrity of
different portions of the atmosphere, was not exactly as is gene-
rally represented, or as has been stated above. They connected
the purity of the air, not so much with the presence of oxygen,

* ExperimenU and Ob$ervatiotu on Air (1775), vol. i., introductioxi, p. zx;
and Tol. h, (1779), introduction, p. zzix., also p. 280.

t Ejtperimentt and Ob^ervadoM (1779), toI. Iy., p. 269.

w I

218 CATENDISH AS A CHEMIST*

as the absence of phlogiston. Their object, as stated by them-
selves, was to ascertain, not the degree of oxygenation, but the
degree of phlogistication of the atmosphere. This fact must
not be overlooked. It is difficult to guard successfuUy agidnst
the tendency to represent the chemists of a former time as
holding our views exactly as we hold them. We are apt to
think of Priestley and his contemporaries as apprehending as
distinctly as we do, that air consisted, in greater part, of
unequal measures of two unlike gases. For some ten years,
however, after the discovery of nitric oxide, and its application
to the analysis of the atmosphere, chemists explained its com-
position otherwise. The early conception of air as essentially
one and indivisible, was not easily thrown aside, nor did it seein
necessary that it should be. Dephlogisticated air (oxygen) was
air minus phlogiston; phlogisticated air (nitrogen) was air plus
phlogiston. Both were equally air; and the principle, which by
its presence or absence altered their properties, was imponder-
able, intangible, and unknown.^ The atmosphere could thus
be represented as consisting, not of two unlike airs, but of one
air, and of phlogiston. The latter term was used in a wider
sense than when first introduced by Stahl, and signified the
principle common to all those bodies which, when left in con-
tact with the air, lessened its respirability, and its power of
sustaining animal life and flame, whether this vitiation of the
air was accompanied by the combustion of the phlogisticating
body or not. Nitrogen was specially distinguished by the name
Phlogisticated Air » We are apt on this account to conceive
that the phlogistication of air must always be synonymous with
the abstraction from it of oxygen, so as to leave nitrogen.
Paradoxical, however, though it may seem, air phlogisticated
was not necessarily nitrogen, although phlogisticated air was.
The former might, besides being nitrogen, be air with its oxygen
replaced in whole, or in part by carbonic or sulphurous acid, as
well as by other bodies; or air with its normal amount of oxy-
gen and nitrogen, but containing the gases named above, or any
other irrespirable or poisonous gases diffused through it in such
quantity as to render it noxious to life, and unfit to support

* Cavendish^ for example, in another paper, speaks of "the dephlogisticated
part of common air." Phil, Tram., i7Hi, p. 123.

THEO&T OF THE EUDIOMETER. 219

combustion. Pure nitrogen ; a mixture of nitrogen with irre-
spirable gases; and a mixture of common air with these, were
thus all air pMogisiicated. To ascertain the extent of this phlo-
gistication^ or vitiation, was the object of the early analysts of
the atmosphere, who did not at first propose to themselves the
task of determining the relative volume of constituent gases in
it; although in the end their inquiry unavoidably merged in such
a research. They did not accordingly name their instruments
Pneumatometers, Aerometers, Gasometers, or the like. They
should have named them Phlogistometers : *^ Measurers of the
badness of the Air.'* They preferred, however, the more eupho-
nious title of Eudiometers, or measurers of its goodness ; a title
still retained, and curious as the only fragment of the Phlogiston
Nomenclature which has survived to the present day.*

These instruments were constructed on the assumption that
the atmosphere was liable to the greatest variations, as to its
degree of phlogistication, or amount of impurity. When air
moreover, previously respirable, was phlogisticated, it was
observed to undergo a diminution of bulk, which was great in
proportion to its original purity. Thus, liver of sulphur, solu-
tions of the alkaline persulphurets, a mixture of sulphur and iron
filings, and nitric oxide, as well as other substances, were known
to diminish the bulk of air while they phlogisticated it. We
should err gready, however, if we assumed that the chemists who
first employed eudiometers, distinctly apprehended that this re-
duction of volume resulted from the conversion of the oxygen of
the air, into a liquid or soluble compound. Cavendish's paper,
recounting experiments which ** were made principally with a
view to find out the cause of the diminution which common air
is well known to suflFer by all the various ways in which it is
phlogisticated,'^ was not read to the Royal Society till a year
after his communication on the eudiometer.f Scheele's im-
portant treatise on Air and Fire, which discussed the same ques-
tion, did not, according to Hoefer, appear in its original form
till 17773 and was not generally known in France or England *
till I781.t

* The name Eadiometer appears to haye been introduced by T^n ^ finn ^, Black's
Leeturet, toL ii., p. 523.

t Bjeperiments on Air, Phil. 7V0fw., 1784, p. 119.
i HUiaire de la Chimie, t. ii., p. 460.

220 CAVENDISH AS A CHEMIST,

The majority of chemists of the period accepted as an ulti-
mate^ or for the time, inexplicable fact, the diminution in bulk
of air, when it was phlogisticated, or as we should now say,
de-oxidised. A theory on the subject was not essential to the
employment of the eudiometer. It was enough that air con-
taining no phlogiston (oxygen), suffered a great reduction in
volume when it was phlogisticated; and that air, saturated with
phlogiston (nitrogen), suffered a much smaller reduction, or none
at all. It could thus be assumed, that air diminished in volume
when it united to phlogiston, and that the less of that principle
it contained before it was phlogisticated, the more of it woidd it
combine with, and the greater would be the reduction of volume
which occurred.

It must further be noticed, before discussing Cavendish's
paper at greater length, that the opinion, natural enough, that
a body exposed to so many vitiating influences as the atmo*
sphere is, could not be uniform in composition, appeared to the
early observers completely confirmed by their analyses. No
precise endeavour, accordingly, was made to ascertain even the
average quality of air, although hundreds of analyses were per«
formed, the mean of which would have given at least an
approximation towards it. Priestley had no fixed standard to
which he referred different specimens of air, when he was ana*
lysing them. His habit was to examine two specimens at once,
the one ex hypoihesi good, the other bad, and to mark the dif*
ference between them. In this way, as we have seen, he com-
pared the foul Birmingham air with the ^^ very good'' air of
Wiltshire; but the quality even of the latter was assumed to be
variable, and was not reckoned as a constant quantity. How
great the variation in quality seemed to be, when tested by
those who expected variation, and employed an imperfect appa-
ratus to measure it, will appear from the statement of one of the
analysts of the period. Signer M. Landriani, after making an
eudiometrical tour through Italy, writes to Priestley in Novem-
ber, 1766: ^^I have had the satisfaction of convincing myself^
that the air of all those places which, from the long experience
of the inhabitants, has been reputed unwholesome, is found to
he 80y to a very great degree of exactness^ by this instrument qf
mine; so that the theory seems to correspond very well to obser-

fontana's eudiometeil 221

vation. In the mountains near Pisa I made trial of the air at
different heights^ beginning on the plain, and proceeding to the
highest summits; and found a remarkable difference in the state
of the air^ every stratum being purer in proportion as I as-
cended.* • • • • • The air of the Pontine Lakes^ that of the
Sciroccho at Rome (so very unwholesome)^ that of the Cam-
pagna Romana, of the Grotto del Cane, of the Zolfatara at
Naples, of the Baths of Nero at Baja, of the sea-coast of Tus-
cany, were all examined by me, and found to be in such a state
as daily experience led me to expect/^f No one who reads this
will feel surprised that Landriani should have been the person
who introduced the word Eudiometer.

Cavendish begins his paper by observing that ^^ Dr. Priest-
ley's discovery of the method of determining the degree of
phJogisticatum of air by means of nitrous air (nitric oxide), has
occasioned many instruments to be contrived for the more cer*
tain and commodious performance of this experiment; but that
invented by the Abbe Fontana is by much the most accurate of
any hitherto published.^':]:

The great improvement in Fontana's eudiometer over pre-
vious instruments, consisted in the graduated tube in which the
diminution of the mixed gases was measured, being long and
narrow, and provided with a wide-necked funnel, through which
the air and nitric oxide were rapidly passed. The gases rose
in one continuous column ; ^' so that," adds Cavendish, ^^ there

* Sanssore, on the other hand, inferred from his experiments, *'that the air of
the valleys among the Alps and at Geneva is better than that on the tops of high
mountains." Voyagu daru lei Alpes (1779), t. i., p. 517.

t Experimewtt and Obtervaiiotu on Air, vol. iii., appendix, p. 380. Similar
statements wiU be found from other correspondents in the appendix to Priestley's
fourth volume. One of these writers is Dr. Dobson, who found " marine air to be
one-eighth of a measure better than common air." (P. 469.) The air here referred
to was that procured from sea-water, raised to the temperature of 212® P., and
doubtless contained more oxygen than atmospheric air does. It was naturally
enough assumed that the air above the sea would be identical in composition with
that found dissolved in it, and in this way the salubrity of marine districts was
accounted for.

X Besides Fontana and Priestley; Magellan, Dobson, and Landriani are referred
to by the second as devisers of nitric oxide eudiometers. At a later period, A. Hum-
boldt endeavoured to improve them, as well as Thomson, Dalton, and Davy (Thom-
son's 8y9tem qf Chemiitry, 6th ed., vol. iii., p. 167) ; but they have long been
abandoned by all chemists.

222 CAVENDISH AS A CHEMIST.

is time to take the tube off the fannel^ and to shake it before
the airs come quite in contact; by which means the diminution
is much greater and much more certain than it would otherwise
be/^ The diminution in volume, also, reached a maximum, in
the short time during which the mixture was shaken, so that the
latter was not sensibly altered in bulk subsequently, however
long it was left over water.

Cavendish referred those phenomena to the opportunity
which was afforded by Fontana's instrument for each small por-
tion of the nitrous air being in contact with water, either at the
instant it mixes with the common air, or at least immediately
after. He thought it, accordingly, worth while to try whether
the diminution would not be still more certain and regular, ^^ if
one of the two kinds of air was added slowly to the other in
small bubbles, while the vessel containing the latter was kept
continually shaking;'^ and finding his anticipations fulfilled, he
constructed an instrument by means of which the gases might
be mixed in the way which yielded the most accurate and con-
stant results. His eudiometer consisted essentially of three
parts: 1. A small glass jar with a handle, which served as a
measure; 2. A hollow glass globe with a wide neck, in which
the combination of the gases took place. This was suspended
in a trough of water, with its mouth downwards, so that it
could be readily shaken backwards and forwards, and one of
the gases (air or nitric oxide) was measured into it, at the com-
mencement of the experiment; 3. A glass cylinder, provided
above with a cap and stopcock, and open below, but made to fit
a brass socket or stand, with a small aperture in its centre. A
measured portion of the other gas was introduced into this vessel,
which was then placed in its socket, with the nozzle of the stop-
cock within the neck of the suspended globe. When the stop-
cock was opened, water entered by the aperture in the socket
of the cylinder (3), and the gas contained in it slowly ascended
into the globe which was kept constantly agitated.

In using this apparatus, it was in the option of the experi-
menter to add the common air slowly to the nitric oxide, or the
nitric oxide slowly to the common air. Cavendish generally
did the former. He was not satisfied with the measurement of
the gases, as errors were occasioned by more water *^ sticking to

ACTION OF WATEK IN EUDIOMETER. 223

the sides of die measure and tube at one time than at another/'
He preferred, accordingly, to determine the quantity of air and
nitric oxide used, and the diminution which followed their mix-
ture, by weighing the containing vessels under water. It is not
necessary to enter minutely into a consideration of this process.
The final weighings gave the weight of a volume of water
exactly equal to the volume of mixed gases which had been
expended, and likewise the weight of a volume of water, repre-
senting the space through which the gases had contracted.

When air was in the cylinder, and nitric oxide in the sus-
pended globe, a measure of the former was taken to 1^ of the
latter.. Equal measures would have sufSced, but it seemed well
to take a slight excess of the nitric oxide, lest it should be
impure. In the case supposed, the air was slowly added to the
nitric oxide, and the 2^ measures suffered a contraction of 1*08,
which number was what Cavendish called the test of air tried
in this way.

When the nitric oxide was added to the common air, a
measure of each was taken, and the diminution was only 0*89.
The cause of this difference will be considered presently.

Priestley, it will be remembered, marked the purity of the
air by the volume of mixed gases remaimnff after contraction
had ceased ; Cavendish, on the other hand, noted the volume
which ditappeared during contraction.

A point neglected by all previous experimenters, was the
quality of the water with which their eudiometers were filled.
Cavendish made trial with water of different degrees of purity,
from distilled water to ^' water fouled by oak shavings,^' and
found that though the other conditions of the experiment
were the same, the result varied materially according to the
quality of the water employed. He remarks in reference to it,
''this difference in the diminution, according to the nature of
the water, is a very great inconvenience, and seems to be the

chief cause of uncertainty in trying the purity of air

It shows plainly, how little all the experiments which have
hitherto been made for determining the variations in the purity
of the atmosphere can be reliecLon, as I do not know that any
one before has been attentive po the nature of the water he has
used, and the difference proceeding from the difference of waters

224 CAVENDISH AS A CHEMIST.

is much greater than any I have yet founds in the purity of air.'*
He recommends accordingly^ as the best way of obviating the
inconvenience complained of, the employment in all cases of
distilled water. This water, however, he found to absorb
di£ferent quantities of nitric oxide at different times, partly as it
appeared, owing to its temperature not being the same at each
experiment, partly in consequence of the proportion of oxj'gen
present in the water varying, so that different quantities of
nitric oxide were withdrawn by different specimens of water.

The greater number of these and of the other eudiometrical
experiments were made by adding air to nitric oxide ; the order
of mixture which Cavendish preferred. Many, however, were
tried with the gases mixed in the reverse order, when it always
appeared, as already mentioned, that the contraction in bulk
was much less than when air was added to nitric oxide.

It is curious to notice Cavendish's explanation of this phe-
nomenon as illustrative of the difficulty which he and his con*
temporaries experienced, in accounting for the diminution in
bulk which attended the deoxidation (or phlogistication) of air.
" When nitrous and common air,'* says he, '^are mixed toge*
ther, the nitrous air is robbed of part of its phlogiston, and is
thereby turned into phlogisticated nitrous acid, and is absorbed
by water in that state, and besides that, the common air is phio*
gbticated and thereby diminished/^ Here it will be seen, two
causes are assigned for the contraction of the air. — 1. The nitric
oxide, an insoluble gas, by losing phlogiston which it commu-
nicates to the air, becomes a substance soluble in water, which
alisorbs it. — 2. The phlogiston, transferred to the air, compels
it, in virtue of an unexplained power which it possesses, to
diminish in bulk. Cavendish's own words, in continuation of
those already quoted, are — " The whole diminution in mixing is
equal to the bulk of nitrous air, which is turned into acid, added
to the diminution which the common air suffers by being phlo-
gisticated." The diminution owing to the latter cause was, as
Cavendish knew, a constant quantity, but not that, as he be-
lieved, owing to the former. Nitric oxide, according to our
chemist, could part with variable quantities of phlogiston to air
according to the relative proportion of the two gases. When
a small quantity of nitric oxide was added to a large volume of

PURITY OF NITRIC OXIDE. 22 J

air, it parted with a larger amount of phlogiston to the air, than
it did when the nitric oxide was in excess. A smaller volume
accordingly, of nitrous air sufficed to effect the maximum con-
traction of common air, if it were added in successive small
quantities to the latter, till it ceased to contract, than was suffi-
cient for that purpose if the common air were let up bubble
by bubble into the nitric oxide.

Cavendish details a long series of experiments demonstrating
the truth of these statements, but they do not call for minute
reference. It is important, however, to notice that his views
concerning the variable phlogisticating (i. e, deoxidising) power
of nitric oxide over common air, are in exact accordance with our
modern views, provided only (as in translating the statements
of a writer of the phlogiston school we are generally justified in
doing) we always substitute for addition of phlogiston to air,
^abstraction of oxygen from it. We may then understand Caven-
dish as imperfectly teaching, what Dalton, Gay Lussac, and
Humboldt afterwards announced more fully, viz., that the same
volume of oxygen, according to circumstances, can combine with
one or more volumes of nitric oxide, occasioning in each case
a different amount of contraction, although the volume of oxygen
withdrawn is the same*.

Having settled this point, Cavendish proceeds to record a
aeries of experiments made to prove the superiority of his
method to Fontana^s, which need not be detailed.

After giving reasons for assigning to his own the preference,
he enters on the consideration of the question, how far the
accuracy of the *^ nitrous test" is affected by the quality of the
nitric oxide employed. In discussing this he makes a distinc-
tion, at first sight not very intelligible, as to two modes in which
the gas may differ. *' First, it may vary in purity, that is, in
being more or less mixed with phlogisticated or other air ; and,
secondly, it is possible, that out of two parcels equally pure,
one may contain more phlogiston than the other.*' The first
cause of difference, unless the impurity were fixed air, which
could, however, be excluded. Cavendish did not think of much
importance, as the presence of foreign matters would only in-

* Dalton's paper is contained in Phil, Mag,, zzriii^j 351 ; quoted in Thomson'^
Sftlem qfChemiiiry, m,, 169.

226 CAVENDISH AS A CHBMIST.

orease the proportion of nitric oxide required to produce fuU
contraction, and he always employed an excess sufficient to
cover any probable amount of impurity*

The second cause of difference would, if considerable, de-
stroy the whole value of the test, as two different specimens
of nitric oxide would give different results, though employed in
exactly the same way. To determine this, he prepared nitric |

oxide from quicksilver, from copper, from brass, and from iron^ !

and tested common air with each. He found no appreciable I

difference between the three first, but the '^ air from iron'' occa- 1

sioned a greater diminution of bulk than the other specimens of j

nitric oxide did, when added to a nearly equal volume of com*- ' '
mon air, and a smaller diminution than they occasioned, when •

mixed with four measures of air. From this Cavendish con- \

eluded that the nitric oxide prepared from iron was both impure,
and contained ^^ rather less phlogiston than the others,'' so that |

more of it was expended in condensing the same amount of |

oxygen ; or, what came to the same thing, a given bulk of this
nitric oxide caused less contraction in a large volume of air, than i

the same measure of the other specimens of the gas did. It I

is not unlikely that the gas from iron contained hydrogen, and
perhaps also nitrous oxide, so that its deoxidising effect on air,
would be less, bulk for bulk, than that of pure nitric oxide.
Cavendish recommended the gas prepared with copper, as con-
stant in properties, and easily procured.

Having by those careful trials certified the value of his i

eudiometer, Cavendish proceeded to apply it to the deter-
mination of the important question. Is the atmosphere constant <
in composition ? The following is his account of the result of |
his researches on this subject. '' During the last half of the
year 1781) I tried the air of near sixty different days, in order
to find whether it was sensibly more phlogisticated at one time
than another, but found no difference that I could be sure of,
though the wind and weather on those days were very various^
some of them being very Aur and clear, others very wet, and
others very foggy." This conclusion was founded on a very
extensive series of experiments, for seven or eight analyses
were made in different ways of the air of each day. From first
to last, indeed. Cavendish cannot have made fewer than 500

SUDIOMBTBICAL SCALE« 227

quantitative determinations of the composition of atmospheric
air. The result of his protracted observations he stated thus —
^ On the whole, there is great reason to think that the air was
in reality not sensibly more dephlogisticated on any one of the
sixty days on which I tried it, than the rest. The highest test
lever observed was l-OOO, the lowest 1-068, the mean 1'082/'*

This was the result of researches into the quality of the air
from day to day. Cavendish made other experiments ** to try
whether the air was sensibly more dephlogisticated at one time
of the day than at another, but could not find any diflFerence/'
Trials were also made " with a view to examine whether there
was any difference between the air of London and the country.**
Slight difierences appeared sometimes in favour of the purity of
the London air, sometimes in favour of that of Kensington; ^^but
the difference was never more than might proceed from the
error of the experiment; and by taking a mean of all, there did
not appear to be any difference between them. The number of
days compared was twenty, and a great part of them taken in
winter, when there are a great number of fires, and on daya
when there was very little wind to blow away the smoke."

The settlement, by those ample trials, of the uniform com-
position of the atmosphere, enabled Cavendish to suggest what
till then was wanting, viz., a common scale of graduation appli-
cable to all nitric oxide eudiometers. Atmospheric air he
proposed to call 1*00. Nitrogen supplied the zero, or was 0*00 ;.
and for those who agreed with Scheele and Lavoisier in sup-
posing that common air ^' consists of a mixture of dephlogis-
ticated and phlogisticated air,** oxygen was the maximum, and
was marked by Cavendish 4*8.t Those numbers he refers to
as the standards of the several gnses mentioned, in contradis-
tinction to their tests, which were the numbers representing the

* This number, it win be remembered, signifies that a measure of air being
slowly added to a measure and a quarter of nitric oxide, contraction occurred through
a space equal to one measure and iJJs^hs, or -the two measures and a quarter of
mixed gas and air left 9*18 parts of one measure (of nitric oxide and nitrogen)
unabsorbed.

t Cavendish does not imply in this paper, an undoubting agreement with Scheele
and lAToisier, as to the distinct nature of dephlogisticated, and phlogisticated air.
His concurrence in this opinion is much more explicitly announced in his Bxperiment$
on Air, {Pfril. Trant., 1784, p. 141.)

Q 2

228 CAVENDISH AS A CHEMIST. i

contraction which occurred when those gases were mixed with j

nitric oxide in the eudiometer. Thus the standard of air was
I'OO^ but its test, as we have seen^ was I'086.^' The standards
for specimens of air less oxygenated than common air, were
found by taking the test of the air under examination, and then i

making an artificial mixture of similar composition^ of common '

air and nitrogen. In adjusting this mixture, one measure of i

the former was taken^ and variable quantities of the latter, till
a mixture was obtained which suffered the same amount of I

contraction in the eudiometer, or had the same test as the air
under examination. If, ex. gr.; " its test was the same as that .

of a mixture of 1 part of common air and w of phlogisticated

air (nitrogen), its standard was .'' If the specimen were ^

more oxygenated than atmospheric air, then the quantity (j?) of
nitrogen which must be mixed with it to reduce it to the purity
of common air, determined its standard, which was 1 + ^r.
Oxygen would thus^ in round numbers, be 1 + 4 = 5. '

I have described the principle of Cavendish's graduation^
because he does not directly give in this paper his estimate of '

the relative quantities of oxygen and of nitrogen in atmospheric
air. As he called it unity, or 1, and assumed it as a constant
quantity, he made no reference to the factors of this unit,
although it was a middle point in his scale. The standard,
however, of oxygen was a number obtained thus: — Let O parts
of oxygen added to N parts of nitrogen, form a mixture iden- j

tical with common air, then the standard of oxygen is -— — ^ ^

Cavendish gives 4*8 as the standard for oxygen. If we divide
100 by this, we shall obtain the per-centage by volume of
oxygen in air, or 20-83. t Air, therefore, according to him, had I

the composition by volume: —

* A somewhat aimilar method of gradBation was foUowed by Dr. Dobson, of
livcrpool, in 1799. He made "good common air" the middle point of his scale,
but called it 0. From this he counted upwards 22 degrees to oxygen, and down-
wards 20 to nitrogen. The numbers above represented degrees of goodness, or of
supcnonty in purity to common mt; those below were degrees of badness. Priest-
ley's Experiments and ObtervaiioM m Air, vol. iu., app., p. 470.

t Taking 100 volumes of air O + N = iOO. By Cavendish's fonnuk^i? =

4-8; «*«^oresubstituting>hevalueofO + N;l^= 48 and = i^= 2083.

O 4-8

IMPERFECTION OF EUDIOMETER. 229

Oxygen 20-833

Nitrogen 79-167

100-000

According to Dumas' recent analysis^ the numbers are: —

Oxygen 20-90*

Nitrogen 79*10

10000

The approximation is very close. Scheele made the amount
of oxygen 25 per cent.5t Lavoisier made it 27 per cent.;t Saus-
sure 22 per cent. § Cavendish's analysis, therefore, was much
more accurate than those of his illustrious contemporaries.
In his later essay (*^ Experiments on Air*') he announces the

. result of his analysis of air more fully. Referring to an expe-

riment, he says, " some dephlogisticated air was reduced by
liver of sulphur to ^th of its original bulk; the standard of
this air was 4*8, and consequently the standard of perfectly
pure dephlogisticated air should be very nearly 5, which is a
confirmation of the foregoing opinion, for if the standard of

\ pure dephlogisticated air is 5, common air must, according to

this opinion, contain ^th of it, and therefore ought to lose -|^th
of its bulk by phlogistication, which is what it is actually found
to lose.*' II

I The part of the paper immediately succeeding that last

^ discussed, is occupied with references to the best mode of pre-

paring nitrogen for the graduation of the eudiometer, and in

I explaining the cause of the slight contraction which attends the

addition of nitric oxide to pure nitrogen. Cavendish refers
this to the solution of the nitric oxide in the water, but it was

* Regnault, by a nnmber of determinationB made firom the 24th to the 31st of
December, 1847, foand the proportion of oxygen in the atmosphere to Tary between
\ 20*90 and 21*00 vol. per cent. In January^ 1848, it varied between 20*89 and

\ 20*99 ; the results obtained in the analysis of air at various hours of the same day were

found to oscillate between the same limits. R. F. Marchand found the quantity of
oxygen in the air in ten experiments to vary from 20*90 to 21*03, the mean being
20-97 vol. per cent Liebig and Kopp's Annual Report qffheProffregi o/CAemUtry,
1847-48, part ii.« pp. 298, 299.

t Hoefer, Traiii de Chimie, t. ii., p. 463.

X BlemenU of ChemUtryj translated by Kerr, p. 86.

% Black's Lectures f ii., p. 524.

il Phil. Trane.y 1784, p. 141.

230 CAVENDISH AS A CHEMIST.

probably also owing to the nitrogen being mixed with a little
oxygen derived from atmospheric air displaced from the water.
It was so appreciable^ that although the standard of nitrogen
was 0; its test was 0*7*

The paper concludes with an estimate of the nature and
extent of the information supplied by the eudiometer^ of great
value. Cavendish shows that etymologically the name had no
significance^ and this in a twofold way: for^ 1. In so far as the
instrument takes cognizance of the degree of phlogistication^
or impurity of the atmosphere^ it betrays no difference between
one specimen of air and another^ so that apparently there are
no degrees of goodness to be measured; 2. Even when the
atmosphere is certainly phlogisticated, as by the addition of
some ounce measures of nitric oxide to the air of a large room^ I

their '^effect in phlogisticating the air must be utterly insensible '

to the nicest eudiometer.^^ ^^ In like manner^ it is certain that
putrefying animal and vegetable substances, paint mixed with i

oil^ and flowers, have a great tendency to phlogisticate the air;
and yet it has been found" that such air ^^ was not sensibly (

more phlogisticated than common air.'' The general inference i

from this is^ '' that our sense of smelling can^ in many cases,
perceive infinitely smaller alterations in the purity of the air
thzii can be perceived by the nitrous test/*

The nitric oxide eudiometer has long been abandoned, but
the constant results which Cavendish alone among chemists
obtained with it, remain a lasting monument to his unique skill,
which converted a most imperfect analytical instrument into a
delicate and accurate recorder of the relative proportions of the
more abundant constituents of the atmosphere.

It need scarcely be noticed, that we are still as much in
need of an eudiometer, properly so called, as the contemporaries
of Priestley and Cavendish were. There cannot be two opinions
as to the atmosphere being as little entitled to be considered a
perfectly homogeneous mixture, as the ocean is; nor does any
other obstacle stand in the way of the analysis of the air, than
that presented by the comparatively small quantity of many of
the substances which must be sought for in it. Ldebig, however,
has taught us how to overcome this difiiculty, at least in part,
by analysing rain and snow, which bring down to the earth the

. EXPERIMENTS ON AIR. 231

soluble substances of the atmosphere which they have encoun-^
tered in their fall; and Dumas and others have shown how
much may be done by forcing large volumes of air through
solutions of substances which combine with, and detain certain
of its ingredients. Medicine, as well as meteorology and chem-
istry, have the deepest interest in such inquiries, and we may
anticipate the period when a laboratory will form an essential
part of our meteorological observatories, and systematic and
continuous analyses will be made of all the accessible consti-
tuents of the atmosphere.

Experiments on Air.*

The paper now to be considered contains the record of
experiments made in part in the summer of 1781, before those
on the analysis of air which have just been commented on.
The earlier researches, however, could not be successfully pro-
secuted without a knowledge of the composition of the atmo-
sphere, and Cavendish, accordingly, interrupted the original
inquiry, after it had made some progress, till he had completed
the protracted eudiometrical investigation, which was made public
a year before the ** Experiments on Air/* That title conveys a
very imperfect idea of the nature of the researches which were
carried on by its author. Its most important section, according
to our modem estimation, is that which treats of the synthesis
of water from its elements. The production, however, of water,
or the determination of its composition, was not the special
object of the inquiry, which was undertaken with a view to
ascertain what were the products of the deoxidation of atmo-
spheric air by the ordinary combustibles, and some other bodies
having a great affinity for oxygen. Little notice has been
taken, even by the professed historians of chemistry, of the
igeneral scope of the paper, but much criticism has been ex-
pended on those parts of it which relate to the Water Contro-
versy, It is desirable, accordingly, in order to avoid repetition^
to limit the present abstract to an unpolemical analysis of the
contents of the " Experiments on Air,'* the disputed portions
of which will be considered in detail, when the claims of Watt

* P;if7. Tran*,, 1784, p. 119j read to the Royal Society January 15, 1784.

232 ' CAVENDISH AS A CHEMIST.

and Cavendish^ as the discoverers of the composition of water,
are under discussion.

Cavendish begins by observing that ^^ the following expert*
ments were made principally with a view to find out the cause
of the diminution which common air is well known to suff^ by
all the various ways in which it is phlogisticated^ and to discover
what becomes of the air thus lost or condensed/' He then
mentions that many have supposed that '^ fixed air is either
generated or separated from atmospheric air by phlogistication^
and that the observed diminution is owing to this cause.''
His first experiments, therefore, were made with a view to
ascertain the truth of this opinion; and he began by excluding
from consideration all animal and vegetable substances, which,
as they ** contain fixed air,"— or as we should now say, yield it
by combustion — could not be employed to phlogisticate the air
in such experimenta crucis as he desired to make. He then
proceeds to detail the only methods known to him which are
not liable to objection, viz., ^^the calcination of metals, the
burning of sulphur or phosphorus, the mixture of nitrous air»
and the explosion of inflammable air." He doubts about the
electric spark, which he thinks may phlogisticate the air only
by igniting some combustible matter present in the containing
vessel, so that he does not deem it necessary to experiment
with it. The unexceptionable methods of investigation are then
discussed seriatim. He states, in the first place, that '^ there is
no reason to think that any fixed air is produced during the
calcination of metals." Priestley, he observes, found none, and
Lavoisier only a very slight and scarcely perceptible turbid
appearance, when lime-water was shaken in a glass vessel full of
the air in which lead had been calcined. A statement of
Priestley's, that impure quicksilver is changed by agitation and
exposure to the air, into a powder containing fixed air. Caven-
dish sets aside as not unexceptionable in reference to the
question he was considering, because this gas may have been
contained in the impure mercury before its agitation with the
air was commenced. '' I never heard," he continues, *' of any
fixed air being produced by burning sulphur or phosphorus;
but it has been asserted, and commonly believed, that lime-
water is rendered cloudy by a mixture of common and nitrous

PIILOGISTICATION OR DBOXIDATION OF AIR. 233

air, which, if true, would be a convincing proof that on mixing
those two substances, some fixed air is either generated or
separated." He showed, however, that if the gases are washed
with lime-water before being mingled, no carbonic acid can be
detected after mixture.

Cavendish then tried whether fixed air was produced by
the explosion of inflammable air from metals (hydrogen), with
either common or dephlogisticated air (oxygen). The gases were
washed with lime-water before the electric spark was passed, and
*^ the event was, that not the least cloud was produced in the
lime-water when the inflammable air was mixed with common air,
and only a very slight one, or rather diminution of transparency,
when it was combined with dephlogisticated air/' The general
conclusion from the whole experiments was, that ^^on the
whole, though it is not improbable that fixed air may be gene-
rated in some chymical processes, yet it seems certain that it is
not the general effect of phlogisticating air, and that the diminu-
tion of common air is by no means owing to the generation or
separation of fixed air from it/'

Having thus disposed of carbonic acid as the constant pro-
duct of the phlogistication or deoxidation of air. Cavendish
proceeded to try whether, as some of Priestley's experiments
seemed to render probable, ^^ the dephlogisticated part of com-
mon air might not by phlogistication be changed into nitrous
or vitriolic acid." For this purpose, he burned sulphur in air
over milk of lime, filtered and evaporated the resulting solu-
tion, and found that ^^ it yielded no nitrous salt, nor any other
substance except selenite; so that no sensible quantity of the
air was changed into nitrous acid." He tried also " whether
any nitrous acid was produced by phlogisticating common air
with liver of sulphur." For this purpose he boiled sulphur
with milk of lime, and then shook the solution with large
quantities of air, till the liquid lost its yellow colour, ^^ a sign
that all the sulphur was, by the loss of its phlogiston, turned
into vitriolic acid and united to the lime, or precipitated; the
liquor was then filtered and evaporated, but it yielded not
the least nitrous salt."

Cavendish calls the salt of lime produced in both these ex-
periments, "Selenite." In the first it must have been sul-

234 CAVENDISH AS A CHEMIST.

phite, in the second^ hyposulphite of lime. It did not escape
his observation, however, that the salt he encountered in both
these trials differed from ^' Common Selenite^' (sulphate of
lime). Unlike it, the salt '^was very soluble, and even crys-
tallized readily, and was intensely bitter/' By exposure to the
air, however, and evaporation to dryness, it lost its great solu-
bility, and ceased to interfere, as it did at first, with the search
for a nitrous salt/

The same negative result was obtained with sulphur and
milk of lime, when oxygen was substituted for common air.
Cavendish then proceeded to try whether any vitriolic acid was
produced during the phlogistication of air. For this purpose he
caused a large quantity of nitric oxide (the phlogisticating agent)
to combine with the oxygen of common air confined over distilled
water. The acidulated water was then distilled, saturated with
salt of tartar, and evaporated to dryness. He procured in this way
87i grains of nitre, which was unmixed with vitriolated tartar
(sulphate of potash), ^^ consequently no sensible quantity of the
common air with which the nitrous air was mixed, was turned
into vitriolic acid.'^ He then records an erroneous conclusion
as to the relative acidity of nitric oxide and nitric acid, which
he founded upon the experiments last detailed; and thereafter
proceeds thus: ^^ Having now mentioned the unsuccessful
attempts I made to find out what becomes of the air lost by
phlogistication, I proceed to some experiments, which serve
really to explain the matter .''

The account which follows is so brief and clear, and the
passage is so important, that it does not admit of conden8atidii«
** In Dr. Priestley's last volume of experiments is related an
experiment of Mr. Warltire's, in which it is said that on firing
a mixture of common and inflammable air by electricity in a
close copper vessel holding about three pints, a loss of weaght
was always perceived, on an average about two grains, though
theyessel was stopped in such a manner that no air could escape

* Although Cavendish did not distinguish between the salt of the first experi-
ment (CaO,SO^ and that of the second (CaOyS'O^y he spoke of them considered as
one, as owing their peculiarity to the sulphur-acid which they contained, "being
very much phlogisticated." In his nomenclature, phlogisticated vitriolic acid was
our sulphurous acid; and vitriolic acid "very much phlogisticated" was a body sach
as hyposulphurons acid.

SYNTHESIS OF HYDROGEN AND OXYGEN.

236

by the explosion. It is also related^ that on repeating the expe-
riment in glass yessels^ the inside of the glass^ though clear and
dry before, immediately became dewy; which confirmed an
opinion he had long entertained, that common air deposits its
moisture by phlogistication. As the latter experiment seemed
likely to throw great light on the subject I had in view, I
thought it well worth examining more closely. The first expe-
riment, also, if there were no mistake in it, would be very
extraordinary and curious; but it did not succeed with me; for
though the vessel I used held more than Mr. Warltire's, namely,
24,000 grains of water, and though the experiment was repeated
several times with different proportions of common and inflam-
mable air, I could never perceive a loss of weight of more than
one-fifth of a grain, and commonly none at all. It must be
observed, however, that though there were some of the experi-
ments in which it seemed to diminish a little in weight, there
were none in which it increased.*' Cavendish then goes on to
mention that ^^ in all the experiments the inside of the glass globe
became dewy, as observed by Mr. Warltire; but not the least
sooty matter could be perceived.*'

In the experiments detailed, the hydrogen (generally from
2inc) and the common air were mingled in known quantities,
and the amount of diminution in bulk which followed explosion,
ascertained in each case. The test of the air, including its
standard, in other words the amount of oxygen (if any) re-
maining after detonation, was also observed. The mode in
which this was done is not described, but it cannot be doubted
that it was by means of the nitric oxide eudiometer, as the
nomenclature made use of is that explained by Cavendish in
describing that instrument. (Ante, p. 2270* These quantita-
tive results are given fully in a table, of which the fourth entry
is the following.

Common Air.

InflammAble
Air.

•423

Diml&ntloii.

*612

Air remaining

after the

ezplodon.

•811

Tettofthls

Airinfint

method.

•097

Standard.

•03t

♦ Phil. Trans., 1783, p. 131.

f The fifth oolumft records the amoant of contraotion which occurred when
the air uncondenaed by the explosion was mixed with a little more tlum an eqnal

236 CAVJBNDISH AS A CHEMIST.

On this entry Cavendish makes the following important
comment^ which contains his earliest formal announcement of
the synthetical determination of the composition of water:
'^From the fourth experiment it appears that 423 measures
of inflammable air are nearly sufficient to completely phlogis*
ticate 1000 of common air; and that the bulk of the air remain*
ing after the explosion is then very little more than four-fifths
of the common air employed; so that as common air cannot be
reduced to a much less bulk than that by any method of phlo^s-
tication^ we may safely conclude that when they are mixed in
this proportion^ and exploded, almost all the inflammable air,
and about one-fifth part of the common air, lose their elasticity,
and are condensed into the dew which lines the glass.^^ This re-
markable passage shows how clear and precise were the views of
Cavendish. It distinctly intimates, not a hesitating supposition
concerning the fate of the common air and hydrogen which
had disappeared, but an assured conviction that they were con*
verted into the dew, i. e., that the dew was the very gases in the
liquid state.

Having ascertained in this way the connection between the
disappearance of the gases and the appearance of the dew.
Cavendish proceeded to investigate the nature of this dew. For
this purpose, he so arranged as to burn together 500,000 grain
measures of inflammable air with 2^ times that quantity of
common air, within a glass cylinder, and collected the resulting
liquid. *^By this means, upwards of 135 grains of water were
condensed in the cylinder, which had no taste nor smell, and
which left no sensible sediment when evaporated to dryness;
neither did it yield any pungent smell during the evaporation;
in shorty it seemed pure water.''

A short unimportant paragraph then occurs, which is
followed by his conclusion from both sets of experiments,
namely, those with the globe, and those with the cylinder.
" By the experiments with the globe it appeared, that when
inflammable and common air are exploded in a proper propor-
tion, almost all the inflammable air, and near one-fifth of the
common air, lose their elasticity, and are condensed into dew.

measure of nitric oxide ; and the sixth the amount of oxygen in that air according to
a eudiometrical scale, which made nitrogen 0, air 1, and oxygen 4*8. Phil, TYtrnt.,
1/83, p. 131. Ante, p. 227.

PRODUCTION OF NITRIC ACID. 237.

And by this experiment it appears that this dew is plain watery
and consequently that almost all the inflammable air^ and about
one*fifth of the common air, are turned into pure water/*

Having thus assured himself that the gases had changed
during the phlogistication of the air into a liquid^ and that the
liquid was pure water, Cavendish proceeded to make similar
experiments with oxygen and hydrogen, in the proportion of
19,500 grain measures of the former to S7>000 of the latter^ or
rather less than two volumes of hydrogen to one of oxygen. An
exhausted glass globe was filled with the mixture^ ^^ and the
included air fired by electricity^ by which means almost all of
it lost its elasticity .'* As Cavendish wished, however, to examine
the liquid product of this combustion also, he replenished the
globe with fresh supplies of the mixture, and ^^ by this means,
though the globe held not more than the sixth part of the mix-
ture, almost the whole of it was exploded therein, without any
fresh exhaustion of the globe.*' By an ingenious contrivance,
'^ the whole quantity of the burnt air was found to be 2,950
grain measures; its standard was 1*85.** In other words, the
residual uncondensed gas contained more oxygen than common
air, in the proportion of 1*85 to 1*00. On proceeding, to
examine the product of combustion, it was found that the
^^ liquor condensed in the globe, in weight about 30 grains, was
sensibly acid to the taste, and by saturation with alkali and
evaporation, yielded near two grains of nitre ; so that it con-
sisted of water united to a small quantity of nitrous acid. No
sooty matter was deposited in the globe. The dephlogisticated
air used in this experiment was procured from red precipitate.**
This is the first mention of the appearance of nitric acid, which
compelled Cavendish to undertake an additional and difficult
inquiry, and delayed the publication of his observations on the
synthesis of the elements of. water.

At first. Cavendish ^^ suspected that the acid contained in
the condensed liquor was no essential part of the dephlogisti-
cated air,** but was derived from the basic nitrate of mercury
contained in the red precipitate from which the oxygen was
prepared. He accordingly repeated the experiment, with oxy-
gen from the same source, but which had been well agitated
with water. The product of combustion, however, was still

238 CAVENDISH AS A GHBMIST.

acid. The experiment was also tried with oxygen from red
lead and oil of vitriol, but with the same result, only the nature
of the add was not ascertained. Then oxygen was procured
from the leaves of plants, and exploded with inflammable air as
before; the '' condensed liquor still continued acid and of the
nitrous kind/'

There appeared thus to be no difference in the action of the
oxygen depending on its mode of preparation. Could the pro-
duction or appearance of acid depend on the proportions in
which the gases were mixed ? Cavendish proceeded to try this*
He observes: ''In these experiments the proportion of inflam*
mable air was such, that the burnt air was not much phlogia*
ticated; and it was observed, that tlie less phlogisticated it
was, the more acid was the condensed liquor.'' In other words
hydrogen was exploded with more than a combining measure
of oxygen, in the experiments where acid appeared; and the
greater the excess of oxygen, the larger the proportion of acid
produced. The latter proposition, it need scarcely be noticed,
can only be true within certain limits. Cavendish then repeated
the experiment with oxygen from plants, and a large proportion
of hydrogen, so that '' the burnt air was almost completely phlo-
gisticated," f . 6., a very slight residue of oxygen remained uncom-
bined, and '' the condensed liquor was then not at all acid, but
seemed pure water; so that it appears that with this kind of
dephlogisticated air, the condensed liquor is not at all acid when
tiie two airs are mixed in such a proportion liiat the burnt air
is almost completely phlogisticated, but it is considerably so
when it is not much phlogisticated." Cavendish then tried
whether a difference in the proportions would in the same way
affect the production of acid when the oxygen made use of was
procured from red precipitate. He found the same law to hold:
the more oxygen the mure acid; and he adds, ''there can be
little doubt but that the same rule obtains with any other kind
of dephlogisticated air."

Oxygen procured from Turbith mineral (basic sulphate of
mercury SHgO + S(y) was then made use of, along with less
than a combining volume of hydrogen, with a view to ascertain
whether the acid produced would still be the nitrous, which it
was found to be. Cavendish further shpwed that the non-

GBNBRAL GONCLUSIOKS. 239

ftppearanoe of acid, when common air was detonated with
hydrogen, in such proportions as to condense nearly all the
oxygen, was quite consbtent with the results he had procured
when the latter gas was substituted for air* For when a mix-
ture of oxygen and nitrogen, in the proportions in which they
form common air, was employed along with less than a com-
bining measure of hydrogen, no nitrous acid appeared.

From the experiments recorded, Cavendish drew four con-
clusions, which, as stated by himself, are in effect as follows : —
1. When hydrogen and oxygen are exploded, with the latter in
excess, a small quantity of acid is developed. 2. From what-
ever source the oxygen is procured, ^* the acid is always of the
nitrous kind.^^ 3. If the hydrogen and oxygen are in such
proportions as to leave, after combination, only a slight residue
of the latter, ^^ the condensed liquor is not at all add, but seems
pure water, without any addition whatever.^' 4* It follows from
3, that '^almost the whole of the hydrogen and oxygen is con-
verted into pure water.^'

Cavendish, it will be observed, with the accuracy which as
mudi characterizes his statements, as his experiments, contents
himself with saying, that '^ almost the whole^^ of the mixture of
gases became water. The uncombined residue, however, he
goes on to show, was very small, '^ not more than -^ of tihe
dephlogisticated air employed, or -f^fth of the mixture.'^ The
existence of a residue, which he did not attempt to reduce to a
minimum or absolute zero, he refers to the impurities present
in the hydrogen and oxygen ; ^* consequently,^' he adds, ^ if
those airs oould be obtained perfectly pure, the whole would be
condensed.^ The absence of acid firom the liquid obtained by
detonating hydrogen with air, or with a mixture in atmospheric
proportions of oxygen and nitrogen, had still to be explained.
Cavendish thought it probable, that when the other conditions
for the production of acid were secured, namely, a small volume
of hydrogen to a large one of air, '^ the explosion is too weak,
and not accompanied with sufficient heat.'' Modem chemists
have acquiesced with its originator in the justness of this view.
A parenthetical passage then occurs, which it will afterwards
appear was added to the paper after it was read, and before it

240 CAVENDISH AS A CHEMIST.

was printed.* In this addition, Cavendish states that ^^ all the

foregoing experiments on the explosion of inflammable air with

common and dephlogisticated air, except those which relate to

the cause of the acid found in the water, were made in the

summer of the year 17BI, and were mentioned by me to Dr.

Priestley, who in consequence of it made some experiments of

the same kind, as he relates in a paper printed in the preceding

volume of the Transactions.f During the last summer also, a

friend of minej: gave some account of them to M. Lavoisier, as

well as of the conclusion drawn from them, that dephlogisticated

air is only water deprived of phlogiston; but at that time, so far

was M. Lavoisier from thinking any such opinion warranted, that,

till he was prevailed upon to repeat the experiment himself, he

found some difficulty in believing that nearly the whole of the

two airs could be converted into water. It is remarkable that

neither of these gentlemen found any acid in the water produced

by the combustion, which might proceed from the latter having

burnt the two airs in a different manner from what I did; and

from the former having used a different kind of inflammable air,

namely, that from charcoal, and perhaps having used a greater

proportion of it.'^

Without anticipating what will af):erwards be considered, it
may be noticed here that Lavoisier was of opinion, as he himself
tells us, '^que I'air inflammable en brQlant devoit donner de
I'acide vitriolique, ou de I'acide 8ulfureux.''§

He acknowledges his acquaintance with Cavendish's experi-
ments in the following terms: — ^^Ce fut le 24 Juin, 1783, que
nous fimes cette experience, M. de la Place et moi, en presence
de MM. le Roi, de Vandermonde, de plusieurs autres Acad€-

* The passage in question is the paragraph commencing, "All the Jbregoing
ejtperimentSt** and ending " a greater proportion of it** Phil, Trane., 1784, pp.
134-135. It is placed within brackets, in Mr. Mnirhead's reprint of the "Experi-
ments on Air." Watt Corr,, p. 129.

t For 1783, pp. 426 and 434.

X Dr., afterwards Sir C. Blagden.

§ Memoiree de VAcademie dee Sciencee pour 1781 (printed in 1784), p. 473,
reprinted by Mr. Muirhead in the Watt Corr,, p. 173. The pages of the original
memoirs of Lavoisier, Mensnier, and Monge, on the composition, &c., of water,
published in the M/moiree de VAcadSmie, 1784 and 1786, are copied here and
elsewhero from Mr. M airhead's convenient reprint, for the sake of those to whom
the French work may be more accessible.

LAVOISIER'S ALLEGED PLAGIARISM. 241

miciens^ et de M. Blagden^ aujourd'hui Secretaire de la Society
Royale de Londres; ce dernier nous apprit que M. Cavendish
aroit d^jii essay^^ a Londres^ de briUer de Pair inflammable dans
des Taisseaux fermes, et qu'il avoit obtenu ime quantity d'eau
trcB sensible/'*

Sir Charles (then Dr.) Blagden protested against this account
of matters, as concealing a part of the truth. His state-
ment is quoted here, as it shows who the friend was to whom
Cavendish referred, and what were the grounds on which he
made the reference to Lavoisier's knowledge of his experiments
and conclusions ; but to avoid repetition, the discussion of the
question of priority between Cavendish and Lavoisier is re-
served for another place. '*He (Lavoisier) should likewise have
stated in his publication, not only that Mr. Cavendish had
obtained ^ une quantity d'eau tr^s sensible,' but that the water
was equal to the weight of the two airs added together. More-
over, he should have added, that I had made him acquainted with
Messrs. Cavendish and Watt's conclusions ; namely, that water,
and not an acid or any other substance, arose from the com-
bustion of the inflammable and dephlogisticated airs.''t

From this episode. Cavendish returns to the consideration
of the subject which mainly interested him. Before doing so,
however, he thinks it *^ proper to take notice, that phlogisti-
cated air appears to be nothing else than the nitrous acid united
to phlogiston ; for when nitre is deflagrated with charcoal, the
acid is almost entirely converted into this kind of air." This
view was quite consistent with the phlogiston doctrines. When
nitre, which contained nitric or nitrous acid, was heated alone,
it gave off dephlogisticated air, i. e., oxygen ; when heated with
charcoal, it yielded phlogisticated air, i. e.^ nitrogen. Cavendish
was well aware, as he states lower down, that "a small part of the
acid, however, is turned into nitrous air, and the whole is mixed
with a good deal of fixed, and perhaps a little inflammable air,
both proceeding from the charcoal." As he did not, however,
know the composition of carbonic acid, or the nature of the
inflammable gas which charcoal yielded when heated, his mere

* Afim. de VAead. for 1781, p. 472, reprinted in Watt Corr,, p. 176.
t Letter from Dr. Blagden to Dr. Crell, published in Crell'a Chemische
Annalen, 1786, translated by Mr. Muirhead in Watt Corr., p. 73.

242 CAVENDISH AS A CHEMIST*

acquaintance with the fact of their production, gave him no
assistance in explaining the changes which occurred, and he
appears to have conceived that the nitrous acid obtained phlo-
giston from the carbon, and so became phlogisticated air. He
explains this more fully in the subsequent paragraphs, which
are too important from their bearing both on his views in
reference to the nature of water, and the production of nitric
acid by the action of the electric spark on air, to be omitted
here. He observes that ^^ it is well known that the nitrous
acid is also converted by phlogistication into nitrous air, in
which respect there seems a considerable analogy between that
and the vitriolic acid.'* The analogy is then illustrated in a
way which for brevity's sake I omit, and he proceeds : — ** In
like manner the nitrous acid, united to a certain quantity of
phlogiston, forms nitrous fumes and nitrous air, which readily
quit their phlogiston to common air; but when united to a
different, in all probability a larger quantity, it forms phlogis-
ticated air, which shews no signs of acidity, and is still less dis-
posed to part with its phlogiston than sulphur/'

In these explanations Cavendish, true to the guiding prin-
ciple of the phlogiston school, supposes an addition of phlo-
giston, where the present chemistry teaches that there is a loss
of oxygen. Vitriolic or sulphuric acid by combining with a
certain amount of phlogiston forms sulphurous acid, and by
taking up still more phlogiston constitutes sulphur. Nitric
acid united to a certain quantity of phlogiston forms nitrous
acid or nitric oxide ; united to still more phlogiston, it forms
nitrogen.

This being premised. Cavendish proceeds to apply the
views suggested, to the explanation of the acidity of the water,
obtained by exploding apparently pure hydrogen and oxygen
together. He indicates "two ways by which the phenomena
of the acid found in the condensed liquor may be explained;
first by supposing that dephlogisticated air contains a little
nitrous acid, which enters into it as one of its component parts,
and that this acid, when the inflammable air is in a sufficient
proportion, unites to the phlogiston, and is turned into phlo-
gisticated air, but does not when the inflammable air is in too
small a proportion ; and secondly, by supposing that there is no

CHANGE or VIEW CONCERNING PHLOGISTON. 243

skrous acid mixed witb> or entering into the composition of,
dephlogisUcated air; but that, when this air is in a sufficient
proportion^ part of the phlogisticated air with which it is
debased, is, by the strong affinity of phlogiston to dephlogisti-
cated air, deprived of its phlogiston and turned into nitrous acid;
whereas, when the dephlogisticated air is not more than suffi-
cient to consume the inflammable air, none then remains to
deprive the phlogisticated air of its phlc^iston, and turn it into
acid/^

It depended upon which of these views was adopted, what
opinion should be held concerning the nature of water as well
as of oxygen. The inseparability of the questions ; what was
the source of the nitrous acid } and what the composition of the
water which it accompanied ? has generally been overlooked in
the criticisms which have been published of Cavendish's views,
but in his apprehension it was essential that they should be
studied together* In this spirit he proceeds to say — ^^ If the
latter explanation be true, I think we must allow that dephlo-
gisticated air is in reality nothing but dephlogisticated water, or
water deprived of its phlogiston ; or in other words, that water
consists of dephlogisticated air united to phlogiston ; and that
inflammable air is either pure phlogiston, as Dr. Priestley and
Mr. Kirwan suppose, or else water united to phlogiston ; since
according to this supposition, these two substances united
together form pure water.'' It has already been noticed that
when Cavendish experimented on hydrogen in 1766, he regarded
that body as phlogiston (see p. 197). He now gives Priestley
and Kirwan the credit of this opinion which he abandons, in
flavour of the view that hydrogen is, what we should now call a
hydrate of phlogiston. In a note he assigns as his reason for
this change of belief, the observation, '^ that common or dephlo-
gisticated air do not absorb phlogiston from inflammable air,
unless assisted by a red heat, whereas they absorb the phlogiston
of nitrous air, liver of sulphur, and many other substances,
without that assistance; and it seems inexplicable, that they
should refuse to unite to pure phlogiston, when they are able
to extract it from substances to which it has an affinity;
that is, that they should overcome the affinity of phlogiston
to other substances, and extract it from them, when they

244 CAVENDISH AS A CHEMIST.

will not even unite to it when presented to them/^ According
to this yiew^ hydrogen could not be a simple substance because
it did not combine at ordinary tem])eratures with air or oxygen
although they have a great affinity for it, but must be regarded
as a compound of phlogiston more difficult of decomposition,
by oxygen^ than nitric oxide or the alkaline sulphurets were*
The question is further enlarged on by Cayendish, but its con-
sideration is reserved till the discussion of the water controversy
is reached.

Such was the view of the nature of oxygen^ hydrogen and
water^ if the second supposition were adopted, as in fact it was
by Cavendish. '^ On the other hand, if the first explanation be
true, we must suppose that dephlogisticated air consists of water
united to a little nitrous acid, and depriyed of its phlogiston ;
but still the nitrous acid in it must make only a yery small part
of the whole, as it is found that the phlogisticated air, which it
is converted into, is very small in comparison of the dephlogis-
ticated slit"

The second of these explanations Cavendish adopts as
^'much the most likely,^^ and he assigns three significant
reasons for the preference.

1. " It was found that the acid in the condensed liquor was
of the nitrous kind, not only when the dephlogisticated air was
prepared from red precipitate, but also when it was procured
from plants or from turbith mineral ; and it seems not likely
that air procured from plants, and still less likely that air pro^
cured from a solution of mercury in oil of vitriol, should contain
Any nitrous acid.^*

2. " Another strong argument in favour of this opinion is,
that dephlogisticated air yields no nitrous acid when phlogisti-
cated by liver of sulphur; for if this air contains nitrous acid,
and yields it when phlogisticated by explosion with inflammable
air, it is very extraordinary that it should not do so when phlo-
gisticated by other means.^^

3. ^^But what forms a stronger and, I think, almost decisive
argument in favour of this explanation, is, that when the de-
phlogisticated air is very pure, the condensed liquor is made
much more strongly acid by mixing the air to be exploded
with a little phlogisticated air, as appears by the following

COMPOSITION OF WATER, 245

experiments/^ Cavendish then gives the details of two experi-
ments, in each of which the same amount of oxygen and
hydrogen from the same specimens was employed^ but in the
second trial a certain volume of nitrogen was added, which had
the effect of increasing the proportion of nitric acid^ as ascer-
tained by a quantitative analysis of the liquid in both cases. ^' It
must be observed/^ he remarks^ " that all circumstances were
the same in these two experiments, except that in the latter the
air to be exploded was mixed with some phlogisticated air,
and that in consequence the burnt air was more phlogisticated
than in the former; and from what has been before said [t. e.,
in the £rst explanation, as to the presence e:^ hypoihesi of
nitrous acid ready formed in the oxygen], it appears that this
latter circumstance ought rather to have made the condensed
liquor less acid; and yet it was found to be much more so,
which shows strongly that it was the phlo^sticated air which
furnished the acid/^

Another pair of trials leading to the same inference is then
given, but it does not call for minute consideration, and the
general conclusion from the whole is stated thus : ^' From what
has been said, there seems the utmost reason to think that
dephlogisticated air is only water deprived of its phlogiston,
and that inflammable air, as was before said, is either phlogisti-
cated water, or else pure phlogiston; but in all probability the
former/^

This is the last direct reference to the composition of water
which the paper contains, but in the interval between its com-
munication to the Royal Society and its publication, an essay
on the same subject by Mr. Watt was read at one of the public
meetings of that body. Cavendish in consequence added the
following passage, which^ as it has been much animadverted on,
is here given entire.*

*^ As Mr. Watt, in a paper lately read before this Society,
supposes water to consist of dephlogisticated air and phlogiston
deprived of part of their latent heat, whereas I take no notice
of the latter circumstance, it may be proper to mention in a few

* The passage in question is the paragraph beginniDg **A» Mr. Wail,** and
ending "than it U worth." {Phil. Trans, for 1784, pp. 140, 141.) It is marked
.by brackets in Mr. Muirhead's reprint. {Watt Corr., pp. 135, 136.)

24^ CAVKNDtSH AS A CHEMIST.

words the reason of the apparent difference between us. If
there be any such thing as elementary heat, it must be allowed
that what Mr. Watt says is true; but, by the same rule, we
ought to say that the diluted mineral acids consist of the con-
centrated adds united to water, and deprived of part of their
latent heat; that solutions of sal-ammoniac, and most other
neutral salts, consist of the salt united to water and elemen-
tary heat; and a similar language ought to be used in speaking
of almost all chemical combinations, as there are very few
which are not attended with some increase or dimmution of
heat. Now I have chosen to avoid this form of speaking, both
because I think it more likely that there is no such thing as
elementary heat, and because saying so in this instance, without I

using similar expressions in speaking of other chemical unions,
would be impioper, and would lead to false ideas; and it may \

even admit of doubt whether the doing it in general would not
cause more trouble and perplexity than it is worth.''

Cavendishes views on the nature of heat, which he did not
regard as a material entity, will appear when his papers on
Freezing Mixtures are under review. In practice, his succes-
sors have for the greater part agreed with him in discarding \
Watt's special reference to the evolution of latent heat, as not
more called for in the case of hydrogen and oxygen^ than in
that of other combining substances. But no theory of the
synthesis of water can be complete, which does not account for
the evolution of heat which accompanies the union of its ele-
ments. In so far, therefore, as Mr. Watt attempted to explain
this, his view is more perfect, but it is now-a-days confessedly ^
insufficient to account for the phenomena, which still await a
satisfactory explanation. *

The remainder of Cavaidish's paper is chiefly employed
with speculations on the nature of common air and oxygen, to
which his discoveries led him. He begins by stating, more
fully than he had done in his paper on the eudiometer, his con-
viction that Scheele and Lavoisier are right in supposing that
^^ dephlogisticated and phlogisticated air are quite distinct sub-
stances, and not differing only in their degree of phlogistication;
and that common air is a mixture of the two." The proof of
this which he adduces, is one which they did not give, and

THEORY OF PBOCESSES FOR OXTGEN. 247

could not have given ; namely^ " if the dephlogisticated air is
pretty pure^ almost the whole of it loses its elasticity by phlo-
gistication^ and, as appears by the foregoing experiment, is
turned into water, instead of being converted into phlogisticated
air/' He goes on to say, that though in his experiments the
whole of the mixture of oxygen and hydrogen was not turned
into water, at least -If ths were ; and that by liver of sulphur he had
reduced oxygen to less than ^tb of its bulk, and other parties
had reduced it further; ^^so that there seems the utmost reason
to suppose that the small residuum which remains after its phlo-
gistication proceeds only from the impurities mixed with it/'
From this it should seem that Cavendish believed the diminution
of air by liver of sulphur to result from the oxygen of the former
combining with hydrogen derived from the alkaline sulphuret,
and so changing into water. He then introduces a passage
quoted at length in the notice of the paper on the eudiometer
(p. 229), as to the proportion of oxygen present in air, which
he states to be " one-fifth of its bulk;'' and continues: " From
what has been said, it follows, that instead of saying air is phlo-
gisticated or dephlogisticated by any means, it would be more
strictly just to say it is deprived of, or receives, an addition of
dephlogisticated air." This passage is interesting, as an ap-
proach to the language of the Lavoisierian school, which teaches
that oxygen is added, where the believers in phlogiston held
that the latter was removed, but Cavendish clung too closely to
his faith in phlogiston, to abandon willingly its nomenclature,
and he immediately excuses himself for making the proposed
change, ^^as the other expression is convenient, and can scarcely
be called improper, I shall (says he) still frequently make use
of it in the remainder of this paper." This remainder will not
require so full an abstract of its contents, as the portions already
considered.

The light which Cavendish had now obtained on the true com-
position both of water and of air, changed of necessity his views
concerning the theory of the methods in use for preparing oxygen.
Priestley had supposed that both the vitriolic and nitrous acids
were convertible into oxygen, as this gas could be prepared in
the greatest quantity from substances containing those acids,
especially the nitrous. Cavendish, however, thought that his

243 CAVENDISH AS A CHEMIST.

researches seemed '^ to show that no part of the acid is con-
verted into dephlogisticated air/' In corroboration of this view,
he refers to experiments proving that red precipitate^ though it
yields oxygen abundantly, contains no nitrous acid, ^^ and con-
sequently that, in procuring dephlogisticated air from it, no acid
is converted into air j and it is reasonable to conclude, therefore,
that no such change is produced in procuring it from any other
substance/*

He then considers in what manner those acids act in pro-
ducing dephlogisticated air, and after adducing certain consi-
derations which seem to warrant the inference, states it as his
conclusion, ^^ that the red precipitate may be considered, either
as quicksilver deprived of part of its phlogiston, and united
to a certain portion of water, or as quicksilver united to depUo^
gisticated air ; after which, on further increasing the heat, the
water in it rises deprived of its phlogiston, that is, in the form
of dephlogisticated air, and at the same time the quicksilver
distils over in its metallic form/' The passage marked above
in italics, shows how clearly Cavendish apprehended the possibi-
lity of reversing the form of explanation current among be-
lievers in phlogiston, but he showed no preference for this
view. In a note he observes, ** It would be ridiculous to say,
that it is the quicksilver in the red precipitate which is deprived
of its phlogiston, and not the water, or that it is the water and
not the quicksilver; all that we can say is, that red precipitate
consists of quicksilver and water, one or both of which are
deprived of part of their phlogiston/' This passage is import-
ant as showing, what many other passages in the papers of
Cavendish and his contemporaries also show, that the disco-
very of the composition of water would not, in the hands of
the disciples of the phlogiston school, have materially altered
the aspect of chemistry. The difference it introduced was
little more than this, that where formerly transferences of phlo-
giston, ffom one body to another, were assumed to take place,
now water instead of phlogiston was shifted backwards and for-
wards, and decomposed and recomposed as the exigencies of
theory required. The whole of the concluding part of Caven-
dish's paper, and of Watt's "Thoughts on the Constituent
Parts of Water," amply illustrate this, and the former distinctly

EVOLUTION OF OXYGEN BY LIVING PLANTS. 249

enounces it in the continaation of his remarks : "Mercurius Cal-
cinatuSy" he observes, ^'appears to be only quicksilver, which has
absorbed dephlogisticated air from the atmosphere during its
preparation; accordingly, by giving it a sufficient heat^ the de-
phlogisticated air is driven off, and the quicksilver acquires its

original form ; but yet, as uniting dephlogisticated air

to a metal comes to the same thing as depriving it of part of its
phlogiston^ and adding water to it, the quicksilver may still be
considered as deprived of its phlogiston/' After the same man-
ner he accounts for the evolution of oxygen from nitre by the
assumption^ ^^ that the acid absorbs phlogiston from the water
in the nitre, and becomes phlogisticated, while the water is
thereby turned into dephlogisticated air/^* A little further
on, however, he suggests it as not unlikely "that part of
the acid in the nitre is turned into phlogisticated air, by absorb*
ing phlogiston from the watery part/'

A good deal of space is then devoted to the theory of the evo-
lution of oxygen from the sulphate of the red oxide of mercury,
and turbith mineral, which need not be minutely dwelt on, as
the general conclusion is " that the rationale of the production
of dephlogisticated air from turbith mineral, and from red pre-
cipitate, are nearly similar/' Some remarks follow on the
mode in which vitriolic acid acts in the production of dephlo-
gisticated air, which do not call for special notice, and Caven-
dish proceeds, "There is another way by which dephlogisti-
cated air has been found to be produced in great quantities,
namely, the growth of vegetables exposed to the sun or day-
light; the rationale of which in all probability is, that plants,
when assisted by the light, deprive part of the water sucked up
by their roots, of its phlogiston, and turn it into dephlogisti-
cated air, while the phlogiston unites to, and forms part of,
the substance of the plant/' It need scarcely be mentioned,
that the view suggested here by Cavendish, that plants can de-
compose water and retain its hydrogen is probably well founded,
but that the chief source of the oxygen which plants so
largely evolve, is not decomposed water, but decomposed car-
bonic acid. In confirmation of his view. Cavendish adduces

* Watt held an exactly similar view. Phil, Trans, for 1784, p. 336.

250 CAYENDISH AS A CHBMIST.

many proofs ^' that light has a remarkable power in enabling
one body to absorb phlogiston from another/' He cites as
illustrations of this: I. the bleaching by sunlight of a spirituous
tincture of green leaves, when contained in botties only par-
tially filled, so as to contain, besides the liquid, a portion of
air; 2. the yellow tint which colourless nitric acid acquires by
exposure to light; and 3. the effect of the same agent in reducing
to the metallic state moist salts of silver and gold. In refer-
ence to the last example, he adds, ^' There is the utmost reason
to think, that, in both cases, the revival of the metal is owing
to its absorbing phlogiston from the water/'

Having in this way shown the probability of plants possess-
ing the power he attributes to them, he observes that '^ v^^e-
tables seem to consist almost entirely of fixed and phlogisticated
air, united to a large proportion of phlogiston and some
water, since, by burning in the open air, in which their phlo-
giston unites to the dephlogisticated part of the atmosphere,
and forms water, they seem to be reduced almost entirely to
water, and those two kinds of air. Now plants growing in water
without earth, can receive nourishment only from the water
and air, and must, therefore, in all probability, absorb their
phlogiston from the water/' The use of light in promoting the
growth of plants. Cavendish thus refers to its enabling them to
absorb phlogiston from the water. He alludes also to the
fact that plants yield more oxygen when growing in water im-
pregnated with fixed air, than they do in plain distilled water,
but he did not understand the function of the carbonic acid,
of the composition of which he was ignorant, nor did he sus-
pect that it was decomposed. His explanation of its action
was that ^^as fixed air is a principal constituent part of vege-
table substances, it is reasonable to suppose that the work of
vegetation will go on better in water containing this substance
than in other water."

This is the concluding passage of the paper as it was origi-
nally sent to the Society, but its author added a passage between
the reading and the printing of his essay, which is of great im-
portance, as containing a comparison of the merits of the phlo-
giston hypothesis of which he was certainly the ablest English

CRITICISM OF LAVOISIER'S VIEWS. 251

adTOcate^ and the antiphlogiston theory which the great French
chemist was now enforcing by irresistible arguments.* ^ There
are several memoirs/^ he observes, "of Mr. Lavoisier, pub-
lished by the Academy of Sciences, in which he intirely dis-
cards phlogiston, and explains those phenomena which have
been usually attributed to the loss or attraction of that sub-
stance by the absorption or expulsion of dephlogisticated air;
and as not only the foregoing experiments, but most other phe-
nomena of nature, seem explicable as well, or nearly as well,
upon this as upon the commonly believed principle of phlogis-
ton, it may be proper briefly to mention in what manner I
would explain them on this principle, and why I have adhered
to the other.'* Cavendish then gives a short but very perspi-
cuous sketch of Lavoisier's views, which it is not necessary te
quote at length, but one statement demands notice. *' Accord-
ing to this [Lavoisier's] hypothesis, we must suppose that
water consists of inflammable air united to dephlogisticated air."
It seems at first sight singular that Cavendish should use these
words to express Lavoisier's opinion, for they are identical with
those in which the former expressed his own view. The differ-
ence, therefore, must have lain in the meaning attached to the
words by the French and English chemist, and it is important,
in reference to the assignment of the due share of merit to all
parties, to notice that Cavendish imputed a different opinion to
Lavoisier from that which he held himself. The difference lay
solely in the view entertained concerning inflammable air.
According to the former, it was a principle common to many
combustibles, and either alone, or in combination with water,
assumed the form of an elastic fluid. According to the latter it
was a substance sui generis, present in some of the compound
combustibles, but not contained in those which were simple,
such as sulphur, carbon, phosphorus, and the metals. The
diffierence of opinion went for little, in reference to water con-
sidered alone, but it was of the greatest importance in refer-
ence to the theory of chemical changes in which combustibles
and oxygen took a part.

* The added passage begini, " There are several memoire,'* and ends * those
three substances.** {PhU. Trans, for 1784, pp. 150—153.) It is inclosed in
brackets in Mr. Muirhcad's reprint. (Watt Corr., pp. 147—150.)

252 CAVENDISH AS A CHEMIST.

After stating Lavoisier's view of the composition of water^
the oxyacids, and the metallic calces, and his rationale of the
evolution of oxygen from red precipitate and nitre, as well as
from growing plants. Cavendish comments upon it thus : — ^^ It
seems, therefore, from what has been said, as if the phenomena
of nature might be explained very well on this principle without
the help of phlogiston; and, indeed, as adding dephlogisticated
air to a body comes to the same thing as depriving it of its phlo-
giston, and adding water to it, and as there are perhaps no
bodies entirely destitute of water, and as I know no way by
which phlogiston can be transferred from one body to another,
without leaving it uncertain whether water is not at the same
time transferred, it will be very difficult to determine by expe-
riment which of these opinions is the truest; but as the com-
monly received principle of phlogiston explains all phenomena,
at least as well as Mr. Lavoisier's, I have adhered to that."

This view of Cavendish's was referred to before. It shows
strikingly the necessity under which a false theory lies of multi-
plying falsities, when contradicted by an experimenium crucis.
Instead of an appeal to the Balance, which would have at once
shown that only one of the theories consisted with truth, or
an endeavour to collect and exhibit the water, which according
to Cavendish was produced at every oxidation, the necessary
experiments were declined on the score of their difficulty, and
the production of water assumed, instead of demonstrated. At
a later period, a similar line of defence was adopted by those
who advocated the doctrine of the composite nature of chlorine,
and named it Oxymuriatic Acid, but it was brought to the test
of direct trial, and water was shown not to be produced. In truth.
Cavendish's defence of his antiphlogiston views, has all the
appearance of an after-thought, and was added after his opi-
nions had been made public^ as a justification of what could not
then be withdrawn. He is more successful in disputing the
truth of Lavoisier^s doctrine ^^ that dephlogisticated air is the
acidifying principle," which he acknowledges to be true if no
more be meant than that the addition of phlogiston (t. e., the
loss of oxygen) deprives certain acids of their acidity. This^
however, he did not think was a universal phenomenon^ for
'^ as to the marine acid and acid of tartar, it does not appear that

kirwan's remarks. 25a

they are capable of losing their acidity by any union with
phlogiston/' or as Lavoisier would have stated the fact, by any
loss of oxygen. This shrewd observation, it need not be said,
has been confirmed, so far at least as marine or hydrocliLoric
acid is concerned, but there is no reason to suppose that
Cavendish was aware of the true nature of that acid.

The ^^Experiments on Air*' were immediately subjected to
criticism, in so far as they were adduced by their author, as
disproving the doctrine that the phlogistication of air was
always attended by the production of carbonic acid.

Richard Kirwan, one of the most accomplished chemists of
the time, held this view strongly, and published ^^ Remarks on
Mr. Cavendish's Experiments on Air.'' The paper was read
to the Royal Society on February 5th, 1784, about three weeks
after that on which it commented, and it follows Cavendish's
in the vol. of Transactions for that year.* Kirwan endeavours to
prove that during the calcination and amalgamation of metals,
and by the action of nitric oxide, and of the electric spark
on atmospheric air, fixed air (carbonic acid) is produced.
It is unnecessary to discuss the paper at length, as its views,
for the greater part, are now discredited, and the objections
adduced against Cavendish's conclusions are known to be
unfounded. For the same reason, it is unnecessary to give
Cavendish's ^^ Answer to Mr. Kirwan's Remarks upon the
Experiments on Air," which was read to the Royal Society,
March 4th, 1784, and follows Kirwan's paper in the Transac-
tions.t An extract from the conclusion (p. 175) wiU sufil-
ciently illustrate the nature of Cavendish's reply. ^^ There
are five methods of phlogistication considered by me in my
paper on air, namely: first, the calcination of metals, either
by themselves, or when amalgamated with quicksilver; secondly,
the burning of sulphur or phosphorus ; thirdly, the mixture of
nitrous air; fourthly, the explosion of inflammable air; and

fifthly, the electric spark As to the first method,

or the calcination of metals, there is not the least proof that
any fixed air is generated, though we certainly have no direct
proof of the contrary ; nor did I in my paper insinuate that we
had. The same thing may be said of the burning of sulphur

♦ Vol. IxxiY., p. 154. t Ibid., p. 171.

254 CAVSNDISH AS A CHEMIST.

and phosphorus. As to the mixture of nitrous air and tiie
combustion of inflammable air, it is proved that if any fixed air
is generated, it is so small as to elude the nicest test we have.

So that out of the five methods enumerated^ it has

been shown, that in two no sensible quantity is generated, and
not the least proof has been assigned that any is in two of the
others; and as to the last [the electric spark] good reasons
have been assigned for thinking it inconclusiye, and therefore
the conclusion drawn by me in the above-mentioned paper
seems sufficiently justified ; namely, that though it is not impos*
sible that fixed air may be generated in some chemical pro-
cesses, yet it seems certain that it is not the general effect of
phlogisticating air, and that the diminution of common air by
phlogistication is by no means owing to the generation or sepa-
ration of fixed air from it.''

This paper was followed by a ^ Reply to Mr. Cavendish's
Answer," read to the Royal Society, March 18th, 1 784,* in which
Kirwan contests some of Cavendish's opinions, but very briefly^
because he considers, as he states at the beginning, the greater
part of his reply to Cavendish as still unanswered. The latter^
however, entered uito no further discussion as to the general
question, what causes the diminution of air during its deoxida-
tion.

In one respect, Cavendish's answer was incomplete. He
acknowledged that he could not give any proof of the truth of
his opinion, that the diminution of the sur by the electric sparky
^ is owing to the burning or calcination of some substance con-
tained in the apparatus*" He proceeded, however, soon after,
to test the justice of this theory^ and found that it was only
partially true, and that the production of nitric acid was the
chief cause of the diminution of sir by the electric spark. This
leads directly to his next paper which contains the announce-
ment of his great discovery of the composition of nitric acid.
It bore the same title as his previous communication.

* Pkil, TVoiM., vol. Izziv., p. 178.

ACTION or ELBCTRIC SPAEK ON AIR. 255

Experiments on Air.
Second Series.

This paper* is professedly a continuation of that recording
the first series of Experiments on Air. In it Cavendish had
conjectured, as he now repeats, that the diminution in bulk
which air undergoes, when phlogisticated by the electric spark,
''was owing to the burning of some inflammable matter in
the apparatus ^ and that the fixed air supposed to be produced
in that process, was only separated from that inflammable
matter by the burning.'^ He now records experiments made
with a view to test the truth of this conjecture, which showed
that the real cause of the diminution was very different from
what he suspected, and depended ''upon the conversion of
phlogisticated air into nitrous acid.^' The experiments are
then related which led to the important discovery of the nature
of that acid. The first part of the paper is occupied with an
account of the apparatus made use of, which need not be
minutely described. It consisted essentially of a glass syphon
fiUed with quicksilver and inverted, so that each of the limbs
stood in a glass of the same fluid. Air was then passed up into
the syphon tlirough the quicksilver, and soap-lees, or any other
liquid employed during the experiment, was introduced in the
same way. To transmit the electric spark, an insulated ball
connected with the quicksilver in one glass, was placed at such
a distance from the conductor of a Friction Electrical Machine,
as to receive a spark, whilst the quicksilver in the other glass
communicated with the ground.

In the first experiment, " When the electric spark was made
to pass through common air included between short columns
of a solution of litmus, the solution acquired a red colour and
the air was diminished, conformably to what was observed by
Dr. Priestley.'^ Lime-water was then substituted for the
litmus, and the spark passed till the air could be no further
diminished. No cloudiness appeared in the liquid, but the air
was reduced to two-thirds of its original bulk, ^' which is a
greater diminution than it could have sufiered by mere phlo-

* Read Jane 2iid, 1785, Phil, TVafu., 1785, p. 372.

256 CAVENDISH AS A CHEMIST.

gistication, as that is very little more than one-fifth of the
whole." The experiment was then repeated with imptire
oxygen, but no cloudiness of the lime-water appeared, and it
was not precipitated by the addition of carbonic acid, but
yielded a brown sediment on the further addition of ammonia.
From these phenomena. Cavendish infen^ed that "the lime-
water was saturated by some acid formed during the operation/'
but that no fixed air was produced. Soap-lees, however,
(solution of caustic soda or perhaps potash) were found to cause
a more rapid diminution of air than lime-water did, so that they
were employed in the subsequent experiments. The first of
these was made with a view to determine " what degree of
purity the air should be of, in order to be diminished most readily,
and to the greatest degree.'^ Oxygen suffered a slight diminu-
tion^ doubtless owing to the presence of impurities, such as
nitrogen and traces of combustible matter. When nitrogen
was used, "no sensible diminution took place; but when
five' parts of pure dephlogisticated air were mixed with three
parts of common air, [or, according to Cavendish, three
volumes of nitrogen with seven of oxygen,] almost the whole
of the air was made to disappear.''* This ascertained, a mixture
of nitrogen and oxygen, in the proportions stated above, was
exposed to the electric spark whilst confined over soap-lees.
As fast as the air was diminished, fresh quantities were sup-
plied, till no further diminution took place. The soap-lees
were then poured out of the tube, separated from the quick-
silver, and examined as to the change they had undergone*
They " seemed to be perfectly neutralised, as they did not at
all discolour paper tinged with the juice of blue flowers. Being
evaporated to dryness, they left a small quantity of salt, which

* Cayendiflh's calculation of the relative amonnt of oxygen and nitrogen present
in the mixture of these gaaes, which he entirely conyerted into nitric acid« is erro-
neous. The mistake must have been accidental, for he states^ as the basis of his
calculation, "that common air consists of one part of dephlogisticated air (oxygen),
mixed with four of phlogisticated (nitrogen)/' Three measures, therefore, of air,
could contain only six-tenths of a measure of oxygen, instead of two measures,
which he assumed to be present. The actual composition of his mixture of 5 parts
of oxygen with 3 parts of air, was — oxygen 5*6 yolumes, nitrogen 2*4. He should'
hare taken 5 measures of air to 9 of oxygen, or what is the same proportion, 2 mea-
sures of nitrogen to 5 of oxygen. The mixture he did employ nearly approached in
composition to this.

SYNTHETICAL PRODUCTION OP NITRIC ACID. 257

was evidently nitre^ as appeared by the manner in which paper
impregnated with a solution of it burned.^^ The experiment
was repeated on a larger scale^ and with still more decisive
results. '^ The liquor^ when poured out of the tube, smelled
evidently of phlogisticated nitrous acid, and being evaporated
to dryness, yielded 1-^ gr. of salt^ which is pretty exactly equal
in weight to the nitre which that quantity of soap-lees would
have afforded^ if saturated with nitrous acid. This salt was
founds by the manner in which paper dipped into a solution of
it burned, to be true nitre.'^ It contained a trace of vitriolic
acid^ but not more than the soap-lees originally did^ and ^^ there
is no reason to think that any other acid entered into it^
except the nitrous.^^ At firsts indeed^ it appeared that some
hydrochloric acid had been produced^ for the saturated soap-lees
precipitated nitrate of silver; but Cavendish set this apparent
evidence of hydrochloric acid aside^ referring the precipitation
to the acid produced having been much phlogisticated^ in
other words^ to its having been one of the lower oxides of
nitrogen.

The production of nitric acid being thus beyond doubt^
Cavendish proceeded to form a theory as to its generation. In
his previous paper, detailing ^^Experiments on Air/' he had
referred to the results obtained when charcoal is detonated with
nitre as proving that ^^ phlogisticated air [nitrogen] is nothing
else than nitrous acid united to phlogiston.'^ He now considers
what are the logical consequences of this view^ if it be followed
out. The statement is best given in his own words. ^^ Accord-
ing to this conclusion, phlogisticated air ought to be reduced to
nitrous acid by being deprived of its phlogiston. But as
dephlogisticated air is only water deprived of phlogiston, it is
plain, that adding dephlogisticated air to a body, is equivalent
to depriving it of phlogiston, and adding water to it; and there-
fore, phlogisticated air ought also to be reduced to nitrous acid,
by being made to unite to, or form a chemical combination
with^ dephlogisticated air; only the acid formed this way will
be more dilute than if the phlogisticated air was simply de-
prived of phlogiston." In other words, nitrogen is a compound
of nitrous acid and phlogiston, which is a hydrate of inflam-
mable air, and when oxygen is added to nitrogen, it unites with

361 CAVENDISH AS A CHEMIST.

the phlogiston of the latter^ producing water^ which dilutes the
nitrous acid separated or set free at the same time*. Caven*
dish then continues^ '^This being premised^ we may safely con-
clude^ that in the present experiments the phlogisticated air
was enabled, by means of the electrical spark, to unite to, or
form a chemical combination with, the dephlogisticated air, and
was thereby reduced to nitrous acid, which united to the soap-
lees and formed a solution of nitre; for in these experiments
those two airs actually disappeared, and nitrous acid was actu-
ally formed in their room ; and as, moreover, it has just been
shown, from other circumstances [namely, from the action of
the charcoal on nitre referred to in the previous paper, AfUe^
p. 241], that phlogisticated air must form nitrous acid, when
combined with dephlogisticated air, the above-mentioned opi-
nion seems to be sufficiently established. A further confirma-
tion of it is, that, as iar as I can perceive, no diminution of air
is produced when the electric spark is passed either through
pure dephlogisticated air, or through perfectly phlogisticated
air; which indicates the necessity of a combination of these
two airs to produce the acid. Moreover, it was found in the
last experiment that the quantity of nitre procured was
the same that the soap-lees would have produced if saturated
with nitrous acid; which shows, that the production of
the nitre was not owing to any decomposition of the soap-
lees.'*

Nothing can be clearer than the interpretation here given of
the immediate phenomena concerned in the production of nitrous
acid, although the more remote reactions were obscured by
Cavendish's false notions as to the nature of inflammable air
which have frequently been referred to. The bearing of these
results on those obtained in the previous experiments, on the \

occasional production of acid along with water, when hydrogen !

and oxygen were detonated together is evident, and was fully i

recognised by Cavendish. In reference to his earlier researches, ^

he remarks, « I also gave my reasons for thinking, that the small |

a hld^X^i T'^^ J^'" ^"^^ ^** *^y^^» according to our present nomenclature. \

set free. »c»nlting water combined with the nitric acid, which was j

NATURE OF NITROGEN. 259

quantity of nitrous acid, produced by the explosion of dephlo^
gisticated and inflammable air^ proceeded from a portion of
phlogisticated air mixed with the dephlogisticated^ which I sup-
posed was deprived of its phlogiston, and turned into nitrous
acid by the action of the dephlogisticated air on it, assisted by
the heat of the explosion. This opinion, as must appear to
every one^ is confirmed in a remarkable manner by the fore-
going experiments ; as from them it is evident^ that dephlogis-
ticated air is able to deprive phlogisticated air of its phlogiston,
and reduce it into acid, when assisted by the electric spark; and
therefore it is not extraordinary that it should do so, when
assisted by the heat of ^he explosion/^

The discovery of the power of hydrogen to produce water
when detonated with oxygen, led directly to speculations on
the nature of hydrogen and oxygen ; and the derivation of nitric
acid irom nitrogen, led in like manner to theories concerning
its intrinsic qualities. Cavendish, accordingly, followed up his
discovery of the composition of nitric acid by an inquiry into
the nature of nitrogen. He points out in the sequel of his paper
that the known properties of this gas are all negative, and that
it might fairly be doubted *^ whether there are not, in reality,
many different substances confounded by us under the name of
phlogisticated air.'^ He ''therefore made an experiment to
determine, whether the whole of a given portion of the phlo-
gisticated air of the atmosphere could be reduced to nitrous acid,
or whether there was not a part of a different nature from the
rest, which would refuse to undergo that change/^ On trial, he
found that so small a quantity of nitrogen escaped conversion
into nitric acid, " that, if there is any part of the phlogisticated
air of our atmosphere which differs from the rest, and cannot
be reduced to nitrous acid, we may safely conclude that it is not
more than T^th part of the whole.^^

The remainder of the paper is occupied with a detail of
experiments, instituted with a view to ascertain whether, "when
any liquor, containing inflammable matter, was in contact with
the air in the tube, some of this matter might be burned by the
spark, and thereby diminish the air." To determine this point,
oxygen was confined over distilled water, soap-lees, and infusion
of litmus respectively ; so that while the conditions essential to

260 CAVENDISH AS A CHEMIST.

the production of nitric acid were wanting^ those condocing to
combustion were most fully realised. Very slight diminutioa
occurred with the first two liquids, owing, doubtless^ to the
presence of a little air, and therefore of nitrogen. When a solu*
tion of litmus was employed, it became first red, and^ by^and^
by, as successive sparks passed, paler and paler, till it was
qtute colourless and transparent* When lime-water was let up
into the tube, it became cloudy ; '^ therefore the litmus was, if
not burnt, at least decompounded^ so as to lose entirely its purple
colour, and to yield fixed air/^

The account which Cavendish gave of his experiments was so
full and explicit, that apparently no one could fail of success
who repeated them. It was otherwise, however ; and in conse-
quence he published the following paper, nearly three years after
the first :—

On the Conversion of a Mixture of Dephlot/isticated and Phlo-
ffisticated Air into Nitrous Acid by the Electric Spark*

After a reference to the nature of the experiments recorded
in the first communication on nitric acid to the Royal Society,
Cavendish says : " As this experiment has since been tried by
some persons of distinguished ability in such pursuits without
success, I thought it right to take some measures to authenticate
the truth of it. For this purpose I requested Mr. Gilpin, clerk
to the Royal Society, to repeat the experiment, and desired some
of the gentlemen most conversant with these subjects to be
present at putting the materials together, and at the examination
of the produce.^*

From a later part of the paper (p. 271), it appears that the
parties ^'who have endeavoured to repeat this experiment are,
M. Van Marum, assisted by M. Pacts Van Trootswyk ; M. La-
voisier, in conjunction with M. Hassenfratz; and M. Monge.'^
" I am not acquainted,'^ adds Cavendish, '^ with the method
which the three latter gentlemen employed, and am at a loss
to conceive what could prevent such able philosophers from
succeeding, except want of patience."

As later chemists have found no difficulty in repeating
Cavendish's experiments on the production of nitric acid by

« Kead to the Royal Society, April 17, 1788; Phil. Tram., lixvul, p. 26]|
1788.

BEPETTTION OF EXPERIMENTS ON NITRIC ACID. 261

the synthesis of nitrogen and oxygen^ it is not necessary to ana«
lyse this paper minutely.

Two repetitions were made with the same apparatus as
Cavendish himself employed. The tedious transmission of
electrie sparks was managed by Mr. Gilpin alone^ but the com-
mencement and conclusion of the experiment were witnessed by
others. ^' On December 6^ 1737^ in the presence of Sir Joseph
Banks, Dr. Blagden, Dr. Dollfiiss, Dr. Fordyce, Dr. J. Hunter,
and Mr. Made, the materials were put together.'^

^^On January 28 and 29, the produce of this experiment
was examined in the presence of Sir Joseph Banks, Dr. Blagden,
Dr. Dollfuss, Dr. Fordyce, Dr. Heberden, Dr. J. Hunter, Mr.
Maeie, and Dr. Watson.^' The result of this repetition was, in
the words of Cavendish, '^ that the mixture of the two airs was
actually converted into nitrous acid ; only the experiment was
continued too long, so that the quantity of air absorbed was
greater than in my experiments, and the acid produced was
sufficient, not only to saturate the soap-lees, but also to dissolve
some of the mercury .'' A second repetition accordingly was
commenced on February 29, 1788; and ^^ on March 19, the
produce was examined, in the presence of Dr. Blagden, Dr.
Dollfuss, Dr.Fordyce, Dr. Heberden, Dr. J« Hunter, Mr. Macie,
and Dr. Watson.'' It yielded the same results as before; only
less mercury was dissolved, so that the characteristic properties
of the nitre were more easily perceived.

It may be noticed here, that the great test on which Caven-
dish relied as a proof of the conversion of the soap-lees into
nitre, was the deflagration of paper dipped into a solution of the
latter. The presence, however, of nitrate of mercury interfered
with the characteristic combustion of the touch-paper, and ren-
dered the demonstration of the production of nitric acid less
satisfactory. It would have been better had mercury been dis-
pensed with in the experiments, and the spark passed over the
surface of the soap-lees, which could have been introduced into
a syphon or Yolta's eudiometer, with its wires so arranged as to
be a little above the liquid. Faraday has shown that, if a piece
of paper dipped in a solution of caustic potash be stretched a
little below and between two brass balls, from one of which
electric sparks are passing to the other, the alkali will be rapidly

262 CAVENDISH AS A CHEMIST.

converted into nitrate of potassj and the paper become touch-
paper. Van Marum, who employed in his repetition of the
nitric acid experiments the celebrated Teylerian electric machine,
failed^ notwithstanding — at least as he supposed — in obtaining
the same results as Cavendish. He ^^ used a glass tube, the
upper end of which was stopped by cork, through which an iron
wire was passed and secured by cement, and the lower end was
immersed into mercury ; so that the electric spark passed from
the iron wire to the soap-lees.'^ At the conclusion of the ex*,
periment, the alkali seemed unsaturated. Cavendish made two |

objections to the mode in which Van Marum and Trootswyk '

experimented : the first, that there was a risk ^^ of the iron wire j

being calcined by the electric spark, and absorbing the dephlo*
gisticated air ;'^ the second, ^^ that the only circumstance from j

which they concluded that the alkali was not saturated, was the
imperfect marks of deflagration that the paper dipped into it '

exhibited in burning, which, as we have seen, might proceed as
well from some of the mercury having been dissolved, as from \

the alkali not being saturated.^^ ^

Van Marum supposed that the difference between his results
and those of Cavendish arose from the latter having used oxygen
prepared from the impure black oxide of mercury, and reproached
him with having declined to explain how this oxide was pre-
pared. Cavendish, however, made no mystery of the process^
which simply consisted in agitating mercury mixed with lead,
as he informed Van Marum in a letter printed in his paper. In
the letter he also points out that he had likewise employed
oxygen from turbith mineral, and that there was no reason for ^

supposing that the source of the oxygen made any difference as
to the success of the experiment. Van Marum^ after all, how-
ever, only doubted whether the alkali could be entirely saturated* ^
As to the possibility of this, he and Trootswyk would not speak
positively : " Nous contentant pour le present d'avoir vu, que
Punion du principe d'air pur et de la mofette produit de Pacide
nitreux, suivant la decouverte de M. Cavendish.''*

* Von Mumm'f aeooont ii quoted by CaTendish in hU paper, and the antwer
whieh the Utter Bent to a letter from the former reqnetting information is given
in fall.

IMPORTAirCfi OF CAVENDISH'S DISCOVERY. 268

The failures in repeating the nitric acid experiment were
soon forgotten^ and the value of Cavendish's discovery was uni-
versally acknowledged. Dr. Black says of it : '^ This discovery
by Mr. Cavendish is one of the most important in the whole
science of chemistry .''* Black here referred chiefly to the new
light which it threw on the theory of chemistry ; and the unani-
mous opinions of later authorities on this point need not be
detailed. Neither Cavendish^ however, nor any of his contem-
poraries^ anticipated the importance which his discovery would
be found to possess in relation to natural phenomena. We now
perceive^ that not only must every lightning flash convert a
certain quantity of atmospheric air into nitric acid, but that this
compound, either in its free state, or as nitrate of ammonia,
must be brought to the earth by the rains that accompany or
follow thunder-storms. Much discussion has been carried on,
as to the importance of this atmospheric nitric acid in furnishing
plants with nitrogen, especially in tropical regions, where
thunder-storms are frequent. The subject cannot be considered
at length here : it is fully discussed in the last editions of Lie-
big's " Chemistry of Agriculture/* and of Johnston's ^^ Agricul-
tural Chemistry/'

Cavendish's discovery must also be considered as supplying
the explanation of the production of nitrates in the soil, which
was so long a perplexity to chemists, inasmuch as it demon-
strated that, although nitrogen appears, when carelessly exa-
mined, to be a quite neutral or indifferent substance, it is in
reality a combustible body. It was left to the ingenuity and
skill of Cavendish's successors, and especially to Liebig, to show
in what way exactly nitrogen undergoes combustion, during the
generation of nitre in a mixture of animal matter and alkali,
and to point out the great probability of ammonia derived from
the decaying animal matter being the compound of nitrogen
which is burned by the oxygen of the air during nitrification.
When Cavendish, in 1 781 and afterwards, burned a mixture of
hydrogen, oxygen, and nitrogen into water and nitric acid, he
established a truth which, more than half a century later, was

• Lectures, yol. ii., p. 108.

264 CAVENDISH AS A CHEMIST.

to famish the rationale of the process by which nitre is generated
wherever moist organic matter containing nitrogen^ along with
an alkaline, or other powerful base, and free oxygen, meet under
the influence of a suitable temperature. It would not be easy,
accordingly, at the present day, to over-estimate the importance
of Cavendish's discovery of a method by which common air may
be converted into nitric acid.

The paper of 1788 was the last of Cavendish's published
chemical researches. I now, accordingly, consider in detail his
claims as the discoverer of the composition of water. These
have already been referred to in the ** Life,** but only dogma-
tically. I here enter upon a critical inquiry into the relative
claims of all the alleged authors of that discovery. This cannot
be prosecuted without some repetition of what has been stated
already ; but as far as possible I shall avoid going over ground
already traversed. The statements in the Life are addressed to
general readers. What follows is intended for students of
science.

A CRITICAL INQUIRY INTO THE CLAIMS OF ALL
THE ALLEGED AUTHORS OF THE DISCOVERY
OF THE COMPOSITION OF WATER.

PRELIMINARY DISCUSSION.

The remarkable discovery which Cavendish announced in his first
" Experimentfl on Air/' that a mixture of hydrogen and oxygen can be
bnmed into its own weight of water^ had scarcely been made public in
January^ 1784^ before James Watt claimed the announcement as having
been previously made by him ; and Lavoisier declared that he had disco-
vered the compound nature of water before, and independently of either.
A controversy accordingly arose, in which Cavendish and Watt disputed
with each other the priority of the discovery, whilst they were at one in
disallowing Lavoisier's claim to take precedence of either. The dispute
was not settled during the lifetime of the claimants, at least so far as the
English rivals were concerned, and for a long period after their deaths the
controversy excited no public interest. In 1839, however, it revived,
and from that period to the present it has occasioned the keenest discussion
among some of the most eminent literary and scientific men of England,
Scotland, and France. I shall refer to it in future as the Water Con-
TROYBBST. It wiU take its place in the history of science side bv side
with the discussion between Newton and Leibnitz, concerning the inven-
tion of the Differential Calculus; and that between the friends of Adams
and Leverrier, in reference to the discovery of the planet Neptune.

For the sake of perspicuity I shall divide the following discussion into
several sections, and first consider those who have taken part in it.

1, Disputants in the Water Controversy.

The Water Controversy commenced in March, 1784, when De Luc
made known to Watt the contents of Cavendish's '* Experiments on Air,"
read to the Royal Society on the 15th of January of that year. Watt in
consequence transmitted to the Society his *' Thoughts on the constituent
parts of Water, Ac." which formed the subject of three communications
read to that body in April and May, 1784; and in the first of these his
claim to the disputed discovery was carried back to April 26th of the
preceding year. Lavoisier in the meanwhile laid before the French
Academy of Sciences his alleged discoveries on the same subject, sharing the
merit to some extent with his countrymen. La Place, Meusnier, and Monge,
but awarding no credit to Watt, and «carcely any to Cavendish. An
interval, however, of several months elapsed between the reading and
publication of the " Experiments on Air," during which Cavendish made
three additions to his paper, which have been already referred to, and
will be noticed again. In one of these, which alone concerns us at

266 CAVENDISH AS A CHEMIST.

present' he implicitly asserts a claim to priority over Watt, by stating
that many of his experiments had been performed in 1781, and commu-
nicated to Priestley, before the latter made certain researches whose latest
date is April 21st, 1783; and in the same paragraph he explicitly de-
clares, that Layoisier was made acquainted with his views in tae sammer
of the same year, before the French chemist had formed any opinion of his
own as to the composition of water. Here, so far as the principals were
concerned, the controversy ended. Cavendish and Watt contented them-
selves with putting on permanent record the evidence which seemed
necessary to justify their claims to the discovery in dispute, but neither
made any public attack on his rival, although each in private asserted
priority over the oth0r. No long time, moreover, elapsed before these
illustrious men were sufficiently reconciled to cultivate each other's
acquaintance. I have conversed with more than one scientific man, who
remembers to have seen them together at Sir Josephs Banks' conver-
saziones, and we have it on the authority of Mr. James Watt in reference
to his father, that ^' after becoming in 1785 a fellow of the Royal So<uety,
he formed the personal acquaihtance of Mr. Cavendish and lived npou
good terms with him."*

It was not likely in these circumstances that the dispute would be
revived by those whom it most concerned. Neither Cavendish nor Watt
accordingly appears to have directly or openly stirred in the matter again,
although the friends of each were careful to maintain his claims, when
opportunity for so doing offered. One such interference only caUs for
notice. In 1 786 Blagden, the intimate Mend of Cavendish, addressed a
letter to Crell, the editor of a well known German scientific journal, in
which he accuses Lavoisier of having misstated the nature and amonnt of
the information which Blagden communicated to him in reference to
Cavendish's experiments. This letter, which will be fully discussed
further on, avoided any consideration of the question of prionty between
the English philosophers, and is the last public allusion to the Water
Controversy which was made during their lifetime. Lavoisier made no
reply to the serious charges preferred against him, and the question had
ceased to be one of general interest long before the close of last century.

Cavendish died in 1810, Watt in 1819, and the next year witnessed
the decease of Sir Joseph Banks and Sir Charles Blagden, the last
survivors of those who had taken part in the controversy of 1784. From
1820 till 1839 nothing occurred to revive the dispute, nor did any one
during the interval systematically endeavour to a(^'ust the merits of the
several claimants, although various opportunities offered for attempting
this. Many references however, were made in scientific treatises to the
great discovery of the composition of water, but generally of a dogmatic
kind, and consisting chiefly of ascriptions of the merit of the discovery
entirely to one or other of the claimants, though sometimes they recom*
mended a division of the merit, in different ways among them. I do not
detail those imperfect notices of the Water Controversy for two reasons.
In the first place, the question under discussion is not one which ca/i be
settled by an appeal to authorities, or the balancing of great names, such
as those of Berzelius and Davy, Arago and Brewster, Dumas and Peacock,
against each other, according to the different views which each of them has

* Watt Corr. p. iv. Ante, p. 161, where Cayendish's imt, in 1785, to Watt at
Birmingham is noticed. To prevent confiuioni I shall hereafter refer to the celebrated
engineer hj his surname as Watt, and to his son, who died, recently, as Mr. Jamei
Watt.

THE WATER CONTROVBESY. 267

taken. The question is one of eyidence, wMch these and the other dis-
tinguished men who have taken part in the dispute, can assist in deciding
only hy their skill in the analysis and exposition of fskoUi, the value of
which may be appreciated by persons of ordinary intelligence. In the
second place, it is only since 1839, that the documents essential to the
settlement of the controversy have been even partially before the public,
and only since 1846, when the Watt Corretpandence was published in
foil, that the means of coming to a satis£Mtory conclusion have been in
the hands of all. It may be doubted, indeed, whether we are yet in a
condition to judge fairly of Lavoisier's claims, but we know enough to
assure us that they are of posterior date to those of Cavendish and Watt,
and the question I have chiefly to consider affects the rival claims of the
English philosophers.

I pass therefore to 1839, when the vexed question, who first discovered
that water is a compound of hydrogen and oxygen, was at length, and for
the. first time, deliberately discussed, and pressed to a settlement. Mr.
James Watt, who had been associated with his father in his studies and
was familiar with his views on the composition of water, never lost sight
of his claims as a discoverer in chemistry, and from time to time sought
to interest various distinguished philosophers in a matter which he had
naturally much at heart. He did not succeed, however, in inducing any
of his countrymen to come forward publicly to assert Watt's priority to
Cavendish as the discoverer of the true nature of water, but he ultimately
found in one of the most accomplished French philosophers an able and
willing advocate of his &ther*s claims.

Watt enjoyed the honour of being one of the Foreign Members of the
French Institute, and in 1833, Arago was called upon, in his capacity of
Perpetual Secretary of the Academy of Sciences, to write the ^loge of the
great engineer. Arago accordingly came to this country in the autumn
of 1834, to obtain materials for the projected memoir, and having satisfied
himself that Watt, and not Cavendish, was the discoverer of the compound
nature of water, he announced this in his ''Historical Eloge of James
Watt," which was read to the Academy in December, 1834, but not

Sublished till 1838. Lord Brougham, who had been requested by Mr.
ames Watt to pass judgment on the validity of the evidence adduced in
support of his father's claims, was present at the public meeting of the
Institute at which the iloge was read. On his return to England he
drew up a '' Historical Note on the discovery of the theory of we Com-
position of Water," to which Mr. James Watt appended some notes, and
the double document was published along with Arago s Eloge in the
Annuaire du Bureau de$ Longitudes, for 1839 ; as well as in the
Minwires de VAcadhnU dea Sciences, for thQ same yeaa

Cavendish had no near relative alive to watch over his interests, as
James Watt had watched over those of his father; but his memory was
too sacred in the eyes of his countrymen, to allow of his being long left
without zealous defenders. The Rev. William Vernon Harcourt was
President of the British Association for the Advancement of Science, for

1839, and conceiving that Arago had done great injustice to Cavendish,
devoted a considerable portion of his eloquent inaugural address to the
refutation of what he regarded as unjust and erroneous views.* To these
animadversions Arago replied at a meeting of the French Academy in

1840, and the distinguished chemist Dumas put on record his conviction

1" Report of Brit, AMSoe.for 1839. President's Address, p. 6 to 15.

268 CAVENDISH AS A CHEMIST*

tbat his colleague's history of the discovery of the composition of water
was complete in all respects.*

Sir Darid Brewster sought meanwhile to mediate between the con-
tending parties, but did not succeed in inducing the advocates of the
claims of Watt and Gftvendish to moderate the extreme views urged on
both sides, f The controversy in consequence proceeded without any
abatement of its onesidedness. When the report, accordingly, of th»
British Association for 1839 was published, nearly a year after Mr.
Harcourt*s address was delivered, he added a postscript, containing a
lengthened reply to Arago, Dumas, Brougham, and Brewster, along with
an appendix containing various documents bearing on Cavendish's claim,
as the only party entitled to the honour of being styled the discoverer of
the composition of water. In 1841, Berzelius published a conditional
judgment in favour of Watt's claims.}; In 1845, Lord Brougham animad-
verted on Mr. Harcourt's Postscript, questioning the validity of certain of
the proofs he adduced of Cavendish's priority.§ In the same year the
Dean of Ely (Dr. Peacock) reviewed his Lordship's work, assailing his
conclusions and asserting anew the claims of Cavendish.|| . In 1846,
Mr. Harcourt likewise replied at great length to Lord Brougham, and
reiterated his previous declarations. IT And the year after, Dr WheweU
reasserted his conviction that Cavendish was entitled to the honour of the
disputed discovery.**

The friends of Watt were not slow in reasserting his claims. A most
important addition was made to the literature of the Water Controversy
in 1846, by the publication of the ''Correspondence of the late James
Watt on his discovery of the Theory of the Composition of Water."
This was accompanied by a letter from his son, and an introduction by
the editor, Mr. Muirhead, a kinsman of Watt's, in which the right of their
illustrious relative to the entire merit of the discovery was urged in the
strongest terms. Finally, the Watt Corrapcmdenee was discussed ia
1847 by Sir David Brewster in the North British Review, and in 1848
bv Lord Jeffrey in the Edinburgh Review, both of whom prefer the
claims of Watt to those of Cavendish. Lord Jeffrey's paper likewise
contains an authorised statement from M. Dumas that he held unchanged
the opinion in favour of Watt, '' which he put upon record nearly seven
years ago.'*tt

The preceding list does not include the names of all who have taken
part in the Water Controversy, but defines sufficiently accurately those
by whom it has been chiefly conducted. Its discussion will be facilitated
if I add the exact titles of the works alluded to, with a brief notice of the
nature of each.

* Connie* Retidu9 de PAcadinUe de$ Seieneei, 20 Jan. 1840, pp. 109 to HI.

t Edinhurgh Review, Januaiy, 1840.

t Jahree Bericht, 1841. II Heft pp. 43—51. The advocates of Watt reier very
confidently to Benelina, as on their side. In reality, however, hit opinion ii yery
guarded, and, in 1843, he assigns scarcely any merit to Watt, and little to CaTendish,
between whom, however, and Lavoisier, who receives much the larger share, he divides
the honour of the disputed discovery. (Lehrbueh der Ckemie, 1843, pp. 370-*2.)

§ JAvee qfMen qf Letters, See Life of Watt, p. 400.

II Quarterly Reeiew^ 1845, p. 105.

t Lond, j- Edinr, Phil. Mag, Feb. 1846.

** Hiitorg qf Inductive Seieneet, 2nd edition, pp. 206— >207.

ft Edinburgh Review, Jan. 1848, p. 85.

THE WATER CONTROVEBSY. 269

2. Bibliography of the Water Controversy.

The FkHosophical TraruactioTU, and the M4moirt$ de VAcademU de$
Scieneety are the chief repositories of the early literature of the Water
Controversy. Mr. Muirnead has done a great service^ accordingly^ by
prioting^ in the Appendix to the Watt Correspondence, the chief papers
of Cavendish and Watt referring to the disputed discovery, and those of
liavoisier, Mensnier, and Monge, which are more conveniently consulted
ip that gentleman's accurate reprints than in their original issues. I
shall therefore give the page of Mr. Muirhead's volume, as well as that
of the Mimoires de VAcadimie, when citing the papers already referred to,

1775 to 1786.

£xpfirimenU and Observations on different kinds of Air: in six volumes.

By Joseph Priestley, LL,D., ic,

1700.

JBxpef^iments and Observations 07i different kinds of A ir, <Lc., being the
former six volumes abridged and methodized. By Joseph Priestley,
LL,D,, 4cc,

Priestley's experiments, and those of Warltire, which the former
records, confessedly led to the discovery under discussion. He was
appealed to, moreover, by Cavendish and Watt, as the umpire between
them, so that all his statements concerning the original investigations
into the composition of water, are of the utmost importance.

The later edition of the Experiments on Air, though professedly an
ahridgment of the original work, contains new and important matter, so
that it occasionally demands separate consultation. I shall refer to the
first as Priestley on Air, and to the second as Abridgment of Priestley
on Air,

1784.

ExpeAments on Air, by Henry Cavendish, Esq., F.R.S, its S.A. Phil.
Trans., Vol. Ixxiv., p. 119; or Appendix to Watt Correspondence,
p. 111.

Cavendish's other papers of which abstracts have been given, especially
tkose on Hvdrogen and the production of Nitric Acid, must also be
referred to, but it is unnecessary to repeat their titles here.

1784.

J%otbgIUs on the cofistituetU parts of Water and of Dephlogisticated Air;
mth an Account of some Experiments on that subject. In a letter from
Mr. James Watt, Engineer, to Mr. De Luc, Phil. Trans., Vol. Ixxiv.,
p. 829, or Appendix to Watt Corr., p. 77.

Sequel to the Thoughts on the constituent parts of Water and DeplUogis-
tioated Air, In a subsequent letter from Mr. James Watt, Engineer*, to
Mr, De Luc, F.B.S, Phil. Trans., Vol. Ixxiv., p. 334, or Appendix to
Watt Corr., p. 1 06.

1784.

M^moire ott Von prouve par la dico7nposition de Veau que ce Jluide nest
point une substance simple, et qiCU y a plusieurs moyens d^obtenir en
gf*and Voir inflammable qui y entre comme principe constituant. Par

270 CAVBNDISH Ad A CHEMIST.

MM. Metunier et LawMer. Memoires de TAcademie des Sciences
for 1781 (printed in 1784), p. 269, or Appendix -to Watt Coir.,
p. 151.

1784.

Mhncire daiu lequd on a pour objet de prouver que Veau ned point une
substance simple, un ^Ument proprement dii, mats quelle est susceptible
de decomposition et de recomposition. ParM, Lavoisier. Memoires de
FAcad^mie des Sdenoes pour 1781 (printed in 1784), p. 468; or
Appendix to Watt Corr., p. 171.

1786.

MSmoire sur le risuUaJt de VinJUimmaAon du gaz injtammable et de Pair
dSphlogistiqui, dans des vaisseaux dos. Far M, Monge. Memoires de
TAcad^mie des Sciences pour 1783 (printed in 1786); or Appendix to
Watt Corr., p. 205.

1786.

Letter from Dr. Blagden, Sec. E. S,, to Br. Lorem Crdl. Chemische
Annalen, &c., von Dr. Lorenz Grell, p. 58; or translation by Mr.
Muirhead in Watt Corr., p. 71.

1889.

Eloge Historique de James Watt; par M. Arago, SecrStaire PerpHud de
VAead&mie des Sciences. Paris. Annuaire da Bureau des Longi-
tudes; or, Memoires de rAcad^mie des Sciences, pour Fan 1839.

Two English translations of this work have appeared, namely: —

1. Historical Eloge of i amjsa Watt; by M. Arago, isc. Translated from

the French, with additional Notes and an Appendix, by James Patrick
Muirhead, Esq., M.A., ofBcUliol College, Oscford, Advocate. London,
1839.

2, Life of James Watty by M. Arago, Ac; reprinted from, the Edinburgh

New Philosophical Journal for October^ 1839. Edinburgh, 1839.

Mr. Muirhead's work is the more complete of the two, but the " Life**
is probably more generally accessible. I shall cite the former, however,
as the portion of it relating to the History of the discovery of the Com-
position of Water is reprinted in the Appendix to the Watt Correspond-
ence, where it can be conveniently consulted, side by side with other
important documents bearing on the Water Colitroversy. Both the
translations, as well as the original Eloge, contain Lord Brougham's
?JiP«>fte'°t Historical Note with Mr. James Watt's annotations; and Mr.
Muirhead has added to his work some useful comments in his own
name, as well as a reply to Mr. Harcourt's first discussion of the claims
of Watt and Cavendish, which was made public in the interval between
Uie appearance of the « Eloge Historique," and the publication of the
*.ngH8h transb^tion, so that in order of time, Mr. Harcourt's address
comes before the latter.

♦l.«^^^^'l ^^f^l is unquestionably the most important publication which
coLjT'Tt 1, *^^ ^*^' Controversy has called forth; and when it is
temi^f^?n wwi. T^""^ **^^ *^^« of a discussion is determined by the
the SfeLJiJ^^of ♦i 'S,^"*^"*®^^^* >* ®a»»not b^t ^ deeply lamented that
in so onesS a ^L!"^"*^? Academy should have opened the controversy
so onesided a spirit a« he showedf, and should have urged the claims of

THE WATER CONTROVERSY. 271

Watt rather as an advocate than aa a judge. The charges whichhe brought
by implioation against the good faith of Cavendish, whom his countrymen
universally regaraed as a man of spotless integrity, could not but provoke
a retaliation of harsh dealing towards Watt. The results have been
equally detrimental to the interests of both claimants, on each of whom
judgments have been passed so partial, that if we credited them we
should be compelled to acknowledge that Cavendish and Watt, instead of
being among the most remarkable men of their af e for genius and virtue,

I were wanting alike in generosity and intellectual capacity. Had Arago

been more just to Cavendish, his opponents would have been more just to
Watti and the claims of both would have been more speedilv adjusted.

It is necessary, however painful be the task, to point this out at jpre-
sent, as the dogmatic and partisan spirit which so painfully charactenses
the greater part of the literature of the Water Controversy, must be largely
referred to the extreme views which were urged by Arago in his Eloge
of Watt

^ . 1840.

Heport of the Ninth Meeting of the British Aseoeiation for the Advancement
of Science, held <U Birmingham in August, 1839. Address hy the
Rev, W. Vernon Harcourt, President for the Tear.

A portion only of Mr. Haroourt's eloquent address refers to the Water
Controven^. It was published in the Athenceum and other journals at
the time of its delivery, and was in consequence noticed by Arago and

^ Dnmas before its offickl publication in the report cited above.

Their comments are contained in the Comptes rendus Hehdomadaires
des Seances de VAcad&mie des Sciences, 20 Janvier, 1840, p. 109, of which
a reprint is given in the Appendix to the Watt Correspondence, p. 260.

Sir David Brewster also noticed Mr. Harcourt*s address in an article
on Arago*s Eloge of Watt, in the Edinburgh Review, No. 142, 1840.

In conseouenoe of those critiGisms, Mr. Harconrt added a lengthened
postscript, which follows the address in the authoritative report of it, and
is followed in its turn by an appendix oontuning extracts from nnpub-

jk llshed papers by Cavendish, and a lithographic fiu^imile of the ^^ records
in his Note-book of all his experiments relating directly to the composi*
lion of water." The last is a most important document. Mr. Harcourt's •
entire paper is the most valuable contribution to the literature of the

* Water Controversy, which has appeared on the Cavendish side, and is

not less learned and forcible, than eloquent. It is greatly to be regretted^
however, that its accomplished author should have formed so low an esti-
mate of Watt as a chemist. I cannot believe that impartial critics^

\ putting the Water Controversy aside, will ratify Mr. Harcourt's judgment
on this point. I owe it at least to my own convictions to say, tnat a
careful study of Watt's chemical papers, lias satisfied me that the sagacity,
originality, and perspicuity of oonception, which he displayed in his
purely physical inquiries, did not forsake him when he entered on
chenucal research. The denial of this by Mr. Harcourt has led to a
corresponding depreciation of Cavendish's merits, and is one of the causes

I why the controversy has been so greatly prolonged, and has exhibited so
much of a partisan character throughout.

1840.

y A Few Notes on the History of the Discovery of the Co^nposition of Water.

By J. 0. EaUiweU, Esq., F.B.S.

This is a pamphlet of three* pages, ** intended as a supplement to a

272 CAVENDISH AS A CHEMIST*

paper on the same subject, by Lord Broagham." It is important chiefly
as containing a part of Watt's letter to Priestley, of 26th April, 1783,
which is not printed in the former's '' Thoughts on the Constitoent Parts
of Water, &c. &c."

1845.

Lives of Men of LetUrs and Science, who Jlourished in the time of Geo. III.

By Henry y Lord Brougham,

In connection with the lives of Blacky Watt, Cavendish, and Priestley,
which are contained in this volume, Lord Brougham has republished his
''Historical Note on the Discovery of the Composition of Water ;" and has
added an appendix containing a brief reply to Mr. Harcourt's addiess
and postscript.

His lordship prefers the claim of Watt, but urges it temperately, and
disclaims all doubt as to Cavendish's good faith in the transactions connected
with the early researches into the composition of water. His work is
remarkable also for the impartiality and, as I believe, justice, with which
it insists on Lavoisier, as well as Cavendish and Watt, being recognised
as an important contributor to the discovery of the true nature of Water.
The Historical Note is more accurate in its statements than Amgo's
Eloge, and we are indebted to Lord Brougham for the explanation of
certain apparent anachronisms in Cavendish's ^'Experiments on Air,"
(1784), which, till his lordship consulted the manuscnpt of this paper in
the R. S. Archives, were inexplicable. The least satisfactory part of
his statement is the reference to Blagden's share in the transactions which
led to the original controversy.

1845.

Article in Quarterly Beview, December, 1845, entitled *^ Arago/k Btvugharn

on Black, Cavendish, Priestley, and Watt"

This paper is interesting as the production of the learned and accom-
plished Dean of Ely, the Rev. Professor Peacock. {Edinr, Bev. Jan.
1848, p. 83.) It is unnecessary, however, to refer to it at length, as it
avowedly takes up the same ground as Mr. Harcourt occupies in his
address and postscript, and is obnoxious to the charges preferred against
these productions, of being too unfavourable to Watt. Dr. Peacock's
views on the relation of heat to chemical combination, which are strongly
expressed in this article, differ from those entertained by the majority of
chemists at the present day.

1846.

Letter to Henry, Lord Brougham, F.B,S. Containing Bemarks on certain
Statements in his Lives of Black, Walt, and Cavendish. By the Bev,
W. V. Harcourt. Lend. Ed. & Dub. PhiL Mag. 1846.

This is a very learned and interesting paper, which will be welcome to
all students of the history of science for the information which it sup-
plies in reference to many important chemical discoveries. It is chieny
occupied, however, not with the Water Controversy, (on which Mr. Har-
court contents himself with reiterating his former opinions), but with
criticisms of many of the statements made by Lord Brougham, in his lives
of the chemists named above. It is foreign to my present purpose to
refer to those, except to say that the Water Controversy has indirectly
led in this, and other productions, to much discussion concerning the
respective merits of the founders of pneumatic chemistry, which has been
of essential service to the cause of science.

THE WATER CONTROVERSY. 273

1840.

Corrupondenee ofiJie late James Wait on his discovery of the theory of the
GompoHtion of Water, with a Letter from his Son, Edited, with
Introductory Remarks and an Appendix, by James Patrick Muirhead,
Ssq.f F»R,S*E,

I have already referred to the valuable nature of the reprints and
oilier contents of the Appendix to this work, and cannot speak in too
high terms of the painstaking and conscientious way in which Mr. Muir-
h^kd has edited the whole volume. The infirmities of advanced life pre-
vented Mr. James Watt from superintending the publication of his father^s
correspondence, but he has contributed an introductory letter of much
interest. The correspondence extends from the close of 1782 to 1786 ;
the more important letters belonging to 1783 and 1784. The extracts
are partly taken from letters written by Watt to Dr. Priestley, Dr.
Joseph Black, J. De Luc, Mr. Gilbert Hamilton, Mr. Smeaton, Mr. Fry,
Mr. Kirwan, Sir Joseph Banks, and Dr. Blagden, partly from certain of
those gentlemen's letters to Watt, chiefly replies. The latter are printed
from the autograph originals which, at the period of their publication,
were in the possession of Mr. James Watt j the former from copies taken
by Watt himself by means of his ingenious copying machine.

It would have prevented much discussion, if this correspondence had
been published in 1839 along with Arago's Eloge ; but no blame can be
attached to Watt's relatives for the delay which has attended its appear-
ance. For some six years, however, it was accessible only to the advo-
cates of Watt, who tantalised the public with partial quotations from it,
whilst the defenders of Cavendish had no means of effectually replying to
the masked battery brought against them, and were even taunted with
their ignorance of unpublished facts which they could not possibly know.

The publication of the correspondence has done as much service to the
canse of Cavendish as to that of Watt, and has greatly enlarged our
means of bringing the Water Controversy to a satisfactory conclusion.

The correspondence is prefaced by a series of lengthened introduc-
tory remarks from Mr. Muirhead, of which I wish I could write in as
favourable terms as I have done of his labours as an editor. That he has
overpraised Watt, who is so deserving of praise, is a fault for which no
ono will greatly blame him, seeing that he is a kinsman of the great engi-
neer's, and wrote in the name of his son. But there was no one from whom
depreciation of Cavendish's merit could come with a worse grace, than
from a relative of his rival. Mr. Muirhead, moreover, disparages not
only Cavendish, but nearly all who have advocated his claims, and is sel-
dom content with the evidence adduced by the defenders of Watt, which
he qnalifies or extenuates till it accords with his own view of the deserts
of his ffreat client. The perusal, accordingly, of his introductory remarks
is a painful task, which is not lightened by the discovery that he has ex-
pended so much of his zeal in attacking Cavendish and his friends, that
he has not perceived what was essential to the defence of Watt, so that a
considerable part of Lord Jeffrey's discussion of the Water Controversy is
devoted to supplementing Mr. Muirhead's argument. It is only justice,
however, to the latter, to acknowledge that liis work everywhere exhibits
proofs of an earnestness and sincerity in his estimate of Watt's merits,
which deserve all praise.

274 CitVENDISH AS A CHBMIST.

1847.
Dr. WheweWi History qfthe Inductive Sciences, ^nd edition^

In this well known work Dr. WHewell reiterates on grounds similar
to those urged by Mr. Harcourt and Dr. Peacock^ the opinion he had
before expressed m favour of Cayendish,

1847,

I^arth British Review, article entitled, Watt and Cavendish, Oon ir ^ve r^

respecting the Composition of Wa$er,

This vigorous and eloquent essay is from the pen of Sir David
Brewster, as appears from the Edinburgh Review tot January^ 1848,
The claims of Watt are much more strongly asserted than in Sir David*9
paper on the same subject, published in 1840, and a very unfavourable
view is taken of the good faith of Blagden, as well as of the generosity of
Cavendish, and the fair dealing of the office-bearers of the Royal Society
in 1783 and 1784.

1848.

Edinburgh Review, article entitled, The Discoverer of the Ccmposition of

Water; Wait or Cavendish f

This contribution to the literature of the Water Controversy, from the
pen of Lord Jeffrey, is without question the ablest analysis of the dis- ^

cussion which has been given to the public. It may be read with equal ^

pleasure by the friends of Cavendish and of Watt, for it treats both
throughout as men of great intellectual power and blameless integrity.
His Lordship prefers the claims of Watt, and urges them with great
earnestness, out he does not conceal what has been so unwisely denied by
previous advocates of the claims of both candidates, that there are diffi-
culties in the way of an exact settlement of merit, and room therefore for
difference of opinion between equally impartial critics. It is a remark^
able circumstance, that ten years should have elapsed before the olaima
of Watt as a discoverer of the true nature of water, were discussed ia 4

such a way as to enable even his well-wishers to make out a logically
consistent case for him. Down to 1848, one-half of the auction m dis-
pute, namely, Did Watt hold that hydro&^en was one of the elements of ^
water? was either assumed or assertea on grounds that involved % i
petitio principii. Lord Jeffrey, for the first time, meets the difficulMr,
and at least offers the proof which had been so long demanded, Hia
paper accordingly contains the ablest defence of Watt's claims which has
appeared, and as such will be frequently referred to in the following
pages.

1848.

A Hiatory of the Royal Society, Sc, S^c, By Charles Richard Weld, Esq.

In this interesting work, Mr. Weld has given a brief account of the
Water Controversy (vol. ii., p. 170), and has contributed from the manu-
scripts in the arcnives of the Society, several references, especially fac- <
similes of one of the interpolated passages in the MS, of the ' Experiments
on Air* (1784), and of Cavendish's and Blagden's handwritings which
add to our means of deciding the questions in dispute. I

THE WATEB CONTROVERSY. 275

In conolnding this bibliograpbical notice, I would refer to two
natters which can be best disposed of here. Mr. Harcourt has been cen-
sored for bringing the que/9tion of the rival olaims of Watt and Cavendish
before the British Association^ in his inaugural address, without apprising
M. Arago of his intention to call in question his statements ; to which
moreover Ara^o, even if present, could not have been permitted to reply.
I cannot see the justice of the accusation. M. Arago might as fairly be
blamed for not informing Cavendish's relatives that he intended to bring
certain charges against him before the Academy of Sciences, where no
answer could be given to the Secretary's statement. It has always been
the practice of me President of the British Association to review the
important scientific events of the past year, and Mr. Harcourt did not
exceed the prerogative of his office in re^uxling the revival of the Water
Controversy as one of them. Neither party in truth deserves any cen-
sure. The Secretary of the French Academy would have done a great
wrong if he had concealed his conviction that Cavendish had acted un-
justly to Watt ; and the President of the British Association did not err
in claiming a right to vindicate in the most public way the honour of the
aoonsed. It is another question, which of the officials was in the right.
That, however, does not concern us at present, and I shall be glad if I
can persuade the reader, before entering on its discussion, to dismiss from
his mind any doubt he may have entertained as to the fairness of the
interference of Arago and Harcourt in the Water Controversy.

The other remark I have to make refers to a more important matter.
It is an interesting feature in the controversy under discussion, that, on
both sides, manuscripts unpublished during the lifetime of their authors
have been adduced in support of the claims of Cavendish and Watt. On
the one side we have Cavendish's laboratory Note-book, on the other the
greater part of the Watt Corretpondence, This fact will have much
weight in settling the disputed point, how far manuscripts are admissible
as evidence in deciding questions of priority. Arago and Sir David
Brewster have objected to the validity of MSS. in reference to Cavendish's
claims. It is curious that such an objection should have come from the
Watt side, for his case would suffer more than that of Cavendish, if manu-
script evidence were refused on both sides. There seems no valid reason,
however, why it should not be accepted in favour of all the candidates.
The genuineness, authenticity, and integrity of the Cavendish and Watt
MSS., which have been adduced in evidence during the controversy, are
unimpeachable. In other words, they certainly are the productions of the
parties to whom they are attributed ; they were written at the periods
indicated by the dates affixed to them ; and they have not been altered by
their originial writers, or by others, or in any way tampered with, since.
Documents of such a character cannot be passed over ajs inadmissible on
the plea of their informality as private papers of the claimants. On the
other hand, the fact that they were not intended for publication, greatly
adds to their value. I shall accordingly consider the Watt Letters and
the Cavendish Note-book as unquestionable authorities, and quote from
them unreservedly.

In connexion with this point, it seems desirable also to notice that
as the law of patents has been held by Mr. Muirhead to determine the
way in which the Water Controversy should be decided ; and as other
statutes have been referred to as of authority in the matter, it may savo
any lengthened defence of the way in which the case is argued in the
foUowing pagesy if I adduce here the striking and authoritative statement

276 CAVENDISH AS A CHEMIST.

of Lord Jeffrey, whose decision is of peculiar value in a case like tke
present, where the one party demands the application of the formal mlei
of courts of justice, and the other refuses to he bound by them.

*' We can by no means adopt/' says his Lordship, '' those narrow and
jealous canons of evidence, derived from the rigid maxims of law, or tbe
precedents in cases of patent, by which both M. Arago and Sir D.
Brewster seem anxious to limit the inquiry. Courts of law must pro-
ceed upon inflexible rules, and can make no distinction of persona; and
are forced therefore peremptorily to reject all evidence proceeding from
the parties concerned, or from those having interest in the issue ; though
it is certain that by so doing they must occasionally decide against tho
truth and the conviction of all unprofessional observers. The question in
a court of law, in short, is never really what tbe truth of a case is, accord-
ing to the actual and conscientious belief of the judges Tor jury), after oon-
sidering every atom of producible evidence that is in existence, but merely
what the import is of the evidence that is legally admissible. But in a case
like the present, where the only judge and jury is tbe intelligent pub-
lic, and where there is neither any motive for excluding any proof which
can at all affect the ultimate conviction of that multitudinous tribunal,
nor any power by which the parties and their advocates can be restrained
or limited in the production of it, it would evidently be mere affectation
and absurdity in those who may assume the office of summing up for their
assistance, to leave out of view anything that is so produced or producible ;
or to pretend to shut their eyes upon those parts of the evidence which,
though strictly inadmissible in a court of law, they yet know to be pro-
ducing, and to be entitled to produce, the greatest possible effect upon
the honest and intelligent minds from which the decision must proceed.
In all questions before the public, in short, no evidence is incuimissible, nor
can there ever be any question, except as to tJlie credit to which it is
entitled ; while the court is always open for as many appeals and new
trials as the parties may choose to venture on.

" The analogy of the law of patents is still more palpably inapplicable.
A patentee gets his monopoly as a reward for being tne first to make the
public participant of a useful discovery ; and therefore it is most just that
he should not be excluded from this reward by any rival claimant, who only
offers to prove that he had previously made the same discovery, but admits
that he had never disclosed it. But where the competition is merely for
the intellectual glory of the discovery itself, it is plain that no such prin-
ciple can apply, and that the palm of priority in the invention may be
justly awarded to one who has been forestalled in the publication;
whilst that of original and independent invention may be shared between
both."*

In this decisive opinion, moat persons I think will concuri and the
remark may once for all be made, that it is vain to attempt to lay down
stringent rules for the adjustment of disputes as to priority between dis*
coverers in science. We cannot settle for others, nor can they for them-
selves, the way in which discoveries shall be effected, so that our canons
shall at once suffice for disposing of disputes. All that we can do, is to
enforce certain rules which may prevent collisions occurring; but if they
do occur, our prospective rules will seldom afford the means of adjusting
the claims of rival competitors, and we must dispose of each case upon its
own merits. The scientific societies of Europe have recently adopted

* Edinhurgh Revisw^ Jan, 1S48, p. 87.

THE WATER CONTROVERSY. 277

dxpedieni after expedient to prevent disputes^ but with a success so
paFtial, that the discorery of the Planet Neptune, and of the ansBsthetic
properties of sulphuric ether, have already led to much keener contro*
voTsies tlian that which originally attended the discovery of the compo-
•ition of water; and these, I need not say, are not the only contested
discoveries of the present day. When we have learned all the laws in
obedience to which Genius unconsciously works, we shall be able to bind
it in fetters, which it will not refuse; but for the present, at least, we
must regard each great discovery as teaching some of these laws, not as
supplying a case wliich is to be decided by them. New truths will often
oome to light, as new stars appear in the heavens, when no one is expect-
ing them, and we shall search in vain in the records of the past for any
precedent by which to dispose of the difficulties that attend their recog-
nition. The discovery of the compound nature of water was such a
tmth, and will be studied to most advantage, if all prepossessions as to
how it should have been made, or should have been announced, are
thrown aside.

3. Questions in dispute between the Princtpals in the

Water Controversy.

It cannot, I think, but strike every dispassionate reader, who has
glanced at the literature of the Water Controversy, as something not a
little singular, that a matter, to appearance, so simple as the question
who first made a single chemical discovery, should have been found so
difficult of decision. The announcement of the discovery in the original
papers of Cavendish and Watt does not occupy many hues, and both of
these able reasoners were very clear and perspicuous writers. Their
principal commentators, too, have been men universally admired for their
genius and talents, and referred to as famous alike in literature and science.
How then has it happened, that a question which apparently any twelve
moderately intelligent men were competent to decide, should have been
discussed with so indecisive result, so far as unanimity is concerned, by the
special jury who have had it before them 1 AViiatevcr'be the explanation
of the conflicting judgments which have been passed on the subject beforo
US, I think it is impossible to avoid the conclusion, that a problem which ten
years' discussion has not solved, cannot be one very easy of solution. The
prolongation of the Controversy has, no doubt, been owing in part to the
absence, at its commencement, of all the data requisite for its conclu-
sion, and partly also to the polemical spirit which nas been displayed on
both sides, but chiefly to the real, but generally unconfessed and perhaps
nnperceived difficulties which are inseparable from the whole question.
The love of justice and fair dealing which induced so many distinguished
men to come forward spontaneously to defend the rights of the several
claimants was too deep-seated to have permitted them to conceal their
eonvictions, had the evidence adduced shown them that their views were
erroneous. Sir David Brewster's fi-ank acknowledgment of the change in
view which the publication of the Wat Correspondence induced in him,
was worthy of so distinguished a philosopher, and would assuredly have
been followed by others, had their views undergone a similar change. As
it is, MM. Arago and Dumas* unaltered confidence in the opinions they
published in favonr of Watt in 1839, and Harcourt, Whewell, and Pea-
cock's equally unchanged advocacy of Cavendish, show that the evidence

278 CAVENDISH AS A CHEMIST.

to which both parties make their common appeal, instead of compiling
a unanimous judgment, is held to justify two exactly opposite conclusions.

In truth, it must be conceded by all dispassionate inquirers, that it is
impossible to base the claims of either Cavendish or Watt on eyidenoe
which is at once direct and unexceptionable, in reference to eyeiy con-
tested point. Certain links in the chain of proof on both sides haye been
beyond cavil from the first; but others can bo supplied only conjectantlly
from indirect and sometimes vague evidence, to which critics, equally
candid and impartial, wiU attach very difierent values, and no two will
perhaps attach the same. Thus, it is now acknowledged, that Cavendish's
experiments on the production of water from its elements were of earlier
date than the observations from which Watt drew his conclusions. Bot
the period ai which Cavendish inferred that water is a compound of
hydrogen and oxygen, is not recorded in any existing document, more
precisely than as lying between the summer of 1781 and that of 1783, and
we must be guided in our choice of one or other of three years by con-
siderations which will weigh differently with different persons. So also the
date at which Watt drew his conclusion concerning the composition of
water is quite certain, but what that conclusion was is by no means
certain ; for it is not precisely stated in any paper hitherto made public;
and its import must be gathered from the comparison of many passages
in his writings, and those of Priestley, which admit of more than one
interpretation.

Where such is the state of matters it would be unwise in any author
to seek to compel his readers to concur in all his conclusions. I wish on
the other hand to urge on all interested in the Water Controversy, that
unanimity of opinion in reference to many of the disputed points, cannot
be expected and should not be demanded. I shall be content, therefore,
to state my own opinion, without insisting at great length on its justness;
and shall devote myself chiefly to as impartial and unpolemical a state-
ment as the circumstances of the case will permit, of the grounds on
which Cavendish, Watt, and Lavoisier, claimed the discovery of the com*
position of water.

These claims, as they were originally urged, were not founded upon
an analysis of water into its elements, which was first effected at a later
period by Lavoisier; but rested on the discovery that two gases could be
made to combine and produce water. The common claim, however, of
the three rivals, if reduced to its simplest elements, involved two points.
Each affirmed (1) that he had discovered for himself the composition of
water; and further (2), that he had made the discovery before the others
did. The question of priority of discovery is the one which has chiefly been
discussed, especially by the advocates on the Watt side; but the question of
reality of discovery is of as much importance, and must be considered first.
There are in addition certain accusations of plagiarism against Cavendish
and Lavoisier; but these are best reserved till the claims asserted by each
for himself have been disposed of. It must further be noticed, that
although the original claims were publicly founded on similar experiments,
Cavendish, Watt, and Lavoisier arrived at their conclusions whilst pur*
suing very different trains of research. Cavendish was investigating the
products of combustion; Watt was speculating on the changes which a
vapour would undergo, if all its latent heat became sensible; and Lavoisier
was seeking in the combustion of inflammable gases for additional proofs
of the truth of his view, that oxygen is the great acidifying agent. Those
researches in their turn sprang out of earlier investigations^ and the whole

THB WATER CONTROVERSY. 279

iiltimi^tely oonrerged to one line of inquiry, wblch led to the discorery
nndor disonssion. A brief reference, accordingly, to these preliminary
ohflervations and to some other9> will make the whole argument more
— ii- followed.

4. Researches which led to the discovery of the Composition of

Water,

From the earliest dawn of scientific speculation, through ages of intel-
lectnal liffht and darkness, down to the days of the first French Revolution,
the simple, uncompounded, or elementary nature of water had been re-
garded as an unquestionable fact. Physical philosophy had for centuries
bosied herself in dictating to Nature a simplicity which she disowned.
Air, earth, fire, and water, were the elements of all things, and the dogma
had been repeated so frequently, that its very echo was mistaken for a
new utterance and confirmation of its justness, and no one doubted its
truth. Yet if it ever were a wise thing to spend much time in wondering
that a discovery was not made more speedily than it was made, we might
wonder here. The phenomena of vegetation, if they had been watched
with attention, would have shown what a questionable doctrine that of
the elementary character of water was. Nor did it escape the sagacity of
Van Helmont and other observers, that the development of a tree trans-
ferred from earth to water, implied the derivation of even the most solid
constituents of plants, from the so-called indivisible liquid. But this
inference lost all its value by being extended too for, and ended in the
implicit assertion, that all the elements of vegetables are contained in
water. Such a conclusion did nothing to further the progress of science,
but rather retarded it, for it was founded on inaccurate experiments,
which left totally out of consideration the infiuence of the atmosphere,
ftnd of the substances dissolved in water, on the growth of plants. Nor
did the observation of the functions of animal life — ^to the maintenance
of which endless decompositions and recompositions of water are so
essential — prove more instructive to the earlier chemists than the study
of the simpler vegetable life had done. Not a weed grew but was in the
secret of the composite nature of water ! The smallest animalcule knew
how to decompose the so-called element, and daily divided the indivisible
that it might change it into its own substance ! No one, however, under-
stood their language, or tried to interpret it, and hieroglyphics which
seem to us pictures which tell their own story, revealed nothing to those
who had already decided that they had no meaning.

The mere observation of natural phenomena thus taught nothing,
and as little did direct experiment. From the earliest times speculative
and practical chemists, whether engaged in economical processes such as
ohiefly occupied the Greeks and Romans, and the ancient Egyptians, or
in mystical investigations like the Arabian and mediseval Alchemists,
mnst have formed and decomposed water hundreds of times, but they
were not aware of the feat they had performed. Nor can we wonder at
their blindness, for long after their later successors were warned that
water was probably not simple — and after they had its elements in their
hands for years-^they failed to detect the significance of phenomena,
which we are apt to conceive carry their interpretation with them.
Water accordingly was reputed for some thousands of years a simple
sabftaace.

280 CAVENDISH AS A CIIKMIST.

We may begin the modern history of its decomposition^ with Newton's
celebrated inference, from its optical characters, that water consisted of
ingredients which were unlike each other, and one of them (or one claas of
them) inflammable. This conclusion is often referred to, as if it had been
greatly more precise than it certainly was, and popular authors write aa
if Newton had predicted, in so many words, that water would be found
to consist of two gases, one of them inflammable. His own words, how-
ever, certainly do not warrant any such inference. In the course of his
observ'ations on the refractive indices of various bodies, he noticed that
whilst transparent, uninflammable substances refracted light more power*
fully the denser they were, there was an exception in favour of combus-
tibles, such as camphor, the oils, turpentine, &c., whose refractive indices
were much higher than their density would account for. " Water," he
goes on to observe, '* has a refractive power in a middle degree between
those two sorts of substances, which consist as well of sulphureous, fat,
and inflammable parts, bb of earthy, lean, and alcalizate ones."*

It is impossible to gather from a statement so general as this, what
Newton's precise opinion was as to the nature of water, and though we
may look back at it as a prediction that one of the constituents of that
liquid would prove to be inflammable, it may be doubted whether Newton
intended to affirm this; and it seems quite certain that his contempo-
rarie?, and immediate successors, did not put that interpretation on his
words. His prediction, in truth, was not called to mind till long after
the detection of hydrogen as a constituent of water had amply fulfilled
it. It went for nothing, accordingly, in leading to the discovery of the
composition of water.

Newton*s observation was made in the beginning of last ccntary.
Some fifty years passed on, and water was still an element. At the
end of this period, men had become familiar with a powerful new
engine for eflecting chemical decomposition in the Friction Electric
Machine and the Ley den Jar. Beccaria, taking advantage of this, exposed
water to the electric spark, but though he resolved the reputed element
into its constituent gases, he was not aware of what he had done.f

Some ten years later, viz. in 1766, Cavendish gave the first detailed
account of the properties of hydrogen, but though, in the course of his
experiments, he must have burned that gas into water many times, he
does not, as already noticed, once mention that he so much as saw a liquid
produced.

Ten years more passed away, and at length the threshold of the dis-
covery was reached. In 1776 John Warltire, an English natural philo-
sopher, observed that when a jet of hydrogen is allowed to bum under a
bell jar, closed below and containing air, till the flame goes out, " imme-
diately after the flame is extinguished there appeai-s through almost the
whole of the receiver a fine powdery substance luce a whitish clovd, and the
air in the glass is left perfectly noxious.'*t In the same year Macquer, a

* Optic», 1704. Book Second, p. 75.

i* Beccaria's observations are contained in his Letiere dell* EleUricunto, published
at Bologna in 1758. Beccaria's experiments, which must not be confounded with those
made nearly half a century later with the voltaic battery, were repeated in England by
Pearson {Phil. Trans. 1797, p. 142), and in Holland by Trootswyk, after the compound
nature of water had been discovered. In this century they have again been carefully
repeated oy WoUaston, Faraday (Electr, Rea. series 3, par. 328), and Grove {PhiL
Trana. 1846). See also Chem, Soe. Mem, toI. iii. p. 340.

% Prietiley on Air, vol. iii. 1777, App. p. 367. Warltire calls the gas he used
"inflammable air,*' but from the description he gives of the mode in which be prepared

THE WATER CONTROVERSY. 281

French ohemist, without any knowledge of Warltire^s observations, and
anxious only to ascertain whether the flame of hydrogen evolved smoke
or soot, made an experiment similar to his, but to better purpose, and
saw the uncertain " whitish cloud" which the English observer mistook for
a fine powder, condeuso into visible drops of a clear liquid like water,
which accordingly he assumed it to be.*

Macquer prosecuted the matter no further, nor did he draw any con-
clusion as to the origin of the water he saw deposited ; and we must pass
on to 1781 before we find anything additional worthy of notice. In that
year, Dr. Priestley made, what he was fond of making, ^' a random experi-
ment," as, with characteristic candour, he calls it. It proved in the end,
like the chance shot of an uncertain marksman, of more real service to the
progress of science, than many of its performer's carefully aimed experi-
ments have done.

The random experiment consisted in exploding a mixture of inflammable
air (apparently hydrogen^ and common air, in a close glass vessel by means
of the electric spark, in tne way first practised by Yolta in 1776.t When
the spark had passed and the explosion was oyer, the sides of the glass
were observed to be bedewed with moisture, but to this latter phenomenon
Priestley paid no attention. Warltire, however, who, as I have mentioned
already, had had his attention previously directed to the appearance of a
powdery deposit, or '^whitish cloud," as attending the combustion of
inflammable and common air, was now struck by the appearance of moisture,
and said to Priestley, whose experiments he witnessed and repeated, that
it confirmed an opinion he had long entertained that common air deposited
moisture when phlogisticated.;]: Similar experiments were made with
oxygen and inflammable air in glass vessels, and the appearance of moisture
was probably noticed, although this is not mentioned. The only remark
Priestley makes, is, that '^ little is to be expected from the firing of inflam-
mable air, in comparison with the effects of gunpowder."

Warltire, however, though he drew the conclusion mentioned above,
paid little attention to the appearance of moisture. The experiments, which
he and Priestley performed, interested him chiefly, because he " had long
entertained an opinion, that it might ba determined whether heat is heavy,
crnot, by firing inflammable air mixed with common air, and applying them
to a nice balance." To avoid the risk of injury from the explosion, he

it, it most haro been h3'clrogen. His account of the experiment, which he considered
"very curioos," is dated Jan. 3, 1777. It has been overlooked by the historians of tbo
Water Controversy.

* Dictionnaire de Cht/tnie, t. ii. p. 314, quoted in Wait. Corr. p. zxviii.

f Arago ascribes to Warltire the merit of first pas)>in^ the electric spark through
gaseous mixtures confined in glass vessels. (Annuaire du Bureau dea Longitudea pour
1 839, or Watt Corr. p. 225.) Warltire, however, states that he borrowed the device
from Priestley {Prieailey wi Air, vol. v. 1781. Appendix, p. 395); and the latter, in
his turn, refers to Yolta as having kindled inflammable air by the electric spark before
he did. {Op. eit. vol. iii. 1777, p. 382.) Arago must have awarded this honour to
Warltire inadvertently, for in his Eloge Higiorigue d* Alexandre Volia^ read to the
Vrench Academy ih 1833, after mentioning that the electric spark had been employed
to light combustibles, such as alcohol and hydrogen, in the open air, he continues—
** Yolta was the first who repeated such experiments in close vessels. (1777.) To him
therefore belongs the apparatus which Cavend-sh employed in 1781 to effect the
synthesis of water." (Annalee d2 Chimie et de Phydque^ t. liv. (1833), p. 402.)

X The account of Warltire's experiments "on the firing of inflammable air in close
vessels," with Priestley's observations on the conclusions th;^y warrant, will be found in
the latter's Esptt, and Obt, on Air, vol. v. 1781, App. p. 395. The account is
ivprinted by Mr. Muirhesd, Watt ^orr, p. x^,

282 CAVENDISH AS A CHEMIST.

employed a copper fluk^ wHioli he filled with the mixlnie of common and
inflammable air, and then weighed. When an electric spark wb6 passed
through the contents of the flask, and the mixture of gases exploded,
great heat was eyolyed as the hydrogen and oxygen combined. The AmA
was then cooled, and weighed again, to ascertain whether it had become
lighter by the loss of the heat which had been given off ; and in seyeral
tnals the vessel appeared on the second weighing to have lost weight;
from which Warltire seems to have concluded that heat is a pondeiable
body,

Warltire and Priestley's experiments were made before the 16th of
April, 1781, and it was their repetition in the summer of that year by
Cavendish, which led to the discovery of the composition of water. Their
consideration, therefore, brings ns to the point where the controveny
commences, and I now enter on the disputed ground. It involves two
questions, namely : 1. What is the discovery which Cavendish, Watt,
and Lavoisier claim to have made ) 2. When was that discovery made t
It will prevent confusion if these questions are discussed apart ; at least
at flrst, and so far as their separate consideration is practicaole.

It may seem at flrst sight unnecessary to discuss formally the nainre
of the discovery claimed to have been made by the rivals in this contro-
versy : and it would be needless, if it were certain that both held the
same view ; but as it has been affirmed that they did not, it is impossible
to discuss the question of priority till the reality and nature of tne con*
tested discovery have been determined.

QUESTION OP REALITY. NATURE OF THE DISCOVERY CLAIBfEO

BY CAVENDISH, WATT, AND LAVOISIER, AND

IMPUTED TO MONGE,

^ ^^A^^^^^^^^^^^^

5. Cavendishes Experiments and Conclusions concerning tlie

dnrnjanaitiiin. nf Wnft»T.

Composition of Water

In the abstract of the '^ Experiments on Air,'* Cavendish's observations
and conclusions have been fully considered, so that a brief and dogmatic
reference to their nature will be sufficient here.

The experiments, it will be remembered, were undertaken in the
course of an incjuiry into the products of combustion in air ; whilst the
special trials which led to the discovery of the composition of water, were
professed repetitions of Warltire's process for demonstrating the pondera-
bility of heat. The determination of that question, however, was not
the motive for the repetition, although this nas been strongly asserted
by some of the advocates of Watt's claims. Cavendish's own state-
ment is, that it was the appearance of moisture incidentally observed
by Priestley and Warltire, which seemed to him " likely to throw great
light on the subject he had in view," and, accordingly, " he thought it
well worth examining more closely." The alleged loss of weight he
also thought, " if there was no mistake in it, would be very extraordinary
and curious." He arranged matters accordingly, so that tne same experi-
ment should test the truth of Priestley's statement, that a deposition of
moisture followed the detonation of hydrogen and air in a dose vessel,

THE WATER CONTROVERSY. 283

and the tratli of Warltire's statement that the explosion was attended by
a loss of weight. To secure the means of making the double observation^
it was only necessary that the vessel should consist of glass, so that the
deposition of- liquid within it might be visible ; and that it should be
weiffhed before and after every explosion. Two sets of experiments were
made with this apparatus ; the one with hydrogen and air, the other with
hydrogen and oxygen. When air was used, it was mingled with hydrogen^
in the proportion (in the decisive trials) of 1000 measures of the former
to 423 of the latter j and the mixture was introduced into a glass globe,
(provided with a stop-cock and wires for passing the electric spark), which
had been previously emptied at the air pump, and its weight ascertained.
The spark was then passed so as to bum the gases, and the globe was
weighed again, to ascertain whether or not it had lost weight. No certain
alteration in weight occurred in Cavendish's trials. " I could never," says
he, '^perceive a loss of weight of more than one fifth of a grain, and commonly
none at all." " In all the experiments," he further observes, " the inside
of the glass globe became dewy," and when this dew was subjected to
what appeared to him a 8u£Scient number of decisive tests, " it seemed pure
water." Other trials were likewise made by simple combustion, without
the aid of the electric spark. In these hydrogen and air, in the proportion
of one volume of the former to two volumes and a half of the latter, were
burned together, as they issued from separate tubes lying side by side, and
opening into a common canal in which the resulting water condensed.
The object of these trials was, to collect a larger quantity of water for
analysis, than the globe experiments furnished.

The conclusion which he drew in full, was as follows : " "By the experi-
ments with the globe it appeared, that when inflammable and common air
are exploded in a proper proportion, almost all the inflammable air, and
near one fifth of the common air, lose their elasticity and are condensed into
dew. And by this experiment [that is, by analysis of the larger quantity
of water obtained by simple combustion,] it appears that this dew is plain
water, and consequently that almost all the inflammable air, and about one
fifth of the common air, are turned into pure water." *

The experiments with hydrogen ana oxygen were made in the same
way. The gases were mingled in the proportion of ''19,500 grain mea*
sures of dephlogisticated air and 37,000 of inflammable air," or one volume
of oxygen to less than two of hydrogen. '' The cock was then shut, and
the included air fired by electricity, by which means almost all of it lost
its elasticity." Sometimes the resulting liouid was sour to the taste, and
contained nitric acid, a phenomenon whicn Cavendish showed to result
from the oxidation of the nitrogen of atmospheric air, not entirely removed
by the air pump when the dfobe was exhausted. If excess of oxygen
were used, the liquid was acid ; but if this gas was in such a proportion
as exactly, or nearly exactly, to oxidise all the hydrogen, then ''the
condensed liquor is not at all acid, but seems pure water, without any
addition whatever; and as, when they are mixed in that proportion, Yery
little air remains after the explosion, almost the whole being condensed, it
follows that almost the whole of the inflammable and dephlogisticated air
is converted inter pure water." t Cavendish's most comprehensive conclu-
sion is summed up in the following sentence. "I think we must allow
that dephlogisticated air is in reality nothing but dephlogisticated water,
or water deprived of its phlogiston ; or, in other words, that water consists

• Phii, TYmu. 1784, p. 129. t Op- cit. p. 133.

284 CAVENDISH AS A CHEMIST.

of dephlogisticated air anited to phlogiston ; and that inflamma^ atr is
either pure plilogiston, as Dr. Priestley and Mr. Kirwan suppose, or eljie
water united to phlogiston ; since, according to this supposition, these two
substances united together form pure water/**

Two remarks, only, seem called for here, in reference to Cavendish's
experiments and conclusions. 1 . The combustible gas which he employed,
he calls infiammable air. It is of importance, therefore, to be certain, that
by this title he denoted hydrogen. That he did, there can be no question.
" In these experiments," says he, " the inflammable air was procured from
zinc, as it was in all my experiments, except where otherwise expres8ed.'*t
There is no expression 'otherwise' in reference to the trials, from which
the conclusions quoted above were drawn. Cavendish's " inflammable lur **
then was hydrogen, and it is worth while to notice that it was so, not
merely in the experiments specially referred to, but even iu those which
he excepts ; only in the latter it was prepared by the solution of iron,
not zinc, in diluted oil of vitriol. I 2. The mode in which he obtained
his results was singularly beautiful and simple, but the elegance of the
process may easily be overlooked ; and more than one critic of the Water
Controversy has altogether misapprehended it. It is necessary, therefore,
to refer to it a little more at length, and I select for illustration the case
of hydrogen and oxygen as simpler than that of hydrogen and air. Ca-
vendish, then, ascertained by preliminary trials, that when about two
volumes of hydrogen and one of oxygen were fired by the electric spark,
they disappeared as gases, and left, (with the exception of a small amount
of incondensible elastic fluid, which he considered as impurity), no residue
of any kind but a little water. Having ascertained this, be filled the
globe with a mixture of hydrogen and oxygen in the proportions men-
tioned, weighed the vessel with its contents, and fired the gases by the
electric spark. When the vessel cooled, it was hung up again to the
balance without being opened, and found not to have changed in weight.
Nothing ponderable, then, was lost. All the gas (a trace of impurity ex-
cepted) had ceased to be gas, and in its place was so much liquid, occupying
a space or volume immensely smaller than the gas had occupied, but
possessing a weight exactly identical. Weight for weight, then, the liquid
had replaced the gas, so that all the ponderable matter of the one was
contained in the other. In other words, |the gas was " turned into^ the
liquid, as Cavendish himself phrases it ; and it only remained to repeat
the globe experiment several times, and to burn considerable volumes of
the gases by direct combustion, so as to procure a sufiicient quantity of
the liquid, to admit of its being analysed, and shown to be pure water.

The refinements of modern chemistry have not devised a more elegant
or demonstrative process. As the whole operation was carried on in the
same vessel, not the fraction of a drop of the liquid could be lost. It was
saved alike for weighing and analysis. The method, which it will pre-
sently appear was practised by Priestley, of sucking up the moisture by
blotting paper, required tbe sacrifice of the liquid without analysis, and
made accurate weighing impossible. Lavoisier's process gave results which
fell far short of demonstrating that the gases burned, and the water pro-
duced, were identical in weight. Only Cavendish could show identity
between the weight of liquid which had appeared, and that of gas which
had disappeared. §

* Op, cit p. 137. + Ibid. p. 127. t Op. et he, eii,

§ The identity w:is of course not abvohtte, in the strict sense of that term, bnt,
dedacting the small aniorLnt of incondensible imparity, approached as nearly to it,

THE WATER CONTROVERSY. 285

I kaowy in trath, but one objection to Cavendisb's process. He did
not dry the gases be employed. Tbat he should not bare done so is
curioQS, for be knew and employed, in 1766, (as be records in his paper
on hydrogen,) the process for drying gases by passing them through a tube
containing a bygrometric salt, which is followed at the present day.* Had
he endeayoured to determine the quantitative composition of water by
weight, this neglect would have involved him in error, but it did not
vitiate bis conclusion as to its qualitative constitution, and would not
have sensibly affected an inference as to the quantitative composition of
water by volume.

The French experiments were no better in this respect than Caven-
dish s, for the gases were not dried by Lavoisier, Meusnier, or Mouge ;
and Priestley led Watt and himself into a serious error, by the process
which be adopted for rendering '' inflammable air" anhydrous.

Cavendish's results, then, are unexceptionable, so far as the materials
made nse of and the mode of experimenting are concerned; and the
inference was warranted and was just, that water consists of dephlo-
gisticated air (or oxygen), and of the inflammable air of the metals (or
hydrogen). Whether the substitution of the word phlogiston for inflam-
mable air, in the more general statement of his results which Cavendish
gave, rendered his statement more vague and uncertain than it was in its
more simple original shape, I shall consider in another section.

6. Priestley's Experiments, and Watfs Conclusions from them

concerning the Composition of Water,

Watt*s claims to be considered a discoverer of the composition of
water, are based by his advocates on a twofold ground. He is reported,
1. to have long entertained the belief, that water was convertible into
air or ga«; and 2. to have inferred, from certain experiments of Priest-
ley s, which their performer did not understand, that water consists of
the particular gases, hydrogen and oxygen. His more general views
will be considered in the sequel; bis special conclusion demands careful
consideration here; and first, of the experiments from which it was
drawn.

In the earlier half of March, 1783, Priestley repeated Cavendisb*s
globe experiments on the convertibility of a given weight of inflammable
and depblogisticatcd air into the same weight of water. With the date
of Priestley's experiments, and the fact that they were a repetition of
Cavendish's, we are not specially concerned at present ; but this passing
reference to these points will render more intelligible many of the
allusions in the succeeding statements.

The earliest account hitherto published of Priestley's experiments, is
contained in a letter from Watt to Gilbert Hamilton, of date 26th March,
1783, in which this passage occurs : — " Ho (Priestley) puts dry depblo-
gisticatcd and dry inflammable air into a close vessel, and kindles them by
electricity. No air remains, at least if the two were pure ; but he finds
on the side of the vessel a quantity of water, equal in weight to the air
employed, "t

probably, as the limits of accuracy in experiment permit; at all events, it came nearer
to it than the results obtained by any of Cavendish's rivals.

* Ante, p. 201.

t ]YaH Corr, p. 17. An earlier reference is given, ante, p. 94«

286 CAVENDISH AS A CHEMIST.

The exaet nataie, object, and yalae of tiie experiment tlins deeeribed,
haye been matters of the keenest diacossion between the friends of Watt
and Cavendish, and particnlarlj in reference to the kind of inflam*
mable air which Priestley employed. Watt describes the experimenta at
greater length in various of his letters, and more fnllj in his " Tboaghts
on the Constitnent Parts of Water, &c.;*' bnt in none of his writinca haa
he stated what kind of inflammable air Priestley employed. The mtter,
however, is more explicit in his own account of his experiments^ and aa j

it confessedly contains the only direct reference to the quality of the '

combustible gas which Watt held to be one of the elements of water, I \

quote Priestle/s statement in full. It is contained in his paper entitled,
" ExperimenU relating to phlogiston, and the eeeming eonvemon of water
into air,** which was originally accompanied by a commentary £rom the
pen of Watt, in the shape of a letter to Priestley, containing an exposi-
tion of the views of the former concerning the composition of water.
For reasons which will be afterwards considered, that commentary vraa
withdrawn before Priestley's paper was read to the Royal Society
(June 26, 1783); so that Watt's conclusions, as they were ultimately
published in the succeeding year (1784), appear quite disconnected from
Priestley's own account of his experiments. It is necessary, therefore,
to notice that the following quotation is from one of the most important
parts of the text on which Watt commented.

" Still hearing of many objections to the conversion of water into air, j

I now gave particular attention to an experiment of Mr. Cavendish's con- '

corning the re-conversion of air into water, by decomposing it in con- |

junction with inflammable air. And in the first place, in order to be sure |

that the water I might find in the air was really a constituent part of it, ,

and not what it might have imbibed after its formation, I made a quantity
of both dephlogisticated and inflammable air in such a manner as that
neither of them should ever come into contact with water, receiving
them, as they were produced, in mercury; the former from nitre, and in
the middle of the process (long after the water of crystallization was /

come over), and the latter from perfectly-made charcoal. The two kinds
of air thus produced I decomposed, by firing them together by the electric j

explosion, and found a manifest deposition of water, and to appearance,
in the same quantity as if both the kinds of air had been previously
confined by water.

** In order to judge more accurately of the quantity of water so depo-
sited, and to compare it with the weight of the air decomposed, I care-
fully weighed a piece of filtering paper, and then having wiped with it ^
all the inside of the glass vessel in which the air had been decomposed,
weighed it again, and I always found, as nearly as I could judffe,
the weight of the decomposed air in the moisture acquired by the i
paper.

" As there is a source of deception in this experiment, in the small ^

globules of mercury, which are apt to adhere to the inside of the glass <

vessel, and to be taken up by the paper with which it is wiped, I some-
times weighed the paper witn the moisture and the mercury adhering to
it, and then exposing it in a warm place, where the water would evapo-
rate, bat not the mercury, weighed it again, and still found, as nearly as
I could pretend to weigh so small a matter, a loss of weight equal to that
of the air.

" I wished, however, to have had a nicer balance for this purpose; the
result was snch as to afford a strong presumption that the air waa recon-

THE WATER CONTROVEBST. 287

yerted into waier^ and therefore that the origin of it had been water.'**
The account which Priestley gives of hie experiments, (here spoken of as
onOj) is exceedingly defective. It is explicit enough^ however, to show,
that in two important points^ his mode of experimenting differed from
that of Cavendish. 1 . The inflammable air he employed was not hydrogen,
but a gas procured by heating perfectly-made charcoal. 2. He removed
the water from the vessel in which it was produced, before he ascertained
its weight. The first of these variations on Cavendish's process was
introduced with the laudable purpose of employing ankydrctta inflam«
xnable air, so that no ready-formed water might be present (at least in a
state of mechanical mixture) in the gas which was to produce water by
its union with dephlogisticated air. Priestley manifestly regarded the
inflammable mrfrom charcoal as identical with the inflammable air from
metaUf and preferred the former because it was freer from moisture : in
truth, as it was procured from red-hot charcoal, it was naturally enough
assumed to be anhydrous. His great anxiety to procure dry gases, pro-
bably arose from his recollection of the explanation which Warltire had
Siven of the source of the water which appeared when inflammable and
ephlogisticated air were exploded together. The latter, as mentioned
previously, held that air deposits its moisture when phlogisticated — an
opinion which Priestley recorded as at least worth notice. According to
this view there was no '' conversion of air (gas or gases) into water,* but
a mere precipitation in the liquid fonn of the aqueous vapour, which was
diffused through the gases before their combustion. If, however, they
could be deprived of this aqueous vapour, or prepared ab initio without
it before they were burned, Warltire's explanation would plainly be inap«
plicable, and the proof that conversion of gas into water had occurred,
would be rendered more complete.

In this way Priestley escaped one fallacy, only however to fall into
a more serious one. We are indebted to Mr. Haroourt for first pointing
out the important fact, that the inflammable air which the former em-
ployed was not hydrogen. Its more abundant constituent (by weight)
must have been carbon, but its exact composition cannot be ascertain ed,
for Priestley merely states that he prepared the gas from perfectly-made
charcoal.

There can be little doubt, however, that he heated the charcoal in an
earthen retort This at least may be inferred, from the following pas-
sage which occurs in the paper containiug the account of his repetition of
Cavendish's experiments. *' Wood or charcoal is even perfectly destruc-
tible, that is, resolvable into inflammable air, in a good earthen retort and
a fire that would about melt iron. In these circumstances, after all the
fixed air had come over, I have several times continued the process during
a whole day, in all which time, inflammable air has been produced
equally, and without any appearance of a termination.'*t Perfectly-made
charcoal, if It were in reality pure carbon, could yield nothing in an air-
tight retort, free from moisture, but a small quantity of carbonic oxide,
and a little carboDic acid, formed by the combination of the charcoal with
the oxygen of the air which filled the unoccupied spaces in the retort.
The best charcoal, however, frequently retains in combination a little
hydrogen derived from the wood from which it was prepared, and always
contains, if it has been exposed to the air, water- vapour and other gases,

• PkU. TVoM. 1783, pp. 426-427} or PrintUy onAir, vol. vi. (1786) p. 50.
t PM. TVmi. 1783, p. 412.

288 CAVENDISH AS A CHEMIST.

which it ahsorhs with great avidity from the atmosphere. Priesilej's
charcoal, accordingly, would probably yield hydrogen, or carbaretted
hydrogen, or both, as well as carbonic oxide and carbonic acid, when
heated in an air-tight retort

From the way, however, in which that ingenious chemist refers to the
evolution of inflammable air from heated charcoal continuing for hours
'* without any appearance of termination," it seems exceedingly doubtful
whether he employed a close vessel. It is much more probable, if not
certain, that the retorts he made use of, were Wedgwood-ware vessel^ of the
kind he constantly refers to in the paper from which I have been quoting.
These were, or appeared to be, air-tight at ordinary temperatures, but
when made red-hot became sensibly porous, so as to permit the diffusion*
of gases through them.* If charcoal were exposed to a high tempera-
ture in such retorts, the gases which it evolved would, to a certain extent^
exchange places with those resulting from the burning fuel by which the
retorts were heated. In this way the carbon, even if pure, would be
exposed to carbonic acid and steam, besides other gases, which pene-
trating the porous walls of the retort, would become converted in whole
or in part into carbonic oxide, hydrogen, and carburetted hydrogen.
Except on this supposition, it seems impossible to account for the endless
evolution of inflammable air from charcoal to which Priestley refers ; and
it can scarcely be doubted that his so-called dry gas contained ready*
formed water.f

It is not necessary, however, to insist at length on this point. It is
not denied by any party, and does not admit of dispute, that Priestley^s
charcoal-gas was not hydrogen. It must have contained as large a
volume of carbonic oxi(le as of hydrogen, on the view most favourable
to the account which he gives of the products which it yielded on com-
bustion ; but it probably also contained carburetted hydrogen, carbonic
acid, and water vapour. On either view it could not, when detonated
with oxygen, afford the results which Priestley conceived that it yielded.
In two important particulars his statement is irreconcilable with the
account of his experiments which he furnishes himself. 1. There is no
proportion in which the charcoal-gas can be burned along with oxygen,
so as to yield a quantity of water equal to the weight of the gases con-
sumed. 2. The product of the combustion of the charcoal-gas is not
water alone, but water and carbonic acid. It is further to be noticed,
that Priestley does not explain how he ascertained that the weight of
water was equal to that of the gases burned. Besides employing a

* See, in illustration of this, Priestley's " experiments relating to the seeming
conversion of water into air." PAi7. TVoiw. 1783,p. 414; and Prof. Graham's ftonejito
^Chemitiry, 2nd edit. vol. i. p. 85.

t Mr. Harcourti assuming that Priestley's retorts were airtight, has calcnlated tho
P^^™^® composition of the latter's charcoal-gas. {Rep, Brit, Assoc, for 1839. Pres.
Address, p. 27.) For the reasons, however, given above, I question the value of any
jaicolation founded on such an assumption; and I think it needless to offer another in
ff f!?^ ^**^ ***® ^*^* **^ ^ imperfect. The quality of the gas, rx, gr,^ would be
anected by the shape and size of the retort, the mode in which it was heated, the con-
nguration and dimensions of the pieces of charcoal, and the extent to which the retort
WM occupied with them, or filled with air at the commencement of the process. The
Quantitv"'*^?*** ^^y^^^ progressively alter in quality, as the charcoal diminished in

fluid. Wh" ^iSi"*k**^ " ^^'^^^ P*^ °^ ^^^ ^^^ °^ ^® ^^^^ ^ ^® ^"^ ^^ ^^^
the auantif«H»« ^ ^^ variable points are unrecorded, a hypothetical calculation of

called ioTh^lL^^^^^l^^^^^^ *^e charcoal-gas, could be of no value; and it is not
oy ine demands of any of the parUes in the Water Controversy,

THE WATBR CONTROVERSY. 289

diflferent gaa from that which Cayendish employed^ he adopted another
method of proving that the whole burnt gas was converted into water.

This required that the absolute weight of the mixed gases fired hy the
electrio spark, should be ascertained at the beginning of the experiment^
and the weight of the resulting water at the end; but Priestley makes
no reference to the preliminarj weighing of the gases. To make his
account consistent (not to say credible), we must suppose that he ascer-
tained or calculated the weight of a globeful of the mixed charcoal-gas
and oxygen before he passed the spark. The resulting water, we have
seen, he absorbed by blotting-paper, which he weighed whilst wet, and
again after it was dry. The loss in weight gave the amount of water
which the gases had yielded.

I need-not expatiate on the rudeness of this process, which could not
have given accurate results in the hands even of a Berzelius or a Faraday,
much less in those of Priestley, who, though singularly ingenious and
inventive, was far from an accurate observer in quantitative investigations.
Defective as his method essentially was, it was rendered additionally im-
perfect, as Priestley himself acknowledges, by the absence of a delicate
balance, and by the mixture with the water of globules of mercury, with
which, it should seem, the globe had been filled, before tho gases were
introduced into it. Priestley, moreover, implicitly acknowledges that the
weight of the gases, and that of the water, were not identical ; though to
what extent the one fell short of the other, is concealed from us by the
total absence of numerical statements which characterises the whole
aecount of this important experiment.

Lastly, it is to be noticed, that Priestley makes no reference to any
examination of the liquid which resulted from the combustion of the gases.
He cannot have analysed it carefully, or he would have found carbonic
acid in it, wjiich no one knew better how to detect than he did ; and he
probably did not analyse it at all, for his method of procedure (unlike
Cavendishes) dissipated the water in the process of weighing.

It appears, then, that Priestley cannot have obtained the results he
professed to have got. He employed the vn'ong gas; he must have weighed
inaccurately; and he either did not analyse, or failed to analyse suffi-
ciently, the so-called water. In this way he unconsciously deceived him-
self; and for a long period he undesignedly led others astray.

Into the geneial consideration of the result of these errors, I have not
occasion to enter; but it is manifestly of the greatest importance to deter-
mine whether Priestley's charcoal-gas experiments were those on which
Watt founded his conclusions.

When Arago revived the Water Controversy, he assumed that
Priestley's experiments were made with hydrogen and oxygen, and that
they were unexceptionable. Those, on the other hand, who disallowed
Watt's claims, referred to Priestley's account of his own researches, as
proving that they were made with improper materials, and were not
trustworthy. Mr. Muirhead left this important point almost entirely
anconsidered, although the testimony of the WaU Correspondence in
reference to it called for special notice. His elaborate defence of Watt's
claims was, accordingly, pronounced defective in a most important parti-
cular; and Lord Jeffrey, acknowledging the force of the objections which
Harcourt and others raised against the validity of inferences drawn from
the charcoal-gas experiments, sought at great length to show that other
unrteorded experiments were made by Priestley with the proper materials,
and in a trustworthy way, and that i\\e^e were the foundation of Watt's

290 CAVENDISH AS A CHEMIST.

conclusions. His Lordship is thus the only advocate of Watt's claims,.
who has endeavoured to show that the Utter was entitled to his conclu-
sions in virtue of the sufficiency and significancj of the experiments which
^ve birth to them, and as his discussion of the question is as complete as
it is forcible, I shall consider it at some length. If Lord Jeffreys view,
indeed, is well-founded, he has succeeded in establishing a claim for Watt
SDcb as none of his other advocates have been able to sustain. What that
claim implies, will be best understood if Wattes conclusion be stated before
any criticism is offered on the experiments from which it was drawn. In
the letter of April 26, 1783, which he addressed to Priestley, the following
passage occurs : — *' Let us now consider what obviously happens in the
case of the deflagration of the inflammable and dephlogisticated air. These
two kinds of air unite with violence, they become red hot, and upon cooling
totally disappear. When the vessel is cooled, a quantity of water is found
in it equal to the weight of the air employed.

^' Tnis water is then the only remaining product of the process; and
water, light, and lieat are all the products, unless there be some other
matter set free which escapes our senses.

'' Are we not then authorised to conclude, that water is composed of de-
phlogisticated air and phlogiston, deprived of part of their latent or elemen-
tary heaff (Jc. ?^** The concluding portion of the paragraph, which refers
to the nature of dephlogisticated air, and the relation of latent heat to the
gases, will be considered in another section. In the passage quoted. Watt,
it will be observed, infers from Priestley^s experiments that water conaista
of dephlogisticated air or oxygen, and of another substance which he
names inflammable air or phlogiston. He does not limit the term injiam- i
mable air (with which alone we are at present concerned) in the para-
graph under notice (nor, as will afterwairds appear, in any other of his
published writings), so that, if his conclusions were founded on Priestley's
recorded experiments, he must have intended by inflammable air thai
from charcoal, and his inference was, that the components of water are
oxygen and the charcoal-gas, not oxygen and hydroiren.

It is not a little singular, that the serious difficulty which the char-
coal-gas experiments throw in the way of Watt's claim to be the discoverer
of the true composition of water, should have been passed over so lightly j
by nearly all his advocates. Lord Jeffrey meets the difficulty in the \
following way. '* Watt himself," says he, " makes no mention of this
charcoal-gas; and nowhere refers to this paper of Priestley's as containing
toe experiments on which he proceeded, but states these for himself in
very minute and particular detail. There is not an atom of evidence,
indeed, to show that, before writing his letter, he had ever seen this paper, '
which was not read in London till 25th May, 1783, nor printed till the
very end of that year; and we think it by far most probable that he knew
nothing of its particular contents till after that publication. It is a ffreat
and fundamental mistake, also, to suppose that the main object and subject
of that paper was the same, or even very much connected with that of i
Watt's letter. Its first and longest division consists of a dissertation on
the nature of Phlogiston generally; and the other on ' the supposed con-
vertibility of water into air,' by which he explains he means into the
common air of the atmosphere. Almost the whole of the experiments
detailed in it, accordingly, are referable to this analytical process; and
there is but a slight notice, extending in all to little more than a page, of

* PhiL TVOM. 1784, p. 333.

THE WATER CONTROVERSY. 291

the synthetieal proceedings of Cayendish, and the few experiments he had
himself made in connexion with them; not so mnch, we think, to test or
confirm the general results reported by Cavendish, as to eliminate one
particular source of possible error. Watt's letter, on the other hand, pro-
fessed only to embody his own 'thoughts on the constituent parts of
water;' ana had, therefore, no material bearing on the general disquisition
of Priestley's. He> had plainly receiyed a full and complete account of all
the experiments on which his own conclusions were founded some time
before the 26th of March, and therefore before either his own letter or
Priestley's paper of 21st April were written; and had no occasion, there-
fore, to look to that paper, or concern himself about its contents, in order
to prepare that exposition of his important theory which he then proposed
to make public. He had, of course, long before communicated hurgely
with his learned friend and neighbour on the nature of that theory, and
made himself minutely acquainted with every material particular of the
experiments on the faith of which it was grounded; and our own firm con-
viction is, that he had been distinctly told, and told truly, that all those
experiments in which the quantity of missing air was carefully measured,
and the freedom of the water proanced from acid ascertained, were made
with the injlammable air Jrom the metals, or the hydrogen of our modem
technology; and that if any mention at all was made between them of
the employment in later experiments of the gas firom charcoal, it was only
for the purpose of showing that the (supposed^ perfect dryness of that air
did not interfere with the general success of the processes." *

Before making any remarks on Lord Jeffrey's argument, I think it of
great importance to notice that his Lordship, as will be seen from the

Suotation, implicitly acknowledges that neither Priestley nor Watt has
escribed experiments with hydrogen, and that no direct evidence of any
kind can be produced to show that they employed that gas. The proof
that they did, if it can be supplied at aU, must be gathered from the com-
parison of many separate passages in the writings of both.

From such a collation of passi^fes, Lord Jeffrey seeks to show : 1 . That
Watt's original exposition of his views concerning the composition of water
had little or no reference to Priestley's paper, in which the charcoal-gas
experiments are recorded; so that Watt is relieved from any share in the
errors into which Priestley fell. 2. That, from various statements and
allusions, it must be inferred that Priestley performed experiments with
hydrogen, and that these formed the groundwork of Watt's conclusions.
It is thus held that Watt did not signify, by injiatnmahle air, the charcoal-
gas, and that he did signify hydrogen.

With neither of these conclusions can I agree, for the following
reasons : — It seems to me to admit of direct proof, that the connexion
between Priestley's paper and Watt's original letter was of the most inti-
mate and essential kind; so that the former was the text, on which the
latter was the commentary. Priestley, it will be remembered, drew up,
in March, 1783, a paper for the Royal Society, entitled '' Experiments
relating to Phlogiston, and the seeming Conversion of Water into Air,"
which was read to that body on June 26, 1783. Watt^ who was aware
of his intention, sent him a letter dated 26th April, of the same year, with
the request that Priestley would present it to the Society, if he thought
proper. Priestley, accordingly, brought it before the Society, and it was
eventually read publicly, April 22, 1784. A second version of it also,

* Edinburgh Review ^ Jon. 1848, pp. 99, 100.

292 CAVENDISH AS A CHBMIST.

containiDg large parts of tbe first repeated verbatim, was addressed in ibe
form of a letter to De Luc, and read to the Royal Society, April 29, 1784.
This later letter forms, with some additions, the paper printed in the
Phil. Trans, for 1784, with the title '' Thoughts on the Constituent Parts
of Water," &c.* It is not to this triple document that we must turn for
evidence to show what the relation was that subsisted between Prie8tle3r*s
paper and Watt's first letter ; for the omitted portions of the latter are
exactly those which are most important in reference to the particular
question immediately before ue. The MS. of Watt*s letter to Priestley
of April 26, 1788, is preserved in the archives of the Royal Society ;
and through the courtesy of Mr. Weld, Assistant Secretary, I have
obtained an authenticated copy of it. It begins thus : '^ On consi-
dering your very curious and important discoveries on the nature of
phlogiston and dephlogisticated air, and on the conversion of water into
air, and vice versa, some thoughts have occurred on the probable causes
of these phenomena, which, though they are mere conjectures, seem to
me more plausible than any I have heard on the subject, and in that view
I have taken the liberty to communicate them to you."t

In this passage it will be seen that Watt, so far from confining his
attention solely to Priestley's experiments on the convertibility of a mix-
ture of inflammable air and oxygen into water, announces his intention
of commenting on various of his fnend's recent discoveries, of which he
proceeds to enumerate four: — 1. The nature of phlogiston. 2. The nature
of dephlogisticated air. 3. The conversion of water into air. 4. The
conversion of air into water. To each of those topics Watt refers in his
letter, which concludes with the following passage, which I quote here
because of its importance as proving that Watt was aware that JPriestlej^s

Saper, which his letter was to accompany, would contain an account of the
iscoveries on which that letter commented : — '' If my deductions have
any merit, it is to be attributed principally to the perspicuity, attention,
and industry with which you have pursued the experiments which gave
birth to them, and to the candour with which yon receive the commanica-
tions of your friends. If you shall think that a hypothesis so hastily com-
piled deserves to have the honour of being communicated to the Rojal
Society, or published in any other way, along with the account of your
experiments, I will be obliged to you to present it to the Society, or to the
public, as you shall see proper.'*

The passage I have marked by italics shows most plainly that Watt
relied on Priestley's account of his experiments, and wished his letter to
be published along with it. It seems impossible, therefore, to contend
that the former was ignorant of the contents of Priestley^s paper, and is

* Watt himBelf gives the following aoooant of the fortunes of hb fint letter : —
'< This letter [April 26, 1783] Dr. Priestley received at London; and after showing it
to several members of the Royal Society, he delivered it to Sir Joseph Banks, the
President, with a request that it might be read at some of the public meetings of the
Society; but before that could be complied with, the author, having heard of Dr.
Priestluy'is new experiments, begged that the reading might be delayed. The letter,
therffore, wan reserved until the 22nd of April last, when at the author's request it
vfttii read before the Society. It has been judged unnecessary to print that letter, as the
essential parts of it are repeated, almost verbatifH, in this letter to M. De Luc ; but, to
authenticate the date of the author's ideas, the parts of it which are contained in the
present letter are marked with double commas." (Phil. TVans, 1784, note, p. 330.)

t The first part of this sentence (down to "phenomena") has already appeared in
print, in A Few Notee on the History (jf the Dieeovery of the Compoeition qf Water,
by J. O. HalUweU, p. 1. (Ante, p.26y.)

THE WATER CONTROVERSY. 293

not to be considered cognisant of the description which it contained of the
experiments relating to the conversion of air into water. I do not
wish to be understood as implying that Watt had read the MS. of
Priestley. It seems, on the other hand, certain, from the date of
Priestley's letter to Sir Joseph Banks (April 21, 1783), which accom-
panied his paper, as well as from internal evidence, that Watt cannot
nave read the latter part at least of Priestley's MS. when he wrote his
letter.

The internal evidence to which I refer, is the fact that, at the date
of his letter (26th April, 1783), Watt believed, on the authority of
Priestley, that by distilling water in clay retorts raised to a high tem-
perature it could be converted into air. Priestley, however, before he
completed his paper, discovered that he had been mistaken in this con-
clusion, and that the apparent conversion of water into air was owing
to the steam of the boiling liquid, and the elastic fluids of the atmosphere,
changing places through the porous walls of the retort. He did not
inform Watt of his mistake, however, till April 29th, 1783, three days
after Watt had written his letter; so that, whilst he refers to the
conversion of water into air as a certainty, Priestley alludes to it as a
mistake, entitling the part of his paper which re^rs to it, " On the
teeming conversion of water into air." It is impossible, therefore, that
Watt can have read the MS. of at least the latter part of Priestley's
paper, and he probably did not peruse any part of it.

Let it then be conceded that Watt was not acquainted with the tpsu-
sima verba, or minute details of Priestley's paper; nevertheless he must be
held to have been generally aware of the account which Priestley has
given of his experiments, for this is implied in the passages of his letter
which I have marked in italics; and, in truth, the statements about to
be made by Priestley to the Royal Society were only repetitions of state-
ments already made orally or in writing to Watt. And further, it
must not be overlooked that Priestley, although informed by Watt that
the latter left to him the task of describing his experiments minutely,
referred to the charcoal-gas, and to it aloney as the inflammable air
which he exploded along with oxygen, in his experiments on the con-
version of air into water. Neither did Watt at any later period disclaim
Priestley's account, or object to the reference to the charcoal-gas, or
affirm that hydrogen should have been named instead of that mixture of
elastic fluids.

The concluding sentence of Watt's letter to Priestley has not been
published before, so far as I am aware; and the advocates of Watt,
therefore, may not have had its contents brought under their notice.
There are passages, however, as pertinent, in the Watt Correspondence,
On 21 St April, 1783, Watt writes to Dr. Black : " Dr. Priestley has made
many more experiments on the conversion of water into air, and I believe
I have found out the cause of it; which I have put in the form of a letter
to him, which will be read at the Royal Society vnth his paper on the
subject"* Here Watt refers to the conversion of water into air, as ^'the
subject" to which his letter and Priestley's paper alike referred ; the one
recording experiments proving (or apparently proving) the conversion,
the other pointing out its cause.

To the same efiect he writes to Mr. Gilbert Hamilton : '' Dr. Priestley
has made many discoveries lately in relation to the conversion of water

• Watt Corr. p. 18.

294 CAVENDISH AS A CHBMIST.

•

into air; and I have from them made ont what water is made of and
what air is made of, which theory I have given him in a letter to be read
at the Rojal Society, along vnth the accoanta of his discoveries.^^*

Similar statements occar in letters from Watt to Smeaton (April 27),
and to Fry (April 28);t and in truth, till the extent to which Watts
claim was perilled by the connexion of his conclusions with Priestley's
erroneous charooal-gaB experiments had been forced upon the attention of
Watt's advocates, thev did not deny that Priestley's paper contained the
account of the researches on which Watt's conclusions were chiefly based.
Mr. Muirhead, ex, gr., says, " Mr. Watt's letter to Dr. Priestley, dated
26th April, 1733, gives the statement of his theory, to be read at the
Royal Society, at the same time as Dr. Priestley's paper, containing the
experimetUs upon vfhich that theory was in great measure founded, ^*X Mr.
Muirhead's testimony on this point is peculiarly valuable, as he is the
most zealous of all VE^att's advocates, and his statement may satisfy the
reader that I have only put a just construction on the passages I have
quoted from Watt's letters, it seems to me impossible, therefore, to
acknowledge the justice of Lord Jeflrey's statements, already quoted, that
'' it is a great and fundamental mistake to suppose that the main object
of Priestley's paper was the same, or even very much connected with that
of Watt's letter;" and that ^' Watt's letter, on the other hand, professed
only te embody his own ^ thoughts on the constituent parts of water,' and
haJ, therefore, no material bearing on the eeneral disquisition of
Priestley's." Those statements are at variance with Watt's own acknow-
ledgments, for the passages I have adduced from hi? correspondence show
that he intended his letter te be a commentary on the whote of Priestley's
paper; and that, instead of limiting himself to the consideration of the
constituent parts of water, he has enumerated four topics ; namely, the
nature of phlogiston; the nature of dephlogisticated air; the conversion
of water into air; and the conversion of air into water; to all of which he
refers. The title {Thoughts on the constituent parts of Water) which his
letter ultimately bore, no doubt conveys a different impression, and is
referred to by Lord Jeffrey as proving the limitation of Watt's speculations
to one branch of Priestley's experiments which Watt took care himself
to describe fully. That title, however, was not added to the paper till
May, 1784, more than a year after the letter was written,§ and after that
portion of it had been withdrawn, which treated of the conversion of
water into air. The first letter, in truth, had no title, and the &ct is
significant, for it needed none if it were to be read after a paper of
Priestley's, on which it was a commentary.

It is unnecessary, however, to dwell at greater length upon this; for
the extent to which Priestley's paper and Watt's letter go over the same
ground, can be determined by any one who will compare the two docu-
ments. The agreement is very close, notwithstanding the alteration
which Priestley made in his paper, after he discovered that he was
mistaken in his supposition, that water can be distilled into air.
Thus Priestley enters at great length into the properties of phlogiston,
and details a long series of experimente demonstrating that ''pure
inflammable air" can reduce metallic and other calces or oxides, and
revive the metal or the combustible (such as sulphur or phosphorus)
which they contain. Then he records the experiments made by heating

* Wait Corr. p, 20. + TTaW Cbrr. pp. 23— 24. t Watt Corr. p.2\.

i Watt Corr. p. 63. Letter from Watt to Blagden, May 27th, 1784.

-',

TUB WATER CONTROVERSY. 295

water in porous clay retorts, which appeared to him, when he communi-
cated his observations to Watt, to prove that water can be converted into
air; and thereafter he describes the repetition of Cavendish's experi-
ments/ which seemed to him to establish with more or less certainty the
converse of his own (supposed) discovery, namely that air (sas or elastio
fluid) can be converted into water. Such is a very brie? analysis of
Priestlej's paper. Watt's (MS.) letter of April 26, 1783, will show to what
extent it is occupied with topics similar to those discussed by Priestley.
Addressing Priestley, Watt says: " Ist. You have shown by the experiment
of reducing the calces of metals in inflammable air, that the latter is
either phlogiston itself, or that it contains a very small quantity of any
other matter. 2nd. You have informed me, that when you mix together
quite dry inflammable air and quite dry dephlogisticated air, and fire
them by means of the electric spark in a close vessel, you find that a
quantity of water very nearly or quite equal in weight to the whole air,

is deposited on the sides of the vessel 3rd. That when yon expose

to heat porous earthen vessels previously soaked with water, or make
steam pass slowly through a red-hot tobacco pipe, that the water or
steam is converted into air, dcc."t This it will be seen is an abstract of
Priestley's paper. I have taken it from the original letter as it exists in
the archives of the Royal Society, because it is the document on which
Watt's claim to priority over Cavendish is mainly based. That paper,
however, is not generally accessible, and for the present may seem less
authoritative than those already before the public. It seems well, there-
fore, to notice that abstracts of Priestley's paper nearly identical, will be
found in a letter from Watt to Black, April 21> 1783, and in one to
Gilbert Hamilton of April 22, of the same year.}: I make but one
further remark on this topic. In Watt's paper as it was ultimately
published, with the title. Thoughts on the con^UtterU parts of Water, the
connexion between his views and Priestley's is less apparent than in the
original letter, but the very difference, in the amended paper, only adds to
the force of the conclusion I am urging. For the difference mainly
consists in the omission of all reference to the power of porous clay
vessels to convert water into air, which Priestley had discovered to be a
delusion, and which fell to the ground along with all that Watt had
foanded upon it. That it should have been referred to, however, at all,
shows how unwarranted are the statements which represent Watt as only
interested in Priestley's experiments on the synthesis of the elements of
water. Priestley himself, an unexceptionable authority in the present
case, thought very differently; for, assured of the importance which Watt
attached to the conversion of water into air, Priestley wrote to him,
informing him of the mistake he had made, in the following terms : —
'* Behold with surprise a'nd with indignation the figure of an apparatus
that has utterly ruined your beautiful hypothesis. "§ To this Watt
replied as if he cared little for the new observations of Priestley: ''I deny

* Quoted in foil, ante, p. 284.

i* The remainder of the letter, which is occupied with Watt's oonclnsionfl concern-
ing the composition of water ; the relation of elementary heat to the production of that
liquid; the nature of oxygen; the mode in which a porous clay vessel acts when it
(apparently) converts water into air, &c. will be referred to in another place. They
form the commentary on the text of which Watt has given the abstract, and are not at
present under discussion.

t Wait Corr. pp. 18—21.

( Watt Corr. p. 25. Priestley to Watt, 29th April, 1783.

296 CAVENDISH AS A CHBMIST.

that your experiment ruins my hypothesis. It is not founded on so brittle
a basis as an earthen retort, nor on its converting water into airj I
founded it on the other factfl, and was obliged to stretch it a good deal
before it would fit this experiment.*'* But that Priestley did not orer-
rate the importance which Watt attached to the porous clay experiments,
is evident from two things: 1. In bis original letter the latter refers to
them thus, ''On considering the last and most remarkable production of air
from water imbibed by porous earthen vessels, (the only case wherein it
appears almost incontrovertibly that nothing was concerned in the pro-
duction except water and heat,) I think," &ct 2. It was the informa-
tion supplied by Priestley, that water was not convertible into air by
porous clay heated, that induced Watt to withdraw his entire letter ft-om
the Royal Society. Writing to Dr. Black on 23rd June (1783), he says,
" i have withdrawn my paper from the Royal Society, on account of an
ugly experiment the said Dr. Priestley tried at my desire, and which
renders the theoiy useless in so far as relates to the change of water into
air by means of porous earthen vessels.*'! I'his fact supplies the best
proof, that the whole of Priestley's paper, and not merely the section of it
referring to inflammable air and oxygen, was before Watt's mind when
he wrote his letter. Had he concerned himself only about the conversion
ofinfiammabU air and oxygen into water, he would not have withdrawn
his paper because water was not convertible into atmaspherie air. Had
his letter accordingly beeu read to the Royal Society at the time of its
receipt, the Water Controversy would either not have arisen, or would
have exhibited a very different aspect from that it has shown. As it ie^
that controversy is a standing record of the intimate connexion that
subsisted between the whole of Priestley's paper and the whole of Wattes
letter.

It should seem, then, that unless very distinct and explicit proof can
be afforded, that Priestley did perform experiments with hydrogen and
oxygen, and that on these Watt's conclusions were founded, it will be
impossible to exculpate him from a participation in Priestley's erroneous
preference of charcoal-gas to hydrogen, or to understand Watt as signi-
fying, by inflammable air, the latter gas. Yet if he did not, he cannot be
considered as having taught that water consists of hydrogen and oxygen.

Lord Jeffrey, as already implied, is the only one of Watt's advocates,
who has seen and acknowledged the necessity of proving that by inflam-
mable air Watt signified hydrogen. But even he does not profess to
have discovered a direct statement in any production of Priestley's or
Watt's, that hydrogen was employed in the experiments of the former;
and he only says that he " cannot but believe that there were other expe-
riments made with hydrogen; and this for a great variety of reasons. "§
These I shall presently notice; but before doing so, the reader will not
fail to observe that in discussing them we necessarily abandon the direct
documentary evidence, on which hitherto all our conclusions have rested.
The friends of Watt have been fond of contrasting the gaps which occur
in the early chain of evidence in favour of Cavendish's priority, with the
direct and unbroken succession of proofis which they allege can be
adduced in support of Watt's claims; but it now appears by the acknow-
ledgment of the ablest of Watt's defenders, that there exists no docu-

* Watt Corr, p. 27. Watt to Priestley, 2nd May, 1783.

t MS. Letter, April 26th, 1783.

I Watt Corr. p. 30.

§ Edinburgh Review, January, 1848, p. 94.

THE WATER CONTROVERSY. 297

ment directly affirming that he believed, or taught, that hydrogen is
one of the elements of water. He called the combustible ingredient of
water inflammahle air or phlogitton, and he has in none of his writings
limited either of these terms to hydrogen. If it can be shown that he did
signify hydrogen by the titles in question, it is only by a lengthened and
circuitous process, inyolving the comparison of many passages in the
writings of Watt and Priestley, and which does not, even in the hands of
Lord Jeffrey, yield a decisive result. The result it does yield, however,
must be ascertained as the only means of doing justice to Watt.

His theory of the composition of water he implicitly announced, as
Cavendish also did, in two ways: — I. As a conclusion from certain expe-
riments. 2. As a formula more general in its character, founded upon
that conclusion. The particular conclusion was, that water consisted of
inflammable air and oxygen. The general formula was, that since
inflammable air is phlogiston, water may be defined to be a compound of
phlogiston and oxygen. I reserve the full discussion of Watt's views
concerning phlogiston, as I have done those of Cavendish on the same
subject, to another section; and limit myself here to the consideration of
his inference from Priestley's experiments.

I have already quoted Watt s conclusion, as given in his " Thoughts
on the constituent parts of Water," &c.; I state it here again in a more
compendious form from his letter to Black (21st April, 1783), as the text
of the following remarks: ''When quite dry pure inflammable air and
quite dry pure dephlogisticated air are fired by the electric spark in a
close glass vessel, he [l^riestley] finds, after the vessel is cold, a quantity
of water adhering to the vessel, equal, or very nearly equal, to the weight

of the whole air Are we not then authorised to

conclude that water is composed of dephlogisticated and inflamn^able
air r**

One, then, of the elements of water, according to Watt, was dephlo-
gisticated air, by which it is not disputed that he signified what we now
name oxygen. The other was " inflammable air," and we are now to
consider whether he denoted by that gas, hydrogen. As this question,
however, is of the greatest importance, and cannot be discussed without
digressing from the direct patk in which our nuun inquiry lies, I shall*
devote a separate section to its consideration.

7. On the siffnificaiion of the term Injlammable Air as used by Watt,
to denote the combustible element of Water.

When Arago revived the Water Controversy, he thought it so certain
that Watt signified by inflammable air, hydrogen, that he considered him-
self at liber^ to substitute the one term for the other, and did so in his
Eloge of Watt. A large part, accordingly, of Mr. W. V. Harconrt's
inaugural address at the meeting of the British Association in 1889, was a
protest against the liberty thus taken as not consistent with the facts of
the case. In 1840, Arago sought to vindicate himself from the charge,
by pointing out that he had not given Watt any unfair advantage over
Cavendish by the change of words he had made, inasmuch as he had
substituted 'hydrogen' for 'inflammable air,' when referring to the latter's
views, as well as when discussing those of the former. This explanation,
however, was manifestly insufficient, and involved apetUio principii; for

• Wait Corr, p. 19.

i9B CAVENDISH AS A CHEMIST.

an important part of the question in dispute was, ^ Did Watt use tbe
word ' inflammable air' in the sense in which Cavendish employed it f **
The advocates of the latter showed that he defined his inflammable air aj»
that from zinc^ and in effect thej asked at the hands of the advocates of
his rival for evidence that Watt employed the word in the same sense.
Arasro thought it a sufficient reply to this request, to refer to an additi<Hi
which he had made to Lord Brougham s historical note which was printed
along with the Eloge of James Watt.

The statement was to the following effect " In 1784, the prepara-
tion of two permanent and very dissimilar gases was known. Some
called these gases» pure air and inflammable air; others, dephlogisticated
air and phlogiston; and lastly, others, oxygen and hydrogen."* If
Arago*s opinion, as stated in the quotation, be just, there is an end to the
Water Controversy; but his view cannot be substantiated. The term
'' hydrogen" was not used even by Lavoisier at the period when Watt and
Cavendish read their papers to the Royal Societv, and could not precede
in time the discovery of the compound nature of water. I shall set that
term therefore aside as irrelevant to the present discussion, and the signi*
fication of the word " phlogiston," in its widest acceptation, has airouijr
been adjourned to a succeeding section ; so that I am now to inquire,
first, whether Arago is ri^ht in affirming that inflammable air, and phlo-
giston in the sense of inflammable air, so certainly signified hydrogen io
] 783 and 1784, that Watt must be understood to refer by these titles to
that gas; and secondly, whether, as Lord Jeffrey urges. Watt did so
limit his use of phlogiston and inflammable air when describing his con-
elusions or the experiments from which they were drawn, that it cannot
be doubted that he denoted by both terms hydrogen.

It was not till the very close of last century, that ehemista
thoroughly awoke to the conviction that difference of property indicated
radical difference of substance, simple or compound, and became satisfied
that the various gases were not modifications of one air or gas, bat
specifically distinct bodies. This length, however, they had not got in
1783; and, accordingly, they included the whole of the combustible gases
known at that perioa under the one title of inflammable air, which
Arago conceives them to have applied only to hydrogen. That they did
not, however, is not difficult to prove. Priestley prided himself on having
clearer views as to the essential differences between the gases than any of
his predecessors, not excepting '* even Mr. Cavendish. "f Yet his defini-
tion of inflammable air, so late as 1790, was as follows: — The term I
inflammable air " sufficiently characterises and distinguishes that kind of
air which takes fire, and explodes on the approach of flame."]: I quote
this passage on account of its precision and brevity, and because Priestley's
" Observations on Air * were regarded as a storehouse of facts, to which '
Cavendish, and especially Watt, but in truth all the chemists of Europe, 1
had recourse for information; however cautious the wiser amongst them J
were, in discriminating between the value of the ingenious author's i
observations and his conclusions. 1

* Watt Corr. p. 252, and 263; or, HUtorical Note by Lord Brougham, appended
to French and English editions of Watt's Eloge; also Compiet Eendw, Jan. 1840,
pp. 109—111.

f Priestley on Air, abridged. Published in 1790. Vol. i. Introduction, p. 6. f

The whole of this introduction is worth perusal, in reference to the subject under i

discusfflon. j

X Op. cit. p. 8. I

THE WATER CONTROVERSY. 299

In the fifth of his six original Yolumes on air, which was published in
178], Priestley gives sammaries of all the ^Busts he had collected from his
own observations and those of others, in reference to the different airs.
Pages 335, 6, 7, and 8, are devoted to *' Facts relating to Inflammable
Air." These pages, consulted in their double character of index and
oammary, supply the most comprehensive account with which I am
acquainted, of what the substances were to which the chemists of the
eighteenth century gave the name of inflammable air. From Priestley's
'* Facts" we learn that that title was applied, 1. to hydrogen; 2. to sul-
phuretted hydrogen; 3. to various definite compounds of carbon and
nydrogen, such as marsh gas and defiant gas; 4. to combustible vapours,
such as those of ether and turpentine; 5. to mixtures of combustible gases
and vapours, such as coal gas; 6. to mixtures of combustible and incom-
bustible gases, which contained so much of the former as to be inflam-
mable, such as the gas from heated charcoal, consisting of carbonic oxide,
carbonic acid, and carburetted hydrogen. To all those elastic fluids
the name of " inflammable air" was given, and by Priestley's contem-
poraries as weU as himself. He did not introduce the name; on the
other hand he tells us, " I found the terms common or atmospherical air,
fixed air, and inflammable air, used by all philosophers, and no person
whatever had objected to them."*

1 have quoted chiefly from Priestley in illustration of the meaning of
the word inflammable air. Testimony to the same effect might be pro-
duced from his contemporaries. It will be enough, however, if I show
that Watt, Cavendish, Lavoisier, and Priestley, who were the parties
chiefly connected with the Water Controversy, aid not consider the word
inflammable air as necessarily synonymous with hydrogen. Watt and
Cavendish's views are fully stated a little further on. Of Lavoisier I
have only to say, that he most certainly did not, as Arago implies, iden-
tify hydrogen with inflammable air, but only with one combustible gas,
which he distinguished by the name of " aqueous inflammable air" (air
inflammable aqueux).t

It appears, then, that every known gas, vapour, or mixture of gases
or vapours (in a word, every elastic fluid), which was combustible in the
atmosphere, was called, in 1783, inflammable air. When the word,
therefore, occurs unqualifled in writings of that period, it signifies neither
more nor less than combustible gas. Hydrogen un^uestionaoly was called
by the chemists of the latest Phlogiston School, *' inflammable air;" bnt
the words are not convertible. The one was a generic, the other a specific
term. Hydrogen was inflammable air, but inflammable air was not
necessarily hydrogen. When, accordingly, the word inflammable air
occurs in a writing of the last century, the canon of interpretation is not
to settle summarily that it signifies hydrogen, but by a study of the con-
text to discover what combustible sas it does denote.

The chemists of the closing half of the preceding century were day by
day realising more clearly that there were different kinds of inflammable
air. But in 1783, not one of those various airs had been subjected to
analysis, so that the views of philosophers as to the nature of the differ-
ences among them were necessarily very vague. They seem, indeed, to
have oscillated between the conception of one elementary air, modified
variously by impregnation and mixture so as to become inflammable, or

* Priestley on Air, vol. ii. 1776, p. 334.

t Mimoires de VAcadhnie des Scieneea pour 1781, p. 468 (printed in 1784).
Reprinted in Wati Corr. p. 171.

300 CAVENDISH AS A CHEMIST.

rather perhaps of one inflammable air, mi 'generis^ bat in like manner
liable to alteration in properties; and the idea of specifically distinct
bodies.* The last view, however, is nowhere unequivocallj expressed,
and was not clearly apprehended, nor could it be, for the mind h&s no
satisfaction in dwelling upon differences, when it can do no more than
doubtfully realise that there are differences without apprehending their
nature or extent. f

Although the chemists of the Phlogiston School, howcTer, waived pre-
cision of definition as to the kinds of inflammable air with which they
were acquainted, they nerertheless distinguished certain of them by dis-
tinctive names. These had not relation, as our titles at the present dav
generally have, to the properties or composition of the gases; but to their
sources. Thus the carburetted hydrogen which rises from stagnant pools,
was '' the inflammable air of marshes^ — a name still retained % The
mixture of gases obtained by heating charcoal, was ^^the inftammabU air
of or from diarcoal.^^ Hydrogen, the body that concerns us most^ waa
'^ the inflammable air of or from the m^etals,^^ It received this name because*
according to the views of chemists before Lavoisier's time, when iron or
zinc dissolved in acidulated water, and hydrogen was given off, it was
tlie metal, and not, as we declare, the liquid which yielded the gas.
This name was introduced after Cavendish's paper on hydrogen in 1766.
He uses the term frequently; Priestley employs it constantly. Watt was
familiar with it, and introduces it in the paper containing his views con-
cerning the composition of water thus: "According to Dr. Priestley's
experiments, dephlogisticated air unites completely with about twice its
bulk of the inflammable air from metals **§

It does not seem necessary to adduce further evidence. The chemists
of 1783 and 1784 certainly did not, as Arago supposes, appropriate the
term 'inflammable air" to hydrogen; and Watt knew that they did not.

It remains to inquire in what sense he used the word '^ innammiable
air." Before doing so, I lay down the following rules for my own and
the reader's guidance : — 1 . When a chemist of the last century employs
the term "innammable air" without qualification or restriction, or any
reference^ in the context or otherwise, to its source or mode of prepara-
tion, he mu^t be understood to include under the title each of the com-
bustible gases or elastic fluids which he can be shown to have called
inflammakble air. 2. When a chemist of the last century is describing an
experiment with inflammable air, though he does not define the latter,
its nature may generally be learned by his account of the process by
which he prepared it. We shall know, for example, if it were hydrogen
which he made use of, by his stating that the inflammable air was
obtained by dissolving iron or zinc in diluted sulphuric or muriatic acid.

* See, in illostration of this, a very interesting letter from Volt a to Priestley (1776),
published by the latter, in his 3rd volume on Air (1777), App. p. 381.

t The justice of this remark will be appreciated, if it be remembered that the com-
bostible gases are analysed by oxidmng them, whilst, at the period of which I am
writing, the produ<:ts of combustion in oxygen were unknown. Till water was demon-
strated to be the oxide of hydrogen, there existed no means of proving that an inflam-
mable gas contained hydrogen; till carbonic add was shown to be an oxide of carbon,
the presence of carbon in tiie majority of the combustible gases could not be proved.
So far, indeed, was a distinct recognition of hydrogen as a specific substance from pro-
ceding the discovery of the composition of water, that it was this discovery that fur-
nished the means of distinguishing hydrogen from other combustible gases.

t PriettUy on Air, vol. iii. (1777), Appendix, p. 382.

§ Pkil. Tram. 1784, p. 349.

THE WATER CONTROVERSY. 301

8. If he desire to limit bis observations to a single kind of inflammable
air^ he will do so by stating its source. Thus he will describe or define
hydrogen as the inflammable air of or from the metals.

To what substances Watt applied the title of inflammable air, will
be understood from the following quotations. In a letter to Dr. Black
(3rd February, 1783), he mentions " that olive-oil, or oil of turpentine in
that earthen retort, produces very pure inflammable air."* In a letter to
Priestley (2nd May, 1783), he refers to 'Hhe inflammable air produced,
from spirit of wine and oils.**t In a letter to Gilbert Hamilton he states,
that " Dr. Priestley makes fixed air from dephlogisticated and inflam-
mable air in the following manner : — He takes mere, precip. ruber., which
yields only dephlogisticated air; and iron, which yields only inflammable
air, and heats them together. They produce only fixed air.*'t

The reference here is to certain experiments of Priestley's, in which
he conceived, that by heating pure iron filings he could make them yield
inflammable air. He describes his experiments thus in the paper
(" Kxperiments relating to Phlogiston, and the seeming Conversion of
Water into Air"), (m which Watt's letter was a commentary. " The
second article that I shall now mention, aflfords an indisputable proof of
the generation of fixed air from dephlogisticated air, and phlogiston or
inflammable air I was firing some shavings of iron in dephlo-
gisticated air confined by mercury, by means of a burning lens. In this
way I quickly fired the iron, and it burnt away in a very pleasing
manner. But what struck me most was, that of the air that remained, a
considerable portion was fixed air, though in the receiver I had nothing
but the purest dephlogisticated air, together with the iron which could

only give inflammable air Afterwards to put this hypothesis

concerning the constituent principles of fixed air, to a more direct proof,
I mixed iron filings, which gave only inflammable air, with red precipi-
tate, which I found to give nothing but the purest dephlogbticated; and
when I heated them in a coated glass retort, they gave a great quantity
of fixed air, in some portions of which, uineteen-tweutieths were sibsorbed
by lime-water; but the residuum was inflamnuible."§ From this account
it appears that the iron must have contained carbon, for it yielded
carbonic acid when heated with red oxide of mercury; and in all pro-
bability the inflammable air referred to, consisted of a mixture of com-
bustible gases (hydrogen, carburetted hydrogen, and carbonic oxide),
resulting from the action of the heated carburet of iron on moisture
entangled among the metallic filings. At all events, it could not be pure
hydrogen.ll

• Wait Corr, p. 14. t Watt Corr, p. 27.

t Watt Corr. p. 17. $ PhiL TVaiu. 1783, pp. 412, 413.

II Cavendish refers thus, to what he calls " Dr. Priestley's experiment of expelling
inflammable air from iron by heat alone." — " I am not/' he continues, '' suffideotly
acquainted with the circumstances of that experiment to argue with certainty about it ;
but I think it much more likely tliat the inflammable air was formed by the union of the
phlogiston of the iron filings with the water dispersed among them, or contained in the
retort or other vessel in which it was heated." (Phil. TVatu. 1784, p. 137, note.)

Priestley adopted Cavendish's views. In the 6th vol. of his Expt*. and Obs, on
Air, published in 1786, he refers to his having previously observed that " iron filings
in a gun-barrel, and a gun-barrel itself, had always given inflammable air whenever I
tried the experiment." (p. 88.) Induced, however, by what he says. Cavendish told
him, he observed matters more carefully, and noticed that ** iron filings are seldom so
dry as not to have moisture enough adhering to them, capable of enabling them to
give a considerable quantity of inflammable air ... . Being thus apprised of the influence

302 CATENDISH AS A CHEMIST.

In addition to the passages I hare quoted, in which the tenn {/n^iaM"
mable air is employed bj Watt, there are other places in which the word
phlogiston is used as s^onymous with that term. Although^ accord-
ingly, the full signification of phlogiston is not at present under discuB-
sion, it is necessary to refer to it where it signifies infiammable air.
Thus, in a quaint letter addressed by Watt to Mr. Fry of Bristol, &fter
referring to the composition of air and water, he says, '^ I will add the
receipt below for making both those elements.

''To make Water.

" R.— Of pure air and of phlogiston q.s., or if you wish to be yerj
exact^-of pure air, one part; of phlogiston, in a fluid form, two parts by
measure. Put them into a strong glass vessel which admits of being
shut quite close; mix them, fire them with the electric spark, &c. &c."*

Here, it cannot be doubted, that Watt signified by phlogiston, infiam-
mable air; and that he always uses the words as synonymous and inter-
changeable, is most strongly held as a cardinal point by the advocates of
Watt, from Arago to Muirhead. I have already quoted the statement of
the former. The latter, when defending Arago from the criticisms of
Berzelios, says, " We have adduced incontestable proof in no less than
eight distinct passages from Mr. Wattes own writings, besides those cited
from Priestley and others on the same point, of his having considered
phlogiston and inflammable air to be identical.'" And again, '' Both in his
paper on the constituent parts of water, and in his correspondence now
published, he repeatedly uses ' phlogiston* and ' inflammable air' as con-
vertible terms; and that, not by implication merely, but in the most direct
and distinct language in which his belief could be stated. *'t

The italics m the preceding quotation are Mr. Muirhead's, and from
both passages, as well as from Arago's statement, it will appear, that the
advocates of Watt themselves insist upon their client being understood
always to signify by phlogiston, inflammable air. Were I writing as a
partisan, I mi^ht content myself with their statement^ and insist upon the
conclusion which I am about to urge on the reader. It is not my object,
however, to avail myself of concessions made by the advocates of any of
the parties in the Water Controversy, unless those are consistent with my
own convictions, and, as Mr. Muirhead did not see the danger to his
client of his concession, and was not alive to the necessity of establishing
that Watt signified by phlogiston or infiammable air, hydrogen, but
supposed that the three names were identical in meaning, 1 cannot build
upon his admission ; especially since Lord Jeffrey acknowledges and
seeks to supply the defect in Mr. Muirhead's argument. After ul, how-
ever, Lord Jemrey is at one with Mr. Muirhead, in holding that Watt sig-
nified by phlogiston infiammable air, only he seeks to limit Watt's employ-
ment of it to the one inflammable air, hydrogen. The whole, then, of the

of uDperoeived moisture in the prodacdon of inflammable air, and willing to ascertain
It to my present satisfaction, I began with filling a gnn-barrel with iron-filings in their
common state, without taking any particular precaution to dry them, and I found that
they gave air, as they had been used to do, aild continued to do so many hours. I even
got ten ounce measures of inflammable air from two ounces of iron filings in a coated
glass retort. At length, however, the production of inflammable air from the gun-barrel
ceased; but on putting water into it, the air was produced again, and a few repetitions
of the experiment fully satisfied me that I had been too precipitate in concluding that
inflammable air is pure phlogiston.'' (p. 90.) Watt refers to those experiments, aa
they were originally tried, in the letter to Gilbert Hamilton. ( Watt Corr. p. 17.)
• Watt Corr, p. 24. + Op, cit, pp. cxiii and cxir.

TUB WATER CONTROVERSY. 303

advocates of Watt contend that he ased phlogiston as a synonyme for at
least combuatible gas. And in this view I believe that they are correct,
at least in interpreting Watt's direct references to the composition of water,
although sach is not Mr. Harcoart's view. One consideration, however,
is sufficient to show that the opinion attributed by Mr. Muirhead to Watt,
was really held by him. His conclusion was certainly drawn from experi-
ments in which inflammable air and dephlogisticated air were employed,
and it represented water as consisting of those two gases. When, there-
fore, he used the word phlogiston to indicate one of the gases of water,
he must have signified by that term either inflammable or dephlogisticated
air, and as he certainly did not intend the latter, for which his synonyme
was pure air, he must have intended the former.

Thus much, then, premised, I quote the following passage, which is
one of the most important statements in Watt's writmgs in reference to
the combustible elements of water. It has not in general been referred
to by his own advocates. He is discussing in the close of his '' Thoughts
on the constituent parts of Water,'* the Question, how much heat is evolved
when phlo^ston is combined with dephlogisticated air, and refers in illus-
tration to certain experiments of Lavoisier and La Place made with sulphur,
phosphorus, and charcoal. After stating the nature of their trials, he
continues, " The weight of the ashes of an ounce of charcoal is very in-
considerable ; and by some experiments of Dr. Priestley's, charcoal, when
freed from fixed air, and other air which it imbibes from the atmosphere,

is almoit wholly CMivertihle into phlogiston It is also worthy of

inquiry, whether all the amazing quantity of heat let loose in these ex-
periments was contained in the dephlogisticated air ; or whether the
greatest portion of it was not contained m the phiogiUon or infiammable
air."* From this passage, I think it indisputobly appears Uiat Watt
applied the term phlogiston in the sense of inflammable air, to a com-
bustible gas into which Priestley had shown (or appeared to have shown),
charcoal could be converted. How important this reference of Watt's is,
in connexion with the much criticised charcoal-gas experiments of Priest-
ley, from which the former is alleged by the advocates of Cavendish to
have drawn his conclusion, I need not insist. To that I shall recur.

At present, it seems only necessary to notice, that although Watt does
not specially refer to the nature of the experiments by which Priestley
effected the conversion referred to, there cannot be any reasonable doubt
as to what the experiments were. They have already been noticed
incidentally as contained in Priestley's paper on " Phlogiston and the
seeming conversion of water into air," which Watt's " Thoughts," in
their original epistolary form, were intended to accompany. The fol-
lowing passage from the first section of Priestley's paper, which is solely
occupied with experiments illustrating that phfogiston and inflammable
air are the same thing, will sufficiently demonstrate what the process was
by which charcoal was converted into pbloffiston : " I shall conclude these
observations on phlogiston with two articles, one of which seems to con-
tradict an established maxim among chemists, and the other a former
opinion of my own.

"" It is generally said that charcoal is indestructible, except by a red
heat in contact with air. But I find that it is perfectly destructible, or
decomposed in vacuo, and by the heat of a burning lens almost wholly
converted into inflammable air ; so that nothing remains besides an ex-

* Phil. Ttans. 1784, pp. 351, 352.

304 CAVENDISH AS A CHEMIST.

oeedindj small qnantitj of white asbes, wbich are seldom visible, except
when in very small particles they happen to cross the sunbeami as ihej
flj aboat within the receiver. It would be impossible to collect or weigh
them ; but, according to appearance, the ashes thus produced from many
pounds of wood, could not be supposed to weic^h a grain. The great
weight of ashes produced by burning wood in the open air, arises from
what is attracted by them from the air. The air which I get in this
manner is wholly inflammable, without the least particle of fixed air in
it. But in order to this the charcoal must be perfectly well made, or
with such a heat as would expel all the fixed air which the wood contains,
and it must be continued till it yield inflammable air only, which in an
earthen retort is soon produced."* The succeeding paragraph, which
refers at length to the heating of charcoal in retorts, has already been
given.

Priestley republished his account of these experiments in 1786, but he
added the following important note, as a qualification of his original
statements. ''Notwithstanding iheae facts, it will appear from my bubse>
quent experiments, that water was necessary to the formation of this
inflammable air from wood as well as of that from tron.*'t

Altogether, then, it appears, that Watt applied the term inflammable
air, or phlogiston as a synonynie of it, to 1. Oil gas ; 2. Light carburett^d
hydrogen ; 3. A gas obtained by heating moist iron containing carbon ;
4. A gas obtained by heating charcoal ; and it may be noticed further
that in the only passage of his paper in which he directly refers to hydro-
gen, he deflnes it as ''the inflammable air from metals.**^

There were thus five gases or gaseous mixtures, which were severally
styled by Watt, inflammable air or phlogiston, and there was one of those
five, which, like his cod temporaries, he distinguibhed as the inflammable
air from metals. It cannot, therefore, be conceded to his advocates, that
when he uses the term "inflammable air" without any qualitication,
he signifies only that one among the five which we now name
hydrogen.

I may now, therefore, take for granted, that it is incumbent on
the advocates of Watt to show which of his hve inflammable airs he
regarded as the combustible element of water ; if in reality he assigned
to one a preference over the rest, as an ingredient of that liquid. I pass,
therefore, to the arguments by which Lord Jeffrey seeks to snow that the
inflammable air specially referred to by Watt as an element of water was
hydrogen. "It seems admitted," says his Lordship, " that if Priestley had
made, and shown, or reported to Watt, other experiments with the proper
hydrogen, which might certainly have given the results which he specifies,
there would have been nothing to say against the accuracy, any more than
against the originality of that conclusion. Now, looking at the whole of
the evidence before us, we have come to be satisfied that Priestley Aac^, in
point of fact, made and shown, or reported to Watt, such other experi-
ments ; and that, though it may be somewhat difficult to account for some
expressions which he uses in speaking of his charcoal experiments in that
paper, it would be immeasurctbly more difficult to believe that there were
no other experiments with hydrogen, and that those two gifted individuala

• Phil. TVflfM. 1783, p. 411.

t Prieutley on Air, vol. tI (1786). p. 24. The italics are the author's own. He
gives a full account of the experiments (tried at Cavendish's suggestiou), which led
him to change his opiuion, in the same vol pp. 87 — 90.

X PkU. TraiiM. 1784, p. 349.

THE WATER CONTROVERSY. 305

were the dupes and victims of an hallucination without parallel or
precedent in the history of the human understanding.'** .

Lord Jeffrey then proceeds to adduce various reasons for helieving that
there were other experiments made with hydrogen. " First of all, the
whole series was professedly entered on as a mere repetition of those of
Cavendish, which were made exclusively with that substance ; and it seems
inconceivable that, when the main object was to test their accuracy, he
should not have begun at least, with the same materials."f It might
suffice as reply to this argument, to notice, that it is not easy to see why,
if Priestley made preliminary trials with hydrogen, he should not have
mentioned the fact ; whilst on the other hand it cannot surprise us, that one
who thought inflammable air from charcoal preferable to that from metals,
because it was drier, should have thought it needless to use the latter.
But it is of more importance to notice that Priestley did not try the ex-
periments under consideration, to test Cavendishes accuracy t ^ut to convince
himself that air or gas (gases) could be converted into water ; and that
neither Priestley nor Cavendish refers to the former's experiments as mere
repetitions of the trials of the latter. Priestley explicitly announces a
device of his own (namely, the substitution of the charcoal-gas for hydro-
gen), which was intended to render the experiment more crucial.]: And
Cavendish does not refer to Priestley's experiments as identical with his
own, but only as ''of the same kind," and points out as important, the
fact of Priestley's ''having used a different kind of inflammable air,
namely, that from charcoal, and perhaps having used a greater proportion
of it." There seems then, no antecedent probability that unrecorded
experiments with hydrogen were made by Priestley. It is possible, not*
withstanding, that such may have been made, and the following are the'
reasons assigned by Lord Jeffrey for believing that they were. I quote
his Lordship's six arguments in full, before commenting on any of them.
" First, as early as March 26th, 1783, Priestley had told him [Watt] that
' he put dry dephlogisticated air and dry inflammable air into a close vessel,
and fired them by electricity, that no air remained when both were pure^
bat that he found on the sides of the vessel a quantity of water, equal in
weight to the air consumed.' Now, this is the very experiment, shortly
recited, from which, a few weeks after, Watt intimated that he had drawn
his famous conclusion, and we have now only to ask whether these results,
or anything at all like them, could have been produced if the gas from
charcoal, or anything but hydrogen, had been the inflammable air
employed 1

" Secondly, on the 2 1st April, Watt writes to Dr. Black and to Priestley
himself, informing the one and reminding the other, that he [Priestley],
after firing the dephlogisticated and inflammable air as above, and opening
the close vessel over mercury, found that the mercury rose and filled the
vessel * to within one two-hundredth part of its whole contents,* and that
there was a quantity of water equal or nearly to the weight of the whole
air employed.

" Thirdly, that sometime before 28th April, Priestley had also told him
that, in order to form water, ' you should take of pure or dephlogisticated
air one part, and of phlogiston (or inflammable air) two parts by measure,

and fire them by the electric spark.'

1 . ,

. * Edinr. Rev. Jan. 1848, p. 94.
"I* Op. et loe. eit,
t See, in illustration, Priestley's account of his ezpeiimentsr quoted in full, p. 284.

306 CAVBKDISH AS A CHEMIST.

'' Foarthlj, tbat before IStli May, Prieetlej had also told him < titat die
water remaining after the explosion U not in the least ocidL*

'« Fifthly, that Priestley had also told him before the 2i8t April, that
by heating the calcee of metals with ' inflammable air/ tbey were rednoed
to the metallic state, the air being absorbed (or disappearing) ao com-
pletely, ' that only two parts out cf one hundred and one remained at tie
end of the operation;' from which he had inferred that this * inflammable
air wot the thing called phlogiston' Now, it is certain, from the detailed
accoant of this very experiment, giren by Priestley himself, in his paper
read to the Royal Society, that the inflammable air there used was the
proper hydrogen; be expressly describing it as ' air extracted Jrom iron
by oil of vitriol; and it was this inflamnutble air, therefore, and nothing
else, that he and Watt were led by this very experiment to consider as
'the thing called phlogiston.'

''Bat, sixthly, in Watt's own paper, given into the Society in Noyember,
1783, and subsequently printed in their Transactions, he distinctly states
that ' according to Dr, Priestley's experiments dephlogisticated air unites
completely with about twice its bulk of the inflammable air from metaU —
the inflammable air being supposed to be wholly phlogiston* Now this ia a
separate and distinct experiment from that in which the calces of metals
were reduced by the same agent; and is, therefore, a second and addi-
tional proof what sort of inflammable air both these philosophers consi-
dered as identical with phlogiston, and on accoant of what properties tbey
so considered it." A passage follows which for brevity's sake I omit^ in
which Lord Jeffrey concedes that in Priestley's experiments on the redac-
tion of metallic oxides by hydrogen, he did not observe that water was
jproduced. "But," continues his Lordship, "in the second experiment
where both airs had been carefully pat together in the proper proportions
for forming water, and were found to 'nnite completely,' or be mutually
absorbed, it seems impossible to doubt that the formation of water mnst
have been expected, and consequently observed : and, accordingly, though
very briefly recorded, we find that it was so; for, in the very same para-
graph, and at the distance of only four lines from the words we have
cited, the learned author proceeds : — " Therefore one ounce of dephlogisti-
cated air will require 120 grains of inflammable air, or phlogiston (that
is, unequivocally, of hydrogen) to convert it into water***

In the preceding quotations Lord Jeffrey has, for brevity's sake, in
one or two places given the signification of Priestley and Watt's state-
ments, without adducing their very words. If taken, however, exactly
as they occur in their writings, none of the passages adduced will be
found to contain an explicit reference to hydrogen as the substance which
Watt intended to be understood when he spoke of inflammable air, or
phlogiston, as a constituent of water. Neither do the whole taken to-
gether warrant this inference. The argument based on the first and
second passages referred to, amounts simply to this; that since Priestley's
experiments are incredible unless we suppose him to have used hydrogen,
it should be conceded that he did employ this gas. This argument mieht
have some weight if Priestley had given no account whatever of his
experiments, and we had no means of judging of their nature except by
learning their alleged phenomenal results. But when he deliberately telJa
us that he used charcoal-gas, and in effect reiterates the statement years
after the Water Controversy commenced; and when he even repeats the

* Sdinr, Rev, /an. 1848, pp. 99, 96.

THE WATER CONTROVERSY^ 307

experiment^ as I shall presently show he did, with the inflammable air
from charcoal,* instead of that from metals, in spite of all that Caven-
dish, Watt, and Lavoisier had published in reference to the composition
of water, it is impossible, in the face of his own statements, to affirm that
he most have used hydrogen. In truth, even if it were certain that he
had used that gas, it would only render the charcoal-gas experiments the
less credible; for it would be still more difficult to believe that one who
had accurately observed the results of detonating a mixture of oxygen
and hydrogen should err as Priestley did with the charcoal-gas, than it is
to understand the mistakes which he committed whilst limiting his
attention solely to the latter.

Lord Jeffrey attaches importance to the fact, that Watt twice states
that if the vessel in which the gases were burned, was opened under
mercury after the explosion, the liquid rose into the vessel, and almost
entirely filled it. That the account of an experiment, however, which
is certainly erroneous in three particular8,f should err in a fourth, cannot
much amaze us. If Priestley could deceive himself, as he certainly did,
in reference to the three points of most importance in his repetition of
Cavendish's experiments, he might well be misled in a particular which
he did not deem of such interest as to deserve mention by himself
at all.

Moreover, it is not impossible that he may have obtained certain of
the results he describes with the charcoal-gas. In the passages to which
Lord Jeffrey refers, Watt states that " if the vessel was opened with its
mouth immersed in water or mercury, so much of these liquids entered as
was sufficient to fill the glass within about i^Trdth part of its whole
contents."}:

The phenomenon here described might occur with the eharooal-gas so
far as water is concerned, if the gas were detonated with a volume of
oxygen, exactly or very nearly sufficient to convert it into a mixture of
carbonic acid and water, both of which would dissolve in the water
admitted into the vessel, and allow the liquid to fill it completely. With
mercury this could not happen in any conceivable circumstances, so that
we must either discredit the account altogether, or may suppose the
experiments to have been made with hydrogen. I am willing to accept
eitner alternative. It is of the greatest importance, however, to notice
that the experiments in which the vessel was opened under the surface of
a liquid, must have been distinct from those in which the water produced
was examined as to its purity, and as to its equality in weight with the
mixture of gases which had been burned. Priestley's method of pro-
cedure, unlike Cavendish's, rendered three distinct sets of experiments
necessary. (1). One in which the water was absorbed by blotting-paper,
and its weight ascertained; (2) a second set (which perhaps was not
made) in which the water was analysed as to its purity; and (3) a third,
in which the total conversion of the burned gases into liquid was ascer'»
tained by opening the mouth of the vessel in which the explosion had
taken place, under water or mercury.

It )s possible that experiments of the three different kinds indicated,
were performed with inflammable air prepared in the same way. In

* PrUiile^ on Air, vol. vi. p. 126.

t Namely, in asserting that charcoal-gas and oxygen (I) yield water only when
burned together; (2) yield a weight of water equal to that of the gases burned; (3) and
do not yield carbonic acid.

t PhU. Trmu. 1784, p. 332; also Watt Corr. p. 10.

308 CAVENDISH AS A CHEMIST.

Priestley's own account, however, only the first set of experiments ii
descrihed, and we are left to conjecture, and to Watt's allusions, to dis-
cover how the second and third series were performed. It is not impro-
bahle that the same inflammable air may not have been used in each
series of experiments. We have Priestley's authority for aflSrming that
the inflammable air from charcoal was preferred for the experiments, by
means of which he sought to establish that there was equality of weight
between the amount of gases burned and of water produced. On the
other hand we have no information whatever as to what inflammable air
was used when the production of nothing but pure water seemed to be
proved. Even, therefore, if we should concede, on the ground of the
incredibility of the account on any other supposition, that hydrogen was
employed in the experiments where the vessel was opened under mercniy
and under water, it does not follow that the much more important
experiments in reference to the weight and purity of the water, on which
Priestley chiefly relied, and which alone he published, were also made
with that gas. Without further comment, therefore, I leave the reader
to decide for himself, whether the alleged rise of the mercury so as com-
pletely to fill the globe after each explosion, is an additional inaccurate
statement in an account otherwise in many respects incredible, or a
trustworthy report of a collateral experiment, in which hydrogen was
employed.

Lord Jeffrey's third argument is founded on the quaint recipe " to
make water," given by Watt to Mr. Fry, of Bristol, and already quoted
in greater part (ante, p. 300,)* In this he tells his correspondent to take
^' of phlogiston, in a fluid form, two parts by measure," a direction which
his Lordship thinks Watt himself received from Priestley. If this suppo-
fiition be well founded, we must inquire, not what Watt but what Priestley
signified by phlogiston, and it is quite certain that the latter did not
restrict the term to hydrogen (ante, p. 97), but used it as sjmonymous
with inflammable air, his definition of which as a gas {any gas or gaseous
mixture) combustible in air, has already been given. I apprehend, how-
ever, that Watt's recipe is rather the general expression, in the shape of a
formula, of his widest view concerning the nature of water, than the mere
report of information given him by Priestley. If this be the true view, the
passage is of no importance to the question under discussion, for it
contains no statement or allusion by which we can discover that Watt
signified by phlo^ton one combustible elastic fluid rather than another,
or that he specially intended hydrogen.

Lord Jeffrey's fourth reason for aflirming that Priestley made experi*
tnents with hydrogen, is, that Watt reports that the water, which resulted
from the combustion of hydrogen, was not in the least acid.f His
Lordship frequently refers to this, and even urges that it proves that
Priestley's experiments were better than Cavendish's, since the latter was
seldom able to get pure water by the combustion of hydrogen and
oxygeu.t Cavendish, nevertheless, knew quite well how to obtain pure
water from hydrogen and oxygen, and has told us how to secure its pro-
duction. § It is quite true, however, that in the majority of his experi-
ments on the detonation of mixtures of pure hydrogen and oxygen, the
water was acid; and the fact is one of great importance in reference to
the question before us. Paradoxical as it may appear, the alleged

* fffl// Corr. p. 24. f Watt's letter to De Luc : ITaW Cferr. p, 30.

t Bdinr. Rev. p. 100. $ P/dl. JVans, 1784, p. 138.

THE WATER CONTROVERSY. 309

purity of the prodaced water in Priestley's experiments, instead of
justiiying the inference that he employed hydrogen, leads to the very
opposite conchision. For although pure hydrogen and oxygen can yield
nothing bat pure water^ when they combine, it is very difficult to exclude
from one or both gases a little atmospheric air, the nitrogen of which
bums along wi|h the hydrogen, when the latter is detonated with oxygen,
(at least if the oxygen is present in a quantity exceeding its combining
measure), and , produces nitric acid. Priestley seems to have detonated
the gases in a glass tube with wires inserted into its sides, resembling the
electric pistol or Volta's eudiometer.* He does not appear to have
emptied this tube at the air-pump as Cavendish did his globe, but to have
filled it with mercury which he displaced by the gases, afterwards detonated
in the eudiometer. This method of procedure, as well as Cavendish's,
rendered inevitable the presence of nitrogen in the apparently pure
hydrogen and oxygen. And accordingly we find that Y^-gdth part of the
gases escaped condensation; " which remainder," observes Watt, " is
phlogisticated air [nitrogen] probably contained as an impurity in the
other airs."t

The conditions necessary to the production of nitric acid were thus
certainly realised in Priestley's experiments, and although it is not abso>
lutely impossible that in a solitary trial pure water may have been ob-
tained, it is in the highest degree improbable that a series of experiments
can have been made with hydrogen and oxygen, in the way Priestley
operated, without nitric acid being detected. It appears, accordingly,
that so soon as Priestley's attention was directed to the fact that pure
Tirater is not the invariable product of the combustion of hydrogen and
oxygen, he not only confirmed Cavendish's results, but went the length of
affirming that acid water is always produced by the combination of
the two gases. The following quotations from his later writings will
speak for themselves. The italics are his own. " Having nerer failed,
when the experiments were conducted with due attention to procure some
acid whenever I decomposed dephlogisticated and inflammable air in close
vessels, I concluded that an acid was the necessary result of the union of
those two kinds of air, and not water only."]:

" There is, therefore, no source of the nitrous acid which I find on the
decomposition of dephlogisticated and inflammable air, besides the union
of those two kinds of air, which therefore do not make mere water ^ as
the antiphlogistians 6uppose."§ These statements were criticised by Ber-
thollet and others, who did not deny the (apparent) facts, but objected to
the conclasions drawn from them. Their objections, however, made no
impression on Priestley, who when he republished his observations in
1790, reiterated still more strongly his previous statements. " I muH
say, as I did when I was myself a believer in the decomposition of water,
that I have never been able to find the full weight of the air decomposed
in the water produced by the decomposition ; and that now I apprehend

* Phil. Tratu, 1784, p. 331. t Watt Corr. p. 19.

t Abridg. qf Priestleg on Air, toI, iii. (1790), p. 54. The paper is entitled,
" On the Composition of Spirit of Nitre from dephlogisticated and inflammable air,"
and is reprinted from Phil. Trans, vol. Ixxriii. It contains a lengthened account of
experiments, like those of Cavendish, but chiefly made in a copper vessel, which was
corroded by the acid, and yielded, after several explosions within it, a marked quantity
of nitrate of copper. The inflammable air used in those experiments was hydrogen,
prepared by passing steam over red-hot iron.

$ Op. cit. p. 63. See also the extracts from the Wedgwood CorresfMmdence, ante,
pp. 97-103.

310 CAVENDISH AS A CHEMIST.

it will not be denied, that the produce of this decomposition is not mere
water, but always some acid.'**

The non-appearance of acid, then, in Priestley's original experiments,
instead of justifying the inference that he employed hydrogen, forbids
the belief that he did. Nor most it be forgotten that CaTendish explicitly
pointed out that the probable cause of Priestley's finding no acid, was his
''haying used a different kind of inflammable air, namely, that from
charcoal, and perhaps haying used a greater proportion of it."t PriesUey
was thus taxed with having used the wrong materials when he found no
acid, and never denied the justice of the accusation ; whilst, on the other
hand, in the only experiments where he certainly employed the right ma-
terials, he found abundance of acid produced, as he was the first himself
to declare.]: It will appear in the sequel, that Monge and Lavoisier also
procured acid water when they experimented as Cavendish did.

Lord Jeffrey's fifth argument refers to certain experiments of Priest-
ley's, in which he reduced metallic oxides by heating them in an atmosphere
of hydrogen by means of a burning glass, which concentrated the sun's
rays on the oxide. As hydrogen was certainly employed in these expe-
riments, and as both Priestley and Watt call it phlogiston, his Lordship
seeks to show that " it was this inflammable air, therefore, and nothing
else, that he and Watt were led by this very experiment to consider as
'the thing called phlogiston.'" I might repeat, in reference to this
opinion, that whatever was Watt's view concerning phlogiston, it is quite
certain that Priestley applied that term to many substances besides hydro-
gen. It is enough, however, to determine what Watt's employment of
the word was ; and I have but to remark here, that the passages adduced
only show that he termed hydrogen, phlogiston, not that he confined that
term to the single gas in question. In truth, if Watt's own words are
taken, they will be found at variance with Lord Jeffrey's interpretation of
them. Watt does not say that he inferred " this inflammable air,"
(namely, a special combustible gas, which he defined) — to be phlogiston ;
nor does he even refer to it as " the inflammable air," so as to limit his
inference to the particular gas which Priestley employed. His words are :
'' He [Priestley] found, that by exposing the calces of metals to the solar
rays, concentrated by a lens, in a vessel containing injlammable air only,
the calces of the softer metals were reduced to their metallic 8tate."§
The words, infiammahU air are three times repeated without limitation or
definition in the succeeding sentences, and the final conclusion is " that
inflammable air must be the pure phlogiston, or the matter which reduced
the calces to metals. "|| To the same effect Watt writes to Black that
^' by reducing metals in inflammable air, he [Priestley] finds they absorb
it, and that the residuum of ten ounces out of the hundred, is still the same
sort of inflammable air ; therefore inflammable air [not this inflammable
air] is the thing called phlogiston."ir

* Ahridg. of Priestley on Air, vol. iii. p. 555.

t Phil, Trans. 17«4, p. 135.

X Priestley's charcoal-gas, when detonated with oxygen, would yield, as we have
already seen, carbonic acid. As water, however, dissolves only its own Tolome of that
gas, the quantity of the latter present in the liquid produced at each explosion, would be
much too small to give it an acid taste, or to invest it with the power of reddeniog
vegetable blues. Had Priestley suspected the presence of fixed air, he would have
tested for it with lime-water.

$ PkU. Trans. 1784, p. 331.

{| Op. et he. cit.

%' Walt Corr. p. 19,

THE WATER CONTROVERSY. 311

It 10 quite true, that on consultiog Priestlej^s account of bis reduction
experiments, we find that hydrogen was the gas employed ; but so little
importance does Watt attach to this fact, that he never once mentions it
in his paper, or throaghout his correspondence. It seems to me therefore^
that Watt's statements warrant a conclusion exactly the opposite of that
which Lord Jeffrey thinks they justify, viz. that provided pure inflammable
air waa employed, it was unimportant from what source it was obtained.
On this, however, I do not insist. It is enough to affirm that the passages
quoted do not prove a limitation by Watt of the term phlogiston to hydro-
gen, but imply a use of it, as synonymous with inflammable air.

Lord Jeffrey's last argument is oased upon the only passage in which
Watt explicitly refers to hydrogen in connexion with the production of
water. The allusion, however, as his Lordship acknowledges, is only inci*
dental, and the passage does not occur in Watt's original letter. It is,
nevertbeless, the most important statement in his writings referring to
hydrogen as an element of water. I shall therefore quote it in full.

" According to Dr. Priestley's experiments, dephlogisticated air unites
completely with about twice its bulk of the inflammable air from metals.
The inflammable air being supposed to be wholly phlogiston, and being -^^^
of the weight of an equal bulk of dephlogisticated air, and being double in
quantity, will be ^.y of the weight of the dephlogisticated air it unites with.
Therefore one ounce (576 grains) of dephlogisticated air will require 120
grains of inflammable air, or phlogiston, to convert it into water. And
supposing the heat extricated by the union of dephlogisticated and inflam-
mable air to be equal to that extricated by the burning of phosphorus, we
shall find that the union of 120 grains of inflammable air with 576 grains
of dephlogisticated air, extricates 9265® of heat."*

The passage just quoted, which occurs near the close of the latest
▼ersion of Watt's " Thoughts," is not the record of observations made by
Priestley, except so far as the combining measure of hydrogen and oxygen
is concerned. Lord Jeffrey seeks to show that that could not have been
discovered without the production of water being simultaneously observed,
but this cannot be conceded. I have sought for Priestley's own earliest
reference to the fact that hydrogen combines with half its bulk of oxygen;
bat I have not been able to find it, and I cannot in consequence offer any
precise criticism as to the nature of his experiments.

There are several references, however, in Priestley's later papers to
the combination of hvdrogen and oxygen in the proportions mentioned by
Watt ; but, singularly enough, he refers to his experiments not as original,
bat as resembling those of Lavoisier; and he further affirms that water was
not always produced when hydrogen and oxygen were detonated together.
^ I also,' says he, " procured water when I decomposed dephlogisticated
and inflammable air from iron, by the electric spark in a close vessel,
which is an experiment similar to those that were made by Mr. Lavoisier
at Paris. t I put 3.75 ounce measures of a mixture of air, of which one-
third was dephlogisticated, and two-thirds inflammable from iron, in the
close vessel ; and after the explosion I found in it one grain of moisture.
* * * * But repeating this experiment with half as much dephl<^
gisticated as inflammable air, / eotUd nU perceive any water after the ex-

* PAi7. TroHB. 1784, p. 349—350.

t This alliuion is rather ohscnre. LaToisier's recorded experiments were made by
direct kindlmg of the gases, which burned tranquilly togetberi not by exploding them
by the electric spark. This was the method of Cavendish and Monge, and indeed had
been employed by Priestley himself, before eidier of those observers used it.

312 CAVENDISH AS A CHBMIST.

pcriment."* Here, then, is exactly tbe case Lord Jeffi'ey rapposee. The
proper materials for the production of water were mingled in the proper
proportion, and the appearance of water was watched for, and yet no water
was seen. Nor is this a solitary statement of Priestley's. (Ante, p. 97.)
In the same paper he asserts that he had never "heen aole to procure any
water when he revived mercury from red precipitate in inflammable air,
or at least not more than may oe supposed to have been contained in the
inflammable air as an extraneous substance ;"f and he makes simifair
statements in reference to the reduction by hydrogen of the black oxide of
mercury, and the oxide of lead.

It IS quite certain, moreover, as Lord Jeflrey acknowledges, that in
Priestley's earlier reductions of metallic oxides by hydrogen (1783),
where much larger quantities of the gas were employed than are likely
to have been used in his detonations, he altogether overlooked the prodofi- '
tion of water, nor did he attach any importance to its appearance in his ,
experiments along with Warltire in 1781. From the Wedgwood Corrt- \
spondence it also appears (ante, p. 97), that Priestley deliberately asserted
that hydrogen and oxygen may be exploded together in their combining
proportions, and yet not produce water. I feel it impossible, therefore,
to concede that when he detonated two measures of hydrogen and one of
oxygen, he must certainly have witnessed the formation of water. The
eudiometers he describes m his " Experiments and Observations on Air,'*
were small, and could contain only a few cubic inches of the mixed gases, |
which would yield on detonation but one or two grains of water. This I
might easily escape observation, even if the experiment were made in a
shut vessel, as the steam produced would not condense till the vessel «
cooled. In truth no one, I thiuk, who has perused the endless con-
tradictory statements regarding the products of the combustion of inflam-
mable air, which are scattered through Priestley's volumes, will feel
disposed to credit him with having seen anything which he does not
explicitly affirm that he did see.

If Watt's statement, moreover, be looked into closely, it will be found
to be quite hypothetical, except in reference to the combining measure of i
hydrogen. His object was to calculate the amount of heat evolved during
the combustion of inflammable air, on the assumption that it is equal to
that *' extricated by the burning of phosphorus." To make this calco- \
lation it was necessary to know the relative weights of equal volumes of
inflammable air and oxygen, so as to determme what weight of the |
former would unite with an ounce (by weight) of the latter, when the '
two gases produced water by their combination. The inflammable air
from the metals, or hydrogen, however, was the only combustible sas
whose density had been ascertained,! and was the only one, accordingly, ,
which could be made the basis of such a calculation as Watt pursued. ^
For this reason he refers specially on this solitary occasion to a particular
inflammable air as the only one he had in view, and that not with direct
reference to the production of water, but as rendering probable the con-
clusion ''that the union of 120 grains of inflammable air with 576 grains
•of dephlogisticated air, extricates 9265"* of heat."

That tbe known (or supposed) density of hydrogen was the main
<»iu8e of Watt's special reference to it, appears further from the continua-
tion of his argument. He is discussing the question whether the phlo- ,

♦ Priestley on Air, vol. ri. (1786) pp. 126, 127.

+ Op. cit. p. 128.

X By Cavendish, PAiV. Trans. 1766.

THE WATER CONTROVERSY. 313

giston of cbarcoal gives ont as mach heat as phlogiston in the form of
inflammable air does ? The facts on which he reasoned seemed to show
*' that the union of phlogiston in different proportions with dephlogtsti*
cated air does not extricate proportional quantities of heat/* For this ho
accounts as follows : — " This difference may arise from a mistake in sup-
posing the specific gravity of the inflammable air Dr. Priestley employed
to have been only -^^r of that of dephlogisticated air; for if it be supposed
that its specific gravity was a little more than one-eighth of that of the
dephlogisticated air, then equal additions of phlogiston would have pro-
duced equal quantities of heat :* this matter should therefore be put to
the test of experiment, by deflagrating dephlogisticated air with inflam-
mable air of a known specific gravity," &io.^ The concluding reference
shows that in defect of better data, Watt referred to the inflammable
air from metals, as the combustible gas whose density was best known,
and the only one therefore in reference to which a calculation founded on
weights could be based.

The only passage, then, in "^hich Watt names hydrogen, is secondary
to the main subject of his " Thoughts," to which it was an addition ma^
some seven months after their first issue. It does not profess to record
experiments on the production of water performed with hydrogen, but
founds on its supposed density an infereuc-e as to the amount of heat it
should evolve when it burns. The statement, accordingly, may most
justly be referred to as proving that hydrogen was a gas (one of the
gases) which Watt denoted by inflammable air, but it cannot serve to
demonstrate that it was t/ie only gas to which he gave that name. The
continuation of his paper, in the passage quoted, is the best evidence that
be gave other gases the same title, for he proceeds without interruption
to refer to charcoal, and in the course of his argument states, as already
mentioned, that it is almost wholly convertible into phlogiston, or injlani-
Tillable air,

I have now to refer to evidence of another kind, illustrating the
nature of the materials which were made use of in the experiments from
which Watt's conclusions were drawn. Priestley, Cavendish, and Watt,
all had occasion to reconsider the data upon which their inferences con-
cerning the composition of water were founded. The two first, some
time after they had given their opinions to the world, commented upon
the original publication of their views; and the third added notes to his
** Thoughts*' before they were printed. Something, accordingly, may be
learned, both from what they said and from what they left unsaid, in
illustration of the point under discussion. I begin with Priestley. His
statement is of great importance; for whatever view may be held as to
Watt having read Priestley's paper before he addressed his first letter to
him, there is not and cannot be any dispute as to Priestley having read
Watt's letter, and knowing exactly to what extent the letter referred to
his experiments. Yet when Priestley returned to the subject of the com-

* Where Watt or Priestley got those numbers does not appear. The want of any
precise statement on this point, adds another element of yaguene&'s to the imperfect
record we possess of Priestley's experiments. If we are to understand that the latter
had ascertained, by direct trial, that the inflammable air he employed in his repetition
of Cavendish's experiments was only 8 or 9 times lighter than oxygen, we could hare
no better proof that hydrogen vaa not one of the gases made use uf, for its density is
only i^th of that of oxygen.

t PhiL Trant. 1784, pp. 350—351.

314 CAVENDISH AS A CHEMIST.

position of water^ long after the publication of CaTMidish and Watt's
papers, he declared in effect, that he had pablished full j to the world the
experiments which were the ground-work of Watt's oonclosion. ** In the
experiments," says he, '^ of which I shall now giv^e an aoconnt, I wu
principally guided by a yiev to the opinions which have lately hew
advanced by Mr. Cavendish, Mr. Watt, and M. Lavoisier. Mr. Caven-
dish was of opinion that when air is decomposed water only is pro-
duced; and Mr. Watt concluded from some experiments, of which I gaxe
an account to the [Royal] Society, and also from some observations of his
own, that water consists of dephlogisticated and inflammable air," &e.*
The passage I have marked in italics seems of itself sufficient to negative
the supposition, that Watt's conclusion was based upon experiments
which Priestley did not publish, although he made them knowni to Watt.
Nothing can be more explicit than the declaration of Priestley, that not
unreported trials, but the very experiments which he had detailed to the
Society, were the occasion of his friend's " Thoughts on the constituent
parts of Water. "f Besides this reference to the intimate connexion
between his experiments and Watt's conclusions, Priestley published
various disavowals of his original assertions concerning the production of
water. I have already quoted pajasages from his later writings, in which
he retracted his affirmation in 178S, that pure water was the only product
of the detonation of inflammable air and oxygen (ante, pp. 97 — 103). In
certain of these passages, he also retracts his early assertion that the
weight of water produced, equalled that of the gases burned. The one
retractation indeed necessitated the other, for it would have been a con-
tradiction in terms to have affirmed that the burned gases changed entirely
into water, and yet partly into acid. The conclusion, accordingly, to
which Priestley came when he repeated his experiments was the follow-
ing — '' That water in great quantities is sometimes produced from burning
inflammable and dephlogisticated air is evident from the experiments of
Mr. Cavendish and Mr. Lavoisier. I have also frequently collected consi-
derable quantities of iivater in this way, though never quite so much as
the weight of the two kinds of air decomposedL"}

Mr. Harcourt drew the attention of Arago to these retractations of
Priestley, as neutralising the force of his earlier and opposite declarations.
Arago in reply contended that the latest date in the history of the Water
Controversy is 1784, and that he was not bound to consider statements
which had not been made till 1786 or 1788.§ I must urge, however,
that no critic of the Water Controversy can excuse himself from giving
evety attention to Priestley's affirmations. His earlier statements regard-
ing the synthesis of the elements of water are inexplicable and incre-
dible; and therefore worthless as the foundation of any conclusion. This
is in effect the opinion of Lord Jeffrey, who holds that if Priestley did
in reality employ the charcoal-gas, he and Watt were " the dupes and

* Priestley on Air, vol. vi. (1786) p. 71. The paper u reprinted from Pkil'
Trans, vol. Ixxv. p. 279.

+ In the qnotation given in the text, Priestley refers to ohservations of Watt's

own. ^ It seems needless to demonstrate at length, that these observations were not

^periments with inflammable air (of any kind) and oxjgea, as all the dispntaats in the

^ater Controversy are agreed in holding that Watt made no such trials. The experi-

^«ntB which he did make are recorded in his "Thoughts/' and consisted driefly of

^>ervations as to the evolution of oxygen from various nitrates when raised in tempera-

*'+*^ *"* *^® conversion of latent into sensible heat.

I ^'^••'^ on Air, vol. vi. p. 138.

* ^-^P^es Rendus, 20 Jan. 1840.

THE WATER CONTROVERSY. 315

victims of an hallucination without parallel in the history of the human
understanding.*'

To refuse in these ciroumstances to listen to an explanation which
acknowledges the hallucination, and to insist on asserting that Priestley
mast have obtained impossible results, although he took great pains to
explain that he had been mistaken in thinking he had obtained those
results, might possibly be the duty of a partisan, but would be a great
fault in a historian. Had Priestley's original trials been made with materials
which could have yielded the results reported, we might have supposed
that he succeeded once, though he failed ever after; but when we find his
early statements inconsistent and inexplicable, and his later statements
consistent and quite credible, there -surely cannot be two opinions, as to
which are to be believed.

A statement of Priestley's not less important than those already
adduced, remains to be given. His charcoaUgas experiments have been
represented by Lord Jeffrey as secondary and subsidiary to more perfect
trials made with hydrogen. Yet long after Cavendish, Watt, and
Lavoisier had published their views upon the composition of water,
Priestley repeated his charcoal-gas experiments, with a view to test the
theory which they had published. In his sixth volume on air, published
three years after the date of Watt's first letter, he details the following
observation : —

" Using more precautions to exclude all water from either of the two
kinds of air before the experiment, (both the dephlogisticated air, which
was from nitre, and the inflammable air, which was from charcoal, being
from the first received in mercury, and always confined by it,) I stiu
found a little water after the explosion.

** I varied this experiment by producing the inflammable air in the
dephlogisticated air as follows. Into a vessel containing dephlogisticated
air confined by mercury, I introduced a piece of perfect charcoal, as hot
from the fire as I could bear to handle it, and threw upon it the focus
of the lens, so that a quantity of the air was imbibed; but I could not

perceive that any moisture was formed The

phlogiston the charcoal contained uniting with the deplogisticated air,
free from moisture, formed, I presume, the fixed air that was found after
this process."*

This passage is remarkable, first, as showing how deliberate Priestley's
selection of the charcoal-gas was, and why he preferred it to hydrogen,
namely, because it was drier; and secondly, how defective his early
analysis of the product of combustion of the charcoal-ffas must have been,
for he had no difficulty now in detecting carbonic acid.

It thus appears, (1) that Priestley explicitly affirmed that he had pub-
lished to the Royal Society the experiments from which Watt drew his
conclusion; (2) that he retracted his ori^nal declaration that he had
obtained a weight of pure water equal to the weight of the inflammable
air and oxygen he exploded to&ether; (3) that he regarded the charcoal-
gas as positively preferable to hydrogen, when the production of water by
synthesis of its elements was the object of experiment; and (4) that he
believed fixed air, or carbonic acid, to be an oxide of phlogiston, which
last term, as synonymous with inflammable air, he applied to the
charcoal-gas as well as to hydrogen. Priestley thus testifies against
Watt having drawn conclusions from experiments made by him with
hydrogen. I turn now to Cavendish,

* Prietiley on Air, vol. n. (1786) pp. 127—128.

316 CAVENDISH AS A CHEMIST.

The criiicism which Cavendish published on Priestley's experiments
hajB been quoted more than once already. It is contained in one of the
additions which he made to his paper between its being read and printed,
in consequence of the public reading of Wattes paper. In this he refers
to Priestley " having used a different kind of inflammable air, namely,
that from charcoal."*

The interest of this reference lies in the fact that Cavendish taxed
Priestley with having used the charcoal-gas; and as the passage implies
only that gas, and not hydrogen in repeating the experiments of the
former, and Priestley, as we have seen, not merely acknowledged the
justice of the statement, but used the same gas again in testing the trath
of Cavendish's fully published views.

It remains to inquire whether Watt himself ever detected the fallacy
of the charcoal-gas experiments, or added an3rthing to his paper after
he became acquainted with the views of Cavendish and Lavoisier, to
qualify the vagueness of his original references to phlogiston and inflam-
mable air, and limit these terms to hydrogen. It appears that he did not
introduce any such qualifications, and the fact is of great importance, for
his " Thoughts on the constituent parts of Water, &c.," was not, as some
have affirmed, a hastily written letter, but (at least as it was ultimately
published) a document which had been very carefully considered and
frequently revised. Thus the first version of the "Thoughts** was a
letter to Priestley, of date, 26th April, 1783. Two days later, how-
ever, he recalled this letter and forwarded another copy, assigning as
his reason for so doing that he had "discovered some inaccuracies in
language, and some inconsistencies in the theoretical essay** he had
previously sent his friend, t

This second letter was not publicly read till the succeeding year;
before this, however, viz. in November, Watt drew up a third version of
his views, which he addressed as a letter to De Luc, of date, November
26, 1783.J This letter, nevertheless, satisfied its author no better than
the first to Priestley, and on April 17th, he writes to De Luc,§ stating
that he had made certain alterations on what he had sent him. On the
same day he writes also to Sir Joseph Banks, '' I have, however, revised
the letter itself, and by this post send a corrected copy to him [De Luc],

which he will deliver to you, I have also added some

notes, &c.'*|| A postscript, also, which he had not been able to finish in
time to add to this letter, was sent as a separate communication to
De Luc, of date, April 30, 1784.7

Finally, after the letter to Priestley, and the letter and postscript to
De Luc had been read to the Royal Society, Blagden wrote to Watt to
know in what shape these papers should be published.** Watt desired in
reply that the letter to Priestley, and that to De Luc, should be incorpo-
rated m a way which he pointed out, and at the same time he furnished a
title to the double document to which he also added an explanatory
note.tt There were thus no fewer than five versions of Watt's " Thoughts,"
besid^ a postscript or sequel, which he himself styles "an explanatory
Jetter. XX Between the issuing of the first and last of those versions, more
than a year elapsed, during which the subject which they discussed was

7:»o4>lt/;« /r^,^«^Jp?,i^ ^^«*; P- l^\ '\ » -"titled, S^iuel to tke

p. oz. ft Op, ctt. p. 63. tt Op. eii. p. 56.

THE WATER CONTROVERSY. 317

thoQght over a^in and a^in by Watt, and varioas amendments were
made in the statement of his views.* Yet in the last version, as well as
in the first, he is satisfied with calling the combustible element of water
inflammable air or phlogiston, and nowhere informs the reader that he
desires him to understand bj these titles only the inflammable air from
metals, t. e, hydrogen. In this respect, there is a great contrast between
him and his rivals, for both Cavendish and Lavoisier were careful to
define that their conclusions had reference to experiments made solely
with hydrogen.

Again, the WaXt Corre9p(mdence embodies some seven separate
accounts of Watt's views on the composition of water, addressed by him
to different friends '.t and in truth there are few of the letters which do
not more or less refer to the subject. But in none of them is there a
more precise definition of the combustible element of water, than that it
is phlogiston or inflammable air. Can it be imagined that if Watt had
intended hydrogen to be understood by these terms, he would not have
said so? or is it conceivable that throughout a whole year he should have
been meditating on the nature of the elements of water, and discussing it
among his ablest friends, and yet never once signify to one of them, that
of all the inflammable airs he referred to in the course of his remarks, he
desired to be understood as intending only the one from metals^ when he
described the constituents of water ? The total absence of any limitation
of inflammable air or phlogiston, is irreconcilable with the supposition
that such was his intention.

That I do not wrong Watt in saying this, may be proved by a yery
simple, yet sufficient and fair test. Let us suppose that a contemporary
of Watt, anxious to verify the truth of his theory, and ignorant of all that
Cavendish and Lavoisier had published on the subject of water, consulted
•Priestley's paper " On the seeming conversion of Water into Air, &c.,"
and Watt's '' Thoughts," as well as his Correspondence, with a view to
discover what two substances he should fire together by the electric spark
in order to produce water. In Watt's writings, he would see one of the
two substances precisely enough defined as dephlogisticated air or oxygen.
The other he would find called " phlogiston in the fiuid form," or inflam-
mable air; but without any directions as to the source from which it
should be prepared, or special reference to its qualities, except that it
should be dry and pure. If he sought through Watt*s writings to ascer-
tain whether he limited the word inflammable air to one gas, or to several,
he would find him applying it to five different combustible gases; and if
be turned to Priestley s paper he would find charcoal-gas the only inflam-
mable air specified. Would such a student come to the conclusion that
Watt connected his inference merely with hydrogen, or guaranteed the
verification of his theorVy only if that gas were employed 1 Would he
not rather, either take the charcoal-gas on the authority of Priestley, or
following Watt alone, consider himself free to use any of his ^ve inflam-
mable airs, and among others the " very pure inflammable air" obtained

* In referring to these reriBions, I have no purpose of disputing Watt's claims, as
based even upon the earliest Tersion of his ''Thoughts," In truth, the changes which
lie made were of no great importance, except in so far as he omitted the reference to
Priestley's alleged transmutations of water into atmospheric air, and were rather for
the worse than the better, in some respects. All that I wish to show, is, that if there be
any want of precision in his definitions, it cannot hare been the result of haste, or
Inadvertence, as he had many times conned oyer the record of his yiews before he gave
it to the public.

t Wad Corr. pp. 18—58.

318 CAVSNDISH AS A CHBHIST.

inm oil, ar tke phlogiston in » flaid form, which could he procnTed Vy
h ^»^ng charcoal I In trathy fiye different oiiidentB of Watt's wiitbga,
mi^t each have used a different inflammable air^ and yet haye josdfiei
his choice by a reference to Watt; nor conld an umpire, to whom they
might have appealed, hare found in Watt's papers the means of deciding
which gas alone deeenred approvaL

Another point deserres a moment's attention. The adrocates of Watt
are naturally solicitous to disconnect his conclusion from Priestle/i
inexplicable charcoal-gas experiments, and haye laboured strenuously to
represent him as either ignorant of their nature, or indifferent to them.
I naye said enough already in abatement of the plea of ignorancej I
would now refer to that ot indifference. No reason can be giyen why
Watt should haye been indifferent to the charcoal -gas experiments; and
reasons can be giyen why he should haye preferred them to thoee with
hydrogen. He relied entirely upon Priestley's account of his experiroenti^
and he must haye possessed a spirit of diyination to haye disooyered that
his friend had deceived himself and misled him, when he declared that the
eharooal-gas yielded the results which Priestly affirmed that it did.
A priori, charcoal-gas was quite as likely as hydfrogen, to produce water
when detonated with oxygen. Nothing but direct trial could determine
whether the one or the other was a water-producer, or true hydrogen.
When Priestley, therefore, asserted that charooal-gas, when it umted wiUi
oxygen, produced nothing but water, he was as much entitled to credit^
so £Etf as the mere assertion was concerned, as if he had stated that oar
hydrogen is the <mly body whose oxide is water.

But further, the very considerations which induced Priestley to
substitute the charcoal*gas for hydrogen, in repeating Cavendish's expe*
riments, are likely to have had equal weight with Watt The former gas
had the character of being dry or anhydrous, so that any water resulting
from its combustion, or at least any water exceeding in weight that of the
combustible gas burned, must have been produced, not merely deposited
from pre-existent vapour (at least in the charcoal-gas); whereas hydrogen,
as prepared by the solution of iron or sine in a diluted acid, was certain
to nave diffused through it much aqueous vapour; and the water which
appeared after its combustion might be supposed to be only that which
accompanied the gas, from the aoid which yielded it Tlutt Watt was
alive to those considerations, is evident from all his writings on the com-
position of water. In the MS. of his letter to Priestley, he says, " Yon
have informed me that when you mix together quiU dry inflammable air
and dephlogisticated air, Sic, In the printed version this is changed
into "pure dry inflammable air."t He informs G. Hamilton tnat
Priestley ** puts dry dephlogisticated air and dry i^/iammable air into a
close vessel, &c."^ To the same gentleman he writes again, " Pure dry
dephlogisticated air, and pure dry inflammable air fired together, &c.' §
To Dr. Black he repeats the phrase of his first letter, " quite dry pure
inflammable air."|| All these references occur in descriptions of Priestley's
experiments, and warrant the conclusion that Watt, unaware, as he along
with all his contemporaries certainly was, of the difference between the
inflammable air from charcoal and that from metals, would indubitably
have preferred the former, as '' quite dry pure inflammable air."

The conclusions, then, to which the entire discussion prosecuted in
this section conducts us, are: — 1. That the experiments on the synthesis

* Ante, p. 290. f PMl, Tratu. 1784, p. 331.

t Watt Corr. p. 17. h Ibid. p. 20. |i Ibid. p. 19.

THE WATER CONTROVSRST. 319

of the elements of water, from which Watt drew his conolnsion eoneeming
the nature of the latter, were made by Priestley.

2. That Priestley stated that he published to the Royal Society these
experiments; and that Watt did not object to the statement; whilst from
the published account it appears that the ffases burned together to
produce water, were oxygen and the inflammable air from charcoal.

3. That Cayendish drew the attention of his readers to the hkct that
Priestley had employed the charcoal-gas and not hydrogen, and that
Beither Priestley nor Watt found fault with the statement.

4. That Watt was acquainted with Priestley's experiments, and has
himself recorded his belief in the latter's statement that charcoal can be
oonyerted into phlogiston or inflammable air by heat.

5. That Watt nowhere describes experiments on the synthesis of the
elements of water, made with hydrogen, or in any of his stotements of the
composition of water defines its combustible element otherwise than as
phlogiston or inflammable air, which titles he applied to five different
combustible gases.

6. That Watt, like his contemporaries, had a special name for
hydrogen, viz. inflammable air of the metals; and would haye used that
term u he had wished to specify hydrogen as the inflammable air which
was an element of water.

When, therefore. Watt stated his opinion concerning the nature of
water, as a conclusion from Priestley*s experiments, in the words, ^ Are we
not then authorised to conclude that water is composed of dephlojristicated
and inflammable air f "* he must be understood to refer to the * inflammable
air from charcoal,' if he be held to signify one infiammahle air more thtm
another. But as he generalised his conclusion, and announced that ^pure
inflammable air is phlogiston itself," and that '^ water is dephlogisticated
air deprived of part of its latent heat, and united to a large dose of phlo-
giston,"f he cannot be held to have limited his reference to the charcoal-
gas, but must be considered as including under the titles phlogiston or
inflammable air, at least all the combustible gases to which he gaye either
or both of these names. Whateyer, therefore, be the merit of Watt, a
question which I shall afterwards consider, he has not the merit of having
inferred or announced, either before or after Cayendish and Lavoisier, that
water is a compound of the gases we now name oxygen and hydrogen.
In other words, he was not a £scoyerer, and afiyrtwri^ not the discoverer
of the true composition of water.

8. On the full significatum of the term Phlogiston, as employed by

Cavendish and Watt.

In the two preceding sections it has been shown that Cayendish and
Watt both employed the term phlogiston, as a title for the combustible
element of water, and so far, therefore, as a synonyme for * inflammable
air.' Their precise opinion, however, concerning the nature of water,
cannot be learned without a further inquiry into their views concerning
that mystical entity, phlogiston. To ascertain this, nevertheless, lb a
somewhat difficult matter. We are too prone at the present day to speak
as if all the chemists of the Phlogiston School, from Stahl to Priestley, held
precisely the same doctrine concerning phlogiston ; whereas, in reality,
the later disciples held nothing, almost, in common with their predecessors,

♦ Wait Corr. p. 19. t Wstt Corr, p. 21.

320 CAVENDISH AS A CHBMIST.

except the name. On this point M. Dumas, referring to tbe period when
the Water ControYorsy arose, says, *^ At this epoch Macquer, Baum^, and
many other chemists, had each contrived, in order to meet the new exi-
gencies of the science, a phlogiston of sach a kind as best suited himself.
Lavoisier had no longer to deal with the phlogiston of Stahl, but with a
crowd of entities of that name, which had no quality in common, unless
that of being intangible by every known method."* To the same effect
Lavoisier himself declares that " the chemists have made a vague principle
of phlogiston which is not strictly defined, and which in consequence
accommodates itself to every explanation into which it is pressed ; some-
times this principle is heavy, and sometimes it is not ; sometimes it is free
fire, and sometimes it is fire combined with the earthly element; some-
times it passes through tbe pores of vessels, and sometimes they are
impenetrable to it. It explains at once causticity and non-causticity,
transparency and opacity, colours and the absence of colours. It is a
veritable Proteus which changes its form every moment.'*t

In short, every orthodox chemist of the 18th century, not favonrinc^
the Lavoisierian schism, considered it his duty to make confession of his
belief in phlogiston ; but what it was he believed in, he was by no means
so particular in declaring.

From this charge neither Cavendish nor Watt can be exempted.
Both avoided giving any definition of phlogiston, yet both imputed to it
properties the very reverse of those which were ascribed to it by Stahl,
who certainly would not have understood many of the references to it,
contained in the later writings of the so-called disciples of his own school.
Thus, vague though his description of phlogiston was, he defined it with
a certain precision as a combustible principle, or essence of inflammability,
present in all combustibles, which parted with it when they burned, and
lost in consequence their combustibility. According to this view, the
calx or oxide which is left as the residue of ordinary combustion, is the
combustible minus the phlogiston, which it has given off. In consistence
with this doctrine, Stahl, hsid he been aware of Cavendish's results, would
have afiSrmed that hvdrogen consisted of the water which appeared when
the gas was burned, and of ]>hlogiston, which during the combustion
pasted away. Cavendish and Watt, on the other hand, held that the
water (a certain minimum hypothetically present in the inflammable air
excepted) did not pre-exist in the gas, out had been jyroduced, and that
the phlogiston was all present in the water, and was essential to its exist >
ence. Their phlogiston, therefore, was the very opposite to that of Stahl.

Watt departed still further from the ancient faith, for he dispensed
with phlogiston as a means of explaining combustion, accounting for it
by the theory of an ''elementary heat" present in all bodies, and
given out when they burned. Ihus although inflammable air was,
according to him, phlogiston, it needed an addition of elementary heat,
before the phenomena of combustion could be explained. Watt's phlogis-
ton came in this way to contain StahVs phlogiston, for the functions of
the " elementary" or " latent heat," were exactly those which Stahl attri-
buted to his phantom, and the original fire-essence was for a moment
saved from its approaching extinction by Lavoisier, by being made the
depositary of an inner quintessence of fire.!^

* Le^OM iur la Philotophie CMmique, par M. Dumas, p. 161.
t Ibid. Ibid. p. 162.

X Watt's views were i^eculiar in another respect. The more ancient plilogistiaus
held that phlogiston was a ''principle of levity/' and conferred positiTe lightness on

THE WATER CONTROVERSY. 321

In this singular way^ both Cavendish and Watt man*ed the force of
their own conclusions^ and whilst using language which was irreconcilable
with a belief in phlogiston^ kept using the word, as if they were the
orthodox representatives and successors of Stahl. After all, liowever,
the historian will not wonder much at a phenomenon which so constantly
appears in the history of mankind. So marvellous is the fascination that
names exert, and so deeply are we all imbued with a conservative spirit,
that even the greatest reformers in politics, arts, literature, and science,
are found tenaciously clinging to a word long after they have counted it
their highest triumph to have swept away the reality which it represented.
It was so with Cavendish, Watt, Priestley, Kirwan, Scheele, and in truth
nearly all the chemists of the last centory. They took the greatest liberties
with tho entity phlogiston, but with the name, no one before Lavoisier
Tfould meddle. Nay, they even strove to the very last to frame a
nomenclature for the gases, and, as far as possible, for all bodies, which
should include only references to the one charmed word, although they
never were able to construct more than three derivatives from it, viz.
pMogiMiccUedy depMogisiicaUd, and super-pkiogisticated,

A nomenclature so scanty and so barbarous, compelled its employers to
Oeal in perplexing and often contradictory statements. Neither Cavendish
nor Watt, for example, could find better terms in which to state his view
of the composition of water, than that it consisted of dephlogisticated air,
united to phlogiston ; in other words, of dephlogisticated phlogisticated
air. A plus B, mintts B, is equal, it should seem, to A; but according to
the literal interpretation of this statement, it is equal to A B ? A sub-
stance, apparently, could not be both phlogisticated and dephlogisticated ;
possessed of, and deprived of phlogiston, at the same time ; yet water,
according to the formula of Cavendish and Watt, was in this predicament.

The defects of their nomenclature, however, were not the only sources
of obscurity in the writings of these philosophers, when they alluded to
phlogiston. We are too apt to represent them, as the extreme advocates
on both sides have done, as regarding phlogiston in the same light as we
do hydrogen, viz. as a special ponderable suostance. This, however, was
not exactly the view of either Cavendish or Watt. From a period even
earlier than that of Stahl, an endeavour had been constantly making to
identify the supposed common cause of inflammability, with some one
combustible. In the writings of Beccher (StahFs teacner), of Sir Isaac
Newton, and Stephen Hales, there are many indications of a disposition
to embody the principle of combustibility, (which with them was name-
less), in sulphur.* After 1766, however, the date of Cavendish's expo-
sition of the properties of hydrogen, and in consequence, apparently, of a
suggestion of hist in reference to hydrogen, (although he did not after-
bodies. Watt, on the other hand, says, ''It appears to me yery probable» that fixed air
contains a greater quantity of phlogiston than phlogisticated air does, because it has
a greater epedfic gravity,*' &c. {Phil, Trans, 1784, p. 335) ; and again, '* if we reason by
analogy, the attraction of the particles of matter to one another in other cases is increased
by phlogiston, and bodies are thereby rendered tpedfieally heavier," (Jbid, p. 352.)

* The statement in the text must be taken with an important qualification. The
fbunders of modem chemistry, especially Boyle, Hooke, Mayow, as well as Rey, under-
stood the true nature of combustion much better than their immediate successors did.
Stahl's interregnum was the dark middle-age of the science, which intervened between
the imperfect announcement of a consistent theory of combustion by the early disciples
of Bacon, and its revival, extension, and verification by Lavoisier. {British Quarterly
Review (February, 1849, p. 239), and Brande's History qf Chemistry, pp. 61^67.)

t Ant^ p. 197.

322 CAVENDISH AS A CHEMIST.

wards take the credit of it,) infiammmble air in the widest sense of Uie
term, with its appropriate qualities of elasticity, levity, and oombosiibilityi
was fixed upon as the more probable embcKiiment of the subtle fire-essenoe
or phlogiston. In spite, however, of its identification witb iaflannnaUe
ur, phlogiston never entirely lost its phantom character. It assumed,
indeed, new relations as a material entity ; but it retained its old attri-
butes as an ideal existence. It was, at once, an ' airy nothing ' as phlo-
giston; and a ponderable something as infiammable air.

This double view of the nature of phlogiston, which was common to
Cavendish and Watt, has not been recoi^nised by the advocates of eithtf.
Uuch needless discussion, accordingly, has been carried on r^;arding tbe
relative superiority of their views, in so far as those relate to the pre-
sence of water as an essential ingredient in phlogiston or inflammable air ;
and mueb difliculty has been experienced in accounting for either philo-
sopher entertaining so singular an opinion. Before attempting to di^MMS
of this difficulty, I have to remind the reader that it as much embarraases
Watt's exposition of his views, as it does that of Cavendish. The advo-
cates of the former do not deny that he ultimately taught that one of the
elements of water, contained water as one of i£« elements ; but they claim
for him that at least he originally taught that inflammable air without
water is pure phlogiston, whereas Cavendish always held the leas precise
view, that inflammable air is a hydrate of phlogiston.* Mr. Muirhead
has even gone the unwise length of affirming, that the Dean of Ely
'' has said, among many other things equally incorrect and absurd, that
Cavendish had from the first adopted the conclusion that hydrogen^ or
infiammable air, was the real phlogiston of the popular theory, f It
is Mr. Muirhead, however, who is in error, not Dr. Peacock. I have
already pointed out that it is a mistake to imagine that Watt's view
differed from Cavendishes, in so far as the presence of water in inflam-
mable air was concerned. If any preference, in truth, is to be given to
the views of the one over those of the other, it must be assigned to the
statement of Cavendish. Watt and he alike commenced by holding that
anhydrous infiammable air is phlogiston, and ended by believing that the
latter, when free, should rather be regarded as a compound of inflammable
air and water. Cavendish's original opinion, however, goes back to 1766,
and was founded on his own researches ; whilst Watt's earlier view dates
only from 1783, and was based upon the observations of Priestley and
Kirwan. Cavendish, moreover, did not commit himself absolutely to
either view, but only gave the one a preference. Thus after stating that
" inflammable air is either pure phlogiston, as Dr. Priestley and Mr. Kir-
wan suppose, or else water united to phlogiston," he adds in a note,
" Either of these suppositions will agree equaliy well with the following
experiments ; but the latter seems to me much the most likely."]:

In reality, then, Cavendish and Watt were at one in their final belief,
as to the necessity of water to the existence of phlogiston " in the aerial
form,*' whatever difference there might be in their views, in reference to
the kind of inflammable air of which phlogiston was the hydrate. To
this conclusion they appear to have come independently of each other,
although this has been denied; and Cavendish induced Priestley to
adopt the same view.

Thus much then premised, it remains to determine what were the mo-
tives which induced those observers to hold a view which only added

* Edinburgh Review, Jan. 1848, p. 103.
+ Wait Corr, Introd. Rem. p. dii. t -PAiV. Tram, .1784, p. 137.

THB WATER COMTROVERST. 323

obseaiitj and contradiction to their enunciation of the nature of water*
Nothing, we may be certain, but a strong conviction of its truth, or at
least of its great probability, can have led such clear thinkers as Caven*
dish and Watt to a conclusion, the awkwardness of which was apparent to
themselves. On this last point, Priestley, commenting on their views,
writes thus,-**' Inflammable air seems now to consist of water and
inflammable air, which however seems extraordinary, as the two sub-
etances are hereby made to involve each other, one of the constituent
parts of water being inflammable air, and one of the constituent parts of
inflammable air being water; and, therefore, if the experiments would
favour it (but I do not see that they do so,) it would be more natural to
sappose that water, like fixed air, consists of phlogiston and dephlogisti-
oated air, in some difierent mode of combination."*

The reason assigned by Cavendish himself, for preferring the view
which represented phlogiston as a compound of hydrogen and water, has
been referred to in the abstract of his " Experiments on Kit ^ first series J.
He thought that if hydrogen were pure phlogiston, it would spon-
taneously combine with oxygen, whereas it required to be made red-
hot before it united with it ; unlike liver of sulphur and nitric oxide,
the phlogiston (hypothetically present) in which needed no elevation
of temperature to cause its union with oxygen. It thus seemed as if
pure piklogiston, when free, phlogistieated (deoxidised) air with more
difficulty than the phlogiston which was locked up or contained in a
state of combination in nitric oxide and the alkaline sulphurets. Swayed
by this consideration. Cavendish thought that, in all probability, there
was something present in hydrogen besides phlogiston, which prevented
the latter exhibiting the intense affinity for oxygen which it would have
shown had it been free.f

Having come to this conclusion. Cavendish at once fixed upon water as
the substance which prevented phlogiston from exhibiting its characteristic
affinities, and this for a very significant reason, which seems, however, to
have escaped the notice of the critics of the Water Controversy. When
phlogiston (hydrogen) and oxygen were exploded together, they condensed

* Printley on Air, vol. vi. p. 406.

'\ Kirwan criticised this opinion in the following terms: ''Mr. Cavendish is
inclined to think that pure inflammable air is not pure phlogiston, because it does not
immediately unite with dephlogisticated air, when both airs are simply mixed with each
other. This reason seems to me of no moment, because I see several other sub^tanoes,
that have the strongest affinity to each other, refuse to unite suddenly, or even at aU,
through the very same canse that dephlogisticated and inflammable airs refuse to unite,
viz. on account of the specific fire which they contain, and must lose, before such union
can take place.'' {Phil. Trans, 1784, p. lf>8.) This criticism, however just in the sense
in which we now understand it, was no reply to Cavendish's statement, and had in truth
no meaning in the mouth of a phlogistian, unless, like Watt, he supplemented phlogiston
by Black's latent or elementary heat. To tell Cavendish, who disbelieved in the mate-
riality and latency of heat, that phlogiston must part with specific fire (t.e. an entity
identical in functions with phlogiston) before it would unite with oxygen, was only to
ny> that phlogiston most part with phlogiston, before it can exhibit the affinities of
phlogiston. This suggestion, moreover, left unexplained why the (hypothetical) phlo-
giston of liver of sulphur, and nitric oxide, so readily gave off its specific fire, whilst ihit
phlogi^ton in (or identical wish) hydrogen retained its elementary heat so obstinately.
It is not surprising, therefore, that Cavendish should have attached no weight to
Kirwan's arguments.

Priestley gave the true explanation of the indifference of free hvdrogen and oxygen
to each other, at ordinary temperatures. His opinion is quoted by Watt in the following
passage, which is remarkable as containing one of the earliest examples of the word
noiceni being used in that peculiar and ill- defined but expressive sense, in which it is

324 CAVENDISH AS A CHEMIST.

into water. Had any non-aqueous hodj, however, been united with
phlogiston in the hydrogen, it would have been found in the condensed
water ; but the water was quite pure, so that unless a portion of it had
pre-existed, and been united to the phlogiston, the latter alone must hare
formed the inflammable air. That one or other of these altematiyes must
be accepted by every interpreter of his experiments, was the belief of
Cavendish, as his own words, I think, unquestionably show : '' inflam-
mable air is either pure phlogiston, or else water united to i

phlogiston since these ttoo substances united together form pure I

water," and for the reasons already given he preferred the latter altema- ^
tive.

In this conclusion Watt practically concurred, probably in conse-
quence of pursuing a similar train of reasoning. On this point, however,
we have no information from himself or others.* In the earlier version
of his ''Thoughts" (April, 1783) he stated that Dr. Priestley had found
by some experiments made lately, that inflammable air '' is either wholly
pure phlogiston, or at least that it contains no apparent mixture of any
other matter."t In the later version (November, 1783) he adds, " In my
opinion, however, it contains a smaU quantity of water and much ele-
mentary heat.*']:

So far, then, all is certain enough, but something more is needed to
account for the tenacitv with which Cavendish and Watt, as well as
Priestley, held by the belief that inflammable air contains water as an
essential ingredient. Watt has left us in the dark as to the grounds of
his faith in the doctrine ; and the motive which Cavendish states to have
induced him to adopt it, he refers to as the principal, not as the only, reason
which weighed with him. His principal reason, moreover, was a mere hypo-
thesis, which evaded rather than solved the difficulty. There were many
other ways of accounting for the refusal of hydrogen to combine with
oxygen at ordinary temperatures, besides the supposition that the former
gas was united to water. This hypothesis, in truth, involved a second
quite gratuitous assumption, viz. that water resisted the escape of phlo-
l^iston from it. Every theorist, moreover, on phlogiston, which was " vox
ct preterea nihil," felt at liberty to attribute new properties to It, when-
ever difficulties arose in the explanation of phenomena, in which it was
supposed to take part. Cavendish, accordingly, had many hypotheses to
choose among, and must have had some forcible reason for selecting that
one which he preferred; and so must Watt, who was equally free to
invest phlogiston with whatever properties the exigencies of his theory
required.

The true explanation of their peculiar views may be found, I think, in
that double view of the nature of phlogiston which has already been xe-

still employed. Watt mentions that " a mixture of dephlogiaticated and inflammable
air, will remain for years in close vessels, in the common heat of the atmosphere, with-
out suflering any change," . . . and then adds, "These facts the Doctor [Priestley]
accounts for, by supposing that the two kinds of air, when formed at the same time and
in the same vessel, can unite in their fiatcent state ; but that, when fully formed, they
•are incapable of acting upon one another, unless they are first set in motion by external
heat." (Phil. Trans. 1784, p. 334.)

* It is proper to notice, that Watt was as much perplexed as Cavendish, to account
for the refusal of hydrogen and oxygen to unite at ordinary temperatures, and was led
in consequence to peculiar views concerning phlogiston, which will be noticed a little
f^her on. He does not include among these, however, his opinion that inflammable
Air contained water.

t PMl. TYane. 1784, p. 330, $ Ibid. ibid.

THE WATER CONTROVERSY. 325

ierred to, as hold by its later advocates. Cayendisli, Priestley, Kirwan,
and Watt, all spoke freely, and even confidently, of inflammable air as
pblogisttn ; thus transforming the latter from the ethereal, imponderable
entity which it was in the eyes of the older chemists, into a ponderable
substantiality. It is evident, however, that to conceive it a substance pos-
sessing the ordinary properties of matter, was more than they could succeed
in doing. To this I ascribe the addition to the gas, of water, which
should embody the phlogiston and act as a vehicle, or medium by means
of which that entity, in itself insusceptible of isolation, might be trans-
ferred from substance to substance. Cavendish and Watt demanded only
" a little water," but could not dispense with that little. It might be an
infinitesimal minimum, but that minimum must not be wanting, otherwise
the subtle fire-essence would vanish into nothing !* The phlogistians, in
truth, were accustomed to consider phlogiston as an ethereal entity, which
might be transferred from one substance to another ; but they as little
expected to see it isolated, as the ancient believers in the transmigration
of spirits expected to witness the disembodied soul in the course of its
metempsychoses. Thus it happened, that no sooner had the chemists
been led by certain appearances to believe that they had obtained the dis-
embodied phlogiston, than their old prejudices set them to embody it
again; and no one of them, so far as I have read, appears to have, with
entire good faith, identified phlogiston with inflamnmble air. The latter,
in reality, only furnished it with " a local habitation and a name."

Having thus disposed of what was common to Cavendish and Watt in
their conception of the nature of phlogiston, I have to consider in what
respect their views regarding it difiered. I consider this, however, only
in so far as it afiects their exposition of the nature of water.

So far as Cavendish is concerned, I have little to add to what has
been stated already in the abstracts of his paper on hydrogen (p. 108),
and in that of his first '^ Experiments on Air * (p. 231), as well as in the
5th section of the Water Controversy (p. 282). The sum of the matter
may be stated thus. In 1766 Cavendish taught that when iron, zinc, and
tin are dissolved in diluted sulphuric, or muriatic acid, ** their phlogiston
flies ofiT without having its nature changed by the acid, and forms the in"
fiammable air;" so that for some fifteen years before he experimented on
the union of hydrogen and oxygen, he regarded the former as phlogiston.
When he undertook, in 1781, an inquiry into the nature of phlogistica-
tion, he employed hydrogen as an unexceptionable phlogisticating agent,
and in entire consistence with the professed object of his research, which
was not to discover the nature of water, or merely to ascertain the pro-
duct of the combustion of hydrogen, but to determine the change which
air underwent when it was vitiated fphlogisticated) by combustibles, he
declared water (which he believed to be the invariable product of phlogis-
tication) to contain phlogiston as an element. In reference to these
views, what I seek to insist on most strongly is, that Cavendish held
phlogiston to be a substance identical with his inflammable air from
the metals (our hydrogen), and that he believed it to have but one oxide,
namely, water. I cannot, however, unreservedly concur in Mr. Harcourt*s
statement that Cavendish's phlogiston *' was hydrogen and nothing else^f
No chemist, in 1783, was in a condition to say whether the in flam-

* In the statement in the text, I assume, for argument's sake, that Priestley's
experiments were unexceptionable, as Watt believed them to be, and that the former, as
well as Cavendish, obtained pure water alone, as the result of his detonations.

t BHt» At90e. Rep, 1839. Pres. Address, p. 28.

326 • CAVENDISH AS A CHEMIST.

mable air from the metals was identical with other inflammable ain or
not, and Cavendish and Lavoisier were careful to pronounce this an opea
qaestion, contenting themselves with pointing out that their covclasiooii
as concltisions, had reference to one inflammable air only, and leaving it |
for time to determine whether other inflammable airs were identical with ^
that one. If any were, their conclusions might be general ieed to the
extent of that identity. To have asserted more, they must have possessed
the power of divination; for, as I have urged already, the distinction of
the combustible gases from each other was the result, not the precursor,
of the discovery of the composition of water.

It must, however, be added, that though Cavendish called no combostiUe
gas but hydrogen phlogiston, when recounting his own experiments,
nevertheless he generalised much too widely, aud assumed hydrogen to be
present in many bodies which do not contain it. He seems, in truth, to
have been a more faithful disciple of Stahl than is generally supposed ;
and as the latter held that all combustibles contained phlogiston, so
Cavendish taught that they all contained hydrogen. The disputants in the
Water Controversy appear to have overlooked this curious fact, whidi
the advocates of Cavendish might not wish to insist upon, and his de-
tractors do not seem to have discovered ; although in truth an allusion to
it could not much have served their client, for Watt generalised still mom
unwisely than Cavendish did.

The views of the latter are stated too distinctly by himself, to leave it
at all doubtful what they were. When he commenced his " Experiments
on Air,** he found the majority of chemists believing, that when combos-
tibles are burned in air, and when the latter is deoxidised by such bodies as
nitric oxide and liver of sulphur, or, as they phrased it, when air is phlo-
gisticated, the invariable product is fixed air, or carbonic acid. He quickly
proved that this was a mistake; but he erred in one sense as widely as
those he corrected had done, for he affirmed that the invariable product of
phlogistication is water. He was quite aware that most organic com-
bustibles yield carbonic acid when burned, but he conceived this to have
pre-existed in them, and excluded them on that account from his trials
as not suitable for an experimentum cruds. All other phlogisticating
(combustible or oxidable) bodies, including nitric oxide, liver of sulphnr,
and the metals, he believed to contain hydrogen, (as Stahl believed them
to contain phlogiston,) and all of them he thought yielded water, as one of
the products of what we should now call their oxidation. The passages
in his paper which prove this, will be found specially referred to in the
abstract of his " Experiments on Air." Two of the most pertinent may
be noted here. After describing experiments which proved that neither
carbonic, nitric, nor sulphuric acid was the constant product of the phlo-
gistication of air, he adds, "Having now mentioned the unsnccessfol
attempts made to find out what becomes of the air lost by phlogistication,
I proceed to some experiments which serve really to explain the matter.'*
(Ante, p. 234.) The experiments thereafter recorded are those with
hydrogen, which Cavendish plainly considered as explaining phlogistica-
tion in every case. In other words, the phlogistication of air always
diminished its bulk, because it was always attended by the production of
water. To this he constantly refers throughout his paper, and he expli-
citly announces it in his criticism of Lavoisier's views, in which he affirms
that, " Adding dephlogisticated air to a body, comes to the same thing as
depriving it of phlogiston, and adding water to it,*' which was equi-
valent to saying, that the universal result of oxidation is the separation

THE WATER CONTROVERSY. 327

of hydrogen from the body oxidised^ and the addition of water to the
oxide.

From another qnarter, moreover, proof, if possible more decisire, can
be obtained concerning Cavendish's opinions. Kir wan, it will be remem-
bered, published "Remarks on Cavendish's Experiments on Air." In
these the following passage occurs in reference to '' phlogistic processes."
** I selected, as least liable to objection, the calcination of metals, the
decomposition of nitrons by mixture with respirable air, the phogistica-
tion of respirable air by the electric spark, and lastly, that effected by
amalgamation, /n e<zch of these instances, Mr. Cavendish is of opinion
that the diminution of respirable air is owing to the production of wcUer,
which, according to him, is formed by the union of the phlogiston disen-
gaged in those processes with the dephlogisticated part of common air.*'*
To Kirwan's remarks Cayendish replied, qualif3dng his reference to the
electric spark, but finding no fault with the opinion concerning the pro-
duction of water attributed to him.f

Cavendish cannot then be vindicated from the charge of having
generalised too widely in his speculations on the nature of nydro^en and
the production of water. But after this has been fully conceded, we
must guard against depriving him of the great merit that truly belongs
to him. And what was meritorious in his researches may be easily
ascertained by inquiring how much of his views the progress of science
has shown to have been false. Tried by this test, we shall find that
Cavendish's errors lay in his hypotheses, and that time has only confirmed
what he based upon his observations. We no longer believe that every
oxidable body contains hydrogen, and yields water when it is oxidised,
which Cavendish irnagined, but never pretended to demonstrate, was the
case. But we still hold that hydrogen nnites with half its volnnie of
oxygen, and that the resulting oxide is water. And we further concur
with Cavendish in believing, that all bodies which contain hydrogen,
yield water when oxidised. We differ from him, in truth, but in two
particulars. We disbelieve that the simple combustibles contain hydrogen;
and we think it unnecessary to assume that hydrogen contains water.
This latter doctrine, however, was but an opinion liable to correction —
not a settled conviction on Cavendishes part; neither can we, as Arago
has justly urged, demonstrate that there is no water present in hydrogen.]:
The substitution, accordingly, by Cavendish, of phlogiston for hydrogen,
as the title of the combustible element of water, did not in any way alter
the signification of his statement; for he called hydrogen, phlogiston,
from the earliest period of his acquaintance with the gas; and it was the
only substance which he professed to have " turned" into water, by uniting
it with oxygen.§

• Phil. Tram. 1784, p. 154.

t Phil, Trmu. 1784, p. 176. Fnrtiier illastratioii of the same ikct will be found
in a letter of Kirwan to Crell, referred to in a sobseqaent section, and in a statement bj
Priesaey.

t Note by M. Arago to Lord Brtmghom^s Historical Note, reprinted in Sequel to
Watt Corr. p. 253.

$ The advocates of Watt build much upon the fact, that Cavendish accepted
Watt's conclusion as of equal value with his own, although he drew attention to the
fact, that Priestley's charooal-gas differed from the inflammable air from the metals.
This point will be discussed at length in another section. It may suffice, therefore, to
notice here, that all that Cavendish did, was to accept Watt's conclusion as identical
with his own, provided Prieetley'e experiments, from which it was drawn, were trust'
worthy. That they were accurate he doubted, and he afterwards demonstrated that they
were not.

328 CAVENDISH AS A CHEMIST.

I Lave illustrated Cayendisli*s views the more fully, that they have
been misapprehended by Berzelius. The great Swedish chemist was so
excellent a critic of the labours of others, that I feel myself peculiarly
liable to the charge of presumption in disputing the accuracy of the
account he gives of the opinions of Cavendish (as well as of Watt) con-
cerning the nature of water. The following extract from the Lehrbuck
(1843) will show what interpretation he puts upon the language of tho
English chemists. He is discussing the relative merit of Cavendisb,
Watt, and Lavoisier, as alleged discoverers of the composition of water —
a question which does not at present concern us; and in reference to the
first of those philosophers he mentions, that he stated his views thus :—
" Oxygen is water deprived of phlogiston ; hydrogen is water snper^
saturated with the same hypothetical substance. By their mutual com-
bination, water is obtained in its original condition. According to this
explanation, water was still regarded as a simple substance, which con-
stituted the ponderable matter in hydrogen and in oxygen.'**

The view imputed in the above quotation to Cavendish, is founded
upon a misconstruction of his language. When Cavendish called oxygea
water deprived of, or minus phlogiston, or dephlogisticated water, he did
not intend to teach that it was water in any but a negative sense. The
phrases were equivalent to those wo should employ at the present day, if
we chose to say that oxygen is water deprived of, or minus hydrogen, or
dehydrogenated water. Those terms refer to the derivation of oxygen
from water, not to its identity with it, and were natural at a period
when no substances but those which contained water were supposed capable
of yielding oxygen. They seem to us at the present day awkward and
circuitous, when we know that oxygen can be prepared from many
sources, and seldom prepare it from water. They were, nevertheless, in.
an intelligible and significant sense, quite accurate ; and though we are
apt to forget it, we employ exactly equivalent phrases in our present
nomenclature. The word aldehyde, for example, a contraction for
alcohol dehydrogencUuSy by which one of the acetyle compounds (C*HH),HO)
is known, is a term referring to source or derivation, exactly corre-
sponding to the aqua dephlogisticata (dehydrogenata) of Cavendish. No
one interprets aldehyde, i. e. dehydrogenated alcohol, as signifying alcohol
in any other than a negative sense, namely, as alcohol which, by the loss
of so much hydrogen, has become a body distinct in all its properties
from the substance which yielded it. In like manner Cavendish's dephlo*
gisticated or dehydrogenated water was, by its very definition, not water,
but chemically a moiety of it; the half of it, not the whole; the sub-
stance which was left when the combustible ingredient of water was
removed from it. The whole of Cavendish's references to oxygen are in
keeping with this view. I do not know a simple passage in his writings
which even remotely hints, still less one which directly asserts that
oxygen is, in a positive sense, water, or contains it as an essential con-
stituent, and as its only ponderable iugredient. On the other hand,
he contrasts dephlogisticated with phlogisticated air; and observes, that
there is the utmost reason to think that they are distinct substances, and
not diflering in their degree of phlogistication.f

As for hydrogen, Cavendish, we have seen, thought it highly pro-
bable that it contained water; but not because it was impossible that this

* Lehrbuck der Chemie Von J. J. Berzelius. FUnfte umgearbeitete OriginaU
Avflage, 184:^. Er&ter Band, p. 370^2.
t Phil, Trattf. 1784, p. 141.

THB WATER CONTROVERSY. 329

SM conld exist in an anbjdrous condition, but because tbe indifference of
hydrogen to oxygen at ordinary temperatures seemed to imply the pre*
eence of some soDstaoce in the former, which lessened the intensity of its
affinity for oxygen, and this substance be conceived could only be water^
since it was tbe sole residue of the combustion of hydrogen and oxygen.
So &r, however, was Cavendish from holding that the non-aqueous por-
tion of hydrogen is imponderable, that in 1766, years before he thought
it necessary to affirm that water is probably present in hydrogen, he
determined its specific gravity; and whilst referring to it simply as phlo*
^ston, ranked it among ponderable bodies.

In consistence with the views expressed in the quotation, Berzelius
represents Lavoisier as the first who spoke of water as a compound.
We have seen however already, that Cavendish uses such expressions
as these : — *' Almost the whole of the inflammable and dephlogisticated
air is converted into pure water." " Water consists of dephlogisticated
air united to phlogiston." " These two substances united together form
pure water."* In these passages, especially the last two, we have the
compound nature of water as consisting of two things — ^phlogiston and
dephlogisticated air, explicitly announced. Berzelius thinks that we put
an interpretation on these and similar passages at the present day, which
was not intended by their author, and did not occur to any one till after
Lavoisier had published his views. But if we consult Kirwan*s
^' Remarks on Cavendish's Experiments on Air," which were read to the
Royal Society before Lavoisier's papers reached this country, we shall
find him declaring, that " Mr. Cavendish is of opinion that the diminution
of respirable air is owing to the production of water, which according to
him i$ formed by the union of the phlogiston disengaged in those processes,
with the dephlogisticated part of common air."t To this representation of
his views. Cavendish, as we have already seen, made no objection, although
he replied to other statements of Kirwan's.

Blagden is still more precise, for he charges Lavoisier with having
been most reluctant to believe that '^ water was dephlogisticated air
united with phlogiston," when informed that such was the view enter-
tained in England. His prepossessions led him to expect, that hydrogen
and oxygen would produce an acid when burned together; and the state-
ment that water was a compound of these gases was so unwelcome to
him, that, according to Blagden, he insisted that the water was not
" formed or produced out of the two kiuds of air, but was already con-
tained in, and united with the airs, and deposited in their combustion."}:
We have Blagden thus, in the nanie of Cavendish, representing Lavoisier
as protesting against a doctrine almost identical with that which Berzelius
represents Cavendish as holding; and Lavoisier receiving, and at first
discrediting, instead of originating the doctrine of the compound nature of
water.

It may also be noticed here, that Watt regarded Cavendish's views as
identical with his own ; and so also did De Luc, as will appear more
fully in another section. It is certain, however, that whatever were the
defects of Watt's views, he was most explicit in his declaration that water
is a compound. " Air and water," says he, " are not simple elements."

" 1 have found out what water is made of." ^* The ingredients

of water are pure air and phlogiston."§

• Phil, Trans. 1784. pp. 133 & 137. f PhiL lYans, 1784, p. 155.

t Blagden's Letter to Crell, Watt Corr. p. 72.
§ Watt Corr. pp. 24, 25.

330 CAVENDISH AS A CHSIOST.

CaTendish, agft^n, accepted Watt's yiews as to a great extent identiol
vitb bis own; the points ou which they differed having no reference to the
question of the elementary or componnd nature of water. Blagden was
of the same mind; so that Watt, De Luc, and Blagden, as irell as Kirwaa,
testify to having understood Cavendish to teach that wmter is a oont-
pound ; and he was aware that they imputed this doctrine to bim, hot
found no fault with the imputation.

Further, Cavendish, Watt, and Blagden unhesitatingly accused
Lavoisier of plagiarism from the two first, — ^a charge they nev^er would
have preferred, had the doctrine of the compound nature of water been a
novelty to them. This charge has the more weight, that Cavendish and
Blagden at least were alive to the peculiarity of Lavoisier's views in
reference to hydrogen. ''Lavoisier's present theory," says Blagden,
** perfectly agrees with that of Mr. Cavendish ; only that Mr. Lavoisier
accommodates it to his old theory, which banishes phlogiston.***

The considerations adduced above will suffice to show that Benelins
is mistaken in thinking that Cavendish was not understood by bis con-
temporaries to teach that water is a compound body. Great as Lavoisier's
merits are, it was not left to him to deny for the first time the elemen-
tary nature of water; or to teach, that two gases could be burned together
into their joint weight of this liquid. To Lavoisier, however^ I shall
return.

I have now to consider what Watt's opinions concerning phlogiston
were. That he held it, in its state of greatest isolation from matter,
to consist of inflammable air, along with a little water and much
elementary heat, and that inflammable air was a term applied by him
to other gases besides hydrogen, has already been pointed out. My
present object is to inquire whether, on a more full investigation of his
views conceruing phlogiston, it can be shown that he did nat identify
it with hydrogen, which Cavendish did. As the case now stands with
Watt, the only way in which a claim can be established for him as a
discoverer of the composition of water, is by showing that his oonclo-
sions had a primary and special reference to hydrogen. Could that he
shown, it might with justice be argued that, in accepting Priestley's
charcoal-gas as identical with the phlogiston, or combustible element
of water, he erred only to the extent of believing that charcoal, when
heated, evolves hydrogen. It is not a little curious, accordingly, that
none of Watt's advocates should have attempted to demonstrate this,
and so to make out a logically consistent case for him. I shall
assume the possibility of such a view being substantiated, and avoid
doing injustice to Watt, by inquiring if it can.

In pursuing this inquiry, the first point demanding attention is the
important one that, for years before he published his " Thoughts on
the Constituent Parts of Water," Watt had anticipated the probability
of water being convertible into air. "For many years," sajB he, "I
have entertained an opinion that air was a modification of water, which
was originally founded on the feicts that, in most cases wherein air was
actually made, which should be distinguished from those wherein it is
only extricated from substances containing it in their pores, or other-
wise united to them in the state of air, the substances were such as
were known to contain water as one of their constituent parts, yet ne
water was obtained in the processes, except what was known to be

/ Letter to Crell, Watt Corr, p, 73.

THB WATER CONTROVERSY, 331

onlj loosely connected with them, such as the water of the crystalli-
sation of salts.* This opinion arose from a discovery that the latent
heat contained in steam diminished in proportion as the sensible heat of
the water from which it was prodaced increased; or, iu other words, that
the latent heat of steam was less when it was produced under a greater
pressure, or in a more dense state, and greater when it was produced under
a less pressure, or in a less dense state; which led me to conclude that,
when a very great degree of heat was necessary for the production of the
steam, the latent heat would be wholly changed into sensible heat ; and
that, in such cases, the steam itself might suner some remarkable change.
I now abandon this opinion, in so far as relates to the change of water
into air, as I think that may be accounted for on better principfes.^t

It thus appears that Watt conceived that, at a high temperature, steam
would undergo a remarkable change, which, judging from the supposed
fact that only bodies containing the elements of water evolve " air," he
thought (as we learn from a reference of Priestley's)]: would consist in the
water-vapour becoming converted into a permanently elastic fluid. §

It is difficult to be quite certain in what sense Watt used the term
" air'* in the passages quoted. It is manifest, however, that he employed
it either as synonymous with atmospheric air, or as identical with gas or
elastic fluid, in the widest sense of these terms. From his referring in the
next paragraph, in which his remarks are continued, to the evolution of
dephlogisticated air from melted nitre, I am inclined to think that he used
the word ''air" in its widest sense. After all, however, it is quite possible
that Watt himself would have been puzzled to reply, had he been asked
in what sense he employed the ambiguous word. Atmospheric air, as I
have pointed out, in the abstract of Cavendish's paper on the eudiometer^
was not regarded as constant in composition, but might, ex kypotheai, vary
in respirability through a very wide range, of which the extreme limits,
in opposite directions, were oxygen and nitrogen. Anything, therefore,
short of pure nitrogen, was pro tanto respirable air ; and eren nitrogen
"WBa frequently referred to, not as a special gas, but as the phlogisticated
part of atmospheric air. It is of less importance, however, to decide the
absolute signification of the term airy as used by Watt, than it is to show
that his early hypothesis of the convertibility of water into air implied
no precise anticipation of the nature of the air, or elastic fluid, which water
should yield, and still less any expectation that it would consist of two unlike
gases, one of which should prove to be inflammable. The opposite has been
argued by Mr. James Watt, who, in qualification of a statement of Lord
Brougham, adduces the passage on which I have been commenting, as prov-

* It need learcely be noticed that Watt waa in error in this notion, and that nitre,
ex, ffr, which he supposed to yield oxygen hecanae it contains water, is an anhydrona
salt. He as freely and as unwarrantably assumed the presence of water in bodies as
Cavendish did.

t Phil. Tnmt. 1784, p. 335.

X Phil. TVatu. 1783, pp. 415» 416. The passage is quoted at p. 330.

§ Watt abandons this opinion at the close of his remarks, quoted in the text.
Nerertheless, we may justly regard Mr. Grove's beautiful discovery of the power of
white-hot platina to decompose water, as the unexpected fulfilment of Watt's sagacious
conjecture. (Chemical Society* 9 Memoirg, 1847, p. 332.) He seems to have changed
his opinion in consequence of the conversion of air (inflammable air and oxygen) into
water, in Priestley's experiments, being attended with the cliange of latent into sensible
heat, whereas, according to his view, this change should have occurred in exactly the
opposite circumstances, namely, when the liquid was undergoing conversion into gas or

332 CAVENDISH AS A CHEMIST.

ing that ''the idea existed in his [Watt*s] mind previously" to the repetition
of CaYendish*s experiments by Priestley.* Sir David Brewster also seems
to a certain extent to sanction this view, as in his advocacy of Watt's claims
he observes (withoat, however, naming him), '' that to conjecture even
the very improbable fact that water is formed of two different kinds of
air, was a bold and an original idea."t Watt assuredly would deserve
the highest praise if he had entertained so sagacious a thought; but neither
he, nor any other of the claimants of the disputed discovery, predicted,
or professed to have predicted, that water would prove a combination of
two dissimilar gases. All of them, alike in France and England, reached
the discovery as the unexpected result of an ei posteriori investigation.
Whilst, therefore, it would be doing Watt great injustice not to acknow-
ledge that his beautiful researches into the relation of heat to steam led
him to watch for any indications of its becoming a permanent gas, with an
interest and a keenness shared by none of his contemporaries, and pre-
pared him to turn to the best account anything bearing on the converti-
bility of water into air, — it would be doing others equal injustice to affirm
that, before Priestley repeated Cavendish's experiments. Watt had done
more than anticipated that water might be transformed into gas, without
having come to any conclusion as to the probable chemical composition of
the elastic fluid which should be produced.

For several years, as the last quotation from his writings shows. Watt
entertained this idea, but he did not endeavour to realise it by experiment.
Priestley, however, who entertained a somewhat similar, though less precise
notion, made it an object of investigation. '^I imagined,*' says he, ''that
when substances consisting of parts so volatile as to fly off before they
had attained any considerable degree of heat in the usual pressure of the
atmosphere were compelled to bear great heats under a greater pressurej

they might assume new forms, and undergo remarkable changes;

but I hflul no particular expectation concerning the nature of that change.
I was mentioning these ideas to Mr. Watt, in whose neighbour-
hood I have the happiness to be situated, when he mentioned a similar
idea of bis, viz. that of the possibility of the conversion of water, or steam,
into permanent air; saying that some appearances in the working of his
fire-engine had led him to expect this. He thought that if steam could be
made red-hot, so that all its latent heat should be converted into sensible
heat, either this or some other change would probably take place in its
constitution. The idea was new to me, and led me to attend more parti-
cularly to my former projects of a similar nature," &c. &cX Encouraged
in this way by the similarity of his views and those of Watt, Priestley
instituted those experiments with porous clay retorts already frequently
referred to, in which water was apparently, by simple distillation, con-
verted into air. This air was sometimes a little purer, sometimes less pure,
than atmospheric air, but always respirable. In the detail, for example,
of one experiment which did not materially differ from the rest, Priestley
says : " This air was never much less pure than that of the atmosphere.
Sometimes it could not be distinguished at all from it at all [sic in orig.] by
the test of nitrons air, and once or twice I thought it even purer than that
of the atmosphere. "§ In like manner, he observes that " another pre-

* Note by Mr. James Watt to Lord Brougham^a Hittorical Note, appended to
Arago's Eloge of Watt, reprinted in Watt Corr. p. 257.
t North Brit. Ret. Feb. 1847, p. 474.
X Phil. Tratu. 1783, pp. 415, 416,
§ Op, cit. p. 423.

THE WATER CONTROVERSY. 333

sumption in favour of the generation of our atmosphere from water was,
that the purity of the air that I produced from it is so very nearly the
same with that of the atmosphere."* These results were speedily com-
innnicated to Watt, who welcomed them as the probable verification of
his hypothesis. On the 13th December, 1782, he writes to De Luc: —
*' Dr. Priestley has made a most surprising discovery, which seems to con-
firm my theory of water undergoing some very remarkable change at the
point where all its latent heat would be changed into sensible heat.'* He.
then describes one of Priestley's distillations, in which lime and water
were heated in an earthen retort; and adds — '* The air so produced con-
tained a little fixed air, but the greatest part of it was nearly of the
nature of atmospheric air, only somewhat more phlogisticated."f On the
26th of the same month, Priestley writes to Watt : " I now convert water

into air without combining it with lime or anything else The air

is of the purity of that of the atmosphere, and, I think, without any mix-
ture of fixed air."| The final conclusion, then, of Priestley, before he
discovered the fallacious nature of his experiments, was, that water is
convertible into (Umespheric air. In this Watt acquiesced ; for when
Priestley told him that he had unexpectedly discovered that respirable
air was only obtained when the porous retorts were surrounded by pure
atmospheric air. Watt replied conjecturally : " If, after all, this should
account for the production of common air from water," &c.;§ implying
that the experiments had formerly seemed to demonstiate such a change.
When Priestley's experiments appeared unexceptionable. Watt made the
statement still more deliberately. In the letter to De Luc, already

quoted, he says : '' I now believe air is generated from water If

this process contains no deception, here is an effectual account of many
phenomena, and one element dismissed from the list."|| In other words,
water and (atmospheric) air are not distinct elements, but two forms of
one substance, and the liquid can, by change of temperature, be converted
into the gas. It was free to one who held such a view, to consider neither
of the two bodies, which were mutually convertible, as more elementary
than the other, or to prefer one of them as the radical or base of the other.
Watt preferred the latter view, and held that air was '^ generated from
water," which was thus more elementary than air. Watt would not have
argued thus at a later period, but as yet (December, 1782) he had no
evidence that air of any kind was convertible into water, nor does he
make any reference to Priestley *s explosions of inflammable air and oxysen
till some three months (March, 1783) after he had declared his belief that
air is generated from water.

I hare dwelt at length upon this apparently secondary point, because
it is of great importance to notice, that whereas Cavendish reached his
conclusions concerning the nature of water, solely from observations on the
83mthesis of oxygen and hydrogen, Watt's earliest opinion concerning the
nature of that liquid, was founded neither upon synthetical nor analytical
researches into tbe quality of its constituents, but upon observations in
which it seemed to undergo direct transmutation (so far at least as its
ponderable matter was concerned) into atmospheric air. This erroneous
notion, based upon Priestley's delusive experiments, was not entirely
abandoned by Watt, even when he published the latest version of his
'' Thoughts on the Constituent Parts of Water." It confused his specu--

* Phil. Trmu, 1 783, p. 428. t Watt Corr. p. 4. % Op. eit. p. 8.

$ Watt Corr. p. 2T. It Watt Corr. p. 4.

334 CAVENDISH AS A CHSIUST.

lattons oonceniiiig the natuie of water, and seems to bare been the somee
of his belief that phlogisticated air or nitrogen consisted of the eamB
ingredients as water.

Meanwhile Priestley proceeded with his researches, and led Watt
astraj in another particular. The experiments I have been considering
were performed in Ueoember, 1782. On 26tb March of the same jear the
latter writes to G. Hamilton — " Dr. Priestley makes fixed air from depUe*
. gisticated and inflammable air in the following manner. He takes mere.
precip. ruber., which yields only dephlogisticated air; and iron, whidi
yields only inflammable air, and heats Uiem together. They prodnos
only fixed air.'* Thereafter Watt refers for the Jirst time to the prodne-
tion of water by the direct explosion of the gases which he had already
stated produced filed air.* In this way, and at as early a period as Uiat
at which he came to the conclusion that water consisted of inflammable
air and oxygen, he inferred that carbonic acid was identical with water,
as it was also with nitrogen, in qnalitatiye composition. Nor was
Priestley's repetition of Cavendish's experiments undertaken as a speoiai
inquiry into the product of the combustion of inflammable air and oxygen,
but as an incidental justification of his own theory that water is con*
vertible into air. " Still hearing," says Priestley (in a passage already

? noted in full), '^ of many abjections to the conversion of water into air,
now give particular attention to an experiment of Mr. Gavendidi's
concerning the reconversion of air into toater, by decomposing it in ean-
junction with inflammable air."t

It is of importance to notice the mode in which Priestley thus refers to
his experiments with inflammable air. The title of Watt's paper, whidi
was not added till a year after the paper was written, and its general tenor
in its latest version, naturally convey the impression that the main object
of Priestley's researches and Watt's conclusions was the demonstration of
the power of inflammable air and oxygen to produce water. Whereas
Priestley's paper of 1783, studied as a whole, and the several yersions of
Watt's Thoughts, as well as the WaU Correspondenee, show that the

C'imary object of both observers was the demonstration of the transmnta-
lity of water into atmospheric air; whilst the interest which Cavendish's
experiments had for Priestley, lay almost entirely in the fact that they
assisted him in establishing the general proposition that a gas may become
a liquid, and therefore a liquid a gas, and increased the probability of his
particular assertion that water may be transformed into common air. To
Watt again, Cavendish's experiments, as they reached him tbrongb
Priestley's repetition of them, were objects of independent interest : but
he did not originally think the conclusion they warranted of more im-
portance than the one he had agreed with Priestley in drawing from the
tatter's experiments with porous earthen retorts; and the breaking down
of the proof that water might be converted into common air, led him to
withhold his whole paper, and decline to have it publicly read, although
the evidence in favour of the convertibility of a mixture of inflam-
mable air and oxygen into water, remained unafl*eoted by the detection of
a fallacy in the experiments maiie with the earthen retorts.

Discoverers are fond of insisting on their discoveries being studied by
ethers in the order in which they were made, and love to record even
the irrelevant matters which interested themselves during their researches.
Watt was not superior to this foible, but has been at pains to inform us

* WaU Corr, p. 17. t On Phlog^iston, &c. Phil TVant. 1783, p, 398.

THE WATER COl^TROVEKST^ 335

in the onameneemeiii of Uie latest Tersion of his ^^ Thottghts,** ^' I fir$t
thought of tkU waj of solving the phenomena in endeavouring to account
for an experimoit of Dr. Priestley's, wherein water appeared to be con^
Terted into air;"* so that he continued to the last to hanker after giving
the precedence to those experiments which agreed most with his own
^ priori hypothesis. When this is considered, and all that has been
stated alrcMuiy, it seems impossible to accept as tenable the assertion that
Watt thion^oat his paper specially referred to hydrogen when he used
the word phlogiston. It would be more easy to prove the very converse
o£ this for the following reasons : — (1) Watt*s views regarding the nature
of water commenced with a speculation which did not even remotely
contemplate the resolution of water into oxygen and an inflammable air,
mnch less into oxygen, and the inflammable air, hydrogen. (2) He
aooepted Priestley's porous retort experiments, in which water appeared
to undergo conversion into common air, as the full realisation of his
hypothesis^ and he gave the following rationale of the process, which
first appeared in Priestley's paper. "Since pure external air was neces-
sary in order to procure good air, it was concluded by several of my
firiends, and especially Mr. Watt, that the operation of the earthen retort
was to transmit phlogiston from the water xsontained in the clay to the
external air; and that the water thus dephlogisticated was capable of
being converted in respirable [not dephlogiatiecUed] air by the intimate
union of the principle of heat."t What became of the phlogiston lost
by the water appears more clearly from Watt's own account.

'^ On considering the last and most remarkable production of air from
water imbibed by porous earthen vessels, (the only case wherein it
appears almost incontrovertibly that nothing was concerned in the pro-
duction except water and heat,) I think that the earth of the vessel
attracts the phlogiston from the water, and gradually conveys it from
particle to particle, until it transmits it to the external air, which it pro-
bably phlogisticates ; and that, therefore, the same substances moistened
with water, and heated in glass or metalline vessels, can produce only
limited quantities of air, because the earth comes to be saturated wita
phlogiston, which it cannot transmit to the external air, and consequently
will decompose no more of the water than it can retain the phlogiston
of, united to itself. .... I omitted to mention in its proper
place that clay when made hot has a very powerful attraction for phlo-
giston, and in some circumstances becomes quite black with it, but r^ulily
parts with it to pure air, and becomes white again."]:

On these views of Watt, Mr. Harooort comments as follows : — " Here
inflammable gas or hydrogen is obviously out of the question; the phlo-
giiton of the water, which passing through the retort, is presumed to
phlogisticate and vitiate the external air, is nitrogen; and the depblo^is-
ticated air of the water is supposed to retain sufficient phlogiston to make,
with the assistance of heat> good air, of the same purity as the atmo-
sphere.*'^

In the first part of this criticism I cannot concur, for it would represent
Watt as considering phlogiston and phlogisticated air (nitrogen) as the same

* PkU. TVttns. 1784, p. 329. t PAi7. Trant, 1783, p. 431.

% The qaotatioDS in the text are from the unpublished part of Watt's letter to
FriesUey (April 26, 1783) belonging to the Royal Society, and already referred to,
p. 292. llie firrt and last passages hare been published by Mr. Harconrt. {Brit,A»s9c.
Jtep, 1839, p. 24.) The concluding sentence forms a postscript to the letter.

i Brit. Assoc. Rep. 1839, p. 25.

336 CAVENDISH AS A CHEMIST.

thing, whereas he has himself told ns that he regarded nitrogen as '^a
composition of phlogiston and dephlogisticated air." His riew plainlr
was, that the phlogiston passed through the retort as phlogiston till it
reached the external air^ when it combined with oxygen and conrerted it
into nitrogen.

The remainder of Mr. Harcourt*s criticism is quite to the point. Na
property of hydrogen is imputed by Watt to phlogiston, but rather those
of carbon, which Watt believed to be transmutable into that entity; for
the whitening of black clay, which is so familiar a phenomenon in the
potter's kiln, results from the oxidation of the carbon of vegetable matter,
which a lower temperature renders black by charring. If, then. Watt's
ideal phlogiston is to be identified with one ponderable substance rather
than another, it must be with carbon rather than with hydrogen, so
far as we have yet proceeded. On this, however, I am not anxious
to dwell. It is enough if Watt can be shown not to have signified
hydrogen by phlogiston, and that he did not, seems most manifest from
this, that according to his view the walls of the clay retort deprived the
water of part of its phlogiston, which transuded through the vessel, and
parsed off from its outer suriBeMie into the air. Yet* though the retort was
surrounded by burning fuel, and the escaping phlogiston had the little
water and the much heat necessary to convert it into inflammable air, it
did not appear as a combustible gas, neither did it bum nor produce
water, but it phlogisticated the air, i, e. produced nitrogen.

(3) Watt believed thaf dephlogisticated air can unite in certain degrees
with phlogiston without being changed into water."'^ One of the products
which it then yielded was carbonic acid, as Watt inferred from Priestley's
delusive experiment with iron filings and red oxide of mercury, already
dewribed.f (4) According to Watt another oxide of phlogiston, as we
should now call it, was nitrogen. ^'Phlogisticated air," says he, ''seems
also to be another composition of phlogiston and dephlogisticated air;
but in what proportions they are united, or by what means, is still
unknown. It appears to me to be very probable that fixed air contains a
greater quantity of phlogiston than phlogisticated air does, because it has
a greater specific gravity, and because it has more affinity with water.''

Such then were the contradictory views of Watt, whose caution and
clearness as a thinker were less than a match for Priestley's extraor-
dinary blunders as an experimenter. Cavendish, who founded only on
his own admirable experiments, avoided the errors into which Watt fell,
and differed from him in opinion concerning nearly all the gases. The
former regarded carbonic acid and nitrogen as peculiar bodies. The
composition of the first he did not know, but he seems to hare considered
it as a simple body, as he was justified in doing before it was analysed.
Nitrogen he believed to be a compound of nitrous (nitric) acid and phlo-
giston, in perfect consistence with the tenets of the phlogistians, and in
conformity with all the phenomena which were witnessed, so long as the
balance was not employed. Phlogiston, or the inflammable air from the
metals, he represented as a substance whose only oxide vms water.
According to Watt on the other hand, phlogiston had four oxides, viz.
atmospheric air, nitrogen, carbonic acid, and water. The two latter con-
tained the greater amount of phlogiston, water the most, and atmospheric
air the least. In the language of our modem chemistry, therefore, air
would be the peroxide of phlogiston^ and water the suboxide, whilst car-

♦ Phil. Tram, 1784, p. 334. f Op. et he, ciU Ante, p. 301. % Ibid, p. 335.

THE WATEtl CONTROVERSY. 337

bonic acid might be the protoxide^ and nitrogen the deuioxide. Any one
of these sabstances^ therefore, might be changed into the other^ bj the
loss or gain of oxygen or phlogiston; and can)onic acid^ nitrogen, and
atmospheric air only required the union of so much phlogiston to convert
each into water. Water, therefore, might with as much propriety be
regarded as a compound of phlogiston and carbonic acid; or of phlogiston
and nitrogen; or of phlogiston and atmospheric air; as of phlogiston and
dephlogisticated air.

These erroneous opinions were not offshoots from an originally just
conclusion, that water was the oxide of hydrogen. Some of them preceded
the view that water is an oxide of inflammable air : none were derived
from it, nor did it abolish faith in them, although the notion that atmo-
spheric air is an oxide of phlogiston was placed in abeyance.

It cannot be held, then, that Watt's wider speculations concerning
phlogiston, remove the difficulty that attends the interpretation of that
tenn as used in his conclusion from Priestley's explosion experiments. If
phlogiston could not be held to have specially signified hydrogen, when
applying to experiments made with another gas, still less can it be limited
to that substance, when connected with such speculations as I have con-
sidered.

I come, therefore, to the conclusion, that when Watt, in stating his
views on the nature of water, substituted the word phlogiston for inflam-
mable air, he put it beyond question, that he signified by neither phrase
the gas hydrogen.

9. Experiments and Conclusions of Lavoisier concerning the pro*

duction of Water from its Elements,

*" In conformity with the method followed in discussing the opinions of
Cavendish and Watt, I shall avoid at present, as much as possible, all
reference to the claim to priority of discoveiy, contested between the
French and English chemists, and limit myseli in this section to a con-
sideration of the nature of Lavoisier's observations on the synthesis of
hydrogen and oxygen. To him, in association with La Place, belongs
the entire merit of first consciously analysing water ; nor has any one
claimed this great discovery from him. It forms no part, however, of
the present inquiry, to discuss the particulars of Lavoisier's famous
analysis, although in another section, devoted to the consideration of the
relative merits of the claimants in the Water Controversy, I shall endea-
vour to do justice to the genius and labours of the great French chemist.
At present, I confine myself solely to his synthetical researches, which
in their general nature resemble those of Cavendish.

Lavoisier's views are contained in two papers, the exact titles of which
have been given in the bibliographical section, (ante, p. 268). The one is
solely by him, in the other he had the co-operation of Meusnier. Both
were prmted in 1784, but their contents, or at least those of Lavoisier's
own paper, were known in part in England, before their publication in
the Memoires de TAcad^mie des Sciences, and were referred to by Caven-
dish and Watt, in the final versions of their views concerning the nature
of water.

The paper of which Lavoisier was sole author, is of most importance
to the present inquiry. It is entitled " Meuioire dans lequel on a pour
objet de prouver que Teau n'est point une substance simple," &c. It was

338 CAVENDISH AS A CHEMIST.

read^ as its author informs us, '^^ la Rentree pnblique de la Saint Martin,
1 783 ; depuit on y a fait qudqties additions relatives au travail fail en
commun avec M, Meusnier, sur le meme objet. II anroit du se tronver
place avant celui lu par M. Meusnier^ a la Seance pabliqne de Paqaes,
1734. Voyez, p. 269.*

The paper here referred to, is that entitled '' Memoire ou Ton prouTe,
puar la decomposition de I'ean, que ce fiuide n'est point une substance
simple, &c., par MM. Meusnier et Lavoisier. Lu le 21 Avril, 1784."t
In this communication, which is chiefly occupied with the analysis of
water, Lavoisier refers thus to his memoir on the synthesis of hydrogen
and oxygen, " Ce memoire se trouve dans ce mime volume. C^est par
erreur, qu*il a ete imprime post^rieurement a celui ci."}:

It thus appears that Lavoisier was anxious to have his experiments on
the synthesis of hydrogen and oxygen, considered as anterior in time to
those on the analysis of water. As Mr. Muirhead, however, justly ob-
serves, " although M. Lavoisier's paper was in part read before that by
him and M. Meusnier, yet much of it contains express allusions to that
other, and was therefore written later in order of time, and we have in
the Memoires, as printed, no means of determining precisely the extent of
the additions."!

The reader, therefore, must bear in mind, that though the question of
priority between the French and English chemists, is not under considera-
tion in this section, Lavoisier himself acknowledged having altered and
added to his earlier paper after it was read, so that his statements as they
now appear in the Memoires, must not be regarded as written without a
certain acquaintance with Cavendish and Watt's views. What the extent
of that acquaintance was, is one of the problems in the Water Controversy,,
but Lavoisier did not himself deny that he had obtained some information
from Blagden concerning Cavendish's researches to which he refers in this
paper. And without desiring in any way to prejudge the question of
Lavoisier's originality or fair dealing, I must draw attention to the fact,
that he acknowledges that his priority had been called in question, and
though he does not say who had claimed precedence over him, he refers
pointedly in the course of his narrative to Cavendish, and to him alone,
as having through Blagden asserted, that he had at least observed the
production of water by the combustion of hydrogen and oxygen before
Lavoisier did, so that it cannot be doubted that Cavendish was the party
he was most anxious to show had not anticipated him. In order, accord-
ingly, to vindicate his priority, he gives a brief historical account of the
researches which conducted him to his experiments on the synthesis of
hydrogen and oxygen, the contents of which are important in reference
to the merits of all the claimants in the Water Controversy. I give a brief
abstract of his account, accordingly. || From this account it appears, that
before 1777 Lavoisier was of opinion that inflammable air, in burning,
would yield sulphuric or sulphurous acid. M. Bucquet, on the other hand,
thought that fixed air would be the product of this combustion. To de-

* Watt Corr. p. 171 . All my quotations of these French papers are, taken as stated
already, from Mr. Muirhead's reprints appended to the Watt Corretpondence. I
borrow from him also the paging of the Memoirs^ for the sake of those to whom the
originals are more accessible than the reprints. Thus, the quotation in the text ia from
Watt Corr. p. 171; M^m. deVAead.pour 1781, p. 468.

t Watt Corr. p. 151 ; or M^, de VAcad. pour 1781, p. 269.

X Watt Corr, p. 152; or M6m. de VAcad. pour 1781, p. 269.

\ Watt Corr, p. 152.

II Watt Corr, p. 173 ; or Mem. de VAcad. pour 1781, p. 472. j

THB WATER CONTROVERST. 339

tenniDe this pointy Layoisier and Bucqnet set fire to hydrogen^ and
whilst it was hurning at the mouth of a large bottle, poured lime-water
through the flame into the vessel, but without obtaining any precipitation
of the lime, or evidence of the production of fixed air. This experiment
disproved Bucquet's view, but did not establish Lavoisier's. The latter,
aooordinglj, in the winter of 1781-1782, made another ingenious experi-
ment of the same kind. A large bottleful of hydrogen was Kindled, whilst
lime-water was being poured in, and the mouth of the bottle was immedi-
ately thereafter closed by a cork, through which passed a copper tube,
drawn to a small aperture, and conveying oxygen from a gasholder.
When the cork was introduced, the surfskce of the inflammable air ceased
to bum, but at the end of the copper tube within the bottle, a beautiful
jet of very brilliant flame appeared, ''and we saw,'* says Lavoisier, ''with
much pleasure, the vital air [oxygen] burn in the inflammable air [hy-
drogen], in the same manner, and in the same circumstances, as inflam-
mable air bums in vital air."''^ During this combustion the bottle was
constantly agitated, but no change occurred in the transparency of the
lime-water, and when the experiment was repeated with pure water in-
stead of lime-water, no acid appeared, nor was a weak solution of an
alkali neutralised when substituted for those liquids, f

These negative results, Lavoisier tells us, surprised him the more, that
he had already observed that in every combustion an acid was formed.
Sulphur when burned, yielded sulphuric acid ; phosphorus, phosphoric
acid; charcoal, fixed air; and analogy had led him to conclude "that the
combustion of inflammable air should equally produce an acid."! He
returned to the inquiry in 1783, but before he recommenced his experi-
ments on 24th June of that year, Blagden made the communication to
him concerning Cavendish's having obtained water by the combustion of
hydrogen in close vessels, which Lavoisier acknowledges in the following

terms : "M. Blagden nous apprit que M. Cavendish avoitdeja

essay^, k Londres, de briiler de Fair inflammable dans des vaisseaux
ferm^s, et qu'il avoit obtenu une quantite d'eau tres 8ensible."§

The apparatus which Lavoisier employed in June, 1783, was nearly
identical with that made use of at the present day for the oxyhydrogen
blowpipe. Separate gasholders were employed to contain the oxygen and
hydrogen, which were conducted by flexible tubes of leather to a jet
shaped like the letter Y. The stalk of this drawn to a point, formed the
nozzle at which the mixed gases burned, whilst the two limbs in which
the tubes from the gas-holders terminated, were furnished with stopcocks
by means of which the flow of each gas could be regulated. In adjusting
the relative proportion of oxygen and hydrogen, Lavoisier did not
attempt to measure them out by any numerical scale, but proceeded, as
he says himself, " par voie de t&tonnement," and guided himself as to the
extent to which either stopcock should be opened, by the colour and
brightness of the flame of the mixed gases, which appeared at the end of
the nozzle. " La juste proportion," he tells us, " des deux airs donnoit la

♦ Wati Corr. p. 175; or Mim, de F Acad, pour 1781, p. 471.

t The beautiful experiment recorded in the text, in which oxygen is made to bum
in hydrogen^ haa generally, I believe, been attributed to other and later observers than
Lavoisier, as its first performers; but the merit belongs entirely to him, and the whole
account shows how broad was the view which he took of combustion as a phenomenon,
in which each of the opposite bodies essential to its occurrence is equally concerned, and
may with equal propriety be termed the combustible, or the supporter of combustion.

X Wati Corr. p. 175; or Mem. deV Acad, pour 1781, p. 471.

§ Wail Corr. p. 176; AUm. de V Acad, pour 1781, p. 472.

z 2

340 CAVENDISH AS A CHEMIST.

flamme la plus luminense, et la pi as belle." This point haTing been
settled, and the jet set fire to, the nozzle was inserted, so as to fit air-tigfat,
into the upper tubulure of a glass bell-jar standing over mercury, and tbc
gases were allowed to burn till the supply was exhausted. The first
phenomenon observed was the clouding of the bell-jar by vapour;
speedily drops of liquid appeared and ran down the sides of the vessel, so
that in fifteen or twenty minutes the surface of the mercury was covered
by a lighter fluid. To obtain this, a plate was passed under the bell-jar,
without permitting the mercury in it to escape, and its contents were
transferred to a glass funnel. The mercury was then allowed to ran off,
and the water ^' remained in the tube of the funnel. It weighed a little
less than five drachms."*

The liquid thus obtained, " cette oau," was submitted to *' every test
that could be thought of,'' but appeared as pure as distilled water. How
many tests were tried does not appear. Lavoisier mentions only three.
The water did not redden tumsol, nor render syrup of violets green, nor
precipitate lime-water. " In short," he adds, " none of the known
reagents gave the least indication of impurity." He does not say, how-
ever, that he employed any other reagents than those mentioned above.

The conclusion which these results warranted is then stated.
Lavoisier begins by acknowledging that as the flexible (leather) tubes
which conveyed the gases were not absolutely air-tight, it was impossible
to be certain what was the exact quantity of the two gases burned; bnt
as the whole is equal to its parts, and as pure water alone was produced,
he thought himself entitled to conclude that the weight of water was
equal to that of the two gases which had served to produce it.t To this
conclusion, however, he was plainly not entitled, unless he could show by
other experiments (which he could not) that oxygen and hydrogen were
certainly simple substances, not admitting of decomposition. In the absence
of certainty on this point, it was manifestly quite possible, that one or
both gases might be compounds, which when they seemed merely to unite
and produce water were decomposed, so that certain only of their con-
stituent elements were contained in that liquid. In all Lavoisier's expe-
riments, as appears in the continuation of his paper, the weight of water
was less than that of the gases burned. Till, however, this deficiency in
weight could be shown to result from the unavoidable imperfection of the
process, it could not fail to suggest the possibility of some ingredient of
one or other, or of both the gases having escaped, the absence of which
necessitated that the water should difier from the unbumed gases^ both in
weight and in composition.

Lavoisier was alive to the necessity of establishing identity of weight
between the gases burned and the water produced; and, by his own
showing, he should have suspended his judgment till he had repeated
his experiment several times. He was exceedingly desirous, however, to
carry back his conclusions to the earliest possible date, and hence, appa-
rentier, his anxiety to show tbat the single, imperfect, and insufiicient
experiment on which I have been commenting, entitled him to the

* Lavoisier appears to have shown less than his usual ingenuity in tliis process.
It would have been much better to have dispensed with mercury, and burned the mixed
gases within a dry glass vessel kept very cool. At the end of the combustion, this
vessel could have been closed and weighed, and the whole product of water aacertained.
In the process actually followed, a sensible quantity of water must have been left
adhering to the bell-jar, the plate, the mercury, and the funnel.

t Wali Corr. p. 177; or M^m. de VAcad. pour 1781, p. 473.

THE WATER CONTROVERSY. 241

inference which he drew. The arguments, however, by which he seeks to
justify this, are of no value whatever. The axiom that the whole is
equal to its parts, did not necessitate that these should be oxygen and
Lydrogen; and as little did the purity of the water, so long as the whole
weight of gases expended was not accounted for. The only objection,
nevertheless, which Lavoisier acknowledges, is the possibility of the heat
and light evolved during the combustion being ponderable. That they
were not, ho had ascertained, as he tells us, by an independent inquiry
into their ponderability, an unpublished memoir on which had for some
months been deposited with the Secretary of the Academy.''^ The
certainty, however, thus gained, that the deficiency in weight of the
water, as compared with that of the gases which yielded it, could not be
accounted for by a reference to the escape of ponderable matter in the
form of heat and light, rendered it all the more necessary to explain what
had become of the lost gas.f

It must further be noticed, that Lavoisier's method of experimenting
was less accurate than Cavendish's, in many respects.

1. It does not appear that his gas- homers were graduated, so that he
could ascertain how much gas he expended, and a knowledge of the
capacity of his gas-holders, when full, was not sufficient to determine this,
for he was compelled by his plan of procedure to keep the gases burning
for some time before he could begin to collect the product of their com-
bustion.

2. He adjusted the proportion of the gases by the imperfect test of the
colour and brilliancy of their ilame when burning together, a criterion too
uncertain to be relied upon as a proof that neither gas was in excess, or
escaping, to that extent, combustion.

3. As already noticed, he cannot have collected all the water pro-
duced, but only the surplus which did not adhere to the bell-jar, the
mercury, and the plate, and perhaps also to the neck of the funnel.

4. lie tested the water imperfectly. The employment of lime-water
was superfluous, for he had ascertained previously that when hydrogen
and oxygen burned together their product did not affect that reagent.
All, accordingly, that he observed, was that the water did not contain a
free acid or a free alkali. It might, however, have contained one or more
neutral salts, and fixed organic matter. That it did not, in his experi-
ments. Cavendish established by proving that it left no sensible residue
when evaporated to dr3me8s, besides trying Lavoisier's three tests, and in
addition showing that the water was tasteless}: and odourless. I refer to
these points because Cavendish has been accused of imperfectly analysing
the water produced in his experiments, but in reality his analysis was
much more complete than those of his rivals.

* Wati Corr, p. 178 ; M6m, de VAcad, pour 1781, p. 473.

t The recognition by Lavoisier of the necessity of establishing that heat and light
are imponderable, before it could be inferred that water is the oxide of hydrogen, fur-
nishes the best reply to those critics who have affirmed that Cavendish's determinations,
before and after each explosion, of the weight of the globe in which he detonated
hydrogen and oxygen, and his demonstration that the weight did not alter, proved
nothing of importance to the inquiry, except that his globe was air-tight. Lavoisier
would not have said so, for he instituted a separate inquiry to determine this point,
which Cavendish settled without special research, by the same experiments which
furnished the basis of bis other conclusions concerning the composition of water.

X Lavoisier, however, although he does not mention it himself, appears to have
been aware that the water was tasteless, for his colleague. La Place, states that it had
this character. (La Place to De Luc, Walt Corr. p> 41.)

342 CAVENDISH AS A CHEMIST.

It is impossible, indeed, to read Lavoisier's account of his experim^iti
without saiprise. He was in general so sagacioos, so cantious, so inge-
nious, so accurate and painstaking, that one is startled to find him
summarily coming to a conclusion, after trying a single imperfect and
insufficient experiment. He laid it down himself as a maxim, *' C^est an
reste la multitude des faits, bien plutot que le raisonnement, qui doit ^tablir
toute esp^ce de th^orie nourelle;" and he implicitly obeyed it in his
elaborate researches into the methods of analysing water, in his paper on
which he announces this maxim. It is not a little strange, therefore,
that he should have forgotten it, in his not less important observations on
the synthesis of the elements of water. Nevertheless, although by his own
showing, not entitled to give any better account of his experiments than he
gave of Cavendish's, namely, that "he had tried to bum inflammable air
in close vessels, and had obtained a very appreciable quantity of ivater;"
he reported his solitary trial of the 24th June (1783), to the Academy on
the 25th, in these terms: — "We did not hesitate to conclude that water is
not a simple substance, and that it is composed weight for weight of
inflammable and vital air.'** Why Lavoisier should thus have haat«ned
to publish a conclusion, which by his own confession was based, in one
important particular, upon assumed or hypothetical premises, will be
considered when discussing the question of priority. He repeated the
experiment along with Meusnier many times before he finally and formally
published it.f In the record of these repetitions, he states that the
proportion by volume in which the elements of water combine as gases, is
12 parts of oxygen to 22*924345 (or in round numbers 23) of hydrogen.
He mentions however, that there is " quelque incertitude sur Texactitade
de cette proportion;"! and I need not say that they are not accurate.
12 volumes of oxygen combine exactly with 24 of hydrogen. Lavoisier^s
numbers approximate less closely to accuracy than those (1) given by
Cavendish. Priestley also, as reported by Watt, had anticipated Gay-
Lussac and Humboldt in discovering that the combining measure of
hydrogen was 'about' twice that of oxygen.

From the numbers given by Lavoisier, which he thinks could not difier
much from the true ones, Lavoisier proceeds to calculate the composition of
water by weight. In this calculation he assumes that a cubic inch of oxygen
weighs 0*47317 of a grain, and the same measure of hydrogen 0*037 449.§
These numbers are not accurate : that for oxygen should be 0'3419, that
for hydrogen 0*0213; so that the calculation errs considerably. || Accord-
ing to it, 2 ounces and 58*4 grains of hydrogen combine with 13 ounces
7 drachms and 13*6 grains of oxygen to form IG ounces of water; whereas
in reality 2 ounces by weight of hydrogen combine with 16 of oxygen to
form 18 of water.

* Wait Corr, p. 178; or M4m, de VAcad. pour 1781, pp. 473 — 4.

t The account of these repetitions was read to the Academy, " 4 la Rentr^ publiqae
de la Saint-Martin [NoYember] 1783," but additions were made to it between that
period and Aprili 1784, in which year it was printed, although the volume of the
M^moires which contains it, \s entitled "pour 1781." Watt Corr, pp. 171, 151 and
152; or M€m. de VAcad, pour 1781, pp. 468 and 270.

% Watt Corr. p. 179; or M6m. de VAcad, pour 1781, p. 474.

§ Watt Corr, p. 179; or M6m, de VAcad. pour 1781, p. 474.

II From another part of his paper we learn that he and Meusnier estimated hydrogen
to be 12^ times lighter than air, which explains in part this result. ( Wait Corr, p. 172;
or Mim. de VAcad. pour 1781, p. 468.)

THE WATER CONTROVERSY. 343

From these weights^ again^ the relative yolumes of the gases are calcu-
lated to be —

Cubic Inches*

Oxygen 16919-07

Hydrogen 32321*29

49240-36

whereas, as already mentioned, the number for hydrogen should be exactly
doable that for oxygen.

After announcing this result, Lavoisier proceeds to remark tbat " this
single experiment* of the combustion of the two gases, and their conver-
sion into water, weight for weight, does not permit us to doubt that this
substance, hitherto regarded as an element, is a compound body; but one
fact is not sufficient to certify so important a truth : it is necessary to
multiply facts, and after having composed water artificially, to decompose
it.*'f The remainder of his paper, accordingly, is occupied with a state-
ment of the reasons which led him to anticipate that water would prove
susceptible of decomposition; with a criticism of Priestley's experiments
(already referred to) on the revivification of metallic calces by hydrogen,
in which Lavoisier shows that water must have beeti produced, although
Priestley did not observe it ; with a brief account of his earlier attempts
to decompose water by red-hot metals and other bodies, and his success
with iron, zinc, and charcoal; and with some general observations on the
power of plants to decompose water, and on the vinous fermentation as a
process during which the same decomposition occurs.

This part of Lavoisier's paper does not require criticism. It is charac-
terised by the greatest precision, perspicuity, and sagacity; but as none of
Lavoisier's English contemporaries asserted any claim to a share in the
first analysis of water, it is not necessary to say more.

On reviewing his synthetical researches into the composition of water,
considered as original and independent inquiries, and apart from all ques-
tions of priority, they will be found inferior to the similar researches both
of Cavendish and Monge. In the first place, however, it must be noticed
that they are unexceptionable so far as the substances asserted to be the
elements of water are concerned. The one body Lavoisier employed was
* I'air vital/ or oxygen. The other he named, when free, "air inJlammabU
nqueux,'* and when combined, "prindpe injlammahle aqueux^ for hydrogen
was still as strange a word to him as it was to Cavendish, Watt, or
Priestley. This aqueous inflammable air, he tells us, could be obtained
by decomposing water by iron, or by dissolving iron or zinc in sulphuric
or muriatic acid ; but, unlike the English chemists, he did not infer that
the gas was derived from the metal, but that, in all these cases, it came
from the water (pure or acidulated), in contact with which the metal was
placed. He declined to pronounce an opinion on the question, whether
there are different kinds of inflammable air, or only one, liable to altera-
tion, by mixture or combination with different substances which it can
dissolve. It was sufficient for his purpose to inform the reader, that when
he employed the term " inflammable air," he referred solely to that which
could oe obtained by the processes described above ; in other words, to
hydrogen.! Thus far his experiments were unexceptionable; but in four

* The phrase here is evidently not used in the sense of aolitary trial, but in
reference to a series of trials all of the same kind, which^ taken together, form one
experiment.

t Watt Corr. p. 180; or MSm, de V Acad, pour 1781, p. 475.

X Watt Corr. p. 171 ; or M4m, de VAcad. pour 1781, p. 468.

344 ' CAVENDISH AS A CHEMIST.

points, already 60 far referred to, tliey were less tnistwortby. 1. Tbe
gases were not weighed before combustion, but measured, and their weight
calculated from their volume. 2. The mode of measurement is no4
minutely described; but from the large quantities of gas operated on, the
size of the gasholders, ^nd the necesssiry complexity of the arrangement,
including leather tubes whose capacity could not be exactly ascertained,
it is impossible that delicate measurement can have been practised. The
tubes further appear, from Lavoisier's statement, not to have been air-
tight; and when the combustion of the gases was arrested, they must
have been left full of unmeasured gas. 3. The water produced was not
weighed in the vessel in which it condensed, as it was in the processes
followed by Cavendish and Monge, but was transferred to a separate
vessel, in which its weight was ascertained. The transference, hoirever,
cannot have been effected without loss. Even, therefore, if Lavoisier had
employed dry gases, which, like Cavendish, he did not, and had been for-
nished with accurate determinations of their densities on which to base
his calculations, which he was not; the essential faultiness of his method
of procedure forbade the possibility of his demonstrating that, weight for
weight, the gas burned and the water produced were e^nal. To what
extent the weights differed is not ascertainable, for Lavoisier has not sup-
plied us with the means of determining how great the departure from
absolute identity between them was. 4. The analysis of the water was
imperfect; for, besides the few tests tried in the only case where they are
described, it is difficult to believe that nitric acid should not have appeared
in Lavoisier's experiments as well as in Cavendish's, yet, as the latter
pointedly remarked, none was detected.* The apparent purity of the
water, however, made the interpretation of the results obtained all the
more easy. It is impossible, therefore, to exalt Lavoisier's experiments
above those of Cavendish, as some have done. They are liable to the
same objection as those of Priestley, upon which Watt founded his opinion,
namely, that they did not warrant the conclusion based upon them. And
it is of no little importance to notice, that, in effect, Lavoisier acknow-
ledged as much; for when he refers to Monge's experiments on the product
of the combustion of hydrogen and oxygen, which were made contempo-
raneously with his own, he says of them : " He has operated without io$$y
so that his experiment is miich more conclusive than mine, and leaves
nothing to desire."f This amounts to a confession that, in his own re-
searches, there was loss, and that his experiment was not conclusive.

Lavoisier's conclusions, however, whether he was entitled to them or
not, were stated with a precision and clearness to which the announce-
ments of Cavendish, Watt, and Monge cannot lay claim. They are con-
tained in two brief, but most emphatic lines : " Water is not a simple

* The circnmstances of the two experiments certainly differed ; for in the one, a
large Tolnme of hydrogen and oxygen, mixed with a little nitrogen, was exploded at
once by the electric sparky and in the other, a small jet of hydrogen and oxyg^i, mixed
with a little nitrogen, burned comparatively slowly in an atmosphere of air (or, if the
hydrogen in the jet were in excess, in an atmosphere of nitrogen). The difference in
the circumstances, however, was not apparently so great as to account for the absence of
add in Lavoisier and Meusnier's trials, especially as the proportion in which they
burned the gases, viz. 12 vols, of oxygen to 23 of hydrogen, gave more than a com-
bining volume of the former, and secured the conditions essential to the production of
nitric acid. In a later repetition of Cavendish's experiments, Lavoisier had no diffi-
culty in observing the production of nitric acid. See BerthoUet's letter to Blagden,
quoted on page 343.

t Watt Corr, p. 178 j or Mint, de VAcad. pour 1781, p. 474.'

—- M

THE WATER CONTKOVERSY. 345

• »

subetance, but is composed, weight for weight, of inflammable and vital
air.'** Lavoisier, therefore, certainly announced the true doctrine of the
composition of water.

This seems the proper place for mentioning a fact which has not been

preferred to by any of the writers on the Water Controversy, and probably

vrzB not known to them ; namely, that Lavoisier made a second repetition

of Cavendishes experiments after the publication of the latter*s paper. I

have learned the circumstance from a letter addressed by Berthollet to

Blagden in 1785, with the loan of which I have been favoured by the

executor of the latter, R. H. Blagden Hale, Esq., of Cottles, Melksham,

Wiltshire. What follows, is the portion of BerthoUet's letter referring

to the Cavendish experiments; the remainder is occnpied with the account

of the discovery of a comet in the constellation Andromeda, by M. Mi-

cLain ; with certain observations of Abbe Haiiy on Crystallography ;

and with references to a paper by La Place ; to certain chemical analyses

by M. Pelletier ; and to a Cometographie, by M. Pingre.

To Dr. Blagden, Gower street, Bedford Square, London.

Paris, 19 Mars, 1785.

Monsieur, — L'on s'est beaucoup occupe ici ces derniers tems de la belle
decouverte de Mr. Cavendish, sur la composition de Feau. Mr.'Lavoisier
a t&ch^ de porter sur cet objet toute I'exactitude dont 11 est susceptible.
Ayant fait part de son projet k T Academic, elle jugea qu'on ne devait rien
negliger dans une experience qui doit jeter du jour sur un si grand nombre
de faits, et elle chargea la classe de chymie d y assister et de lui en rendre
compte.

Je ne vous decrirai pas les details de cette experience pour laquelle
on n*a epargn6 ni depeuses ni soins ; mais je vais vous en donner une idee:
1/air d^phlogistiqu^ a ^te retire du precipite rouge et. a pas86 trois fois k
travers de Falkali caustique. Le gas inflammable a ete degage de Teau
m( me qu*on a iait couler dans deux tubes de fer qui contenaient des lames
de fer contoumees en spirales ; avant de faire couler Teau qui etait distille,
on a fait le vide dans Tappareil. On a rempli d'air dephlogistique un
Fraud recipient auquel venaient aboutir d*un cot6 un tube qui amenait
Fair d^phlogistiqu^, et de Tautre un tube qui amenait le gas inflammable,
et le gas et Tair 6taient determines par des pressions egales a venir dans le
recipient a mesure que la combustion se faisait.

11 s'est forme cinq onces et demi dVau qui contenaient environ qua-
rante grains d*acide nitreux : 11 n'y a eu oue tres peu de residu. La
quautite d*air d^phlogistiqu^ qui est entree aans la composition de Tcau
a ete en poids k cette du gas inflammable a peu pres conime 81 a 1 9 et
cette de 1 air dephlogistique qui s'est fixe dans le fer dans la decomposition
de Teau a ete dans la memo proportion avec le gas inflammable qui s*est
degage.

Mr. Lavoisier vent repeter Texperience en faisant brCiler Fair dephlo-
gistique dans le gas inflammable, et il y a apparence qu*alors on n aura
point d*acide nitreux, selon les belles observations de Mr. Cavendish. Mr.
De Laplace a preuve d'aprds tout ce qu'on salt deja que Tacide nitreux
etait un compose de gas inflammable et d'une beaucoup plus ^rande
quantite d'air dephlogistique qu'il n*y en a dans Feau, et que le gas
nitreux tlent le milieu entre Tacide nitreux et Teau ; mais d apres ce

• WaU Corr. p. 178; or M€m. de VAeed, pour 1781, pp. 473, ATA.

346 CAVENDISH AS A CHEMIST;

quo Tous m'avez ecrit dans votre demiere lettre, il y a grande appareoce |
que Mr. Cavendieh laissera peu a desirer sur cet objet ; lorsqae^ vous
pourrez m'instruire des demieres experiences que voua m'aTez hat que
m'annoncer, je vous en serai fort oblige. i

Mr. Lavoisier a In nn memoire dans lequel il a explique par les diffe- '
reutes affinites de Fair dephlogistiqu^ les precipitations mutnelies des
metaux de leur dissolutions acides. II a r^pete I'experience de Mr.
Priestley sur la revivification du precipite rouge par le moyen du fer, el
il n*a point retire d*air fixe dans cette operation comme rannonoe Mr.
Priestley.

II attribue tout le gas inflammable qu'ou retire des differentes dissoln-
tions metalliques k la decomposition de I'eau; il croit que tous les metaax
ont besoin d'etre dans Tetat de cfaaux pour etre tenus en dissolution, et
que pour etre reduits en cbaux ils prennent par Tintermede de Tacide, Tair
depblogistique d'une portion d^ean et que de la vient le degagement dn
gas inflammable de cette eau. II attribue cependant dans certains cas
Tair depblogistique qui s'unit au metal a Tacide lui-meme; ainsi 11 croit
que lorsqu*un metal forme de Facide sulfureux avec Tacide vitriolique,
cela depend de ce qu'il s'empare d'une partie de Fair depblogistique qui
est dans Facide vitriolique ; il explique de meme le degagement du gas
nitreux dans les dissolutions par Facide nitreux.

Votre tr^s bumble et tr^s obeissant serviteur^

[Signed] Berthollet.

Tbe main points in tbe letter, it will be seen^ are, that hydrogen and oxy-
geu were burned together; that water was obtained coniaming nitric aculf
and leaving (apparently on evaporation) a slight residue ; or exactly the
results which Cavendish obtained, and which he was surprised, especially
in reference to the acid, that Lavoisier had not obtained. Even the slight
residue was found which has been objected to in the criticisms of certain
of Cavendish's results, as showing that he never procured pure water.

Berthollet declares that the water produced contained its constituents
in the proportion of 81 parts, by weight, of oxygen, to 19 of hydrogen.
The latter number is perhaps a mistake for 10, which would represent tbe
quantity of hydrogen which can unite with 80 of oxygen, to produce 90
of water.

How those numbers were obtained, is not mentioned. It is manifest,
however, from the appearance of nitric acid in the water, that the oxy-
gen, instead of being in the proportion of about half a combining mea-
sure, must have been in excess. This was evidently Lavoisier's opinion,
who proposed to bum oxygen in hydrogen, so as to keep the latter in
constant excess, in the expectation of verifying Cavendish's observation
that no nitric acid would be produced.

La Place's erroneous conclusion that nitric acid is a higher oxide of
hydrogen than water, is a tacit compliment to Cavendish's sagacity in
avoiding an inference which seemed almost forced upon the observer.

The exact period when this second repetition of Cavendish's experiment
was made, is not stated, but it cannot have been very long before the date
of the letter j otherwise Berthollet, who regularly corresponded with Blag-
den, would have sent him earlier notice of it. It should thus seem that a
considerable period elapsed before Cavendish's paper in the Philosophical

THE WATBR CONTROVERSY. 847

Transactions, 1784^ was known in France, a point of some importance, as
he has been accused of sending in great haste misdated copies of his
paper to the Continent.

10. Experiments and Conclusions of Monge concerning the result
of the inflammation of Hydrogen and Oxygen in close vessels,

Monge's experiments are detailed in a paper contained in the Me-
moires de T Academic des Sciences, for 1783 (printed in 1786) pp. 78 to
88. Mr. Mairhead has reprinted the paper along with an illustrative
wood-cut in the "Watt Correspondence," pp. 205-218, and from his re-
print I make my quotations.

Monde's researches are important, as the^ formed an original and
independent inquiry which, had it conducted him to the same conclusions
as it did Layoisier, would have constituted him the discoyerer of the
composition of water in France, and hare placed him on the same level of
merit as its discoverer in England. His paper is entitled, " M^moire sur
le R^sultat do Tinflammation du gas inflammable, et de lair dephlogis-
tique, dans des yaisseaux clos." The experiments recorded in it were
made, he tells us, at Mezieres in June and July, 1 783, and repeated in
October of the same year. " I did not then know," he adds, " that Mr.
Cavendish had made them several months before in England, though on a
smaller scale ; nor that MM. Lavoisier and Laplace had made them about
the same time at Paris, in an apparatus which did not admit of as mnch
precision as the one which I employed."*

Mongers experiments differ from those of Priestley, Watt, Cavendish,
and Lavoisier, in being limited to an inquiry into the nature of the pro-
duct yielded by the combustion of inflammable air and oxygen, and in
being prosecuted without any hypothesis as to the probable result of the
combustion. In what exact sense he used the term inflammable gas, does
not precisely appear, but he did not restrict it to a single substance, for
for when referring to the temperature at which a mixture of oxygen and
*' gaz inflammable'^ takes fire, he observes that this "depends upon the
nature of the inflammable ga8,"f and he refers in illustration, to what
occurs durin? the combustion of a candle, and of boiling oils.

He used, however, only one kind of inflammable gas in his experiments.
This was obtained by dissolving clean iron filings in diluted sulphuric acid,
80 that it certainly was hydrogen. The oxygen he employed was obtained
by heating red oxide of mercury j and he had recourse to many precau-
tions to secure the purity of both gases from admixture with atmospheric

* Watt Corr, p. 206; or Mim. deVAead. pour 1783, p. 79. The originality of
these researches is denied by Blagden, who says, " Mr. Monge's experiments (of which
Mr. Lavoisier speaks as if made about the same time) were really not made antil pretty
long, I believe at least two months, later than Mr. Lavoisier's own [query, Mr. Caven-
dish's], and were undertaken on receiving information of them." (Letter to Crell,
translated in Watt Corr. p. 73.) Blagden's denial, however, is ciutte irreconcilable
with Lavoisier's declaration, that a few days after (qnelques jours apres) the 25th June,
1783, Monge sent an account of his experiments in a letter to M. Vandermonde, who
read it to the Academy. {Watt Corr. p. 178; Mim. de V Acad, pour 11 %\, p. 474.)
As Monge's paper was not printed till 1 786, the year in which Blagden's letter was pub-
lished, it is probable that he had not read the statement quoted in the text when he
wrote to CreU, but referred only to reports which had reached him of the experiments
made in October, 1783, which he supposed to be Monge's earliest researches. The
latter seems to have been quite willing to concede the priority of Cavendish's experi-
ments, and the independence of Lavoisier's.

t Watt Corr. p. 218; or M^. de V Acad, pour 1783, p. 88.

•348 CAVENDISH AS A CHEMIST.

air. Each was collected over water in a carefally graduated bell-jar,
provided above with a tubulure and stopcock. To effect the combastioa
of the gases^ a glass globe or balloon was employed, provided ^rith an
arrangement for passing the electric spark, as in Cavendish's apparatus.
Mouge's balloon, however, had three apertures, one communicating by a
stopcock and tube with an air-pump, by means of which the balloon was
exhausted, and two, communicating respectively by metallic tubes with
jthe stopcocks of the bell-jars containing the hydrogen and oxygen. This
arrangement was rendered necessary by the mode in which he experi-
mented. He was most anxious to prevent any intermixture of the gases
with atmospheric air, and to secure this he provided separate channels for
the hydrogen and oxygen, which were shut off from communication Trith
the outer air, till the experiment was concluded. He did not, moreover,
mix the gases in their combining proportion, and transfer the mixture to
the detonation-globe, as Cavendish did; but having ascertained their den-
sities so that he could convert volumes into weights, he made a vacuum
in the balloon, and transferred a certain portion of each gas into it, bj
opening the stopcock of the bell -jar which contained it. For reasons ^wfaicli
he does not assign, he preferred to introduce at first into the balloon ^7^^
of its capacity of oxygen, and -^ths of hydrogen. An electric spark was
then passed, and an explosion determined. A twelfth part of its volume
.of oxygen, was, thereafter, introduced for the second time into the balloouy
and a second explosion determined; and this process was repeated vrithoat
renewing the hydrogen, till &ve or six explosions had occurred, when the
supply of that gas was replenished, and the process proceeded as before.
At the termination of the 137th explosion, the vacuum was renewed, with
a view especially to empty the balloon of incombustible elastic fluid, wbich
was supposed to be accumulating in it. This emptying of the balloon was
repeated twice, before the conclusion of the experiment. The gas thus
withdrawn was transferred to a receiver and preserved for subsequent
examination, and the explosions were renewed till 370 had occurred, when
Monge found that he had consumed 145 pints ^^ of hydrogen, and 74
pints -^ of oxygen, numbers, which I need not say, make a close approxi-
mation to accuracy, if considered as representing the combining volumes
of hydrogen and oxygen ; but it will presently appear that a certain part
of the gases escaped combustion. Those measures calculated from Monge*s
own determinations of the sp. gr. of hydrogen and oxygen, and corrected
for a change in the height of the barometer, which had occurred during
the experiment, corresponded to 3 ounces, 6 drachms, and 27 '56 grains by
weight.

The amount of water which the combustion of the gases had yielded,
was ascertained by weighing the balloon, first with the liquid which it con-
tained, and a second time after it had been emptied and dried. The differ-
ence gave the weight of the liquid, which amounted to 3 ounces, 2 drachms,
and 45*1 grains. The gas withdrawn from the balloon by the air-pump
was then weighed, and found to amount to 2 drachms 27*91 grains, which,
added to the weight of water found, make up 3 ounces, 5 drachms, 1*01
grains, so that there was a difference between the weight of gases burned
and of water found, of 1 drachm, 26 55 grains.*

* This is Mongers calculation, but will presently appear, that the gas withdrawn
by the air-pump was not entirely hydrogen and oxygen; so that the difference between
the weight of these gases consumed, and of water produced, was a little greater than is
stated.

}

THE WATER CONTROVERSY. 349

Monge remarks that this difference may have resulted^ 1. From his
liaving taken the mean height of the barometer daring the experiment, as
the basis of correction of volume for the whole quantity of gas expended,
instead of correcting each quantity transferred to the balloon, according
to the height of the barometer at the period when it was measured. 2.
From his not having observed the changes in temperature in the bell-jars
containing the gases, occasioned by their proximity to the balloon, which
was heated by the explosions determined within it. 3. From the loss
occasioned by evaporation during each renewal of the vacuum.

The air withdrawn by the air-pump from the balloon amounted to

seven pints, and contained y'^th of its volume of carbonic acid. When

this was removed by lime-water, the residue was fired by the electric

spark, (which shows how needless Mongers repeated renewals of the vacuum

in the balloon were), and was thereby diminished -Jth of its volume, from

'which it was inferred to contain a mixture of hydrogen and oxygen.

The residue was incombustible in the air, but gave ruddy fumes with

nitric oxide, and thereafter underwent contraction " like atmospheric air.''

From this result, Monge inferred that the residual gas contained ;|th of its

volume of oxygen, without, however, showing that he was entitled to

this conclusion.* The exact accuracy, however, of his statement on this

point, is unimportant. From his whole analysis it appears, that the residual

gas consisted of a mixture of carbonic acid, hydrogen, oxygen, and

nitrogen. From its composition, Monge inferre<( that it could not be

regarded as a product of the combustion, but that it was a result of the

impurities contained in the hydrogen and oxygen. These impurities, he

thought, probably came, in part, from the air of the vessel in which the

hydrogen was generated ; in part, from the water of the apparatus, which

was agitated several times during the transference of the gases ; and iu

part from the water employed to dilute the sulphuric acid.

The liquid which collected in the balloon was perfectly transparent.
It reddened blue turnsol paper " imperceptibly;'* which Monge explains
by adding, that the reddening effect was much less than in a previous
experiment,t and was less than that occasioned by saliva. This acidity
was not owing to carbonic acid, for the liquid had no action on lime-
water. It rendered solutions of nitrate of silver and of mercury very
slightly opalescent, from which we may infer the presence of a trace of
some chloride ; but Monge did not draw any conclusion from the faint
action of these reagents. The liquid had no other peculiarity, except
an empyrenmatic odour, like that of distilled water. Monge, accordingly,
inferred that the liquid was pure water containing a small quantity of
sulphuric acid, which the hydrogen had carried over with it from the
solution of iron that yielded it. This ingenious observer, however^ it will

* The groundfl on which it wob founded are doubtful. To prevent mistake, I give
Honge's own words. After stating that the residual gas was not combustible in the
atmosphere, he adds, ** Mais par son melange avec I'air nitreux, il a rutil6, et s'est
encore r^uit comme I'air atmospbcrique. II contenoit done encore ii cette ^poque un
quart de son volume d'air d^phlogistique." ( Wait Corr. p. 215 ; or Mim. de VAcad.
pour 1783, p. 86.) He seems to have inferred that the production of ruddy fumes and
reduction of volume, proved the residual gas to be atmospheric air^ and ihtrrfort to
contain Jth [^th] of oxygen.

t The reference to a previous experiment is important, and is explained by
Mongers statement, that he repeated in October, 1 783, trials which he had made in
Jane and July of the same year. His first experiments were reported, soon after this
performance, to the French Academy, as Lavoisier informs us in his memoir. ( Wait
Corr, p. 178; or Mem, de VAcad, pottr 1781, p. 471)

350 CAVENDISH AS A CHEMIST.

be ohaerved, did not examine the liquid with any test for snlphnric acid, bat
only supposed it to be present. That it was^ may be doubted, consider-
ing the large volume of water with which the hydrogen was washed, and
over which it stood ; but whether or not sulphuric acid was present, it
seems certain that the acidity of the water (which, in the earlier experi-
ment, was manifesty well marked) was owing partly to the presence of
nitric acid. The conditions of Monge's experiments were the same as
those of Cavendish's; and it is certain from Monge*s analysis of the
gas extracted by the air-pump, that nitrogen was mixed with the
hydrogen and oxygen detonated in the balloon. Nitric add, therefore,
must have been produced.*

The presence of this acid, then, we may reasonably infer, was overlooked
by Monge, and the conclusion which he was at liberty to draw from his
experiments was rendered thereby the simpler. He begins by observing,
that a part of the water found was certainly contained in solution in the
gases, but that the whole cannot be accounted for by a reference to this;
for in that case, hydrogen and oxygen would each consist of nothing bat
water on the one hand, and the incondensible matter of fire and of light
(la matiere du feu, et de celle de la lumiere) on the other. He infers,
therefore, that when pure hydrogen and oxygen are detonated together,
the only products are — ^pure water, the matter of heat, and the matter of
light. Beyond this inference, however, he reaches only an alternative
conclusion. ''It remains to determine positively, whether — ^the two
gases being solutions of different substances in the fluid of fire (le fluide
du feu) considered as a common solvent — ^these substances, in conseqnence
of the combustion, abandon the solvent, and combine to produce water,
which will thus no longer be a simple substance ; or whether the two
gases being solutions of water in the different elastic fluids, these aban-
don the water which they dissolved in order to combine and form the
fluid of fire and of lie:ht, which escapes through the walls of the vessels;
in which case fire will be a compound substance. "t

Monge thus would only say, that his experiments demonstrated that
either water or fire was a compound substance; but he thought additional
experiments of another kind were needed to determine which of the two
entities was to be considered no longer a simple body. He points out
that the supposition that water is a compound of (the bases of) hydrogen
and oxygen, would explain the functions of water in vegetation, and
account for many other phenomena, such as the moistening of cold sur-
faces by the flame of vegetable bodies; the condensation of water in the
chimneys of stoves ; and the violence of the detonation of gunpowder,
which he thought must be referred solely to the vaporization of the

* In Monge's process matters were so arranged, that in oertain of his trials there was
a gradually diminishing excess of hydrogen over oxygen during the first five of six explo-
sions whidi were as many as could be performed without replenishing the balloon with
hydrogen. During those five explosions, no nitric acid could be produced , but at the
sixth, the volumes of oxygen and hydrogen were equal, so that there was half an equiva-
lent too much of the former, and this excess could not but form nitric acid vrith the
nitrogen present.

t The original is as follows: " U reste k savoir actuellement si les deux gax ^tant
des dissolutions de substances diff^rentes dans le fluide du feu consid^r^ comme dis-
solvant commun, ces substances, par Tinflammation, abandonnent le dissolvant et se
combineht pour produire de Teau qui ne seroit plus alors une substance simple; on bien
si les deux gaz ^tant les dissolutions de I'eau dans des fluides ^lastiques diffi^rentes, ces
fluides quittent Teau qu'ils dissolvoient pour se combiner et former le fluide du feu et de la
lumiere qui s'echappe & travers les parois des vaisseaux : et alors le feu seroit une
matiere compos^e." {Wati Corr. p. 216 ; or M4m. de VAcad. powr 1783, p. 87.)

THE WATER CONTROVERSY. 361

water produced by its combustion. But whatever weight these con-
siderations had in inclining him to consider water a compound, " this
hypothesis/' as he himself calls it, presented a difficulty which he could not
surmount. Inflammable gas and oxygen were well known to require
only a simple elevation of temperature to determine their combustion; in
other words, according to his view, to determine their separation respec-
tively into so much fire, and so much of two substances (the basis of
oxygen and inflammable gas) which united to produce water. But if
fire were the solvent of these bases, and prevented their precipitation and
combination, why should the introduction of fire into a mixture of the
gases, %,€,, an increase of the solvent of their bases, determine a total
separation of these from their solvent) According to Monge's hypothesis
this must occur ; but it seemed to him totally opposed to all that was
observed in the analogous operations of chemistry. He concludes the
paper accordingly with this remark, which is best given in his own
words. " II nous manque done encore beaucoup de lumi^res sur cet
objet, mais nous avons droit de les attendre, et du temps, et du concours
des travaux des Physiciens."*

It is unnecessary to criticise Monge's conclusion further than to say,
that it plainly excludes him from a claim to be considered a discoverer
of the composition of water. It would be a fault, however, in any his-
torian to pass him by unnoticed and uncommended, and in another
section I shall endeavour to do his merits justice. At present I would
only comment on his experiments, as compared with those of Cavendish;
for, as Lavoisier acknowledged the superiority of Monge*s processes to
his own, it is unnecessary to contrast the methods of the two French
philosophers.

Monge himself appears to have thought his method superior to that
of Cavendish, inasmuch as the experiments of the latter were " plus en
p>etit." Lavoisier thought that his countryman's experiment " ne laisse
rien a desirer ;" and recent critics of the Water Controversy have com-
mended the French experiments as superior to those of Cavendish.
Monge's experiments deserve great praise, and must be placed far above
those of Lavoisier, not to speak of Priestley's inexplicable results.
Whatever exceptions, in truth, a searching criticism may take against
certain parts of Monge's process, no candid reader of his memoir will deny,
that his researches made so close an approximation to accuracy, as fully
to entitle him and every one else to infer, that water is a compound of
hydrogen and oxygen. Nor can it be doubted, that though no experi*
ments had been made in England, Monge's experiments would have
conducted Lavoisier to this conclusion. Nevertheless, Monge's method
of procedure was decidedly inferior to that of Cavendish. 1 . He gained
nothing by the greater scale on which he experimented, for he observed
nothing that Cavendish did not observe, and he overlooked the produc«
tion of nitric acid which Cavendish detected. He was a loser, indeed,
by the magnitude of his operations ; for it is a strange though common
error to imagine that experiments cannot be performed on too large a
scale ; whereas, in quantitative determinations a limit is opposed to all
endeavours greatly to enlarge the scale of operation by the increased
difliculties which attend the accurate measurement of large weights,
volumes, or manifestations of force. The scale of Monge's operations
rendered it almost impossible, that all the measurements and weighings

* Walt Corr. p. 218; or MSm. de I' Acad, pour 1783, p. 88.

352 CAVENDISH AS A CHEMIST.

of gas and liqaid, and all the requisite observations of tbe tbermomcteraad
barometer, should be quite accurately made; and they protracted the experi-
ment over a period of time, which increased the difficulties attending it3
performance. By employing smaller quantities, he could have dispensed
altogether, as Cavendish did, with any consultation of the two latter
instruments, and could have weighed and measured with more precision.

2. The principle on which Monge proceeded was, to make the same
experiment at once qualitative and quantitative, so that, for the sake of
obtaining a large quantity of liquid for analysis, he rendered all his mea-
surements less precise than they would have been, had the scale of his
operations been smaller. Cavendish avoided this, by making two sets of
experiments — one on a comparatively large scale to ascertain the nature
or quality of the product of the combustion of hydrogen and oxygen; the
other on a much smaller scale to determine the amount or qtuintity of
this product.

3. Mongers apparatus was needlessly complex, and its coinplexity
multiplied the chances of error. In Cavendish's arrangement, the
gases were first mixed in their combining proportion, and thus a single
bell -jar served to contain both. The empty globe was weighed before
and after each passage of the electric spark, so that every explosion wa5
a separate experiment, supplying an additional numerical datum. In
Monge's experiments, on the other hand, the gases were mixed in tbe
balloon in various proportions. There was but one (double) weighing,
and the expenditure of 219 pints of gas by 370 explosions, supplied but a
single quantitative result.

4. The analysis both of the gas withdrawn from the balloon, and of
the liquid collected in it, was imperfect. I have already referred to the
obscurity attending Mongers account of the conclusion he drew from the
action of nitric oxide on the residual mixture of nitrogen and oxygen
withdrawn from the balloon, and need not further allude to it.

Monge's analysis of the water, moreover, was also defective, not only
because, like Lavoisier's examination, it included the application of fewer
tests than Cavendish's analysis did; but because a positive imparity,
namely an acid, was detected in the water, and nevertheless the inquiry
was brought to a close without the nature of the acid being ascertained.
Monge assumed it to be the sulphuric, without even applying the tests for
that acid. How important an oversight this was, appears from tbe
valuable results which attended Cavendish's exhaustive inquiry into the
nature and source of the acid which occasionally showed itself in his
experiments, as a product of the combustion of apparently pure hydrogen
and oxygen.

I might make further objections, but I do not wish to be hypercriticaL
It was necessary for the defence of Cavendish to urge that beautiful in
many respects as Monge's experiments were, they were decidedly inferior
to those of Cavendish.

THE WATER CONTROVERSY. 353

THE QUESTION OP PRIOEITY.

11. Who first discovered and taught that Water is a compound of
Hydrogen and Oxygen ? Date of Cavendishes experiments and
conclusions.

The question of priority between Cayendisb, Watt, Lavoisier^ and Monge^
will not require much discussion, so far as the last is concerned, for he frankly
concedes that Cavendish's experiments were of an earlier date than his own,
and he did not draw such a conclusion from these as entitles him to be called
a discoverer of the composition of water. The same will be said of Watt*s
claims, by those who agree in the opinion respecting the erroneous naturo
of his fundamental experiments and conclusions, which has been stated
in Sections 6, 7, 8 ; and it might seem at first sight a needless under-
taking to attempt to settle at what period erroneous conclusions were
first drawn from inaccurate experiments. But Watt's claims cannot be
dismissed in this summary way by those who disallow the validity of
his conclusions, for they were accepted as well founded by his contem-
poraries for many years at least, aud were regarded as identical with
those of Cavendish and Lavoisier. Nor can it be doubted, that Watt's
exposition of the compound nature of water materially conduced to the
adoption of that view, especially in Great Britain. His claim to priority,
accordingly, cannot bo overlookedj and in discussing it, I shall in general
assume for argument's sake, so far as it is practicable, that his views
regarding the composition of water were unexceptionable, and concern
myself chiefly with the date of his conclusions.

As for Lavoisier, he indirectly asserted in his own name a claim to
priority over Cavendish, and cannot therefore be passed by. The
question, moreover, between the English claimants, cannot be aiscussed
without large reference to his proceedings; and whatever opinions may be
formed regarding his originality or fair dealing, all must acknowledge that
after the amplest assignment of merit to Cavendish or to Watt, or to both,
there will remain a large share of praise for Lavoisier, to which no one
else can lay any claim. I begin the discussion with the consideration of
the evidence in favour of Cavendish's priority.

Dates of Cavendish's Experiments and Concltmons,

It will form the best preparation and apology for the length of this
section, if I begin by at once drawing attention to what the friends of Caven-
disli have been too reluctant to acknowledge, namely, that the date of his
conclusions is uncertain. I cannot but concur with Lord Brougham when
he says, '' that there is no evidence of any person having reduced the
theory of composition [of water] to writing, in a shape which now
remains, so early as Mr. Watt."* The vrriting here referred to goes back
to April, 1783, whilst there is no holograph nor other document of
Cavendish's containing an explicit declaration of his conclusions, of earlier
date than the manuscript of his first '' Experiments on Air," read to the

♦ Historical Note. Wait Corr, p. 252.

2 A

354 CAVENDISH AS A CHEMIST.

Royal Society in January, 1784. So far, therefore, as direct documentsiy
evidence is concerned, Watt's claims over Cavendisli can be antedated by
more than eight months. The advocates of Cavendish are thus under the
same necessity of appealing to indirect evidence to support his claim to
priority, which the advocates of Watt are under, when seeking to esta-
blish the validity and accuracy of his conclusions. The indirect evidence,
however, in the case of Cavendish is of a very weighty and important
kind. It consists chiefly of three written documents: 1. Cavendish's
Laboratory note-book, containing the record of all his experiments on air.
2. Two sUktements by Priestley in his paper " on the seeming conversion
of Water into Air:" and 3. The letter from Blagden to Crell, already
referred to. These documents and certain other records, which will be
referred to in the course of the discussion, remove all doubt as to the date
of Cavendish's experiments, and throw much light on the date of his
cancltmons. The advocates of Watt, indeed, at least in England, have
conceded the originality and priority of Cavendish's experiments; (in this
dissenting from Arago, with whom Dumas is understood to agree;) and
have joined issue with the advocates of Cavendish, solely, or almost
solely, as to the originality and priority of his conclusions. According to
Arago, " Priestley records in detail, and as his own, experiments which
prove that the water produced by the combustion of a mixture of hydrogen
and oxygen has a weight exactly equal to that of the two gases which
are burned. Cavendish, some time after, claims this result for himself, and
insinuates that he had communicated it verbally to the chemist of Bir-
mingham."* Sir David Brewster, on the other hand, says, " The friends
of Mr. Watt in England .... have acknowledged the priority of
his [Cavendish's] experiments to the full extent that it has been proved;"
and again, ''we have admitted the originality and the value of Mr.
Cavendish's experiments."t Lord Jefl'rey still more explicitly says, in
reference to the years 1781 and 1782, "It is not, we think, to be denied,
that Cavendish then performed those experiments, and observed those
results from which either he himself at the time, or Watt at an after
time, or he again after Watt drew the grand and momentous conclusion,
that water was not a simple but a compound body. "J Even, however, if
the French Academicians were at one with Watt's Engli«$h friends, in
conceding the priority of Cavendish's experiments to those (whatever
they were) upon which Watt founded his theory, it would be impossible
to separate the question of the date of Cavendish's experiments from that
of the date at which conclusions were drawn from them; the two there-
fore must be considered together. And this seems the proper place for
referring more particularly than has hitherto been done, to Cavendish's
manuscript note-book or journal.

Through the good offices of the Rev. W. V. Harcourt, and the courtesy
of the Earl of Burlington, I have been favoured with a perusal of the
original MSS.; and I can bear testimony to the perfect fidelity of the litho-
graph fac-simile published by Mr. Harcourt in the postscript to his address
to the British Association in 1 839. § This fac-simile includes only a portion
of the Note-Book, but all, however, that refers to the product of the
combustion of hydrogen in air or oxygen. I have found nothing, more-

* Eloge qf Jamet Watt. Tranalation by Mr. Muirhead, Waii Corr. p. 230.
t A'oWA British Review, Feb. 1847, pp. 502, 503. This paper is referred to as
Sir David Brewster's, in the Edinr, Rev, Jan. 1848, p. 84.
X Edinr. Rev, Jan. 1848, p. 72.
$ Brit. Anoc, Rep. 1839. Postscript to President's Address, pp. 22—69.

THE WATER CONTROVERSY. 355

ovev, m t&e omitted passages which could in any way tell against Caven-
dish, so that those who have access only to the lithograph fac-simile may
rest satisfied that nothing has been kept back which would materially
affect the determination of the question under discussion. On the other
band a complete publication of the Note-Book (which, however, its size
forbids) would, incidentally at least, strengthen the case of Cavendish's
advocates, by showing the continuity and consistency of his whole re-
searches from 1778 to 1785. In referring to this Note-Book, I shall
quote from Mr. Harcourt*s lithograph copy, when it contains the passages
"which I wish to refer to, that the reader may have the means of verifying
the quotations. The nnlithographed portion of the MSS. will be quoted
from the original (the property of Lord Burlington), which is paged and
indexed by Cavendish himself. The following description by Mr. Harcourt
will explain the present condition of the Laboratory Notes : — " They fill a
volume of unsewn and single, but paged and indexed, octavo sheets, in
liis own hand, bearing dates from February, 1778, to May, 1785, in the
following proportions; in 1778, thirty-three pages; in 1780, thirty-six;
in 1781, seventy-five; in 1782, forty-five; in 1783, fifty-three; in 1784,
forty-four; in 1785, thirty-three. I found them in a packet entitled
* Experiments on Air.' "*

In judging of the contents of these Notes, it is of the utmost import-
ance that the reader should keep in recollection that Cavendish's obser-
vations on the synthesis of the elements of water did not constitute an
isolated inquiry, but formed an integral portion of a continuous series of
experiments on air extending from 1778 to 1785. This has not been
observed, or at least has not been admitted by the advocates of Watt.
*' If," says Sir David Brewster, " he [Cavendish] had discovered the com^
position of water in July, 1781, is it credible that he would have kept it
secret till January, 1784, and that he would then have brought before the

fublic so great a discovery under the title of ' Experiments on Air V ''t
t was in entire consistence, however, with the whole train of his re-
searches, and with the contents of the paper published in 1784, in which
he discussed much more than the composition of water, that he should have
entitled his communication " Experiments on Air." He gave the same
title to his paper on the synthetical production of nitric acid, and for the
same reason, namely, because the production from their elements of this
acid and of water, was ascertained in the course of an inquiry into the
products of the oxidation of bodies in air. In consequence, however, of
this fact being overlooked. Cavendish's experiments on the production of
water are generally referred to, as if they had originated solely in the
observations of Warltire and Priestley on the Detonation of Inflammable
Air, with a view to test the ponderability of heat, of which they are
alleged to have been an avowed repetition. This they certainly were,
but they were at the same time something more, as we have already seen
in part, in section 5. It was stated there, and also in the abstract of
Cavendish's " Experiments on Air" (Ist series, see pp. 141 and 142) that
his researches contemplated the ascertainment of the changes which
various combustibles, not inflammable air only, produced upon C/Ommon
air when burned in it till they had deprived it of the power of supporting
combustion, or in the language of the day had phlogisticated it. Caven-
dish's own words in his introduction to the first series of " Experiments
on Air," are — ** The following experiments were made principally with a

* Brit Ammoc, Rep. 1839, p. 32.

t North BrUish Rev. Feb. 1847, p. 494.

2 a2

356 CAVENDISH AS A CHEMIST.

view to find out the cause of tbe diminution which common air ie mA
known to suffer by all the various ways in which it is phlogisticated, ud
to discover what becomes of the air thus lost or condensed." The evideiioe
of this has been given already^ so that we are now concerned only with
the date of the researches referred to. They were ultimately embodied,
to the extent they were made public, in the paper on a New Eudio-
meter, as well as m both the series of '' Experiments on Air;" bni many
were never published.

From the lithographed earlier portion of the Note-Book, it appears
that in 1778 Cavendish investigated the action of nitric oxide on air, and
on oxygen, and ascertained, as Lavoisier had already done, that nitne
acid was produced, but not carbonic or sulphuric acid. In this year also
he made many trials as to the best way of employing nitric oxide in the
eudiometer, and executed several approximative analyses of air.*

No experiments are recorded in the Note- Book under date 1779,
but in 1780 they were resumed, and a great part of those observa-
tions was made, which are recorded in the paper on the new eudiometer.
The action also of liver of sulphur on air and on oxygen, and the effect
of various so-called phlogisticating bodies, such as burning sulphur, spirits
of wine, and luna cornea on air, were ascertained, and trials were made
as to the vis inertice of nitrogen;! and the nature of the gas produced by
heating together nitre and charcoal. The whole of these experiments,
without exception, had for their object the determination of the properties
of atmospheric air, and were chiefly intended to bring to light the products
of its phlogistication, and demonstrate the essential nature of that proces.

The most important result of these inquiries was the determination,
with a near approach to accuracy, of the relative amount of oxygen and
nitrogen in common air, without a knowledge of which it was plainly
impossible to experiment to purpose on the effect of combustibles on the
atmosphere. I have given previously the result of Cavendish's analysis
of common air as calculated from his standard for oxygen, but a direct
announcement of it occurs in his Note-Book. He mentions that common
air diminished by liver of sulphur lost i\t of its bulk, and states what its
test appeared to be by his own and Ingenhouszt*s eudiometer. He then
adds, ^' In all probability both methods make the air appear better than
it really is; we will suppose therefore that tbe quantity of pure air in
common air is ]f ."J

Having thus, then, ascertained that the reduction of volume effected on
air by its phlogistication (t.e. deoxidation) was about one-fifth, and having
ascertained also that neither carbonic, sulphuric, nor nitric acid was the
invariable product of the phlogistication of air, he proceeded to the
experiments with hydrogen, the date of which chiefly concerns us.

It will be remembered that in a passage added to his paper between
the period of its being read to the Royal Society and printed. Cavendish
stated that all the experiments recorded in the earlier part of his Memoir,
on the explosion of Inflammable Air with Common and Dephlogisticated
Airs, except those which relate to the cause of the acid found in the

* Unlith, MSS. pp. 1 — 12, also 24 — 31. Mr. Harcourt has drawn attention to
these researches in the postscript of his address to the Brit. Assoc. {Report, 1839,
p. 34.)

t The curious experiments in question may be regarded as the earliest recorded
observation on gaseous efTnsioD, the laws of which have been so beantifally investigated
by Prof. Graham.

t Unlith. MSS. p. 109.

THE WATER CONTROVERSY. 357

"vrater, were made in the summer of tbe year 1T8], and were mentioned
by him to Dr. Priestley.* It is of the utmost importance^ therefore, to
ascertain what those experiments were.

With the extreme aiution and accuracy which characterised his re-
searches. Cavendish hegan by ascertaining " whether there was any pene-
tration of parts on mixing common and inflammable air by means of the
eudiometer. "f In other words, he was anxious to learn whether the
mere mixture of inflammable air with common air would lead to the de-
oxidation of the latter, without the application of flame, or the transmis-
sion of the electric spark. He fouud, on mixing the gases, that the
diminution in bulk did not exceed 7/77^ of the whole; and manifestly
concluded that the diminution was not real, but only apparent. He further
tried whether a mixture of inflammable and common air remained uni-
formly mingled, or separated after standing some hours into a lighter and
bear ier portion, and he came to the conclusion that the gas from the lower
part of the vessel in which the mixture was made appeared to contain
very little, if at all, more common air than that from the top.t

Having thus ascertained that inflammable and common air might be
mixed without acting upon each other, and might be preserved in a state
of nearly uniform mixture for many hours, Cavendish proceeded to repeat
the experiments of Warltire and Priestley on the explosion of inflammable
air with common air and oxygen. These experiments, it will be remem-
bered, were performed by Warltire with a view to settle the question of the
ponderability of heat, and by Priestley, as ^'a random experiment'' to amuse
bis friends. Cavendish's first recorded observation on the subject was per-
formed on the 3rd of July (1781), and is entitled '^ Explosion of inflammable
air by electricity in glass globe, to examine Mr. Warltire's experiment."§
Six different explosions, in which the proportion of hydrogen and common
air was varied, were made, and the results are tabulated. || These experi-
ments were performed on July 3rd, 4 th, 5th, and 6th. The specific
gravities of the residual gasIT were tried on July 17th (p. 118); and on
August 4th and 7th, comparative experiments were made with hydrogen
from zinc and iron (pp. 120 — 122). The first reference to the produc-
tion of water in the globe experiments occurs under date August 4th : —
'' The globe was dry before firing, but was immediately covered with dew
on firing." But the appearance of water must have been observed before;
for the experiments recorded in the published paper of 1784, as made by
burning together hydrogen and air, were made in July; and on a Sunday
in that month, 135 grains of water were obtained by their combustion,
and the liquid ascertained by analysis to be pure.** A result of the same

* PhiL Trans. 1784, p. 134.

t Unlith. MSS. p. 113.

X UnlUh. MSS, p. 114. This experiment was a remarkable anticipation of tbe
similar researches made at a later period, by Hope, Dalton, and Graham, on the homo-
geneousness of gaseous mixtures, and the diffusion and effusion of gases. I need not
saj that Cavendish's observations were quite in keeping with those of Prof. Graham, for
be drew off the gas somewhat slowly by a syphon ; and we know that, in these circum-
stances, hydrogen diffuses out of a vessel in larger volume than a heavier gas, such as
air, does.

$ Liih.MS8,p. 115.

II Lith. MSS. p. 119. Substantially the same table is given in the Experiments
on Air. PhiL Trans. 1784, p. 127.

f Nitrogen, pure, or mixed with a little hydrogen or oxygen, which remained
nncombined after the explosion.

•• TMh. MSS. p. 127.

358 CAVENDISH AS A CHEMIST.

kind, obtained long after, follows this in the Note-Book, bearing date
November 1 8th, 1782.*

A considerable portion of August was spent in making obserratwni
with an instrument for measuring the force of the explosions whidi
occurred, when different proportions of hydrogen and bit or oxygen were
burned together. The apparent object of these experiments was, as Mr.
Harcourt suggests, to furnish " a test of the identity and purity of the
gases, as well as of their combining proportions."!

During June and July, many preliminary experiments were made
with oxygen ;t but the first explosion of oxygen and hydrogen of which
a recortl is given, occurs under date September (1781), when thirty grams
of acid water are said to have been produced, which jrielded nitre when
saturated with alka]i.§ A similar experiment is recorded on September
28th, as having been made with washed oxygen, mingled w^ith a little
less than twice its volume of hydrogen, when fifty grains of add water
were produced. The last experiment of this year was made on October
20th, by exploding oxygen from red lead with a little less than twice its
volume of hydrogen, and the resulting water was found to be acid.

*'Such," as Mr. Harcourt observes, " were the experiments made in
1781 concerning the reconversion of air into water, by decomposing it in
conjunction with inflammable air, which Priestley and Cavendish men-
tion as having been communicated to the former, and repeated, m
consequence, by him in April, )783."||

1 have given these dates somewhat fully, and the reader can easily
follow them in Mr. Harcourt's lithograph of the MSS. in the British
Association Report, 1839. It appears from them that all the experiments
which Cavendish ultimately adduced, as justifying the condosion that
water is a compound of oxygen and hydrogen, were made by him in 1781,
as he asserts in the first passage which he added to his paper of 1784.
He h:id also observed in this year, as he affirms in that memoir, tbe
production of nitric acid when apparently pure hydrogen and oxygen are
detonated together; but he had not, as yet, discovered the cause of this
acid appearing, as the terms of the interpolated passage also imply : so
that his published paper of 1784 exactly tallies with the records of the
Note-Book, so far, at least, as we have yet compared them.

The inquiry seems now to have been laid aside for about a year,
during which Cavendish made a number of additional experiments on tbe
points essential to tbe construction of an accurate eudiometer. These
required many laborious observations on the properties of oxygen,
nitrogen, and nitric oxide, besides various tedious manipulations with
different forms of eudiometer, and a multitude of analyses of different
specimens of air. Tbey are recorded in various parts of the Note-Book
from p. 154 to p. 199, and were, no doubt, preparatory to the publication
of his account of the new eudiometer, which was laid before the Royal
Society in January, 1783. IT

* Liih.MSS.p. 128.

f Postscript to Address to Brit. Assoc. Report, 1839» p. 36. It appears to bsTe
been the indications of this instrument which led Cavendish to his well-founded con-
elusion, that the cause of the absence of nitric acid, when hydrogen and air, instead of
hydrogen and oxygen, are burned together, is the reduction of temperature below the
combustion- point of nitrogen, occasioned by the excess of it present.

t Unlit h. MSS. pp. 54—107.

§ Lith, MSS, pp. 108 and 146.

II Postscript to Address to Brit. Assoc. Report, 1839, p. 37.

^ It will be shown further on, that Cavendish's experiments on air, involved the

]

THE WATER CONTROVERSY. 359

In October^ 1782, Cavendish resumed bis inquiry into tbe cause of the
production of nitric acid wben oxygen and bydrogen are bnmed together;
and his experiments with the oxygen obtained from plants exposed to
sunlight* were made in October, November, and December of this year.f
In the coarse of these experiments with oxygen from plants, the propor-
tion of the gas was varied, and excess of oxygen was found to increase the
production of acid. The early part of January, 1 783, was occupied with
similar experiments on mixtures, in various proportions, of oxygen from
red precipitate and inflammable air, and led to tlie same result. X On
the same day, apparently, January 12th, or at some period between that
and the 18th, nitrogen was purposely mixed with a measure of hydrogen,
and a little more than a combining measure of oxygen, and the resulting
water was found to be evidently acid. These experiments were repeated
on the 18th and 27th of tbe month. Similar experiments are also
recorded out of their order, as made on January 3rd, 4th, and 24th.§
Trials are also recorded, as mad^ with air from Turbith mineral, at p. 240,
but no date is given. The previous sheet is dated July, 1783 ; but the
Kote-Book does not follow an invariable chronological order, for pages
purposely left blank are occasionally occupied with experiments made at
intervals of even two years from each other. Other researches, which
were reported in the paper of 1784, are mentioned under various dates
during 1783 : thus, in March, we have the record of the " Examination
whether air yields any nitrous acid, when phlogisticated by lime-liver of
sulphur. "II This is followed by an account of the investigation " Whether
red precipitate contains any nitrous acid /'IT and that by the description
of the process for preparing oxygen from Turbith mineral,** and a record
of the results obtained when oxygen thus prepared was detonated with
hydrogen.ff A considerable space is then occupied with the account of
experiments on the gas evolved when a mixture of nitre and charcoal is
heated, and also on the gas yielded by the distillation of charcoal.]!]:

The conclusions to which those experiments conducted Cavendish, are
referred to in his published paper of 1784, in his illustration of his views
concerning nitrogen being a compound of nitrons acid and phlogiston. §§
These records are followed by the account of the observations on the
action of sunlight on tinctures of green leaves, and on dephlogisticated
spirit of nitre (colourless nitric acid) to which he specially refers in illus-
tration of his views of the growth of vegetables. || jj They were made in
December, 1783. TIT Pages 270-277 contain notes chiefly on the deflagra-

prosecution of a triple inqairy, viz. Ist, the analysis of air, including the demonstration
of its constancy in composition; 2nd, the discovery of the general product of combustion
in air and oxygen, which led directly to the discovery of the composition of water; and
3rd, the detection of the source of the nitric acid, which occasionally accompanied the
water resulting from the combustion of hydrogen. These three main inquiries branched
off into endless minor ones, and necessitated new investigations into the properties of
all the substances employed. The interval we are considering was, to a great extent,
spent in such an investigation into the properties of oxygen and nitrogen, as materially
conduced to the explanation of the production of nitric acid.

• Phil. Trans. 1784, p. 151.

t Liih. MSS. p. 200. t LUh. MSS. p. 210.

$ Lith. MSS. pp. 216, 217, & 218.

II Uniith. MSS. pp. 232—236; Phil. Trana. U84, p. 141.

^ Unliih. MSS. p. 237; Phil. Trans. 1784, p. 142. x

•• Phil. Trans. 1784, p. 132.
- ft Uniith. MSS. pp. 240—242; Phil. Trans. 1784, p. 133.

tt Uniith. MSS. pp. 244—265. $$ Phil. Trims. 1784, p. 135.

Phil. Trans. 1784, p. 147. f1[ Uniith. MSS. p. 266.

360 CAVENDISH AS A CHEMIST.

tion of nitre with tin and iron filings.* The snooeeding pagvs, item. 278
to 281, record observations on the quantity of fixed air diacemible by lime-
water. Page 282 refers to the evolution of inflammable air £rom zinc in
caustic alkali. Pages 283-285 detail farther experiments on the quantity
of fixed air perceptible by lime-water. Pages 286-291 are on the distilla-
tion of red precipitate with iron-filings.t Pages 292-295 describe a pro-
cess by which red precipitate was prepared, and the next four pagce^
296-299, refer to experiments made on the combustion of sulphur and pfaos*
phorus, on March 6th, 1784. Pages 300 to 301 recount experiments on
the preparation of dephlogisticated spirits of nitre,j: made in April and
November, 1783. This closes the register of experiments referring to
the first series of experiments on air. Page 302 describes a process re-
ferred to in the second series of these experiments (those, namely, on the
production of nitric acid by the combination of nitrogen ami oxygen) ;
and at 306 the record begins of the experiments made on the converti-
bility of common air into nitric acid by sending electric sparks through
it, which are described in the paper published in 1785.§ The remaining
dates all belong to 1784 or 1785.

Such were the experiments recorded by Cavendish up to the period
when he read his paper to the Royal Society. I have reported thera
somewhat fully, comparing them with the index furnished by himself, as
it is a point of importance in reference to the delay which attended the
publication of his experiments of 1781, to show in what way he spent the
interval between that period and January 15th, 1784. bates are not
attached to all the pages of his Note-Book, so that the exact period when
each experiment was performed cannot be ascertained. As late, however,
as January 4th, 1784, an experiment is reported on the deflagration of
tin with nitre, which probably formed part of the researches into the
rationale of the evolution of oxygen from melted nitre, referred to in the
published paper. The latest certain date is December, 1783, when the
experiments on the action of sunlight on vegetable tinctures and colour-
less nitric acid were made. The day of the month is not given in refer-
ence to the tinctures, but the acid is stated to have been "exposed for
two or three days to a weak sunshine, about December 25th, 1783.*'||
It thus appears that Cavendish did not complete the researches which
he published in 1784, till at least Christmas,] 1783 ; and as his paper
was read on the 15th of the succeeding January, and was therefore
probably transmitted to the Secretary of the Royal Society some days
before, and as from its length it must have occupied some considerable
time in drawing up, Cavendish cannot be accused of having been slow in
communicating his results to the public.

Such, then, are the contents of the MS. journal up to 1784. It will
presently be referred to more particularly, when the question is under
consideration ; Was it a journal only of experiments, or of experiments
and conclusions ?

* These experiments are not specially referred to in the published paper, bat it can-
not be doubted that they form part of the series which led Cavendish to the conclusions
that he published concerning the cause of the erolution of oxygen from melted nitre.

t These experiments were made partly with a view to test the truth of Priestley's
declaration, that iron yields inflammable air when heated^ the accuracy of which state-
ment Cavendish doubted (Phil, lYang. 1784, p. 137); partly with a view to furnish the
rationale of the evolution of oxygen from red precipitate, to which he makes so many
references. (Ibid. pp. 142 — 144.)

X Phil. TVans. 1784, note, p. 148.

§ Experiment on Air (2nd series), Phil, Trane. 1785, p. 372.

II Unlith, M8S. p. 267.

THE WATER CONTROVERSY, 361

II. The second docament referred to, as throwing light upon the question
before iis, is Pr. Priestley's " Experiments relating to phlogiston, and the
seeming conyersion of water into air/' read to the Hoyaf Society, Jane
26th, 1783. Two statements occur in the latter section of it which treats
of the conversion of water into air, in which Cavendish's researches are
referred to. The first of these has been quoted already in full, so that a
part only of it need be given here. " Still hearing," says Priestley, " of
many objections to the conversion of water into air, I now gave particular
attention to an experiment of Mr. Cavendish's concerning the reconver-
sion of air into water, by decomposing it in conjunction with inflammable
air."^ Then follows the detailed account of the repetition made with
inflammable air from charcoal, and oxygen from melted nitre ; (see ante,
p. 286) which need not be quoted, except the following sentence, — " the
result was such as to afford a strong presumption that the air was recon-
verted into water, and therefore, that the origin of it had been water. "+
At the close of his paper Priestley again refers to the subject, whilst
describing an experiment of his own, which he says " cannot be explained
60 well on any other hypothesis [than that of the convertibility of water
into air], any more than Mr. Cavendish's experiment on finding water
on the decomposition of air." j:

These, unquestionably, are the passages to which Cavendish refers^
when he states that in consequence of what he told Dr. Priestley, the
latter " made some experiments of the same kind as he relates in a paper
printed in the preceding volume of the Trausactions."§ The acknowledg-
ment that Cavendish's experiments preceded and suggested Priestley s,
was thus made by the latter, before it could be claimed by the former.
It may further be added, that in the abstract of Priestley's paper, drawn
up by Mr. Maty, (Sec. R. S.,) and contained in the journal book of the
Boyal Society, it is stated that '* these arguments received no small confir-
mation from an experiment of Mr. Cavendish, tending to prove the
reconversion of air into water ; in which pure dephlogisticated air and
inflammable air were decomposed by an electric explosion, and yielded a
deposit of water equal in weight to the decomposed air."||

Watt also added a note to his paper before it was printed, in which
he says, " I believe that Mr. Cavendish was the first who discovered that
the combustion of dephlogisticated and inflammable air produced moisture
on the sides of the glass in which they were fired."ir This statement is
not accurate, as the friends of Watt and of Cavendish equally (though for
difierent reasons) acknowledge, but its accuracy is immaterial. I refer to it
only as showing that Watt believed that Cavendish had preceded Priestley
in the performance of certain experiments with inflammable air and oxygen.
It thus appears from the testimony of Priestley and Watt; from the
journal book of the Royal Society ;'. from the public declaration of Caven-
dish at the period when he published his researches; and from the
private record of his experiments at the period when they were per-
formed, that before January, 1783, he- made those observations on the
production of water from the combustion of hydrojjen and oxygen, which
he adduced in his paper of 1784, as showing that water is a compound
of these gases. It cannot be doubted, therefore, that the account given
in Ara^o's Eloge of Watt, which represents Priestley's experiments as
original^ and as having preceded Cavendish's, is inaccurate.

♦ Phil. Trans. 1783, p. 426. t Op. cit. p. 427.

t Ibid. ibid. p. 433. % Phil. Trans. 17S4, p. 134.

II Quoted by Mr. Harconrt, Brit. Assoc. Rep. 1839, p. 44.
H Phil. Trans. 1784, note, p. 332.

362 QAVENDISH AS A CHEMIST.

It remains then onlj to inquire whether the experimentg of tbd
French chemists were of prior date to Cayendish's. But ou this qneettoA
there is no room for doubt. Monge'*^ (ante, 353) frankly conceded that
the English experiments were of earlier date than his own, for which he
only claimed, as he was fully entitled to do, independence and originality.
Lavoisier also acknowledged that Blagden had informed him of Caren-
dish^s experiments before he performed his own (ante, 353).f Bat, ac-
cording to his informant, he concealed the amount of information giTen
him. The following is Blagden's letter in reply to Lavoisier, addressed to
Dr.Crell, and published in his "Chemische Annalen" for 1786. I quote it
at length, as translated by Mr. Muirhead,| because, apart from the liffht
which it throws upon the date of Cavendish's experiments, it is regarded
by tbe advocates of Mr. Watt as perhaps the document of moat impor-
tance to them in disproving the priority of Cavendish's conclusions to
those of Watt.

Translation of a Letter from Dr. Blagden, Sec. R. S. L, to

Dr. Lorenz Crell. Not dated.

** I can certainly give you the best account of the little dispute about the first
discoverer of the artificial generation of water, as I was the principal instrument throogk
which the first news of the discovery that had been already made was communicated to
Mr. Itavoisier. The following is a short statement of the history : —

"In the spring ('Friihjahr') of 1783, Mr. Cavendish communicated to me and
other members of the Royal Society, his particular friends, the result of some experi-
ments with which he had for a long time been occupied. He showed us, that oat of
them he must draw the conclusion, that dephlogisticated air was nothing else than
water deprived of its phlogiston ; and vice versd that water was dephlogisticated air
united with phlogiston. About the same time ('um dieselbe Zeit') the news was
brought to London, that Mr. Watt of Birmingham had been induced by some obser-
vations to form (' fassen') a similar opinion. Soon after this (* bald darauf ') I went to
Paris, and in the company of Mr. Lavoisier, and of some other members of the Royal
Academy of Sciences, I gave some account of these new experiments and of the opinions
founded upon them. They replied, that they had already heard something of these expe-
riments, and particularly, that Dr. Priestley had repeated them. They did not doubt,
that in such manner a considerable quantity of water might be obtained; but they felt
convinced that it did not come near to the weight of the two species of air employed ; on
which account it was not to be regarded as water formed or produced out of tbe two
kinds of air, but was already contained in and united with the airs, and deposited in their
combustion. This opinion was held by Mr. Lavoisier, as well as by the rest of the
gentlemen who conferred on the subject; but as the experiment itself appeared to them
very remarkable in all points of view, they unanimously requested Mr. Lavoisier, who
possessed all the necessary preparations ('Vorrichtungen'), to repeat the experiment
on a somewhat larger scale, as early as possible. This desire he complied with on the
24th June, 1783 (as he relates in the latest volume of the Paris Memoirt), From Mr.
Lavoisier^s own account of his experiment, it sufficiently appears, that at that period he
had not yet formed the opinion that water was composed of dephlogisticated and inflam-
mable airs; for he expected that a sort of acid would be produced by their union. In
general, Mr. Lavoisier cannot be convicted of having advanced any thing contrary to
truth ; but it can still less be denied, that he concealed a part of the truth. For he
should have acknowledged that I had, some days before, apprised him of Mr. Cavendish's
experiments, instead of which, the expression " il nous apprit," gives rise to the
idea that I had not informed him earlier than that very day. In Uke manner, Mr.
Lavoisier has passed over a very remarkable circumstance, namely, that the experiment
was made in consequence of what I had informed him of. He should likewise have

* Wait Corr. {Mim, par M. Monge), p. 206; or Mim. de VAead, pour 1783,
note, p. 79.

t Watt Corr, (M%m. par M. Lavoisier), p. 176; or Mim, de V Acad, pour 1781,
p. 472.

X Watt Corr, pp. 71—74.

THE WATER CONTROVERSY. 363

stated in hit publication, not only that Mr. Cavendish had obtained " nne quantity
d'eau tr^s sensible," but that the water was equal to the weight of the two airs added
together. MoreoTer, he should have added, that I had made him acquainted with
Messrs. Cavendish and Watt's conclusions ; namely, that water, and not an acid, or
any other substance ('Wesen'), arose from the combustion of the inflammable and
dephlogisticated airs. But those conclusions opened the way to Mr. Lavoisier's present
theory, which perfectly agrees with that of ^Mr. Cavendisli; only that Mr. Lavoisier
accommodates it to his old theory, which banishes phlogiston. Mr. Monge's experi-
ments (of which Mr. Lavoisier speaks as if made about tihe same time) were really not
made until pretty long, I believe at least two months, later than Mr. Lavoisier's own^
and were undertaken on receiving information of them.

"The course of all this history will clearly convince you, that Mr. Lavoisier
(instead of being led to the discovery by following up the experiments which he and Mr.
Bacqnet had commenced in 1777) was induced to institute again such experiments solely
by the account he received from me, and of our English experiments; and that he really
discovered nothing but what had before been pointed out to him to have been previously
made out and demonstrated in England."

Cavendish also stated, in one of the additions made to his paper, that
a friend of his (no doubt Dr. Blagden) ''gave some account of them to
M. Lavoisier, as well, as of the conclusion drawn from them, that dephlo-

g'sticated air is only water deprived of phlogiston ; but at that time, so
>r was M. Lavoisier from thinking any such opinion warranted, that till
he was prevailed upon to repeat the experiment himself, he found some
difficulty in believing that nearly the whole of the two airs could be con-
verted into water."*

To neither of these representations did Lavoisier make any reply, and
his account is at variance with that given by his colleague. La Place, in
a private letter to De Luc, which has been brought to light by the publi-
cation of the Watt Correspondence. The letter was written on June 28th,
1783. In it La Place says, " M. Lavoisier and I have repeated recently
before Mr. Blagden and several other persons, the experiment of Mr.
Cavendish upon the conversion into water of dephlogisticated and inflam-
mable airs, by their combustion; with this difierence, that we have burned
them without the assistance of the electric spark, by bringing together
two currents, the one of pure air, the other of inflammable air. We have
obtained in this way more than 2^ drachms of pure water, or which, at
least, had no character of acidity, and was insipid to the taste ; but we do
not yet know if this quantity of water represents the weight of the airs
consumed. It is an experiment to be recommenced with all possible
attention, and which appears to me of the greatest importance. '*t

In the face of Blagden 's, Cavendish's, and in efiect La Place's, denial
of the accuracy of Lavoisier's statements, it is impossible to accept his
version of matters, even to the extent of acknowledging his experiments
to have been original and independent, like those of Monge ; and as the
earliest date, moreover, to which his experiments go back, is 24th June,
1783, he cannot possibly contest priority with Cavendish, whose experi-
ments had been repeated by Priestley in the spring of that year, before
Lavoisier had heard of them.

Whatever, then, is doubtful in the Water Controversy, this at least
is certain, that Cavendish was the flrst who observed that when given
weights of hydrogen and oxygen are burned together in certain propor-
tions, they are replaced by the same weight of pure water. The whole
dispute, so far as priority is concerned, thus turns upon the date of
Cavendish's conclusions, and if they can be shown to have been deduced

• Phil. Trans. 1784, p. 134. f ^«'' (^^rr. p. 41.

364 CAVENDISH AS A CHEMIST.

from his experiments at the period wben thejwere made, the eontroverey
is at an end. It wiU suffice, nioreoirer, if they can be shown to lia^e been
arrived at, at any period before the time when Priestley repeated
Cayendish's experiments, and gave the account of his repetition to Watt^
on which the latter founded his conclusions. Priestley's paper, it wiU be
remembered, on which Watt's letter was a commentary, was accompa-
nied by a letter from the former, dated April 21, (1783),* so that to
establish priority for Cavendish, his conclusions must be shown to have
been drawn at least before that date; and as Priestley's experiments
must have been made some time before that period, which marks the date
of their completion, and the reduction of his conclusions to the written
form in which they now appear, we must go back somewhat earlier in
seeking for the precise time when Cavendish's conclusions were drawn.
From a statement in the Watt Correspondence (p. 17) it appears that we
must go back at least a month; for Watt was acquainted with Priestley's
experiments as early as the 26th of March, 1783. Wedgwood was made
acquiunted with them on the 23rd of the month, (ante, p. 94.) Beyond these
allusions there exist no means of limiting the date more precisely; but it is
probable that Priestley repeated Cavendish's observations in the beginning
of this month; for all the earlier dates of the Watt Correspondence refer
to experiments on the conversion of water into air. Without, therefore,
affecting greater precision^ I shall for the present assume that the question
for consideration is : Did Cavendish come to the conclusion that water is
a compound of oxygen and hydrogen, before Priestley repeated his experi-
ments in March, 1783? The advocates of Watt affirm that he did not,
and they adduce in support of their view the following arguments.

1. They affirm that a passage in Cavendish's paper of 1784, referring
to Watt, contains what is equivalent to a confession of obligation on the

?art of the former to the latter, which amounts to a concession of priority.
f this allegation be true, the controversy is closed ; for if Cavendish con-
fessed himself indebted to Watt for his conclusions, all further proof of
obligation is superfluous. Watt's letter of April 26, 1783, was not pub-
licly read to the Royal Society till the 22nd of April, 1784, that is, several
months after the reading of Cavendish's paper on the 15th of January of
the same year.t No reference whatever was made to Watt in the latter
essay, wben it was read ; but before it was printed, and after Watt's
views were made public. Cavendish added the following passage : " As
Mr. Watt, in a paper lately read before this Society, supposes water to
consist of dephlogisticated air and phlogiston deprived of part of their
latent heat, whereas I take no notice of the latter circumstance, it may be
proper to mention in a few words the reason of this apparent difference
between us. If there be any such thing as elementary heat, it must be
allowed that what Mr. Watt says is true," &c.X

The italics are not in the original ; the remainder of the interpolation
is occupied solely with a vindication of the phraseology which Cavendish
employed in reference to heat, and has been given already in the abstract
of his paper (ante, p. 156). This is the passage relied upon, as containing
Cavendish's acknowledgment of Watt's originality, and, by implication,
priority. Sir David Brewster says of it, — " That he [Cavendish] acted
yuffenerously to Watt, his best friends must admit, — for he has admitted
It himself. The omission, in his ' Experiments on Air/ of all notice of

* PMl, TraM, 1783, p. 398. + Phii, Trans, 1784, p. 330.

t Phil, Trans, 1784, p. 140.

THE WATER CONTROVERSY. 3G5

Mr. Wait and bis tlieory^ was unworthy of a philosopher; and he stands
self-condemned, because he corrected the omission before he printed his
paper/'* Sir David's judgment proceeds upon the supposition that Caven^
dish was indebted to Watt's letter for the views which he published on
the nature of water. But even if he had been (which he was not), it
was impossible that he could have referred to Watt at an earlier period
than he did, for the latter had debarred this by withdrawing his letter to
Priestley, which was not publicly communicated to the Royal Society till
Cavendish's paper had been read. Up to this period it was a private
letter, the public reading of which its author spontaneously desired might
be delayed to an indefinite period. No one, therefore, had a right, even
if aware of its contents, to make public reference to it, nor was it possible
to discover by any public or authoritative document, what portion of it
Watt wished to disavow. Had Cavendish, for example, referred to Watt
as believing that water by distillation in earthen retorts could be con-
verted into atmospheric air, the latter might justly have denounced him
for quoting a private letter of April, 1783, as showing his opinions in
January, 1784. No one, in truth, can blame Cavendish, because he did
not refer to the contents of a letter which by its author's express wish
was not made public.f

Had Cavendish, nevertheless, been personally acquainted with Watt,
and conscious, moreover, that he was indebted to him for his conclusions,
he would have acted dishonestly, not to say ungenerously, had he not
privately communicated with him before publishing his paper. But
these illustrious men were as yet complete strangers to each other, and no
one will affirm that if Cavendish's conclusions were the fruit of his own
unaided researches, he was under the slightest obligation to acknowledge
that another had drawn similar conclusions from a repetition of his experi-
ments; and that the reference to Watt quoted above, admitted no obliga-
tion to him, will be apparent, I think, to every impartial reader. The
passage, in truth, which should be consulted in full, contains a denial
rather than an acknowledgment of obligation to Watt; inasmuch as it is
is mainly occupied with the expression of a refusal to adopt his views.
Watt, as was natural to one who had made the laws of heat a subject of
deep study, attached great importance to the part which caloric played in
transmuting inflammable air and oxygen into water. He gave' great
prominence, accordingly, to the discussion of this in his paper; whilst
Cavendish, who was a disbeliever in the materiality of heat, made no
reference to its evolution as essential to the production of water from its
elements, and referred to Watt's paper simply to explain why he did
not.];

2. It is contended by the advocates of Watt, that Cavendish's Note-
Book contains no record of any conclusion having been drawn by him
from his experiments on the synthesis of hydrogen and oxygen, and that
therefore he cannot be held to have come to a conclusion regarding the
nature of water before Watt's letter was written.

• North Brit. Rer, Feb. 1847, p. 505.

t PhU. Trans, 1784, p. 330.

X It seems well to notice here, that Cavendish's refusal to adopt Watt's language
in reference to heat, was not an opinion expressed then for the first time^ as if with the
purpose of accoanting for an omission in his own paper or lessening the merit of his
rival's views. In 1783, Cavendish, in his commentary on Mr. Hutchin's observations
on the freezing of mercury , had stated at length, that he dissented from Black's views on
the latency of heat, and preferred those of Newton, according to which heat was a state
or condition of matter, and not a substantial entity.

366 CAVENDISH AS A CHEMIST.

When Mr. Harcourt published his lithographed fac-simile of pari of
the Note-Book, he observed regarding it, " Pew rongh day-books of expe-
riments would tell their own tale with such certainty and distinctness as
these; in few, could the consecutive course of reasoning be traced ihoB
clearly from the experiments themselves: there cannot remain a doubt on
the mind of any one who reads them, that in January, 1783, Cavendish I
had not only discovered the certain fact that oxygen and hydrogen in
definite proportions form water^ but likewise the strong probability that
oxygen and nitrogen form nitric acid, two months before Priestley b^an
to experiment, and Watt to speculate, on the notice which Cavendish had
given the former of the composition of water, and four months before
Lavoisier received from Blagden a similar notice.'** To this statement
Lord Brougham replies, " I must add, having read the full publication
with fac-si miles, Mr. Harcourt has now clearly proved one thing, and it I
is really of some importance. He has made it appear that in all Mr.
Cavendishes diaries and notes of his experiments, not an intimation occurs
of the composition of water having been inferred by him from those
experiments earlier than Mr. Watt's paper of Spring, 1783."t

On the same subject Sir David Brewster remarks, ** The publication of
the lithographs of Mr. Cavendish's experiments upon air, which com-
menced in July, 1781, are regarded by his advocates as'establishing his claim
to the discovery of the composition of water."^ He then states his dissent
from this view, observing, *' It is indeed beyond all belief that he [Caven-
dish] could have drawn any conclusions, or formed any theory, seeing that
Mr. Hudson, to whom the Duke of Devonshire had entrusted the whole
of his papers, has declared that he does not find in these journals of the
experiments, anything more than the simple statement of the facts without
any casual mention of theoretical opinions."§ On Mr. Hudson's state-
ment Mr. Muirhead comments as follows: ^* This material fact has since
been placed beyond the possibility of doubt, by the publication of the
journal in question; in the whole course of which Mr. Cavendish does not
make a single inquiry into the cause of the appearance of the water, nor
indicate the most remote suspicion of its real origin: never using any
expressions which could imply an union of the two airs, or which are
inconsistent with the notion which Warltire and Priestley had entertained,
of a mere mechanical deposit of the water."|| Lord Jefirey is much more
guarded in the expression of his opinion on the value of the Note-Book,
than the writers just quoted, but he also is *^ disposed to think that there
is more in the absolute omission of any notice of this conclusion in the
full contemporary journal of the experiments, than Mr. Harcourt is
willing to allow."ir

The advocates of Watt have not overstated matters when they affirm
that Cavendish's Note-Book does not contain any announcement of his
views concerning the nature of water. Mr. Harcourt, indeed, never
asserted that it did, but he, as well as Dr. Peacock and Dr. Whewell,
have perhaps not sulficiently guarded their estimate of the Note- Book as
conclusive of the question of priority between Cavendish and his rivals.
A just criterion by which to test what kind and amount of information
the Note- Book can give, may be found in the answer to the question:
Would the diary of experiments al<me enable us to ascertain what

* Report Brit. Assoc. 1839, p. 38.

t Lives qfMen of letters, ^c. Life qf Watt, p. 401.

X North Brit. Rev. Feb. 1847, p. 493. § Ibid. p. 495.

11 Watt. Corr. p. xxxvi. f Bdinb. Rev, Jan. 1848, p. 125.

THE WATBR CONTROVERSY* 367

Carendish's views concerning water were 1 To this I think there can he
but one answer: It would not. It is only by a comparison of the manu-
script journal with the published paper, that we can make the former
serviceable in illustrating Cavendish's views; but the absence of records
of conclusions is no proof that conclusions were not drawn. If the friends
of Cavendish, indeed, have built too much upon the contents of his Note-
Book, the friends of Watt have as unwisely undervalued it. The use they
make of it, is to say, that had Cavendish drawn conclusions regarding
the nature of water from his experiments, they would have appeared in
bis Note-Book, but as it contains no conclusions, it is incredible that he
can have drawn any. Yet Watt's advocates themselves acknowledge
that Cavendish had drawn conclusions from his experiments (whether
borrowed from Watt or not, does not at present concern us) before he
wrote his paper of January 15th, 1784. The Note- book, however, extends
to 1785, yet there are no conclusions respecting water in those portions of
it belonging to 1784 and 1785, any more than there are in those referring
to the period between 1781 and 1783. Lord Brougham, as we have seen,
says that there is no intimation of Cavendish having drawn his conclusion
earlier than Mr. Watt's paper of Spring, 1783, but in truth there is no
intimation in the journal of his having drawn any conclusion at any time.
Mr. Muirhead observes that Cavendish in his Note-Book never uses any
expressions which could imply a union of the two airs, or which are
inconsistent with the notion which Warltire and Priestley had entertained,
of a mere mechanical deposit of the water. From this it might be
supposed that Cavendish had implied some approval of Warltire and
Priestley's views, and that he had suspiciously omitted a reference to the
combination of the two gases : whereas in reality the Note-Book contains
no expression of opinion of any kind, concerning the origin of the water.
In short, the value of the MS. Note-Book as evidence for or against
Cavendish's priority, must be determined by first ascertaining what the
purport of the book was. If it can be shown that it was intended to
receive the record of conclusions as well as of experiments, then the
absence of the former in the case of water, cannot but be regarded as fatal
to the view that Cavendish had interpreted his experiments at the period
when he registered them. But if the opposite shall appear to have been
the character of the journal, then its silence as to conclusions will be a
characteristic merit, not a suspicious defect. The application of a twofold
criterion will enable us to determine what in reality was the character of
Cavendish's journal.

Firstly. Does it, after the period when he certainly had formed his
conclusions concerning water, contain an account of these ?

Secondly. Does it contain the conclusions which he certainly drew
from the experiments on other subjects which are recorded in it ?

To both these questions the answer must be in tbe negative. This
has been stated in reference to the first already, but admits of fuller
demonstration than has yet been given. The journal is not a bound
volume filled in continuously like a diary, so that unless conclusions were
registered along with the experiments which led to them, they must be
separated from them by references to other matters; and if formed later
in time must appear under a difierent date. The volume consists of a
succession of unsewn and single sheets, each of which is marked with the
four consecutive numbers, denoting the place of its pages in the series.
Many of the pages are only partially occupied with writing, and several
have nothing on them but a number. The same page, also^ not unfre*

368 CAVENDISH AS A CHBMIST.

quentlj contains records belonging to different years, as tbe reader inaj
see bj conBalting Mr. Harcourt's lithographed faivainiile of the Note-
Book. Had Cavendish accordingly designed to enrol his conclusions ia
his journal, he could easily have done so in the blank spaces which still
remain unfilled. Yet whilst the paper of 1784 is a running commeB-
tary on the experiments detailed in the Note- Book, none of the conclu-
sions contained in the former concerning water appear in the journal,
although its index was not completed till at least late in the year 17S5,
and therefore long after Cavendish had published his views concerning the
composition of water.

Had Cavendish, moreover, entertained any deceitful prefect of ante-
dating his conclusions, he could most easily have done this, for any of tho
sheets could have been replaced by a neiv one with fictitious dates,
without disturbing the remaining leaves of the journal. And without
taking even this trouble, the insertion of a few lines on one of the blank
pages would have been sufficient to record his views of the nature of
water. Thus page 127 of the MS. (see lithograph fac-simile) contains
experiments made in 1781; on the combustion of hydrogen and air, and
tho appearance of water as a product of it, and half of page 128 contains
a similar experiment made in 1782; whilst the rest of tlie page and the
whole of page 129 are empty; so that if the conclusion at which Cavendish
had certainly arrived in June, 1783. when Blagden reported it to the
French chemists, had been engrossed on either of the last-named pages,
It would have borne the date of November 18, 1782. The omission,
however, of any such conclusion unequivocally proves that when Caven-
dish indexed his journal in 1785, be deliberately left nothing on its pa^s,
but the simple record of his experiments with hydrogen and oxygen.

Could it nevertheless be shown, that the various experiments not
directly referring to water, were recorded in connection with the conclu-
sions ultimately published as deduced from them, then it cobld not but
appear singular that a solitary exception had been made in the case of
oDservations on the synthesis of the elements of water. It is necessarv,
therefore, to apply the second criterion proposed above. To do this it Is
only necessary to compare the two series of Experiments on Air, as well i
as the paper on the New Eudiometer,* with the MS. journal. Such a j
comparison will show, that scarcely any of the conclusions contained in
the paper are to be found in the Note-Book. Some conclusions there
certainly are, but very few, and these not the great generalizations after-
wards published, but only the interpretations of single phenomena. i

Cavendish's journal, in truth, was neither a day-book containing only the '

bare statement of the experiments which he performed, nor the mere register j

of a completed inquiry, and the conclusions to which it had led. It was '

such a journal as every working chemist, I may venture to say, is accus-
tomed to keep; not the mere jotting down of weights and measures, and
and other numerical quantities; nor the simple record, in the fewest
words, of the phenomena observed during the progress of an inquiry,
and still less a detail of all the motives which led to a particular experi- I

ment being tried, or of all the conclusions which were drawn from it, or ]

of all the changes of opinion which occurred during a lengthened and I

difficult investigation. The peculiar character of such a journal, inter- ^

* The portion of the Note-Book lithographed by Mr. Harcourt refers almost
entirely to the first series of ** Experiments upon Air." The portions omitted contain
the experiments upon wliich ihs conclusions contained in the two other papers men*
tioned above, were founded.

THE WATER CONTROVERSY. 369

mediate betvreeik a day-book,* and a completed memoir or essay, is tbe
fall accoant it renders of phenomena which are described by one fresh
from witnessing them, and by the multiplication of namerical and other
observations, which in the end are only partially made public, but which
never appear excessive to an experimenter, at the period of his researches,
when his object is to fortify himself in his convictions. It is quite con-
sistent with such being the prevailing nature of a journal, that it should
occasionally and sparingly refer to motives and conclusions. This is a
point on which only those who have themselves engaged in scientific
researches, or are familiar with the habits of those who prosecute them,
can fully iudge, and I leave the matter to their decision. I will add,
however, that I believe that every working laboratory would be found, at
any period when it mi^ht be visited, presenting the phenomenon of an
observer freely explainmg to his friends and pupils the motives which led
him to engage in a particular research, and the conclusions to which he
had come, whilst scarcely any and often no reference to either motive or
conclusion would be found in his note-books, although his views would
finally re-appear more or less modified in the record of the completed
inquiry. Let us test Cavendishes journal in this way.

The three greatest generalizations to which his experiments on air
conducted him, were (1) that the atmosphere is constant in composition;
(2) that water consists of dephlogisticated air and phlogiston j (3) that
nitric acid is the product of the union of phlogisticated air (nitrogen) and
oxygen. The Note-Book, however, is not more prolific of conclusions in
reference to the first and third of these discoveries than to the second.
No one, for example, would discover from it what Cavendish's motive was
in trying the majority of the eudiometrical experiments recorded in its
pages, nor what conclusions they were found to justify. Here and there
we get glimpses of the aim of the journalist, nnd sometimes we are told
both his motives and conclusions, as at page 109, which is entitled
'' Comparison of the experiments on the diminution of dephlogisticated air
with Scheeles opinion^ But such statements are the exception : thus
we find no description of his new eudiometer, but only references to its
employment, and we should never discover from these, that Cavendish
had come to the conclusion that the- atmosphere has the same composition
day by day, in town and country, when containing odorous matter and
when free from it. Various experiments also are recorded of which we
can now only surmise the object, as neither the motive which led to their
performance, nor the conclusion which they were thought to justify, are
chronicled in the Note-Book or published paper. Such, for example, are
the trials on the '' vis ineriice of phlogisticated air," (MSS. pp. 80 to 82,)
and those made to determine whether there is any penetration of parts
on mixing common and inflammable air (MSS. p. 113), and those referring
to the persistence of common and inflammable air in a state of uniform
mixture (Ibid. ibid). The immedicUe conclusions drawn from those
experiments are stated, but they are recorded as isolated observations,
and although it is scarcely possible to doubt concerning the motives
which prompted their trial, or the use which was made of them as parts

* In proof of the truth of this, it maj be mentioned that among^ Cavendish's
papers remain the original jottings or "minuttM" as he calls them, of some of his
observations; from these, probably, the Note-Book was drawn up. The following allu-
sion to these /r«/ notes, occurs in his Journal in reference to the height of the thermo-
meter in an experiment on the expansion of common air. " The numbpr is set down
85 in the minutet, but must certainly be a mistake for 80." {IJnlith, MSS, p. 369.)

370 CAVENDISH AS A CHBMIST.

of a continnons 83rBteinatto inquiry, no professed nse was made of dkea
in the published essays, nor is their signification expounded in the Koto-
Book. Similar remarks apply to the notes on which the paper of 17B4
was founded : thus at pa^e 115 we have the heading, *^ Exploeionof
Inflammable Air by electricity in glass globe to examine Mr. Wariiire'f
experiment.'* The result of this repetition we know from Oarendisb'f
paper to have been such as to induce him to deny the truth of Warltire*!
declaration, that a loss of weight attends the detonation of mixtnres of
inflammable air with common air or oxygen in shut vessels; yet we shooU
never be certain of this from Uie Note-Book, for we have it stated ob
four diflTerent occasions, that the globe seemed to lose in weight, whilst in
others it did not. Again; we know from the published paper that oxygen
from red lead, from turbith mineral, and from plants, was used by Caven-
dish, with a view to discover whether the nitric acid which appeared
in certain of his explosions, was derived from basic nitrate of mercoiy
contained in the red precipitate from which his first specimens of oxrjgfi^
were prepared; and that he attached special value to the oxyven froa
plants as most certain to be free from acidi Nothing of all this, however,
appears in the Note-Book. The experiments merely are given as if they
hod been isolated inquines, and the conclusion to which they led, eonoem-
ing the uniform production of nitric acid in certain circumstanoes, does
not appear. Again; we know that Cavendish varied the proportions of
hydrogen and oxygen which he detonated together, and that he came to
the conclusion that in ordinary circumstances nitric acid appears when
the oxygen is in excess ; and nirther, that he believed the source of the
nitric acid to be nitrogen present as an impurity, and that he put this
idea to the proof by adding a little nitrogen to a mixture of hydrogoi
with excess of oxygen, and then exploding it, when his anticipation was
verified. We know aLso, that he inferred from these observations that
nitrogen is a compound of phlogiston and nitrous acid, and that he consi-
dered this view to be corroborated by the results which he obtained when he
heated charcoal and nitre together. Yet although all those experiments
are fully recorded in the Note-Book, the views which they were under-
taken to test, and were thought to justify* are not stated, and no one
probably would have surmised the purport of some of the trials, such as
those with charcoal and nitre, or have felt certain what conclusion onjf of
them justified.

Many more examples illustrating the same thing might be given, but
these may sufiioe. If points on which Cavendish dwells so fully in his
published paper, as the source of the nitric acid which appeared during
Lis repetition of Warltire*s experiment, and the nature of nitrogen, are
not expounded even briefly in his journal, there is nothing at all singular
in no account being given of the conclusions which were drawn frx)m
those detonations in which no nitric acid appeared.

There is nothing, then, suspicious in the circumstance that Cavendish
has not stated his views concerning the nature of water in his Laboratory
Journal. Nor can any weight be attached to the declaration, that the
enormous importance of such a discovery as that of the composition of
water, might be expected to induce him who made it to record it however
briefly, even in his roughest note-book ; for if it can be shown that his
journal was not intended to contain accounts of discoveries however
important, we can find no fault with its omitting all reference to the
theory of the composition of water. It was dealt with as its author's
other theories were. Three of these, which he deemed of the highest

THE WATER CONTROVERSY. 37 1

importance, were, 1. That the nniversal product of the phlogistication of
air 18 not carbonic acid. 2. That the universal prodnct of phlogistication
is water*^ 3. That nitrogen is a compound of phlogiston and nitric acid.
Yet the journal may be searched from end to end, without any statement
bein^ found of any one of these conclusions. Nop need we feel surprised
at this. Cavendish was like a merchant with his day-book, journal, and
ledger: the minutes corresponded to the first: the note-book to the
second ; and the published paper to the third. We might as well blame
the merchant for not recording in his journal that he had made a pro-
fitable speculation, although this was most clearly indicated in his balance
eheet, as blame Cavendish for not recording in his note-book those gene-
ralisations which he purposely left to be announced in his published
paper. If any one, moreover, is disposed to add, that as a very fortunate
merchant is likely to talk or hint of his success long before he balances
his accounts, so Cavendish might be expected to give some utterance to
his satisfaction at making a great discovery long before he gave a detailed
exposition of it in a formal essay, I will say in reply, that Cavendish did
refer to his discovery of the nature of water before he published it, and
this to more than one person. The discussion of this, however, belongs to
another section of our argument, where it will be considered.

Cavendish's Journal, then, is not necessarily a document of no impor-
tance to our present inquiry, because it does not contain the generaliza-
tions which be ultimately based upon it. What, then, is its value f It
may be compared, I think, to a map or chart laid down by a traveller,
who has been engaged in exploring a new and unsurveyed territory. It
lies before us, unaccompanied by the designer's journal, which would
explain to us what object he had in his journey; why he directed his
steps in one direction rather than another; what false movements he
made, and what important goals he succeeded in reaching. But no one
who acknowledges (and all competent judges do acknowledge) that the
designer of the map was a man of rare experience and sagacity; possessed
of all needful accomplishments ; and before the date of this map, univer-

ainiless journey. And even when we cannot discover why he turned to the
east rather than to the west, or to the north rather than to the south, we
will not doubt that some good reason justified the change of direction.
Should, moreover, a record of the traveller's progress afterwards appear,
which recounts the steps of his journey from point to point, as those are
laid down in the map, it will be difficult to resist the impression, that the
dates inscribed upon the chart as marking his daily progress, denote with
more or less precision the periods at wki^ he made tne discoveries, which
he connects with these dates in the completed record of his journey.
And this, in fact, is the conclusion to whicii all the critics of the Water
Controversy have come, in reference to the map taken as a whole. In
other words, when Cavendish is found stating in his published paper a
particular couclusion, such as, that sunlight " enables one body to absorb
phlogiston from another," and cites, in illustration of this, the effect of
BUDsnine in bleaching vegetable tinctures, in rendering nitric acid yellow,
and in reducing salts of gold; whilst experiments of difierent dates regis-
tering the observation of these phenomena, in the course of researches
made purposely to develop them, are recorded in bis journal; few, if any,
will affect to deny^ that his belief in the dephlogisticating power of sun-

2b2

372 CAVENDISH AS A CHEMIST.

light, dated at least from tlie period when he completed his experimeBi&
The u umber of separate researchea, however, chronicled in the journal, i^
rerj large; and the immense majority of them have no conclusions coo- I
nected with them, in tliat record, whilst the clearest interpretation ia pst i
upon each of them (which is referred to at all) in the published paper,
lu these circumstances no impartial critic can avoid the conclusion, that a |
definite interpretation was, sooner or later, put upon each research, at or
near about the period when the latest expenment of each series of obser-
vations was performed. If any otlier view, indeed, were entertained, it
would lead to the incredible supposition, that an immense number of
most significant researches, prosecuted by one of the greatest natoiml
philosophers, taught him, at the period when they were made^ nothing,
or almost nothing, and remained a sealed book to himself for years, till
all at once, or within a very short period, the whole of them suddenly
acquired a meaning, and assumed a shape, which have made them a model
for later inquiries. All the critics of |the Water Controversy wonld
reject a supposition so extravagant as this. Yet if they do so, they mort
concede that the onus probandi lies upon them if they deny that the
probabilities are all in favour of Cavendish, when he in effect declares
that he was as quick in furnishing an interpretation of his experiments
on the production of water from its elements, as in expounding the signifi-
cance of the other researches recorded in his journal. The sagacious
interpreter of ninety-nine phenomena may be safely believed, when he
declares that he had interpreted a hundredth phenomenon, much more
significant than many of those the meaning of which he is acknowledged
to have seen. I apprehend, therefore, that by all those who do not set out
with the hypothesis, that Cavendish was a dishonourable man, (and with
such I enter at present into no argument,) the circumstance that his expe-
riments on the production both of pure water and of acid water from the
detonation of hydrogen and oxygen, were completed in January 1783, will
be regarded as strongly favouring the belief, that his conclusions as ulti-
mately published are at least of as early a date. More than this I am not
anxious to contend for here; but if this can be established, it is idle
to affirm that his journal does not testify in his favour, or that it
testifies against him. And in truth, although the majority of the advo-
cates of Watt have' professedly rejected the evidence of Cavendisli's
journal, they have solicitously repelled the conclusions which his advo-
cates have drawn from its contents.

In reality, therefore, whatever may be said, it is felt upon both sides,
that Cavendish's Note-Book is a very important document. His opponents
accordingly, sensible of this, have struggled hard to represent the labora-
tory journal as testifying only to certain phenomena having been
witnessed, whilst his advocates insist on adducing it in evidence of con-
clusions also having been drawn. This question, which may be best
disposed of here, has arisen in the following way. Mr. Harcourt has
contended in the postscript to his address, that " the experiments which
Cavendish made in the summer of 1781, not only necessarily involved
the notion, but substantially established the fact of the composition of
water.*'* On this statement, Sir David Brewster remarks, " The advo-
cates of Cavendish, thus driven to the wall, take refuge in the allegation
that the experiments of 1781 involve the inference. Were this the
case, the history of science would require to be re-written. The ex-

• Brit, Attoc, Rfp, 1839, p 23,

THE WATEK CONTROVERSY. 373

Eerimenter would thus enter the nicbe of the philosopher, and tho
ighedt efforts of intellectaal power would cease to be appreciated."*
To the same effect Mr. Muirhc»d remarks, " It is thus quite impossible
to say that the experiments necessarily imply the conclusions, or to
consider the right explanation of that most remarkable phenomenon [the
production of water from its elements] as having been included in tho
mere obserration of the fact. To argue the reverse, as Mr, Harcourt has
done, is to betray an ignorance of the writings of the many eminent philo-
sophers who doubted, and even denied the true theory, after it had received
what modem chemists may consider irresistible confirmation."-!* Lord
JefiVey, with judicial impartialii^, takes a middle view of the validity
of Mr. Harconrt's argument. " We do not deny," says he, " that there is
at first sight something plausible and taking in this view of the matter,
especially when addressed to a generation which has always been familiar
with the conclusion, and with the universal assent of mankind in tho
sufficiency of the evidence referred to. Yet it requires but a moderate
acquaintance with the actual history of the progress, even of the most
obvious truths, and of the tenacity and vitality of prejudices and errors,
to make us cease to wonder at the incredulity with which what is at last
felt to be demonstration is often at first received, or at the distrust with
which even the authors of great discoveries have often regarded their
own achievements."]: On this contested point I would remark that there
is justice in the views alike of Mr. Harcourt and of his impugners. Two
Separate propositions, in truth, have been confounded in the discussion;
the one that Cavendish's experiments would have led every one to Caven-
dish's conclusion ; the other, that they led Cavendish to his conclu-
sion. If Mr. Harcourt intended to maintain the first of these proposi-
tions, which I apprehend he did not, the objections of Watt*s advocates
are well founded. It would be idle to affirm that an experiment is self-
interpretative, even to him who appreciates its phenomenal significance.
There is no phenomenon, in truth, of which philosophy has exhausted, or
we may safely say will exhaust, the full meaning. How many lessons a
drop of water would teach him who should observe all the properties
it possesses, no wise man would venture to guess, but the wisest would
acknowledge that an omniscient observer of the whole of its characters
would learn from it much that it has not taught us, perhaps more than
it has yet taught all the generations of men. This remark applies to
every phenomenon without exception, with which the physical sciences deal.
The converse, no doubt, is in a sense equally true. There is no pheno-
menon which does not to some extent interpret itself. All men philoso-
phise on all about them, and learn something from everjrthinff they see;
so that after all the question must ever be, what and how much has each
learned from the things he has witnessed; and if personal testimony be
wanting on this point, we can only hazard surmises. Had we nothing,
therefore, but Cavendish's Journal, from which to argue, we could only say
that his experiments must have taught him something, but what did not
certainly appear; although even then we should feel justified in asserting,
that he was more likely than most men to have been largely instructed
by them. It is quite certain, moreover, that others besides Cavendish
witnessed in as great perfection as he did the phenomena which are sup-
posed by his advocates to have instructed him, and nevertheless, drew no
precise conclusion or drew a false one from them. Sir David Brewster and

* North Brit, Rev, Feb. 1847, p. 495. t Watt Corr, p. xciv.

X Edinr, Rev. Jan. 1848, p 128.

374 CAVENDISH AS A CHEMIST.

Mr. Mairhead have referred to Priestley, Warltire, Maoquer, and Sigwui
de Lafond as Having been In this predicament. Bat these cases are not
in point, for none of those observers witnessed phenomena which would
have justified the conclusion which Cavendish's experiments did, Monge*i
observations, however, were nearly identical with those of Cavendish, as
we have already seen, yet Monge could not decide which of two €X>n^ii-
sions to deduce from them. This fact is of itself sufficient to show that
Cavendish's experiments were not self-interpretative in the sense of com-
pel] ing the conclusion which ha deduced from them. On the other band
the implicit endeavour of the advocates of Watt to show that he displayed
an extraordinary sagacity in putting the right interpretation upon
Cavendish's experiments as repeated by Priestley, must be qualified by a
reference to the fact, that Cavendish's conclusion quickly carried convio-
tion to the minds of Lavoisier, La Place, and Mensnier, as well as to
many other philosophers both in England and abroad, not excepting even
Priestley, who first tested their accuracy, as we shall presently eee. In
short, whilst it must be unreservedly denied that CavendisVs expenments
could not fail to lead any one to the conclusions all now connect with
them, it must with equal plainness be coptended that they were eminently
significant and suggestive.

It is quite a tenable proposition, accordingly, that Cavendish's experi-
ments taught Cavendish something; and that they taught what he affirms
they did teach him. If we credit, in truth, the account which he gives
of his motive for exploding hydrogen and oxygen together, we shall find
a very slight additional exercise of faith sufficient to make us believe, that
his researches led him to the discovery which he professed to have reached
by means of them. His own account of matters in his published paper,
(which the records of his journal confirm,) is, that whilst engaged in an
inquiry into the products of combustion and oxidation, and anxious to
discover what occasioned the diminution in yolume which air underwent
when deoxidized, he received the account of Warltire and Priestley's
experiments, in which a dewy deposit was observed to follow the explO'
sion of inflammable air with common air or oxygon.

He refers to these experiments as twofold, the first made with a
copper vessel, and supposed to show that heat is ponderable; the second
tried with a glass globe, and bringing to light the deposition of dew.
The first experiment he dismisses in a few words as one which did not
succeed with him; of the second he says, that '^as it seemed likely to
throw great light on the subject I had in view, I thought it well worth
examining more closely." It thus appears that the great objeet of his
inquiry was to discover where that portion of the air went, which lost its
elaisticity, and seemed to disappear when bodies were oxidised in confined
portions of it. That " the air thus lost or condensed" was not annihi^
fated, but had passed into some new physical state, or into a condition of
combination in which all its ponderable matter was still present, was
manifestly with Cavendish a fixed conviction from the first; so that the
special question he aimed at answering was, what was the particular
substance or substances in which the lost air existed in a state of conden-
sation. That the lost air was changed into carbonic acid in all caaes was
the prevailing opinion when he commenced his researches. He showed
that this was a mistake in so far at least as many phlogisticating (oxidable)
bodies were concerned. He proved further, that the condensed air had
not passed into the state of either sulphuric or nitric acid, and he was
on the look*out for som^ other body distinct from those acids, in which

THE WATER CONTROVERSY.. 375

the lost fldr rmghi be sbown to exist, when Priestley and Warltire'a
observation of the appearance of moisture when inflammable air and
common air were burned together, so arrested his attention that ho
resolved to repeat it with the utmost care. No one, I think, can
question that it was the appearance of dew, where the prevailing
theory of the day taught that earbonlo aoid should appear, and where
it was possible, as Cavendish evidentlv believed, that sulphurio or
nitrio acid might have appeared, which induced him to repeat the
observation with so much care. Priestley and Warltire had in effect
reported two things; the one that so mucn air disappeared ; the other
that so much water made its appearance] and it was the relation of
the one phenomenon to the other, which Cavendish thought it so well
worth wnile to examine more closely. In truth his statement seems
most plainly to imply, that before he made any experiments on the
subject, he suspected that the water which appeared in Warltire's experi-
ments, was the substance into which the lost air had passed, and in which
it would be found. On this view his experiments have one and all the
greatest significance, and are most happily contrived, so as to discover
what connexion subsisted between the disappearance of gas, and the appear-
ance of water. Unless, indeed, the advocates of Watt will contend that
Cavendish's experiments were aimless trials, prompted by no motive, and
pregnant with no conclusion, they will find it very difficult to explain
why he made such observations as he did, unless the object he had in
view was the one I have stated. They have sought to show that it was
the problem of the ponderability of heat, which mainly interested him;
but this view is totally untenable, because^-

1. Cavendish was a disbeliever in the materiality of heat, and there-
fore, wo may be certain, in its ponderability,

2. He explicitly affirms that the apparent demonstration which Warl*
tire had furnished of the subjection of heat to gravity had not for him,
even if confirmed, more than a secondary interest; the appearance of dew
in glass vessels, not the loss of weight in copper ones, having chiefly fixed
bis attention,

8. Far less elaborate experiments than those ho tried would have
been sufficient to determine whether Warltire's view as to heat being
heavy, was true or not. Warltire did nothing more himself than fill a
globe with a mixture of common and inflammable air, weigh the globe,
send an electric spark through its contents, and then weigh it again; and
Cavendish did not need to do more in repeating the supposed experimentum
crucis. He tells us, however, in his published paper, '' that in all the
experiments the inside of the glass globe became dewy as observed by
Mr. Warltire, but not the least sooty matter could be perceived. Care
was taken in all of them to find how much the air was diminished by
the explosion, and to observe its test.'' From his journal, moreover, it
appears that he tried even more than this; for though he did not publish
the fact in his paper, he determined the specific gravity of the residual,
or uucondensed gas, after each explosion. (Lithograph MSS. pp. 116 —
118.) The three points thus determined, vis. 1. the diminution in
volume which the electric spark occasioned in mixtures in various pro-
portions of inflammable and common air ; 2. the test of the uucondensed
or residual sas, i. e. the amount of oxygen (if any) remaining in it; d. the
density of the unbumed gas, — ^wcre not required, and their decerminatiou
was a most needless waste of labour, if the ponderability of heat was
the only problem which the experiments were intended to solve. The

S76 CAVENDISH AS A CHBMIST*

fiEbct tbat such determinations were made^ is of itself pixxif suffiaent tiat
Cavendish had some other object in view; whilst, on the other hand, tkew
were exactly such experiments as an observer wonld trj whciee object
was to ascertain what connexion subsisted between the a;ppe«raiice of
water and the disappearance of air. In proof of this, it may be notieed
that Cavendish was in advance of the great majority of his contempO'
raries in a knowledge of the composition of atmospheric air* He knew
that it was constant in composition; that it contained between 7^7 th siid
\t]i of its volume of oxygen. He was certain, therefore^ that eirerj
specimen of common air, when completely phlogisticated, (deoxidised)
would undergo a diminution in volume of nearly -Jth. He first, tfaer&-
fore, observed the reduction in bulk which a particular mixture of
hydrogen and air underwent when the electric spark was transmitted
through it. This gave him the amouut of hydrogen and air which had
together disappeared as gases, or lost their elasticity. He then tested the
residual gas in his nitric oxide eudiometer, and thereby discovered how
much oxygen had lost its elasticity during the combustion. This resoit
he further checked by determining the specific gravity of the residoal
gas. Proceeding in this way, and mingling the hydrogen and air in
various proportions, he quickly found one ratio as the mean of six trials,
in which the volumes were such that the hydrogen exactly, or neariy
exactly, phlogisticated or deoxidised the air, so that the latter wms
diminished in volume -J^th, and the residual air was found free from
oxygen by the test of nitric oxide, and answered to the characters, and
had the specific gravity of nitrogen. A table exhibiting these results,
except that no reference is made in it to the density of the residual gas, is
given in the published paper;* and a similar table will be found in the
Journal (lithographed MSS. p. 119). From this it appears that, in July
or August, 1781, he had ascertained that hydrogen, like other oxidable
bodies, diminishes air by -}th of its volume, and that the quantity of
hydrogen required to efi*ect this is (nearly) double the volume of the
oxygen which it withdraws. t

Such are the only experiments recorded by Cavendish in his Journal,
or referred to in his published paper in demonstration of the quantities
of gas which disappeared during the combustion of hydrogen and air.
They taught him that a given measure of hydrogen will always remove a
measure hiaJf as great of oxygen from the air in which it is allowed to
bum; and that, therefore, if we wish to deoxidise a measure of atmo-
spheric air completely, we must bum it along with |ths of its volume of
hydrogen. He had thus settled one-half of the question in order to solve
which he undertook the repetition of Warltire*s experiments. He knew
how much air disappeared (namely, -Jth) when it was phlogisticated by
hydrogen, and what part of the air disappeared, viz. its dephlogisticated
part, or oxygen; and at every explosion he saw this disappearance of
oxygen followed by the appearance of moisture or dew. The nature of
this dew had then to be examined ; and accordingly the account of the
^lobe experiments in his Journal is followed by the record of those com-
bustions of hydrogen and air in a glass cylinder, which he tells us in his
paper were performed with a view to obtain a sufiicient quantity of the
dew for analysis. In this experiment, we find him applying the infor-
mation which the globe detonations had given him, for he causes the

* P«/. TVflfW. 1784, p. 127.

f Cavendish did not reprefent the combining measure of hydrogen as txatUy
double that of oxygen, which it is, but he came very near this ratio.

THE WATER CONTROVERSY. 377

gases to meet in the proportion of one measure of hydrogen to 2} timca
that qnantity of air; and the liquid which condensed he refers to twice
by the name of water, and describes its analysis as related in the pub*
lished paper.* This is the only experiment referred to in his '' Experi-
ments on Air," as demonstrating the pro<laction of pare water from the
combustion of hydrogen and atmospheric air.f

Such, then, were the experiments made in 1781 and 1782, from which
Cavendish inferred in his published paper of 1784, " that when inflam-
mable and common air are exploded in a proper proportion almost all the
inflammable air, and near -J-th of the common air, lose their elasticity and
are condensed into dew," and " that this dew is plain water, and conse-
quently that almost all the inflammable air, and about -J-th of the common
air are turned into pure water."|

When all that has been stated is considered, it cannot be regarded as
an extravagant proposition that Cavendish's experiments had led him to
the conclusion just quoted, long before he published it. There may be

* Phil. Trans. 1784, p. 129.

t In the page of the journal where thii experiment is recorded (Liih.MSS,^.l2B),
a second experiment of the same kind is referred to, of later date, viz. Nov. 18th, 1782,
to which, as wtU as to the original trial. Sir David Brewster refers (North Brit. Rev.
Feb. 1847, p. 494) as inconsistent with the belief entertained by Mr. Harconrt, that
Cavendish could, at the earlier date, have '* established the composition of water,"
because *' the equality of the water and the gases was not and could not be measured, and
the water was not absolutely pure," and also because *' if he [Cavendish] had concluded
from his experiment in 1781, that water was not a simple substance, but composed of two
gases, why was he repeating and recording such experiments as this sixteen months after ?"
**The deduction," continues Sir David, '* is unavoidable, and we hold it to be proved
by thitfaet alone, that he had drawn no such conclusions." Sir David, however, has
overiooked that Cavendish settled the question of equality of weights between the
gases and the water solely by his globe experiments, and that those made with the
cylinder were undertaken only wi& a view to procure a large quantity of water for
analysis. Ko attempt, accordingly, was made to render the cylinder trials quantitative.
In the second experiment. Cavendish distinctly mentions that he did not know what
quantity of gas he burned ; and in the first, the quantities are stated quite generally.
No data are given, as of the height of the thermometer and barometer, by means of
which the volumes expended could be converted into weights; and it is further explicitly
declared, that some of the water produced was lost. The weight of the water procured
in either case is given in round numbers, only to show that it was examined on a
sufficiently large scale to admit of any imparities in it being detected. In Cavendish's
quantitative trials, volumes were not converted into weights, but (to take the simpler
case) a given weight (viz. a globeful) of hydrogen and oxygen was changed by the
electric spark into the lame weight of water. The water certainly was not absolutely
pure, but it made so near an approach to purity, that Cavendish, after testing it very
carefully, said of it, as it was obtained in the first cylinder experiment, that *' it seemed
pure water." It has been shown already (ante, p. 339), that none of his rivals obtained
purer water than he did, or analysed it so carefully.

As for the repetition of the experiment in 1782 showing that no conclusion had
been drawn from Uie trial of 1781, it is sufficient to notice that, from its date, it was
certainly made contemporaneously with the experiments on the influence of variations
in the relative proportion of hydrogen and oxygen on the production of nitric acid, and
evidently with a view to satisfy Cavendish, after he had discovered that nitric, acid was
occasionally produced, that this did not appear when hydrogen and common air were
burned together. This is manifest from the explicit way in which he states that the
water obtained in the second experiment " was not at all acid, nor gave the least red
colour to paper tinged with red flowers." {Lith. MSS. p. 128.) No such special
reference is made to the absence of acid from the water obtained in the corresponding
experiment of 1 781 {Lith. MSS. p. 127), at which period Cavendish had not discovered
that nitric acid was developed iu any circumstances when hydrogen and oxygen are
exploded together.

t PhiL Trans. 1784, p. 129.

378 CAVENDISH AS A CHEMIST.

said to be tvo extreme hypotheses on the subject, the one that imputed te
Mr. Harooart, that the experiments inyolved the oonelusion, and that
therefore their dates are the same; the other, that of the advocates of
Watt, that Oavendish put no interpretation upon his experimentB till he
read Watt's letter of 1788, which first tauffht him their sigqifioanoe.
That neither of these views can be aecepted without qualificatian, all
impartial critics will, I think, acknowledge. But if we are to choose
between the two, we shall find the probabilities immensely in favour of
the former. In judging of this, it must be remembered that we are at
present proceeding solely on the eridence furnished by the MS. journal,
and the published paper as compared together, and that for broFity's
sake we have selected the experiments with hydrogen and air, omitting
those with hydrogen and oxygen. I have chosen those triaU beoanse
the inquiry was macli more complicated than when oxygen was sub-
stituted for air, and the chances of error were therefore much greater.
Moreover, Watt's letter refers solely to experiments with Inflammable air
and oxygen, and could not therefore have furnished Cavendish with the
direct interpretation of the phenomena which appeared when atmospheric
air was employed. His experiments with air, in truth, stand alone, nor
was any repetition of them made before he published his paper. They
were so nicely contrived, that the mean of six trials enabled him to
ascertain to what extent hydrogen diminishes air when exploded along
with it. They were managed in such a way that the same trials proved
that no loss of ponderable matter occurred during the combustion; and
they supplied the demonstration that the nitrogen of the air remained nn-
changed, whilst all its oxygen disappeared, One additional experiment
showed that the dew which resulted from the combustion was pure water.
We are gravely asked to believe, nevertheless, that he who conducted this
beautiful inquiry had been taught by it nothing concerning the on|^n of
the water. Yet if he were not seeking to connect the disappearance of
oxygen and hydrogen with the appearance of water, what, I would ask,
was he in search of 9 He was confessedly examining the products of
combustion with a view to explain why air was diminished in volume
when it supported combustion. The particular combustible which he was
employing was hydrogen, and when it was burned in certain proportions
with air in a shut globe, it diminished the volume of the air, and nothing
was left in the globe but nitrogen and water, whilst the quantity of
nitrogen was exactly that which tne volume of air taken at the beginning
of the experiment contained. The nitrogen of the air thus stood aside, as
a portion of it which remained unchanged, and the only substance standing
in the place of a product of combustion, was a weight of water which of
necessity equalled that of the oxysen and hydrogen which bad lost their
elasticity. All this Oavendish knew; all this he had discovered for
himself, relying upon data which were almost entirely of his own fur-
nishing; and having once made such experiments he never repeated them,
though two years elapsed before he made them fully public, and on them
alone he rested his published conclusions, so that he manifestly regarded
them as needing no repetition or extension in order to justify his views.
Notwithstanding all this, it is declared to be incredible, that he who sought
for the product of the combustion of hydrogen and found water, believed
that water was the product of that combustion. I unhesitatingly afiirm,
on the other hand, that nothing but the strongest evidence to the con-
trary can invest with improbability the belief that at the period of their
pemrmance Cavendish interpreted his experiments as demonstrating that

THE WATBR CONTROYBRSY* 379

hydrogen and oxygen were iuitied or eonveried into pnre water. To this
extent then, at least, I adopt the statement that his experiments bore for
him their oonolnsion in their face. Those who deny this most not stop
short with denying it, but most show what conclusions he drew from his
experiments if not those he recorded.

The MSS. journal then is good evidence, on the lowest estimate of its
value. Its dates are important as earlier in time (so far as the experi«
monts on the production of water are oonoerned) than the oonolusions
considered as aating from the spring of 1783; and we may justly put
upon them the higher value that they mark, within more or less narrow
limits, the probable periods when conclusions were drawn from the expe-
riments the dates of which are given.

It nuiy most justly, however, be nrflod that if Oavendish*s experi-
ments on the combustion of hydrogen and air were of such a nature as to
interpret themselves, it is very nnlikely that the secret which they
revealed in 1781, was never betrayed by Cavendish till 1783. I havo
already acknowledged the probability on other grounds of such a betrajral,
and in truth it is admitted on all hands, as I have heretofore shown, that
Cavendish gave iome account of his researches to Priestley and to Blagden
before he published them. These statements were referred to only in
illustration of the priority of Cavendish's experiments. They are held
however by the impugners of his claims, to nimish the best means of
disproving the priority of his conclusions to those of Watt. I shall con-
sider them, therefore, as coming before us in the shape of objections urged
by the advocates of the latter,

III, Watt's friends ky great stress upon the fact, that Cavendish
communicated to Priestley only his experiments and not his conclusions.
On this point Lord Brougham observes, " Sir Charles Blagden inserted in
the same paper [of 15th Jannarvi 1784], with Mr, Cavendish's consent, a
statement that the experiment had fiist been made by Mr. Cavendish in
summer, 1781, and mentioned to Dr. Priestley, though it is not said when,
nor is it said that any conclusion was mentioned to Dr. Priestley, nor is it
said at what time Mr. Cavendish first drew that conclusion. A fno9t rnate^
rial omtMton."* To the same effect Mr, Muirhead remarks, " Although
in July, 1784, when the Philosophical Transactions for that year were
printed, he [Cavendish] said that his experiments (made in 1781) had
been mentioned to Dr. Priestley, he does not name the precise time, nor
even the year, when the experiments were so communicated. He does
not say that any conclusion was, along with them, mentioned or even hinted
at. He does not even say at what time he himself first drew any con-
clusion on the matter,"! Sir David Brewster takes the same view;
'^ Cavendish," ho observes, " assures us that he eommuuicated his experi-
ments of 1781 to Dr. Priestley, but does not mention when, and we know
that he did not oommunieato any conclusions to that distinguished
chemist."t

It is thus affirmed of Cavendish: 1st That he communicated to
Priestley his experiments, but not the conclusions which he drew from

* HUtorieal Note, Watt Corr. p. 247.

t Watt Corr, p. xxxtU. In another part of hia introductory remarks, Mr. Muirhead
saya that Priestley '* alludes to on* eg^eriment of Mr. Cdvendish as being known to him."
(p. xxxyiit.) This Gallican rendering of m into one ii quite unwarranted, for Priestley,
although he recurred to the subject frequently, and ipade special reference to Cavendish's
experiments after Watt's paper was published, found no fault with Cavendish's deolan*
tion, that " all the foregoing experiments were mentioned by me to Dr. Priestley."

% North Brit, Rev. Feb. 1847f p. 495.

380 CAVBNDISn AS A CHKMIST.

them. 2nd. Tbat be dooa not say wheji he made this eoannanicatioa.
3rd. That he does not say when he first drew the eoDclnsion for hhnselt
From all which it is inferred that there is no evidonee of Caveiidyi
having preceded M^att in his view of the composition of wat«r. I dull
consider these propositions in turn.

I. As to the first, it conld have been wished that the inipngners of
Cavendishes claims had told us what they understood by " an experiment^
and in what terms they supposed Cavendish to have recounted '' all the
foregoing experiments." They cannot surely profess t-o believe tb«t
Cavendish would only tell his brother philosopher tbat he took certain
pieces of apparatus and certain chemical substances, placed them together,
and saw certain phenomena ? Yet they seem to wish ns to understand
as much, so that the revelations of Cavendish would be (to take one ex-
ample) of the following kind: "6140 of inf. air, or 7.^77 of whole con-
tents was put into the bott. ; it did not lose at all on firing, nor on stand-
ing one-half hour. 8615 of water run in on opening, and 230 more on
shaking, in all 8345, or..... its test was .339; its spe. gra. was -^ less
than com. air, 9900 being tried and inc. weight ^ gra."*

I have taken the preceding quotation at random from CavendisVfi
journal, which contains, as we have seen, the simple record of his expe-
riments. Is it credible that this, or any of the other brief statements,
intelligible only to him who recorded them, are examples of the kind of
information which Cavendish gave Priestley ? No one, I am persnaderl,
who has listened to an observer's account of his experiments, will enter-
tain the belief tbat Cavendish would describe his researches to another
chemist, merely as he has recorded them in his journal. The friends of
Watt have, wisely enough, avoided giving their definition of the word
" experiment/' as referred to by them. It is a fundamental rule, however,
of all interpretation, that we are not to put our meaning upon an authors
words, but are to seek to discover in what sense he employed them.
AVhat we have first to ask, therefore, is not how narrow a limitation may
be put upon the word ''experiments," but what meaning did Cavendish
intend and expect it to convey.

That he used it in a much wider sense than as merely the description
of the apparatus which he employed, the manipulations which he prac-
tised, and the phenomena which be saw, is manifest from the place in his
paper where he introduces the reference to his revelations to Dr. Priest-
ley. He has recounted all his experiments on the production of both pure
water and acid water from the combustion of hydrogen with air and with
oxygen, and then he declares that all of them, *' except those which relate
to the cause of the acid found in the water," were mentioned to Dr. Priest-
ley, who was induced in consequence to repeat some of them. If we look
back accordingly to the preceding pages, we shall be able to throw some
light upon the extent of information which was given. Before doing so,
however, I would notice that there was a peculiar significance in Priestley
rather than any one else being made acquainted with Cavendish's observa-
tions. It was Priestley and Warltire's experiments that he had been re-
peating, and he was using an apparatus and materials similar to those which
they employed. There were no parties, therefore, to whom the result of a
repetition of their trials could be so interesting as to them. At the same
time, if the repetition had merely confirmed their results, there would be
little occasion for much detail in recounting the repetition. We know,
however, that Cavendish could not confirm Warltire's results, and that he

• JJik.MSS. p. 117.

_fci

TUB WATER CONTROVERSY, 381

went far beyond Priestley, who in turn repeated Cavendish's experiments
as new to him, and called them, though in a sense extensions of his own
trials, by Cavendish's name. It was thus only, or chiefly, where his pro-
cedure or his results were dlftrent from theirs, that it could be an object
of interest to either party to recount Cavendish's experiments. Thus
much then premised, we may be certain that, limiting himself almost
entirely to the points of difference, he would inform Priestley that he
had not been able to confirm Warltire's alleged observation, that the globe
lost weight, but that he had confirmed the truth of Priestley's statement
ill reference to the appearance of dew after each explosion. He would
then, doubtless, proceed to recount "all the foregoing experiments,"
which Priestley approved of Cavendish's declaring he had communicated
to him. He woula therefore more or less fully explain the results of his
varying the proportions of air and hydrogen, the maximum amount of
reduction which he found could be effected upon air by exploding it with
hydroffen, and the test or quality of the residual gas ; on all which points
Priestley himself had made no researches. He would then proceed to in-
form him, in some such words as he uses in his published paper, that " the
better to examine the dew," he burned comparatively large quantities of
hydrogen and air, and in this way obtained a portion of liquid which he
analysed and found to be pure water. This would be followed by a
description of the experiments which he made '^ to examine the nature of
the matter condensed on firing a mixture of dephlogisticated and inflam-
mable air," in the proportion of a little less than two volumes of the latter
to one of the former, and of his conclusion that almost all of it [the mixture
of gases] lost its elasticity, whilst the liquid obtained was '' water united to
a small quantity of nitrous acid." Less than this cannot well have formed
the subject of Cavendish's communication, so that Priestley must be held
to have been informed by Cavendish of the following facts : — 1st. That the
moisture or dew which Warltire and Priestley had seen deposited on the
sides of the glass globes in which their explosions were conducted, was
pure water. 2nd. That when common air and hydrogen were exploded
in the proportion of a thousand volumes of the former to four hundred
and twenty-three of the latter, the whole of the hydrogen and ^th of
the common air disappeared, or lost their elasticity, and were replaced by
a certain qnantity of water; whilst no oxygen was found in the unburnt,
or residual gas. Srd. That when oxygen and hydrogen were exploded
in the proper proportion, almost all of the mixture disappeared, or lost its
elasticity, and was replaced by a certain quantity of acid water.
4 th. That the disappearance of the gases, and the appearance of water,
were not attended by any loss of weight, as Warltire had stated; the
globe weighing the same before and after the explosion.

Such, I think, is a fair rendering of the kind and amount of information
which Cavendish gave to Priestley. I have interpreted the term experi-
ments as signifying '^ trials and their results," so far as phenomena were
observed, but not so far as conclusions were drawn. Yet every one, I think,
must acknowledge it to be very unlikely that so much information as that
stated above can have been given, and no more. Cavendish, who was not
what Priestley styled himself, a trier of '' random experiments," is not
likely to have reported observations to another, till tney had conducted
himself to some conclusion. No one, indeed, was more slow in commu-
nicating his results till they bad taught him some distinct lesson, nor is he
likely to have volunteered an account of them to another, unless he had
some truth to enforce upon hira by means of their evidence. The fact^

382 CAVENDISH AS A CIIBMIST.

also, mentioned by Cavendish, that he did not commanioate to Priestlej
the obserrations on the oaase of nitric acid appearing in the water, wfaik
he kept back from him none of the other results of his experiments with
hydrogen, stronffljr fttronrs the idea that he withheld an account of the
series of researches which was unfinished, whilst he was readj and will-
ing to rereal the other, which was completed. Moreorer, tbe conunnnt-
cation made to Priestley was in all probability an oral, not a written one;
and the latter, we cannot doubt, would not be a silent and irresponaTe
listener, but would in his own lirely way question Gayendtsfa on tne sub-
ject of his communication. In such a conversation, however, between
two very sagacious men, on a subject in which they were both greatly
Interest^, it could not but be, that some explanation would be given by
Cavendish of the objects which he had in view, in experimenting in a
different way from Priestley, and of the extent to which these objects
were attained by the change of procedure. In other words, some state-
ment would be made regarding the conclusions to which they led.
Whilst, therefore, it is impossible, and would be idle to attempt to define
with absolute accuracy the nature and extent of the information which
Cavendish gave to his brother philosopher concerning all 'Hhe foregoing
experiments," it can still less be conceded to the advocates of Watt (hat
Cavendish himself intended us to understand that he had no conclusions
to communicate, or that he did not choose to communicate those which
he had drawn. Cavendish, in a word, must not be adduced as a witness
against himself^ to the effect that lie revealed no conclusions to Priestley.
Thus much premised, I now turn to Priestley's own account of wiiat
Cavendish told him.

Priestley's statement is meagre and incidental, but nevertheless ot
much importance* According to Lord Brougham, its author ''says
nothing of Mr. Cavendish's theory, though he mentions his experiment;"*
and this ground is generally taken by the other friends of Watt Tiic
whole passage from Priestley's paper has already been quoted in demon-
stration of the priority of Cavendish's experiments. (Ante, p. 284.) A
reference, accordingly, to a part of the first sentence will be sufficient
here. "Still hearing of many objections to the conversion of water into
air, I now gave particular attention to an experiment of Mr. Cavendish's,
concerning the reeonvergion of air into water, by decomposing it in con-
junction with inflammable air."t This passage, historically considered, la
a very remarkable one. It contains the first direct reference which has
reached us of the possibility of converting a mixture of two elastic fluids,
vi«., pure air (or oxygen) and inflammable air into water. It refers, more-
over, not to a phenomenon (for no one ever tauf gases change into water),
but to a conclusion, vis., that in virtue of certain appearances which attend
the "decomposition** of air (oxygen) in conjunction with inflammable air,
it might or should be inferred, that the gases in question undergo conver-
sion into water. This conclusion or inference, moreover, stands connected
with " Mr. Cavendish's experiment ;" and was either drawn by him, or
by Piiestley, before that repetition was made, which supplied the basis of
Watt's conclusions. It is necessary, accordingly, to inquire somewhat
minutely into Priestley's reference, with a view specially to determine
whether Lord Brougham and those who agree with him, are right in
afiinning that no indications of a theory appear in it.

In the first place, then, it may be noticed, that Priestle/s statement

• ffiitoHcai Note, Watt Corr. p. 246. f PM. Tram. 1783, p. 426.

THE WATER CONTROVEESY. 383

coutaina an account, (1.) of an experiment, and likewise (2.) of the
pufpoee for which it was tried, or the conclusion which was drawn from
it. 1. The experiment consisted in decomposing air (respirable air or
oxygen) in conjunction with inflammable air. 2. It was an experiment
'' concerning the reconversion of air into water.**

To clear away a slight preliminary difficulty, it may be obserred,
that Priestley speaks of the reconversion of air into water ; not of the
oonversion simply, for a reason which the previous part of his paper
explains. He was engaged in experiments on the conversion of water
into air at the time when he learned or recalled the account of Caren-
dish^s trials] and regarding the latter as complementary or correlative to
his own, he uses the duplicative particle re to indicate this. It may be
out off, accordingly, and then Cavendish's research will be named, *' An
Experiment concerning the Conversion of Air into Water." Priestley's
words thus imply, that the experiment in question was one of two things.
1 . An experiment or research made to ascertain if air is convertible into
water; or 2. an experiment, which for whatever purpose underta<cen,
led to the conclusion thcU air in convertible into water.

Cavendish 8 researches, in truth, come under both definitions, if his
own account of matters is accepted; but into this it is not necessary to
enter here, as it will be enough if I can show that they come uuder the
second definition. Let it be observed, then, that whilst Priestley does
not give any details respecting the nature of Cavendish's experiment, he
plainly intends us to learn these, from the description which he proceeds
to give of his repetition of it; so that the general steps of their procedure
were the same, and what was observed by Priestley must be re^rded as
having been previously observed by Cavendish. Thus much, indeed, is
conceded by the advocates of Watt, at least by Lord Jeffrey, who says
of Priestley's experiments, " the whole series was professedly entered on
as a mere repetition of those of Cavendish."*

Priestley reported himself to have observed, that when inflammable air
(of a certain kind) and oxygen were exploded together, a weight of water
was found, as nearly as he could judge, equal to the weight of the gases
burned; and the conclusion which he drew from it he thus states : — " The
result was such as to afford a strong presumption, that the air was recon-
verted into water, and therefore that the origin of it had been water."t
In other words, he drew from his repetition, more or less decidedly, the
conclusion, that inflammable air and oxygen are convertible into water;
and that water originates in, or is produced from, or out of these gases.
This conclusion was by no means entertained by Priestley with the
unhesitating confidence which marked its adoption by Cavendish and
Watt, and was by and bye abandoned by him, exactly when the evidence
in favour of it became strongest; but when he wrote the passage on
which 1 am commenting, he thought that there was at least ^' a strong
presumption ' of its truth. The meaning, then, of the word *' concerning^*
as applied to the experiment detailed, will now be apparent. It was an
experiment " concerning the conversion of air into water," in the sense of

* Bdinr, Rep. Jan* 1848, pi 04. It has been ehewn already, however, that
Priestley's repetition was not an exact one, and Carendish pointed out that Priestley
(1) had not employed hydrogen as he did, but inflammable air from diarcoal; (2) that
be probably used a greater proportion of this gas; and (3) that he did not observe the
production of nitric acid. These points of difference, however, notwithstanding their
intrinsic importance, are immaterial to oar present inquiry.

t PAU. 7Van«. 1783, p. 427.

384 CAVENDISH AS A CHEMIST.

an experiment which rendered it highly probable that air is oonvertible
into waiter, and that water has its origin in the gases which can be con-
verted into it. This experiment, moreover, was Cavendishes; not
Priestley's. Before the latter proceeded to repeat it, he recorded it as
"Mr. Cavendish's experiment concerning the reconversion of air into
water,** so that the conclusion which it justified in Priestley's hands, it
had already justified in Cavendish's. In short, before Priestley had made
that repetition of experiments which furnished Watt with the ground of
his conclusions, Cavendish had already furnished Priestley with their
interpretation. The evidence seems to me conclusive; but I wish care-
fully to mark the limits of the interpretation thus furnished. I do not
seek to affirm that Cavendish gave Priestley a lengthened exposition of
his views on the composition of water. All that I wish to claim is, that
when Cavendish reported his experiments of 1781 to Priestley, he accom-
panied the report with thus much of conclusion, that they proved tluU
infiammable air and oxygen could be converted entirely into pure water j
and that water originates in tluse gases.

Against this view of matters the advocates of Watt strongly protest,
affirming that Priestley was at the greatest pains to disavow having
drawn any conclusion from his repetition of Cavendish's experiments ;
and that, on the other hand, he was careful to explain that he was
entirely indebted to Watt for the conclusion which he connected with
them. Lord Jefirey refers at greatest length to this view of the case,
and concludes his protest by asking, " Is it possible to doubt that the

expression as to the experiments ' afibrding a strong presumption'

of the truth of this theory, was never intended to imply that Priestley
had himself originated that theory, or even been strncK with the force
of that presumption ; but only that, after it had been suggested by his
friend [Watt], he could not but acknowledge that there were some
probabilities in its favour f *^

I do not refer to his lordship's arguments in full, because, in tmth, it
seems to be a sufficient reply to thein to point out that they prove too much.
They, in eflect, represent Watt as having interpreted experiments before
they were made, and Priestley as having tried experiments with positively
no object, unless to ascertain what Watt would say of them when they
were reported to him. Watt's friends are not very consistent in main-
taining this, for they require us to acknowledge that he was unacquainted
with the contents of Priestley's paper read June 26th, 1783 ; yet as
Priestley sent that paper to the l^oyal Society on April 21st, he cannot
have been cognizant of the contents of Watts letter of 26th April (the
germ of his '* Thoughts on the constituent parts of water,"^ of the same
year; so that it does not seem yqtj reasonable to affirm that Priestley,
who had not yet received this letter, can have been fully furnished with
an account of Watt's hypothesis. From internal evidence, however, it
is manifest that additions were made to Priestley's paper after the date
(April 2l8t) prefixed to it in the Transactions for 178d.t Before the

* Edinr. Rev, 1848, p. 92.

t This is evident from the fact, that the earliest acoount which Watt received
from Priestley of the delusive character of his experiments on the power of poroos
retorts to convert water into air, was in a letter from London, of 6tAe April 29th, that is,
eight days after Priestley had sent his paper to Sir Joseph Banks, whilst the paper
itself contains an account of the experiments reported in this letter to Watt. The paper
mast therefore have been altered before at least it was printed, and, in all probability,
before it was read in June. The title also supplies evidence of alteration, for it ooatirna

THE WATER CONTROVERSY. 385

additions were made, (before iudeed the 29th of April) Priestley was
iiuDili&r with the contents of Watt's letter, bat this only renders his total
omission of Watt's name in connexion with the experinients on the con<
version of inflammable air and ozwen into water, the more at yariance
with the enpposition that he intended to acknowledge obligation to Watt
for the '' strong presumption " which his repetition of CarendisVs experi-
meots led him to entertain.

It is no part of my arenment, whilst reasoning thus, to represent
Priestley as ignorant or negligent of Watt*s conclusions from the experi-
ments in question. But it is a most material part of my argument to
insist upon the passage under consideration, as not referring to conclusions
drawn by Watt^ but to conclusions drawn by Cavendish. The friends of
the former can assign no reason why his name should have been omitted
by Priestley, when recounting his repetition of Cayendish's experiments.
It could not be from reluctance to allude to Watt, for he refers to him
frequently throughout the paper, and the omission of any allusion to him
when reierring to his own " strong presumption,'* is irreconcilable with
the notion that he wished to be understood as indebted to Watt for the
conclusion he connected with his own experiments.

Lord Jeffrey, indeed, attaches importance to the circumstance that
Priestley disclaims all pretensions to theorise upon the facts he commu-
nicates in the letter to Sir Joseph Banks which accompanied his paper.
" The principal /octe," says he, ''are I think sufliciently ascertained; though
I do not presume to give any opinion with respect to the theory of them."
This, however, was only one of those general disclaimers in which Priest*
ley was fond of indulging, when his habit of hasty generalisation had
involved him in more than ordinary confusion. Never was this disclaimer
more out of place than on the present occasion, for his facts were many
of them most insufficiently ascertained, and were afterwards disproved or
contradicted by himself; and his paper was full of theories. Singular,
also, though it may appear to those who are not familiar with the incon-
sistencies of statement which abound in Priestley's writings, he acknow-
ledged in another place that this paper did contain conclusions, i.e. theories.
In 1785, he published experiments and observations relating to air and
water. In the account of these he says, " In the following experiments I
also had a particular view to a conclusion which I had drawn from those
experiments, of which an account is given in my last communications to
the Royal Society ; viz. that inflammable air is pure phlogiston in the
form of air, at least with the element of heat" This reference shows
that the communication he refers to, is the one adduced by Lord Jeffrey
as containing the disavowal of any purpose of theorising. Yet Priestley
himself furnishes this comment on it, " I shall have occasion to notice my
own mistakes with respect, to condusions, though all the ia^ were strictly
as I have represented them."*

the words, **ieeminff conversion of water Into air;*' whereas, when the paper was writ-
ten, the conTersion was believed to be real. Illastrations of the same fact will be foond
at p. 414 {Pkil. TVoNf. 1783), where reference b made to experiments *' which atjirti
teemed to favour the idea of a conversion of water into air;" and the three condnding
pages of the memoir must, from their contents (pp. 430—434), have been added after
the experiments which brought to light the falladons nature of the conversion had been
made. The fact, we shall afterwards find, has a secondary interest, as demonstrating the
liberty which was conceded to members or correspondents of the Royal Society, to
make additions to their papers after they had been sent to its oflSdals, and this without
alteration of the original date.

• Pkii. TVoiM. 1785, p. 280. Lord Jeffrey quotes another passage to the same

2 C

386 CAVENblSH AS A CHEMIST.

It 18 ainuBiDg, indeed, to bear the great defender of pblogisUMij who
was faithful eyen to death in his allegiance to one of the most risionarj,
and, in the end, clamsj and complicated false theories which the world has
ever seen^ ffravelj declaring himself indifferent to theory. The troth ci
the matter is, Priestley was an inreterate theorizer, only he was constantly
changing his notions, and to a great extent concealed his love of theo-
rizing from himself, by representing the theories he held credible as/octe,
and marking as theoretical only those which he discredited or had abwi-
doned. It is accordingly not at all improbable that he ranked the
conversion of gases into water at the time when he published his paper,
not as a theory, bat as a sufficiently ascertained fietct. It may also be
added that the indifference to theories for which he claimed credit, was in
a certain sense true. It was true in the sense that he loved to change
his theories ; dealing largely in them, but never letting them grow old on
his hands. He had not, certainly, a profound faith in the justice of
Cavendish's conclusions, but he had as little in the justice of "Watt's.
For a short time he believed them, but he speedily abandoned his belief)
and the chief reason, I apprehend, why the tacit appeal of Cavendish and
Watt to Priestley, to act as umpire between them, was never responded
to, is the fact that before the appeal was made, Priestley had oeoome
convinced that both of his friends were in error in their conclusions, and
he naturally thought it a matter of very small importance to settle which
of them had first committed a blunder. The critics of the Water Con-
troversy, accordingly, have in vain sought to claim Priestley as witness-
ing solely in favour of one or other of the English rivals. Of this we
shall immediately have an example. He witnessed clearly in £ftvonr of
neither, and had come to no settled conviction himself, till years afiter
the original controversy was over, when he reached a faith oouoeming
the nature of water wnich no one probably but himself held, and of
which he had to complain that no one would gratify him by offering
to refute it. It is necessary, accordingly, in refemn&f to Priestley's " fiicts
and theories," always to look to their dates, and m application of this
principle, I would here urge that, in April, 1783, he had such faith in
Cavendish's observations, as to publish a repetition of them, and to an*
nounce the " strong presumption " as to the composition and origin of
water which they led him to entertain.

Lord Jeffrey, indeed, seeks to show that '' when reverting two years
after to those speculations of 1783, he [Priestley] takes care in a paper
read to the Royal Society in 1785, to give Watt the sole credit of the
theory of which we are now speaking; and accurately to distinguish
between his own experimentt and the conclusions deduced from them by
his friend — in exact conformity with that partition of credit or of labour
which Watt had publicly adopted at the time. * Mr. Watt,' he says,
* then concluded from some experiments of which I gave an account, and
also from some observations of his own, that water consists of dephlo-
gisticated and inflammable air; in which Mr. Cavendish and M. Lavoisier
concur with Aim.' "* This passage, however, even if we take it merely
as it stands, only shows what in another place I have adduced it to
prove, viz. what were the experiments of Priestley from which Watt
drew his conclusion, and what that conclusion was. It does not declare

effect, from the body of the paper, but not more precise than the one to which I hare
referred. {Bdinr. Rev. 1848t p. 91.) It ia answered, therefore, hi the reply to the
preceding quotation.

* Bdtnr, Rev, 1848| p. 92.

THE WATBR CONTKOVERST. 387

that Priestley drew no conolanon from bis repetition of Cayendiflfa's
experimente before he reported it to Watt, or that Cavendish drew none
from the original trials before he reported them to Priestley; so that the
question whether or not such conclusions were drawn, is left quite unde-
termined bj this reference. The quotation, moreover, given above is a
partial one, and in justice to all piuties the entire paragraph should be
consulted. It is as follows; the passage which is about to be quoted
immediately preceding that already given. ''In the experiments of
which I [Priestley] shall now give an account, I was principally guided
W a view to the opinions which have lately been advanced by Mn
Cfavendish, Mr. Watt, and M. Lavoisier. Mr. Cavendish was of opinion
that when air is decomposed, water only is produced, and Mr. Watt
concluded," See, &o.*

From the entire quotation, then, it appears that Priestley was far
from dividing the merit attaching to the discovery of the composition of
water between Watt and himself. On the other hand, he comwunees with
a reference to Cavendish, and informs us what his views on the subject
were, PriestleVs meaning will be obscure to those who are not familiar
with Cavendish 8 belief that water was an invariable product of oxidation.
I have pointed this out at great length in a preceding section of the
discussion (ante, p. 324), for it has not received sufficient attention from
the critics on either side of the Water Controversy. I need only, there*
fore, say here, that Priestley's statement is to be interpreted as signifying
that Cavendish held that when respirable air (common air or oxyffen) is
decomposed, i. 0. is phlogisticated or deoxidised by combustible or
oxidable bodies, water alone is produced. Cavendish, as I have already
shown, supposed all oxidable or phlogistioating substances to contain
hydrogen and to yield water when burned, whilst their other (hvpothe-
tical) constituents separated, so that nothing but water was called into
being, or as Priestley phrases it "produad; thus nitrogen, ex hypotkeei,
a compound of nitnc acid and hydrogen, produced water and set free the
nitric acid; hydrogen, a compound, ex hypotkeei, of water and anhydrous
hydrogen, when oxidised produced one quantity of water, and set free
another which had pre-existed in it. Into a minute analysis, however, of
the exact meaning of each word it is not necessary to enter. It is
enough if the passage plainly dechires that Cavendish held opinions of his
own concerning the production of water by the action of combustible
bodies on air. He who laid it down as a general proposition that every
combustible when it bums in air produces water, must a forHari have
held this view concerning the particular combustible hydrogen; for it was
the phlogistioating body on wnich he had made the greatest number of
experiments, and the only one which he professed to have seen actually
produce water. Priestley, then, had no purpose of declaring, and never
did declare, that the doctrine of the convertibility of air into water, was
originally taught him by Watt. To assert that he did acknowledge, or
ehould have acknowledged this, proves nothing because it proves too
much, and involves the holders of the opinion in irreconcilable contra-
dictions. According to them Priestley's repetition of Cavendish's ^^cperi-
ments taught him nothing, till Watt interpreted them for him. This
assuredly is an untenable view, for Watt knew nothing of Cavendish's
original trials, nor of Priestley's repetition of them till it had been made,
so that he had no share in inducing Priestley to make the repetition.
Yet certainly Priestley had some very strong motive for trying Caven^-

• Phil. Trtme. 1785, p. 279.

2c2

388 CAVSNDISH AS A CHEMIST.

dish's experiment, and for taking so much pains in preparing the gasei
qoite pnre and dry. Thus much might he urged on any riew of mattos,
hat it acquires peculiar force when we rememher that Priestley had,
earlier than Cavendish, exploded inflammahle air and oxygen in a ghw
glohej had seen dew or moisture after the gases were fired; and had
weighed the glohe before and after the explosion. It was, in troth,
Priestley's experiments including Warltire's, which GaTesdish kad been
repeating, and if this repetition had brought to light no resnlt which they
had not observed, there could have been no occasion for Priestley
repeating a repetition of his own experiments, or for his connecting any
one's name with them, but his own and Warltire's.

There were thtu no parties lees concerned to interest themselves io
what Cavendish had done, if his trials were in no respect novel, than
Warltire and Priestley, so that it is more necessary to seek for a motive
for either of them trying Cavendish's modification of their experiments,
than it would be if the latter had been repeated by any one else. To
Warltire they could give no satisfaction, for they showed that his propoeed
epBpeHmentufn crucis on the ponderability of heat was not crucial; and
that his notion that the water which appeared, was simply deposited from
one or both of the gases which were exploded together, was nnfonnded.
Priestley, on the other hand, had, with more than ordinary discretion,
forborne from theorising upon them, so that he had no prejudice to over-
come on seeing an unexpected conclusion deduced from them; but on the
other hand, he had as little interest in paying attention to the repetition
of what he has styled " a random expenment," and has dismissed with the
incidental notice that inflammable air was not likely to supplant ffun-
powder in warfare. Unless, therefore, his experiments as repeated by
Cavendish had appeared to him in a new and important light, no reason
can be assigned why be should have returned to them in the way he did.
Nevertheless, we find that he went over them a second time with great
attention; that he took great, although misdirected, pains in preparing
the requisite gases; that he thought the results so important that he
reported them to the Royal Society; communicated them to Watt; and
often referred to them years after their performance; whilst not with*
standing all this, he never referred to them as his own, but spoke of them,
considered as one research, as ''Mr. Cavendish's experiment concerning the
reconversion of air into water." All this is inexplicable on the version
of matters favoured by Watt's friends. They miffht avoid acknowledging
that Priestley must have had some motive for following in the steps of
Cavendish, or acknowledge only some trivial reason why he did; but they
cannot escape the obligation under which all critics of the Water Contro-
versy lie, to account for his styling the modified repetition of his own and
Warltire's researches, ''Mr. Cavendish's experiment." Watt's name
must stand altogether aside, for no one can show, or affects to show, that
he had any share in inducing Priestley to try again Cavendish's experi-
ments. Even, therefore, if it could be proved that this second trial
acquired all its significance from the interpretation which Watt put upon
the phenomena which it brought to light, this would only explain why
Priestley recorded and published his repetition, not why he made it.

I set therefore aside as untenable the proposition that Watt supplied
the motive for Priestley's repetition of the " Cavendish experiment" To
assert this is to assert an anachronism. The friends of Watt have justly
enough relieved him from any participation in the blunders which Priest-
ley committed; and have been at no little pains to claim for him a total

THE WATER CONTROVERSY. 889

ignoraaee of what Cavendish had done. They should honestly abide the
coDseqaenees of the position they have assamed, for they cannot with
consistency date Watt's claims earlier than the close of Priestley's repeti-
tion of '^ Mr. Cayendish's experiment."

I return accordingly to the point from which I departed to make this
long digression. My object was to show, that Priestley reveals that a
conclasion, as well as an experiment, was made known to him by Caven-
dish. The latter tells us that it was "in conseqiunce* of certain announce-
ments which he made to Priestley, that the latter made certain trials, " aa
he relates*' in his paper of 1783. On turning to that paper we find that
Priestley has already and spontaneously acknowledged this, and that the
account he gives is to the effect, that being very solicitous to demonstrate
that water could be converted into air, and beset with objections to the
probability of its occurrence, he bethought himself of '^Cavendish's
experiment concerning the reconversion of air into water," and " paid
particular attention to it." He justly thought that if by means of his
experiment he could establish the convertibility of gas into liquid, he
would increase the probability of his own converse proposition that liquid
was convertible into gas. His account, also, of the mode in which he
experimented is in entire conformity with his professed object, however
injudicious his modifications of Cavendish's process were. He spent great
pains in preparing the gases in a state of perfect dryness; he did his best
to collect the water produced during their combustion; and he would
have been glad to have possessed a more delicate balance to have assured
himself that the weight of water produced was equal to that of the gases
consumed. These observations were manifestly made in the light of a
foregone conclusion. The sole object in trying them was the expectation
that they would prove that the gases underwent conversion into water,
and this expectation was entertained because they had already proved
this in the hands of another. That other was Cavendish. He had
reported his experiments to Priestley, not as uninterpreted observations,
but as connected with a doctrine which was new to Priestley, viz. the
convertibility or conversion of air into water.

Two points only call for further notice. The word ''air" does not
signify atmospheric air, but gas, viz. the gas or gases used in the experi-
ments. This is manifest from Priestley and Maty both employing " air,"
when stating the results of trials in which atmospheric air was not made use
of, but only oxygen and inflammable air. It had more special reference,
however, to the oxygen; the inflammable air being most generally referred
to, not as an elastic fluid, but as phlogiston It was intended also, no
doubt, to include atmospheric air so far as its dephlogisticated part or
oxygen was concerned, but Priestley had no occasion to enter into any
precise definition of this, as he referred to Cavendish's experiment only
to prove that respirable air might be converted by phlogisticating it, into
water; and oxygen and atmospheric air were by himself and his contem-
poraries frequently referred to as only different degrees of purity of the
same body, viz, respirable air (ante, p. 216).

The other point is of more importance. What is the exact significa-
tion of the term concerning, as connected with the words, conversion of
air into water ? Some additional light is thrown on this point by the
langua^ used by Mr. Maty in his official abstract of Priestley's jpaper,
which has been quoted already in full. The Secretary refers to Uaven-
dish's experiment as one " tending to prove the reconversion of air into
water." Cavendish's trials were thus brought before the Royal Society

390 CAVENDISH A8 A CHEMIST.

a aeoond time u calonlated to prore snoh a oonreraion. We bare thai
imputed to Gavendishy by botb Priestley and Maty, opiniona ae to the
oonTertibility of certain gases into water. More tban ibis we cannot
witb certainty infer from tbe mere words. But when we take tbem
in connexion witb tbe &ot that Cavendisb's experiments appeared to
Priestley so demonstratiye of conrersion actually occurring^ tbat be made
a careful repetition of tbem to enable bim to' assert that it did ooonr; —
we shall not hesitate to conclude tbat tbe word concerning is intended to
signify that the experiments in Question were regarded by Cayendish as
proving, more or less unequiyocally, tbe conversion which they were said
to concern. I must guard myself, however, against any appearmnoe
of overstating either Cavendish's or Priestley's conclusions. Priestley
assuredly had but a hesitating hAih in the result to which be professed
to have been led. He would not declare that bis repetition irresistibly
compelled the conclusion which he timidly drew; he would not say more
than that his results afforded ''a strong presumption" to that effect; boi
be thought it at the time strong enough to induce bim to make use of it as
an argument in favour of the convertibility of water into air. It is qnite
certain also, that his faith in Gavendisb's results bad been greatly shaken
by his detection, before the publication of his paper, of the f&Uacy of his
own observations on the convertibility of water into atmospberie air.
He cared little, probably, for tbe Cavendish experiment, except as aasist-
ing in the demonstration of his own views; and when they were found by
himself to be hopelessly untenable, be lost his interest in Cavendish's
trials, and probably doubted their truth also. Yet it is curious to notice
bow tenaciously he clings to them even after the abandonment of bis own
speculations. In the close of his paper, after recountins^ an experiment
which militated a^inst his notions, he consoles himself with the reflection
that, though it does not prove the conversion of water into air, still
another experiment '' cannot be explained so well on any other hypo-
thesis, any more than Mr. Cavendish's experiment on finding water on
the decomposition of air;*'* so tbat whilst he would not assert that it
supplied demonstration of tbe transmutability of air into water, and
therefore of water into air, he nevertheless thought that this was the
conclusion naturally deducible from it. He afterwards abandoned this
belief, and in his later papers dissented from Watt, Cavendish, and
Lavoisier, in reference to the nature of water. But he did not do so
because he denied, as some have stated, tbe justice of their reasoning,
but because he had not faith in their data. He had tbe best of all reasons
for losing faith in his own experiments, for, as we have seen already,
they cannot have yielded tbe results he declared they gave; but as rene-
gades generally go to the furthest extreme from their original belief, he
not only denied his own trials, but refused to credit that Cavendish and
Lavoisier had obtained, or that any one else could obtain, from hydrogen
and oxygen their weight of pure water. This change in faith, however,
cannot affect the reality of his former belief, and it was held consistently
enough. He believed the inference just to tbe extent he believed tbe
premises true. No man's belief goes or can go beyond this.

As for Cavendish, if we had nothing but the words employed by
Priestley to guide us to a knowledge of what bis conclusions were, we
should be quite uncertain of their precise nature. But when we take
Priestley's words in connexion with Cavendish's journal, where experi-
ments, exactly fitted to demonstrate the convertibility of gases into water,

• PAiV. TWiif. 1783, p. 433.

THB WATBR CONTROYBRST. 391

are reoorded; and when we find him in hia published paper stating, 'Uhat
almost the. whole of the inflammable and dephlogisticated air is converted
into water; and in another place, '^ that almost all the inflammable air,
and abont -J^th of the common air, are turned into pure water;" and when
we farther find him stating, as the widest generalisation of his entire series
of experiments on air, that water is produced during oxidation, thus oyer^
estimating, instead of undervaluing (as he is accused of doing) the impor-
tance of the appearance of water during combustion: — ^when all this is
considered, we shall not, I think, find much difficulty in believing that
Cavendish employed the word converBum^ when he communicated with
Priestley, in the sense in which he employed it in his published paper.
In other words, Cavendish told Priestley, that if the latter would experi-
ment with hydrogen and oxygen in the way he had done, he would find
that they could be entirely converted into their joint weight of water.
But if davendish could say this, he could also say that he had discovered
the composition of water. As much is acknowledged by Lord Jeffrey,
Tvith his characteristic impartiality. He is referring to the absence of
any conclusions from Cavendish's journal, on which he remarks, ''If he had
even stated in the detail of it [one of his experiments] that the airs were
convertedy or changed, or tumm into water, it would probably have been
enough to have secuied to him the credit of the discovery, as well as to
have given the scientific world the benefit of it in the event of his death,
before he could prevail on his modesty to claim it in public.*'* Such
a declaration he did not make in his journal for reasons which we have
already seen ; but if he made it to Priestley, its force is equally valid.t
I will only add that iome interpretation must be nut upon the words used
by Priestley in reference to Cavendish, by even the most grudging critics
of the latter. They must dispose in some way of this remawable fact,
that before the Water Controversy had or could have arisen, before
Priestley had made the experiments ^whatever they were) from which
Watt drew his conclusions, Cavendish is found entertaining views on the
composition of water. The word conveman could refer onhr to an hypo-
thesis or a theory, to an anticipation or a conclusion. Thus much all
must concede, but with it they must also acknowledge that Cavendish
did not begin for the first time to interpret his experiments, after he
learned the contents of Watt's letter of 1783. The proposition, therefore,
that Cavendish communicated to Priestley only experiments, and not
conclusions, must be set aside; and discussion can only turn on the ques-
tion, what were the conclusions which he certainly revealed?

11. The second statement of Lord Brougham, and those who apee with
him, is, that GavendJsh does not say whenhe mode his communication to
Priestley. This is quite true so far as minute dates are concerned; and

* Sdinr. Rev. 1848, p. 125.

t The journal (Vnlith. MS8.) i» not absolutely silent on thii point. Its index con-
tains the foUowing reference to experiments made between January Srd and February 7th,
1 783. " P. 216. Whether any fixed air is produced by exploding com. and inf. air, and
examinat. wBter produced by exploding com. and inf. air in glass globe." The index is
not dated, and we cannot be certain when it was made. The last portions of it cannot
hmy beetf added till 1785, but, from the appearance of the ink, and the diflPerenoe in the
handwriting, I feel certain that it was filled np from time to time, at intenrals, during
sereral yearf. It is impossible^ however, to fix the date of the quotation, but the words
** water produced" are remarkable as occurring in a reference to experiments made in
January, 1783, and imply that the generation of water from gases was referred to by
CaYen&h as obserrcd by him at that time.

392 CAYENDISII AS A CHBMIST.

in all probability CaveodiBh could not have recalled, when wHiing ia
1784, the precise period to a day or a week, at which^ moie thttn a yetr
before, he reportea his experiments to Priestley. We are apt to forgei
that he could not anticipate the rigid criticism of all his statements, to-
which more than half a centary afterwards the Water Goatrovefsy was
to gire occasion. I make this remark, because conclnsioDs vn&roarable
to his accuracy or fair dealing have been drawn, from the absence on bis
part of minute statements concerning dates, which have aci]«iied all their
importance from long subsequent events; whereas an anxious particularity
about such matters, which at the time could not hare appeared to their
recorder of great importance, would rather have savoured of concealed
design, than have been consistent with the unsuspicious brevity of an
honest and independent observer.*

In truth, however. Cavendish has marked with sufficient precision ike
period when he made his communication to Priestlev. The experiments
reported were made, he tells us, in the summer of the year 1781; and
Priestley, he adds, relates a repetition of them in the PkU, Trans, for
1783. The date of Priestley*s paper is April 2 Ist of that year, so that the
communication must have been made some time between the summer of
1781 and the iq>ring of 1788. By a reference to Cavendish's journal, we
find that the last experiment which he can have communicated to
Priestley was made on September 28, 1781, whilst the majority were
made in July of that year.f

From the WcUt Correspondence, moreover, we learn that Priestley
was experimenting on the conversion of water into air as early as
8th December, 1782;}: and the first reference to the production of water
from the explosion of a mixture of oxygen and inflammable air, occurs
under date 26th March, 1783. So that Cavendish's communication must
have been made to Priestley some time before this date, and probably not
earlier than October, 1781. Such minute references, however, are quite
unnecessary. The only point of any importance is quite certain, vix. that
whatever Cavendish told Priestley, he told him not later than the
sprine of 1783, before the latter made the repetition, and furnished the
data from which Watt drew his conclusion.

III. The third proposition of Lord Brougham and the other advocates
of Watt is, that Cavendish did not state at what time he first drew his
conclusion concerning the nature of water, which his Lordship regards as
a most material omission. This point, however, has been sufficiently
considered in the discussion of his first proposition, where it has been
shown, that Cavendish revealed conclusions as well as experiments to
Priestley when he made the communication, the date of which has just
been determined.

On this point, nevertheless, I would remark, that I have no purpose
of asserting that Cavendish's views on the composition of water, as pub-
lished in his paper of 1784, were fully arrived at, when he made this

* I am very reluctant to wj anything to the prejudice of Lavoisiefi hut it is im*
possible not to be strnck with the needless paiticnlarity with which he records dates, in
an account which all critics seem now ready to acknowledge is a post /actum and
nnjastifiable endeavonr to establish his priority as the discoverer of the Composition of
Water. Watt Corr. M4moire par M, Lawmer, pp. 173—178 ; or Menu de VAead.
pp. 472-474.

t Uth, MSS. pp. 115, 127 & 147, and Brit, Assoc. Report far 1839 p p. 36.

t Watt Corr, p. 3.

THE WATER CONTROVERSY. 393

ooiumimicatioa to Priestley. Ou the other hand, I beliere that they
altered and expanded from 1781 onwards to 1781^ as he became better
and better acqoainted with the composition of atmospheric air, and the
nature and properties of oxygen, hydrogen, and nitrogen. Two distinct
epochs at least can be marked in the progress of his views.

1. The difiooyery that all the oxygen present in air can be converted
by oombnstion with twice its yolume of hydrogen, into their joint
weight of water. 2. The discovery that when the same gases were taken
pore (or apparently pure), they yielded nitric acid as well as water,
which threw a dimcnlty in the way of the condusion drawn from the
experiments made with atmospheric air. The removal of this difficulty
led to protracted researches on the nature of nitrogen, and to an exten-
sion of Cavendish's views concerning the presence of hydrogen in oxidable
bodies, so that he wonld not expound his opinions on the nature of water,
after he discovered the origin of the nitnc acid, in the same way as he
did when unaware of its source. And as the precm period when he
made his communication to Priestley is not known, it would be idle to
attempt minute precision as to how much he told, or could have told the
latter. Thus much, however, as already urged, I contend for, viz. that
Cavendish told Priestley that hydrogen and oxygen could be transmuted •
into pure water, and tha^ therefore, water consisted of these, wliatever
they were.

Could it be shown, nevertheless, that Cavendish was led by the appear-
ance of nitric acid to the conclusion which Priestley, as well as La Place,
afterwards erroneously drew, that the true product of the combustion of
hydrogen in air, or oxygen, is not pure water, but water and (or) nitric
acid, we might suppose that his interpretation of the acid experiments shook
his &ith in the conclusion drawn from his previous observations on the
production of pure water. But we know from the most unexceptionable
anthority, for it is one of the few conclusions recorded in his journal, that
from the first he re^urded the nitric acid as derived from foreign matter.
A record to this effect occurs in his Note-Book, under date S(Bptember,
1781 ;* and he never ceased experimenting till he had shown that this
view was well founded. He made his communication to Priestley, however,
before he had ascertained the cause of the acid, as he tells us himself,
so that it may be considered certain that he confined his statements
to the researches which he made in 1781. To no one, probably, would
he then have announced his views in the form of a proposition, such as,
'' Water consists of dephlogisticated air united to phlogiston,*' as he docs
in his published paper ; for he evidently did not settle the terms in which
he should announce his views, till he had come to a conclusion concerning
the nature of nitrogen, and the source of the nitric acid, both of which
points he disposes of in his published paper, before entering on the expo-
sition of his theory of the composition of water. He was the only one
among all the oMcrvers concerned in the Water Controversy, who en-
countered the perplexing phenomenon of the production of nitric acid. It
might have turned out that this body resulted from the union of sub-
stances present in the oxygen and hydrogen as constituent ingredients.
These might separate from the gases, and form nitric acid, at the high
temperature which attended their combustion when pure, although they
were unable to undergo the same change at the lower temperature which
characterised the combustion of hydrogen and air. In this case, the
ingredients of water would have been proved to be compound substances,

* Lith. MSS p. 147.

394 CATBKDI8R AS A CHEMISt.

which^ by mntnal oombmation prodaced the more complex oompomid,
water. The terms, aoeordingly, in which the oompoeition of water shoeU
be stated, would hare required to be altered, although the fiMst of its ooa-
sisting of hjdrogen and oxygen remained unaltered. An example will
illustrate the justice of this remark. The older chemists were well
aware that when muriatic acid and ammonia meet, they unite and pro-
duce sal ammoniac, and could confidently affirm that the constituents of
sal ammoniac are muriatic acid and ammonia. Their suooeoson prose-
cuted the inquiry further, and discoyered that muriatio add oonnated of
chlorine and nydrogen, and ammonia of nitrogen and hydrogen, 00 that it
could then be stated that the ingredients of sal ammoniac are chlorine,
hydrogen, and nitrogen, and yet uie salt might still with perfect justice be
represented as consisting of muriatic acid and ammonia. So also Carendish,
from the moment that he satisfied himself that when hydrogen and atmo-
spheric air were exploded together, the nitrogen remained unaltered in
quality and quantity, whilst the whole of the hydrogen and oxygen went
into the water, could affirm, once for all, that water consisted of hydrogen
and oxygen, although he might leave unsettled the terms in which he
should announce this, till he had more thoroughly inrestigated the pro-
perties of hydrogen and oxygen. And this is what he actually did; for
whilst to the last he affirmed that hydrogen was one of the ingtedients of
water, he held it probable that the former was in its turn composed of
ingredients, viz. the infiammable air from the metals^ and a little wmter.

IV. I come now to the last, but, as it is considered by those who dis-
allow his claims, the most formidable objection to Carendish's priority.
This is a passage in the letter which Blagden addressed to Crell m 1786.
The letter has been quoted in full already, in illustration of the priority
of Cayendish's experiments to those of the French chemists. We are now
to consider it as throwing light on the date of his conclusions. It is only
valid in BO &r as the disproval of Lavoisier's claim to be an independent,
and the first, discoverer of the composition of water, is concerned. The
friends of Watt, however, take a different view of its value, and attach
great importance to certain omissions which they think they can detect in
it, and which seem to them incompatible witn the claims set up for
Cavendish over Watt. It is thus what the letter does not contain, rather
than what is to be found in it, which is deemed of most importance. The
following passages, selected from the entire letter, as the text of Mr.
Muirhea^'s commentary on it, will also serve to introduce mine : — '' I
[Blagden] can certainly give you the best account of the little dispute
about the first discoverer of the artificial generation of water, as I was the
principal instrument through which the first news of the discovery that
had been already made, was communicated to M. Lavoisier. The follow-
ing is a short statement of the history. In the spring of 1783, Mr. Ca-
vendish communicated to me, and other members of the Royal Society,
his particular friends, the result of some experiments with which he had
for a long time been occupied. He showed us that out of them he must
draw the conclusion that aephlogisticated air was nothing else than water
deprived of its phlogiston, and, vice versd, that water was dephlogisticated
air united with phlogiston. About the same time the news was brought
to London that Mr. Watt of Birmingham had been induced by some ob-
servations to form a similar opinion. Soon after this, I went to Paris,
and in the company of M. Lavoiider, and of some other members of the
Royal Academy of Sciences,* I gave some account of these new experi-

tHB WATBR CONTROVERSY. 895

xaentfl, and of the opinions founded npon them. ........ Bat those

conolnsions opened the way to M. Lavoiser's present theory

He was induced to institute such experiments solely by the aooounts he
reoeiyed from me, and of our English experiments, and he really dis-
coyered nothing but what had before been pointed out to him to haye been
preyioufily mack out and demonstrated in England."* On this passage
Arago remarks :*-''That expression, 'ahotU the game time^ cannot be,
to use Blagden's own words, * the whole truth.^ * About the same time,*
prcvtt nothing; questions as to priority may depend on weeks, on days,
on hours, on minutes. To be preoisely accurate, as he had promised to
be, it was indispensable that he should say whether the yerbal communi-
cation made by Cayendish to seyeral members of the Royal Society, pre-
ceded or followed the arriyal in London of the news of Watt's labours.
Can it be supposed that Blagden would not haye explained so very im-
portant a circumstance, if he could haye brought forward an authentic
date fayourable to his friend f'f To the same effect. Lord Jeffrey says : —
** When he [Blagden] admitted that the news of Watt's concmsion had
come to London about the yery same time with the first reyelation of
Cayendish's, he must haye seen that he had already recognised his right
at least to divide and share the honour of the discoyery with Cayendish ;
and that it wholly depended on the fact of their relative priority, which of
them was entitled to by far the largest share. That he, the client and
partial friend, to say the least, of Cayendish, should hav^e been willing to
Jet their shares appear equal, will be condusiye proof, with most people,
that he yery well knew that a more exact apportionment would haye been
anything but fayourable to his patron. But it is not the less certain that
such an apportionment was due—to truth, to science, and to the parties
themselyes,»-and also that Blagden had the means of making that appor-
tionment; and we fear we must add, that he studiously evaded making it !
He must have known perfectly whether he had first heard of this conclusion
from the one or the other; and, if he had first heard it from Cavendish, is
it possible to doubt that he would haye Said so? After mentioning the
actual communication by that gentleman, it is almost impossible that he
should not (in that case) haye introduced his notice of Watt's by saying,
'Soon after this,' or if it came very soon, ' Almost immediately aifter.'
Even if nothing depended on the priority, this was the natural and
almost ineyitable way of connecting the two notices, if they had really
reached him in that order."]: Similar yiews are expressed by Mr. Muir-
head, who discusses Blagden's letter at great length ;§ and Sir David
Brewster expresses a yery un&yourable opinion in reference to its con-
tents. Blagden's letter accordingly demands a somewhat careful conside-
ration.

It is an essential part of the chief argument urged in the preceding
quotations, to represent Blagden's purpose in writing his letter to have
been to enter into the question of priority, not only as between the
French and English rivals, but likewise as between Cayendish and Watt.
This representation, however, is not, I belieye, well founded; and it is
unportant to show that it is not, and that Blagden restricted himself
to the defence of Cayendish against Layoisier, and purposely ayoided
any minute reference to the priority of the former to Watt. I agree
with the friends of the latter in thinking that Blagden deiigtudly

• lfa« Corr. pp. Ixvii & UviU.

f Bloge of James Watt, Corr. pp. 231, 232.

t Bdinr, Rev. 1848. p. 119. § Watt CoirT. pp. Ixyit— Inil.

396 CAVENDISH AS A CHEMIST.

otooMiei from any attempt to mark aeenmtely tlie date of CtLveadukia
oommunication to him ; bat I differ entirely from tliem aa to tlie motiTes
which led to this* Blagden^s motives^ in trath, in writing to C^IJ, have
been greatly miannderstood and miBrepreeented. He has been aeeoaed ef
officionslj interfering to do Watt a wrong, and of obtnuiTelj making
himself a party in a dispnte with which he had no oeeaaion lo meddle.
These chtfges are not deseryed, and would neyer haye been made had
the circumstances which led to Ids writing been fully known and appre-
ciated. Seyeral facts, howeyer, throwing much light on his oorreapon-
denoe with Crell, have not been referred to by the critics on either side
of the Water Controyersy, whose notice they appear to haye oompleteiy
escaped. I am indebted for a knowledge of them to three papers whicJi
I found among the Cayendish MSS., entrusted to my care by Lord
Burlington.

The first of these is the fragment of a letter in Blagden^s handwriting,
but bearing no address or signature. It was certainly, howeyer^ addressed
to Cayeodish, as appears by the reference in the postscript (b^vafter
giyen) ** to your paper;" and I identified the writing as Blagden's by
comparing it with letters bearing his signature, remaining among the
CayendiBh papers, and with the fiio-simile of his handwriting giyen by
Mr. Weld.* It exactly agrees also with the handwriting of Blagden*s
paper on the ''Cooling of Water below its freezing Point," with the
original MS. of which I haye been fayonred by R. H. Blagden Hale^ Esq.
The following is the entire document :—

*' In a number of Crell's Annals, which I happened not to have looked
over before (May, 1784), I found the following passage: * Mr. Cayendish
in London has imitated (repeated)! the experiments of M. Lavoisier, to
produce water from dephlogisticated and inflammable air by combustion.
He has laid before the Royal Society the result of his experiments, which
confirm that change of the airs, or the new generation of water. His
memoir has met with great approbation, and even the assent of such a
well-informed chemist as Mr. Kirwan. Nothing appears by which it 19
possible to judge from whom Mr. Crell received this information. 't
" Thursday morning, March 10.

<' < It is right to mention that, in the next number of the Annals (for
June), there is a letter from Mr. Kirwan, mentioning your paper in proper
terms, without any notice of M. Lavoisier's name or pretensions. "*§

I have quoted the entire statement as I hare found it amongst the
Cavendish papers, because it proves that Blagden and Cavendish had
encountered CrelFs reference to the latter's researches long before Blagden
published his letter of 1786. It was probably one of his duties, as
Cavendish's assistant, to translate for the latter passages from the foreign
journals referring to chemistry. Various such translations remain among
the Cavendish MSS., especially of papers from the German, in whi^
language Cavendish does not appear to have been a proficient. I gather
as much at least, from a passage in a letter to him irom Blagden^ dated

* Hiit. o/Boyal Society, ?oL ii. p. 175.

t In the original the word " repeated** is written above " iMiiated" evidently as
a synonvme. Tht Germui ia ** Nachgemacht,**

X I have compared this translation, and the two others, which will presently be
given, with their German originals^ and have found them accoiate. I ooald not obtain
personal access to the CkenUeeke Annalen for 1 784 and 1 785. Mr. Weld, however, kmdly
famished me with transcripts of the passsges in question from the copies in the Library
of the Royal Society.

$' Chemieche Annafen Von Dr. Lorenz Crell, May, 1784.

THE WATER CONTROVERSt. S97

** Dover, September 23, 1787," in which he says, " I hope you got Mr.
He^dinger to read GreH's letter; there was something about your sub-*
Bcnption for his journal, which he allows to have been all duly paid, and
an acoount of the freezing of mercury by natural cold in Russia, perfectly
conformable to Mr. Hutchins's experiments. Be so good as to open and
read, or get ready any letters that you think may c<nUain news'** This,
bowever^ is a point of secondary importance.

The account of matters by Crell, contained in the quotation given
above, is as inaccurate as it well could be, and Blagden might well wonder
who could have supplied the information. It represents Cavendish as
having followed Lavoisier, whereas the latter himself acknowledged that
lie had followed Cavendish. It further refers to Kirwan as having
approved of Cavendish's paper, whereas he was the only chemist who at
the time publicly expressed dissent from his views. It was not, impro-
bablv, a desire to correct this mistake that led Kirwan himself to write to
Crell, as Blagden mentions he did in the postscript to his letter. The
following is wat part of Kirwan's communication which refers to Caven-
dish, as I find it translated by Blagden, no doubt for the use of the former.
The words within square brackets are not in the original MS.

*' ' Extract of a Idler from Mr. Kincan in London to Profmor Crell.
{Chem. Aiinals, No. 71. p. 523, June, 1784.) Mr. Cavendbh has laid
before the Royal Society a series of experiments, by which he shows that
water is generated by the combustion of dephlogisticated with inflam-
mable air. And in effect Fin fact, in der that] it is very probable that
in this case dephlogisticated air is converted into water by its combina-
tion with phlogiston [dem Brennbaren] as (since) [da] according to the
experiment, the residuum contains no fixed air, and consists only of a
little phlogisticated air, together with the water. Mr. Cavendish, how-
ever, did not stop at the proof of this important and unexpected pheno-
menon, but went a step further, and endeavoured to prove, that in all
cases in which respirable air is phlogisticated, water is always generated,
and never fixed air^f for which purpose he laboured to invalidate the
proofs which I had given of the generation of fixed air. I answered his
objections in a fortnight, confirming the proofs I had already given, and
adding new ones. He opposed new arguments to this last paper of mine,
to which I again replied, and thus the affair now rests.' "

This communication of Kirwan's probably awakened Crell's attention
to the erroneous nature of the account he had given of Cavendish's
experiments. At all events when he received a copy of Cavendish's
paper he published an abstract of it, along with an apology for the inac-
curacy of his previous acoount of its contents. A translation of part of this
forms the third paper illustrating the correspondence with Crell, which
I have found among the Cavendish MSS. It is marked — " Translation
from Mr. L. Crell's Chemical Annals, 1785, part 4, p. 324. ^ Experi-
ments on Air, and the Water therefrom, by Mr. Cavendish in London.' "j:

* The letter is signed *'C. Blagden/' and addressed to "The Honble. Henry
. Cavendish, Bedford Sqre. London."

t This reference adds another proof to the many given already, that Cavendish was
understood by Kirwan, and allowed himself to be represented, as regarding water as the
invariable product of the phlogistication or deoxidation of air. He made no objection
to such a version of his views being given to the continental philosophers, any more
than to the Royal Society.

X The German is, " Versuche Uber die Luft und das daraus erfolgende Wasser ;
vom Herr Cavendish in London."

SM CAVBNDfSH AS A CHBMI8T.

Crell hafl affixed a note to this title, whioh is the only part of the doei-
ment which ooncenie us. It if translated in the MS. before me, whidiii
not in Blagden's writing, as follows :— <^ * This extraot oontaine the sab-
stance of a paper presented to the Rojal Society in London, by Heniy
Gayendish, Esq., and which has not only been inserted in the Phikwo-
phical Transactions, bat has also been published separate, under the titk
of (Experiments on Air, London, hy J, NichoU, 1784, 4° Maj. p. 37).
Soon f^r Sir Joseph Banks, Bart. (President of the Ro^ Society,) was
so obliging as to send me a copy, for the purpose of mentioning it in these
Chemical Annals. This becomes a twofold duty upon me, becaoae I have
committed the same error as most of my compatriots and other mem of
letters, by iworibing to Mr. Laroisier the discorery of the water resttltin/;
from the different kinds of inflamed air.* (See Ohem. Annals, 17 SS,
Part i. p, 48.) Justice alone therefore demands of me to return to Mr.
Cayendish (whom I take this opportunity to assure of my moat sineere
esteem), the well-earned honour of the Jir$t discovery of this so veiy
important and remarkable phenomenon (which appears dearly from thM
paper), and at the same time to correct some other circumstances in mine
above-mentioned publication.* *'

I have printed these translations as I hare found them in the Caven-
dishMSS. because tfaey are substantially oonect, and they show the exact
amount of knowledge which Blagden and Carendish possessed, oonceming
the different references to the latter's paper whioh were made by Giell.
These points are of special importance as throwing light upon Blagden's
letter of 1786. From the lettm quoted aboye it appears that—

1. It was not Blagden, but CreU, who first referred to Carendish's
experiments.

2. Crell erroneously represented Layoisier as haying preoeded Gayen-
dish, and assigned to the former the honour of the first disooyery of the
composition of water.

8. Neither in Creirs two statements, nor in Kirwan*6 letter, did any
allusion to Watt occur, but only Cayendish and Layoisier were referred
to as connected with the disputed discoyery.

Whilst Crell was thus unintentionally misleading his readers, Laroi-
sier's completed pssearches into the nature of water, which were printed
in 1784, reached England, and Blagden became acquainted with the
representation which Layoisier had giyen of the extent to which he had
made him acquainted with Cayendish's researches. This repreeentaticm
amounted to an impeachment of Bladgen's yeraeity, and to a claim of
independent, if not of first discoyery, in respect of Cayendish. Blagden
accordingly wrote to Crell, denying the accuracy of Layoisier's state-
ments, and asserting in the most decided terms that Cayendish's experi-
ments and conclusions preceded Layoisier*s.

When all that has been stated is considered, the limited object of
Blagden's letter will, I think, be apparent, and he will be freed from any
charge of officiously intermeddling by sending it. There was a peculiar
propriety in CrelFs Annals, rather than any other foreign journiu, being
selected as the channel for communication with the public, becanae that
journal had already published three different papers on the disputed
discoyery, and had commenced by preferring Layoisier's claims, which he
was now publicly asserting in his own name.

* The sense has here been a little mistaken. The German is, '* die Entdeckvngr des
fatten ana d«n angeiftndeten Lnftarten," and shoold be translated, " the diicoverf of
the water from the burned airs," t.e. fh)m the oombii0tio& of oertun gassi.

TH9 WATBR CONTROVBRSY. S99

There WM a poeitive neoesaity also for Blagden being the writer of a
x^plj to LayoUier^ for none but Bladen certainly knew what account he
bad given to the French ohemiats of Cavendish's researches, and the
latter, had he undertaken his own defence, could only have said, as he
did in his paper, that his friend Dr. Blagden had reported to him that he
liad told Lavoisier so and so.

On the other hand, there was no occasion for discussing in CreU's
joamal the claims of Watt, for it had never so much as mentioned his
name, and further it had, by ultimately assigning tiie merit of the first
discovery of the composition of water to Caven£sh, rendered it unne-
cessary to defend him in its pages against Watt. The dispute between
liim and Cavendish had been carried on in England in private, so that its
eiroumstances were probably very little known on the continent, unless
in Paris, where La Place, Lavoisier, Meusnier, and Monge were ready to
divide the honour of the disputed discovery amongst them, without ascrib-
ing anj merit to either of the English ofaimants. An English joumaly
accordingly, was the only fitting place for a discussion of the rival claims
of Cavendish and Watt

Again, whether in an English or a foreign journal, there could have
been no propriety in Blagden adjudicating on the merits of the two
English chemists. He could say nothing for Cavendish, which the
latter could not much better say for himself. And if he would not, in
his own name, enter into public controversy with Watt before the Royal
Society, which was the oidy body qualified to deal with the <|ue8tion, we
may be certain that he would not permit his assistant to vindicate him at
second hand in a foreign journal. For these reasons I contend that the
object of Blaffden's letter was to vindicate his own veracity, and Caven-
dish's originwty, against Lavoisier's implied denial of both. A sense of
justice led him, indeed, to mention Watt as having discovered for himself
what Cavendish first pointed out to all. But he left altogether uncon-
sidered the question of priority of discovery as between them.* Thus much
then premised, I would notice that if the view of the purport of Blagden's
letter which 1 have taken, be the just onOi we must consider him as
having refused for certain reasons to enter into any balancing of merits
between Cavendish and Watt; and we must decline therefore to regard
him as Watt's advocates do, vis. as a witness who said all that he could
say in exaltation of Cavendish's chums, and must therefore be con-
sidered as having had nothing to say on those points on which he said
nothing. As to the motives which induced Blagden to be so sparing in
his reference to the dates which marked the peri^ when Cavendish drew
his conclusions, we can only surmise what they were, and may be quite
wrong in our surmises. We have seen what the hypothesis of the advo«
cates of Watt is, and I have acknowledged its truth to the extent of
admittittg that he deliberately left the dates undefined; but that he did
so because he was conscious that Watt really preceded Cavendish in
drawing his conclusion, I altogether discredit. Other and much more
satis&ctory reasons can be assigned for Blagden's reserve. Before re-
ferring to them, however, I would notice that when he informs us that in
the spring of 1783 Cavendish communicated to him certain results, and
that about tiie same time news was brought to London that Watt had

* I liaTe contended in another place, that Watt'a ooncluiion was Tery different
in signification from CaTendish's, but, at the Mune time, have pointed out that they
paaied for ideatkwl with their coBtempoiBrief.

400 CAVENDISH AS A CHEMIST.

drawn a similar conclusioD^ he mast be ooiindered as intandittg ik ti
understand that he declined entering upon the question of ^iority,
although he tacitly claims it for Caren&h^ by recountin/3^ his coDelvmB
first in order, and referring to Watt's in a more seeondary waj, ibi
similar opinion. It was mainly his belief that Gavendin joeceM
Watt, and he wished that, I think, to be nndentood; bat he entmd M
no justification of this view. It may be impossible for ub to detemuii
why he contented himself with the mere declaration of hia opinion on ^
question of priority, without any attempt to establish it by proof, bat w«
should at least acquit him of the harsh charges which have been broogM I
affaiost him, that he was guilty of equirocation, evasion, or soppresM
of the truth to serve Cavendish, whose assistant he was. For hid
principal and assistant conspired together to misrepresent matters to tke
advantage of the former, which they have been implicitly aeonsed of
doin^, even to the extent of the direct fiftlsification of dates^ noUiiaf
could have been more opportune or less susceptible of detection than the
fabrication of a date in April, 1783, which should nve CSavendish deeidsd
priority over Watt, whose claims date from 26Ui April of that year.
Perjury such as this would have completely escaped detection, unlefl
Blagden chose to make confession of his guilt; out this Cavendiflb,
according to the favourite hypothesis of the advocates of Watt, htA
effectually guarded against, by the liberal gifts and promises of money
which he had made to Blagden. Blagden, then, did not choose is
this case to tell a falsehood to serve his patron, and if he sinned s(
all a^inst Watt, the sin was one of omission. His defence, howcTOv
can be urged on far higher grounds, for as I have sought to show
elsewhere, the reproaches which have been cast on his honour vA
fair dealing, are altogether unfounded. It is highly probable that he
could not establish Cavendish's priority to Watt by reference to a con-
temporaneous record of earlier date than 26th April, 1783. The advo-
cates of Watt argue, as if Blagden could not but possess a memorandom 1
of the very day on which Cavendish had unfolded his conclusions to hinit
so that when the latter's priority was called in question, his assistant
should have been able to produce an authentic document of past dstS|
demonstrating when the communication was made to him. It u foigotteo,
however, in reasoning thus, that Cavendish's priority was not called in
question till the public reading of Watt's paper, April 29, 1 784, ut. *
year later than the period at which Cavencush announced his conclusions
to Blagden. Unless, therefore, they could have divined the occurrace
of the Water Controversy, and had provided themselves with documents
suitable for the vindication of Cavendish's priority, they were not likely
to be prepared with such evidence. It is not to be imagined that Caven-
dish called together a circle of his friends on a particolar day, and for*
mally announced to this select audience that he had formed a conclusion
concerning the composition of water, so that several parties might be
expected to be ready with their note*books to bear testimony to t^
very hour at which the disclosure was made. From the evidence of his
MS. journal, we know that his researches into the source and conditions
of the appearance of nitric acid were not completed till January, 1783,*
so that he was not likely to communicate his conclusions freely, till alter
that date, and yet if disposed to communicate them at all, he might be
expected to do so in the spring months of that year, after the reading of

* IriM. MS8, p. 211, and Brit Amoc, lUportf 1839, p« 37.

THE WATER CONTROV£RST« 401

his paper on ihe New Eudiometer in Janaary, enabled him to prosecute
with undivided attention the other inquiries embraced iu his '' Experi-
meuts on Air/* But in all probabilitj he did not summon a conyocation
of his friends on a certain day, but revealed to them one by one as he
had occasion to meet them, the theories to which his experiments had
conducted him. And to Blagden especially, with whom he was in con-
atant intercourse, his communication was less likely to haye been formal
than to any one else. The principal and his assistant would discuss the
matter together whilst proceeding with the additional researches in which
both were occupied, and the results of all his obseryations would in this
way be conveyed by Cavendish to Blagden, at various intervals during a
period of weeks or even of months. When, accordingly. Watt referred
a year afterwards to his letter of April 26, 1783, as showing the date of
bis conclusions, neither Cavendish nor Blagden miffht be able to produce
any document of earlier date, to demonstrate the utter s priority; since
in ignorance that any controversy would arise, they had taken no pains
to prevent its occurrence; and Blagden might find himself altogether
unable to aflSrm on oath, that one of the many conclusions made known to
him by Cavendish, viz. that concerning the composition of water, had
certainly been made to him before April 26, 1783. It is thus possible
that the silence of both Cavendish and Blagden on this point amounts to
an acknowledgment that they could not produce documentary evidence
£xing the precise date; and even that they could not swear upon oath that
conununioations had passed between them on the point in dispute, before
the date of Watt's letter. But their silence can by no means be inter-
preted as an admission that they acknowledged Watt's letter to have
been known to them before such communication was made. Cavendish
probably found to his regret that he had no means of authenticating his
originality against Watt by extrinsic evidence except by reference to his
communication to Priestley, which for reasons already stated 1 believe he
regarded as -establishing his priority; and beyond this, accordingly, he
contented himself with claiming the conclusions which he published as
bis own, and with acknowledging no obligation to Watt. But to claim
originality was to claim priority, for Cavendish was certainly familiar
with Watt's letter before he published his paper, as Blagden's aJlusion to
Watt in his letter to Crell is in itself sufficient to prove, and nevertheless
when he added a passage to his paper of 1784, commenting on the letter,
be did so only to express dissent from certain of its views without
expressiuff or implying any obligation to its author.

Whether these surmises as to the motives which induced Blagden to
avoid particularity are well founded or not, I leave the reader to deter-
mine. It is of more importance to insist upon two points which are not
hypothetical: ] . No charge of plagiarism was publicly brought by Watt
against Cavendish. We know from Watt's correspondence that he expressed
ill private strong suspicions regarding the fair dealing of his rival: wo
know also, however, from the same source, that the only ground of these
suspicions was the real or supposed identity of Cavendish's conclusions
ana his own, and the fact of his letter to rriestley having been made
known to his rival. But Watt had no knowledge as to the period at
which Cavendish had drawn his conclusions, so that, for anything he
could show to the contrary. Cavendish might have preceded him even by
years. In his published paper, accordingly Watt brought no charge
against his rival, and the only reference he made to him occurs iu a note,
added shortly before the publication of his letter in the shape in which it

2 D

402 CAVBNDISH AS A CHBMIST.

now appears.* This note, moreoyer, is occupied with an aseriptkiii e/
credit, not blame, to Gayendish, and Watt limited faimflelf to what te
could pro ye, vis. that he entertained certain yiews regarding the oom-
position of water in April, 1783. There was thus no public charge prfr-
rerred against Cayendish, nor is it in any degree likelj that he was made
acquainted with the suspicions priyatelj entertained against him hj hit
English riyaL Cayendish contented himself accordingly with adding ts
his paper the first interpolated passage in which he refers to what he had
told Priestley, and what Blagden had told Layoisier conoeming^ his
researches. The reference to Priestley showed that Cayendish preceded
him, and therefore, as I haye already contended, Watt who followed
Priestley. Against Layoisier, on the other hand, he brought a direct
charge, in reply to the latter*s implied denial of his originality.

2. If Cavendish thought it unnecessary to make a formal avowal of
priority to Watt in 1784, there were additional reasons why he shooid
not, either in his own name or in Blagden's, assert it in 1786. We
learn from Watt's son that his father, "after becoming in 1785 s
Fellow of the Royal Society, formed the personal acquaintance of Mr.
Cayendish, and liyed upon good terms with him/'t It would havs
been fatal to this friendly acquaintance to haye reopened the contro-
yersy in CrelFs journal in 1786, and it was eminently in keeping witli
Cayendish*s notorious indifference to fame, that he should haye forbiddea
Blagden to stir the question of priority between him and Watt; and
it was not less in keeping with his nndeyiating loye of truth thai h«
should hayo sanctioned, as perhaps he dictated, the reference to Watt
as an independent obseryer. It was yery different with Layoisier; be
had impeached, by implication at least, the yeracity of Blagden, and
through him that of Cayendish, and the latter, though too modest and
unambitious to be yery solicitous about his intellectual reputation, could
not afford to lose his good name, and took ample measures accordingly
to defend himself against the representations of Layoisier. - The silence,
therefore, of Cayendish and Blagden in reference to Watt's claims^
implied no concession of them, and is explicable on natural and satisfac-
tory grounds.

It cannot after all, howeyer, but be regretted that Watt did not
publicly prefer a charge against Cayendish, which would haye justified
and rendered necessary an equally public reply. As it is, charge and
reply are equally wantmg; the former, it must nut be forgotten, as much
as the latter. The friends of Watt constantly argue as if a charge had
been made but neyer refuted, and carry the reader's sympathies with them,
especially if he forgets, as he is very likely to do, that the Watt Corre-
spondence which he reads in print as a public accusation, was in Cayen-
dish's time a sealed book to nearly all the world, and certainly to him. |
If we adopt the opposite yiew and take for granted that he honestly
discoyered the truths which he taught as his own, and that he supplied
the materials from which Watt drew his conclusion after they passed
through Priestley's hands, who had in part interpreted them, we shall not
find anything strange in Cayendish's simple avowal of his own originality
and priority. The deep consciousness of this would keep him from |
anxiously demonstrating that he had reaped as well as sown the harvest 1
which others claimed, and the absence of any public denial of this |
deprived him of the opportunity of vindicating his capacity and integrity.
He was apparently content that it should be so, and Blagden, probably in
conformity with his wishes, stated that he and Watt had announced their

• PAil. Trans. 1784, p. 332. f Wait Cbrr. p. ir.

5PHB WATER CONTROVBRSY. 403

eonoluflions about the same time^ whilst he was careful to claim for
CaTendish originality, and by giving him precedence, also priority.

The section just completed, has extended to a great length. It will
be observed, however, that it is the most important division of the Water
OontroTersy so far as Cavendish is concerned, inasmuch as it is occupied
"with his defence against the charges of unacknowledged obligations, and
of posteriority to Watt, which have been so largely preferred against him.
Throughout this section I have acted chiefly on the defensive; I now
recapitulate very briefly the chief conclusions contended for, in the form
of direct arguments in favour of Cavendish.

In evidence then of Cavendish's priority to Watt, the following
indirect but important proofs may be announced.

let. It is matter of positive certainty that Cavendish was the first who
converted a given weight of hydrogen and oxygen into the same weight
of water; and that he did this both with hydrogen and the oxygen of air,
and with hydrogen and pure oxygen.

(1.) The experiments of Priestley from which Watt drew his con-
dasion, were, besides being very inaccurate, confessedly a repetition of
Cavendish's, and therefore later in date.

(2.) The experiments of Lavoisier were also, and confessedly, a
repetition of Cavendish's : and

(3.) Monge's experiments, which were original, are acknowledged by
himself to have been of later date than the English researches.

2nd. Cavendish's reference to Watt in his paper of 1784, contains no
acknowledgment of obligation, or concession of priority to the latter.
Cavendish never formally asserts his priority to hia English rival because
it was never formally called in question. He was the nrst to publish his
views, and at the period of publication there was no rival in the field.
When Watt afterwards indirectly sought to establish priority over him,
he made no alteration in his original demands, but contenting himself with
acknowledging Watt as, on his own showing, an independent observer,
he maintained for himself all that he had previously claimed.

3rd. Cavendish's manuscript journal is mainly a record of facts, with
a statement occasionally of particular conclusions from single experiments
or isolated researches, but it does not embody any of the great generaliza-
tions to which those observations, considered as a continuous series, con-
ducted him. Its silence, therefore, in reference to conclusions, supplies
no argument against such having been drawn; whilst a comparison of it
with the published paper of 1784, shows that its recorded facts were
generalized by their observer, and employed by him to establish various
theories announced in that paper. The probabilities, therefore, are all in
favour of the inference that the experiments recorded on the synthesis
of hydrogen and oxygen formed no exception to the others detailed, but
that like them they were more or less fully interpreted at, or soon after, the
period when they were made. The dates, therefore, of the journal may
justly be considered as marking more or less precisely the periods when
the various theories published in the paper of 1784 were formed; and thus
January, 1783, may be considered as the latest date, denoting the time
when Cavendish K)rmed his conclusion concerning the composition of
water; whilst the probabilities are exceedingly great that he was ready to
announce that hydrogen and oxygen are the elements of water, in 1781,
after completing his experiments on the combustion of hydrogen and
cunimon air.

2 d2

40i CAVENDI8H AS A CHSMIST.

4Ui. Cavendish and Priestlej both bear witness to an aceomt
been given bj the former to the latter of his experiments of 1781; ui
from the tenor of both statements, it appears manifest that CstmiM
taught Priestley a process by which the supposed element water cosMlt
proauced out of hydrogen and oxygen, by means of the electric spvka
the application of flame, which determined their conversion into witer.
It is farther certain that the communication was made to Priestlev, befsre
he supplied Watt with an account of the experiments from whidi k
drew his conclusions.

5th. The letter which Blagden wrote to Crell in 1786;, was itttaiM.|
to defend his own veracity against Lavoisier's impeacluneiit of it, sad t»!
claim for Cavendish priority to the French chemist^ who was likewise
accused of plagiarism from his English rival. But it formed no psit cf
Blagden's intention to discuss the question of priority between Cavendisi
and Watt, so that his silence on this point, which the Eng^lish rivals ven
quite competent to settle for themselves, supplies no argument against tkr
priority of Cavendish, and this priority, moreover, Blagden asserts, althoogfc
he enters into no justification of his assertion.

It thus appears that the priority of Cavendish's experiments can k
established beyond the possibility of cavil, and that the priority of bis
conclusions is m the highest degree probable, if not absolutely certain.

12. Date of Priestley^ s Experiments, and of Watt's Conclusions
from them concerning the Composition of Water,

The questions with which this section is occupied, have to a gresi
extent been anticipated in the preceding one; but there are some points
raised by the friends of Watt which call for additional consideration.
According to Mr. James Watt, his father's theory of the composition of
water, was formed long before Priestley instituted his repetition of Ca-
vendish's experiments. On this subject he remarks, ^'It may wilfc
certainty be concluded from Mr. Watt's private and unpublished letters^
of which the copies, taken by his copying machine, then recently invented,
are preserved, that his theory of the composition of water was already
formed in December, 1782, and probably much earlier. Dr. Priestley in
his paper of 2l8t April, 1783, p. 416, states that Mr. Watt, prior to his
(the Doctor's) experiments, had entertained the idea of the possibility of
the conversion of water or steam into permanent air. And Mr. Watt
himself, in his paper, Phil. Trans, p. 33d, asserts that for many years he
had entertained the opinion that air was a modification of water, and he
enters at some lengtn into the facts and reasoning upon which that
deduction was founded.'"*

The "private and unpublished" letters here referred to, are thofe
which, after the note was written, were printed in the Watt Correspon-
dence, and a comparison of it with Watt's published paper will show that
his son's view is quite untenable. Mr. Muirhead indeed, though he re-
prints this note, gives a diiTercnt account of ^matters, and so do Lords
Brougham and Jeffrey, The first of those gentlemen details the progresd
of Watt's views, in the sumniaiy placed at the end of his introductoiy
remarks, in the following terms.

♦ Note by Mr. James Watt to Lord BroughanCa Hhtorical NoU, Wati Cwr.
p. 248.

THE WATER CONTROVERSY. 405

« 1782 :

^' 13th December. — Mr. Watt in writing to Mr. de Luc and Dr. Black,
in en lions an opinion which he had held for many years^ that air was a
modification of water; and that if all the latent heat of steam could be
tamed into sensible heat, the constitution of the steam would be esseu-
tistUy changed^ and it would become air«

*'1783:

" Dr. Priestley having put dry dephlogisticated air and dry inflammable
&ir into a close [glass] vessel, and kindled them by the electric spark,
finds on the sides of the vessel a quantity of water equal in weight to the
air employed.

" 26th March. — Mr. Watt mentions as new to him, that experiTnent of
Dr. Priestley's."*

The italics in the last sentence are my own, and mark the fact

of most importance to our present inquiry, viz. that Watt was unaware

of Priestley's observations on the synthesis of inflammable air and

oxygen, till March, 1783, and as his conclusion, that water consists of

these bodies, was deduced from the experiments referred to, it cannot

be of earlier date, than the period of their performance. It is true that

Watt had speculated on the convertibility of water into air in 1782, and

long before ; but his hypothesis, as the first extract from Mr. Muirhead's

summary shows, and as I have pointed out at length in a previous section,

(ante, p. 329) contemplated only a conversion of water-vapour into gas,

"without deciding anything as to the chemical qualities or composition of

that gas, except that a preference seems to have been given to the idea

that it would be identical with atmospheric air.f Upon this poiut,

however, it is unnecessary to dwell. Lord Brougham justly observes

" that Mr. Watt formed his theory during the few months or weeks

immediately preceding April, 1783, seems probable." J Lord JeflVey limits

the date still more precisely. He contends, as has been stated already,

that Watt drew his conclusion from unrecorded experiments made by

Priestley with hydrogen and oxygen, but he considers the letter to

Mr. Hamilton, of 26th March, 1783, as marking a period even earlier than

that at which Watt drew his conclusion from the experiments in question.

After quoting the letter wLich has just been adduced from Mr. Muirhead's

summary, he remarks concerning it, " Here we have all the essentials

concentrated, and brought to bear upon each other in one view as if

expressly to demand, at once, and supply a solution. Yet that solution ia

not given I and we must therefore hold, had not yet been clearly perceived.

It had presented itself, no doubt, and was already fermenting in the

powerful and capacious mind which had so clearly conceived, and so

lucidly defined the problem. But the fermentation was not completed,

nor the term of incubation expired. Even the penetrating and intrepid

spirit of Watt was baffled and perplexed for a season, and required time

for consideration and circumspection before coming to a decision. Tbe

decision, to be sure, did come as we know within three short weeks after

this date, and perhaps a good deal sooner. "§

• Watt Corr. p. czziv — v.

f Watt, it will be remembered, at first accepted Priestley's poroas retort experi-
ments, where water was apparently converted into atmospheric air, as the realization of
his hypothesis.

X Hittorical Noie^ by Lord Brougham, Watt Corr. p. 257.

§ Edinr, Rev, Jan. 1848, p. 131. The italics are not in the original.

406 CAVENDISH AS A CHEMIST.

On ibis point, then, the majority of the critics on both sides of ike
Water Controversy may be regarded as unanimous, and Wattes ihewj
of the composition of water as a compound of oxygen and inflammable air
or phlogiston, may be considered as dating from 27th March, I7S3, on
few days later.

The particuhir date of Priestley's experiments cannot be ascertained,
nor is it a point of any importance; by a comparison, however, of Priest-
ley's paper of 1783 with the Watt Correspondence, we can surmise pretty
accurately when they were made. The first letter of Priestley to Watt
is dated 8th December, 1782, and probably marks the period, when the
former commenced his researches into the conversion of water into air. It
contains the account of an experiment on this subject, and the remainder of
the Correspondence, in so far as it discusses the nature of water, is ooco-
pied with references to similar trials up to March, 1783; but no allnaioD
occurs, previous to this date, to the conversion of air (gas or gases) into
water, and the first account of an experiment on this point is contained
in the letter to Mr. Hamilton of 26th March, already referred to. It is
preceded in the Watt Correspondence, by a note from Priestley to Watt,
dated March, but the day is not mentioned, in which one of the latest
and most inaccurate of the former's experiments on the distillation of
water into atmospheric air is recorded. It is highly probable, accordingly,
that Priestley repeated Cavendish's experiments early in March, 1 7€&,
(ante, p. 94) for as he communicated his results almost as soon as he
obtained them, to Watt, and as he, in his turn, speedily commanicated
them to his friends, we should have found some earlier allusion to them
in the Watt Correspondence, if they had been made before the period in
question.

The claims, however, of the various rivals in the Water Controversy
are not afiected by the conclusions to which we may come concerning the
very day, or weeK, or even month, in which Priestley's experiments were
made, or Watt's interpretation was connected with them. All most
acknowledge that Cavendish's experiments preceded Priestley's repetition of
them, and that Priestley's repetition preceded Watt's conclusions from it.

13. Date of Lavoisier* a Conchmons concerning the Composition of

Water.

There is not, and cannot be, any question of priority between Lavoi-
sier and the English chemists, if it be conceded that the latter, or at
least that Cavendish, substantially discovered the composition of water,
although he held peculiar views concerning the nature of hydrogen; so
that in truth it is not imperative on a critic of the Water Controversy
to determine when the French chemist came to his conclusions. But it
will render the discussion more complete if this is considered, and it may
be dismissed in a few words. Lavoisier's own claim goes back •nly to
the 25th of June, 1783,* and even this date is not ratified by his ooUeagne
La Place, who wrote to De Luc on the 28th of the month, stating that he
and Lavoisier were not at that time satisfied that the quantity of water
produced in their experiments by the combustion of hydrogen and oxygen,
represented the weight of the gases consumed.f It is certain, however, |
that on the 25th of June, Lavoisier declared himself satisfied that water

* Watt Corr, {M6m. par M. Lavoisier) p. 178; or Mim. de VAcad, pour 1781,
p. 4/3. t Watt Corr, pp. 41, 42.

THB WATER CONTROVERSY. 407

is oempo66d of hydrogen and oxygen^ and we may accept tbis date as at
lea«t the earliest which he coald affix to his conclusions. It places him
later by some two months than Cavendish and Watt, both of whom we
have seen had, in April, 1783, arrived at the conclusions which they
afterwards published. Even, therefore, if Lavoisier's claim to be an
independent discoverer of the composition of water were tenable, which
it is noty he could assert no claim to be its first discoverer.

QUESTION OF PLAGIARISM.

14. Alleged Plagiarism of Cavendish,

In the preceding sections I have avoided as much as possible all refer-
ence to the accusation preferred against Cavendish, of having borrowed^
without acknowledgment, his theory of the composition of water from
Watt; and likewise that preferred against Lavoisier of having borrowed
his conclusions both from Cavendish and Watt. I have sought indeed to
show, that Cavendish certainly formed his theory for himself » so that he
could be under no necessity of robbing Watt or any one else, and this
view I strongly maintain; but in the present section of the argument, I
shall, as far as possible, set this pre-judgment aside, and consider the
charges brought against Cavendish by Watt and his supporters, as if the
question had come for the first time before us. As for Lavoisier, it has
been impossible to avoid treating him as a detected plagiarist, but it
will be necessary to recur to the accusations preferred against him, were
it only to inquire what can be said in his favour.

I begin with Cavendish. The charge against him arose in the fol-
lowing way. I take the account from Mr. Muirhead's introductory re-
marks, as it is quite explicit, though brief : —

''Mr. Do Luc having gone to Paris in December, 1783, and there
passed the month of January, 1784, returned to England in February,
when his letters to Mr. Watt were resumed. In the meantime, on the
1 5th January, Mr. Cavendish had read to the Royal Society the first
part of his celebrated 'Experiments on Air,' of which the second part
was read on the 2nd of June, 1785. In one of Mr. De Luc*s letters, dated
1st March, 1784, he mentions that he had heard some particuUrs of
the paper which Mr. Cavendish had read, but nothing concerning the
conclusions stated in it as to the composition of water, appears to have
been then reported to him. The imperfect account which he thus re-
ceived came from Dr. Blagden. As the paper, however, was said to have
included a thorough examination of the combustion of the two airs^ he
requested Mr. Cavendish's permission to see it, which was granted.

" The consternation into which he was thrown, on perusing it for the
first time, is well depicted in the close of the same letter : — ' Being at this
point of my letter, I have received Mr. Cavendish's paper, and have read
it 1 ! . . • • • Expect something that will astonish you as soon as I

can write to you; meanwhile, tell no one In

short, he expounds and proves yotir system, word for word, and makes no
mention whatever of you'* In a second letter, written four days later,

" Wait Corr, pp. Iviii & lut.

408 CAVENDISH AS A CHEMIST.

De LttO adds, 'That which is, on the other hand, perfeotly dear, pre*
eise, astonishing, is the memoir of Mr. Cavendish. Your oum, temu,
in your letter of April to Dr. Priestley, given as sometking new, by some
one who must have known that letter, which was known to ail the acUve
msTnhere of the Royal Society — to Dr, Blagden above aU, (for he said he
had spoken of it to Messrs. Lavoisier and La Phice) who well knew Mr«
Cavendish's memoir, both before it was read to the Rojal Society, and at
its reading, and who conversed with me about it, as I told yon in my last—
me, whom he knows to be your zealous friend. After strongly recom-
mending caution, De Luc says, ' It ^is yet possible Mr. CJavendish
does not think he is pillaging you, however protMible it is that he does
80 ;* giving as his reasons for desiring to entertain so charitable a hope,
that Cavendish had not objected to let him peruse his paper, and also
the character which both Cavendish and Blagden had previously main-
tained."*

De Luc was an upright, honourable, and accomplished man, and an
attached friend of Watt. His motives are beyond suspicion, but he is
obnoxious to the grave charge of having made a very partiaJ and one-
sided inquiry into the question he so summarily decided, and of having
leaped to the conclusion that Cavendish had acted unfairly to Watt, upon
a mere perusal of the paper of the former, and without making the
slightest investigation into the history of his researches.

The grounds of the charge he preferred were twofold. 1st. The cer-
tainty that Cavendish had read Watt*s letter to Priestley before he drew
up his paper. 2nd. The suspicious identity of the language in Caven-
dish's paper and in Watt's letter. An acquaintance, however, with
Watt's letter, was quite compatible with Cavendish having formed his
theory before he saw the epistle, and the identity of language, even if it
had been absolute, did not of necessity imply borrowing, if the conclusions
to be expressed in the terms of a very restricted nomenclature, were, in
form at least, identical. It was plainly incumbent on De Luc to have
inquired into both these points, before he proceeded to fill Watt's mind
with suspicions against Cavendish. He made no such iuquiry, however;
nor was there a scientific man in London in a worse condition forjudging
of Cavendish's opinions, than De Lnc. He was reader to queen Charlotte,
and from Mr. James Watt we learn that, '^ following the motions of the
court, [he] was not always in London, and seldom attended the meetings
of the Royal Society.'*f He tells us himself that he seldom went to it;$
and Mr. Muirhead informs us '' that Mr. De Luc having gone to Paris in
December, 1 783, and there passed the month of January, 1784, returned
to England in February. "§ He was thus frequently absent from London,
seldom at the Royal Society, and not even in England, when Cavendish's
paper was read. His acquaintance with English, moreover, though con-
siderable, was limited, as I have shown already (ante, p. 79). De Luc
was, therefore, in most unfavourable circumstances for knowing what
researches Cavendish had been prosecuting, and no one was less entitled
to bring a summary charge of plagiarism against him. In truth, De Luc

* Watt Corr. pp. Ix & Ixi. De Luc's letters, from which the preceding extracts
are translated by Mr. Muirhead, are printed in the Watt Corre^Hmdwuee in their
original French, pp. 42-50.

+ Note by Mr. James Watt to Lord Brougham's Historical Note, Watt Corr,
p. 219.

♦ Watt Corr, p. 43.

§ Watt Corr, Introd. Remarks, p. Wiii,

THE WATER CONTROVERSY, 409

betrays his ignorance of Cayendiah's proceedings^ by the explanation
-vtrbioh he saggests as to the origin of his paper. His notion was, that
'when Watt's letter was privately commnnicated to different members of
the Royal Society, it did not excite attention, ''but that some vague idea
of it may have remained in the mind of Mr. Cavendish, which afterwards
germinated and prodaoed this memoir.*'* According to this view. Ca-
vendish's experiments mnst have been post factum trials, made to justify
a foregone eonolosion; whereas, we know that his experiments on tlie
production of water from its elements, were made before January, 1 783,
and that Priestley repeated some of them in the spring of that year, before
IVatt's letter was written, which is supposed to have led to the original
trials. There is thus no one from whom the accusation against Caven-
dish conld well have come with less weight than from De Luc ; and al-
though I cast not the slightest imputation on his motives, I will say that
his zealous friendship for Watt lessened his qualifications for being an
impartial critic of Cavendish's paper. Let us see, however, what the
force of his accusation is.

That Cavendish was acquainted with the contents of Watt's letter is
certain ; so that if the fact of his having read it, or heard it read, is suffi-
cient to prove him a plagiarist, his defence must be abandoned. It is also
certain, however, that he made no concealment of his acquaintance with
its contents, although he acknowledged no obligation to its author. This
appears from Blagden's letter to Urell, of 1786, in which he refers to
Watt's conclusions as known to him and, by implication, to Cavendish,
in the spring of 1783 ; and still more distinctly from a letter from Kirwan
to Watt, in which the following passage occurs : — " Mr. Lavoisier cer-*
taiuly learned your theory from Dr. Blagden, who first had it from Mr.
Cavendish, and afterwards from your letter to Dr. Priestley, which he
heard read, and explained the whole minutely to Mr. Lavoisier last July
[June]. This he authorised me to tell you."f Cavendish, in truth, if not
a plagiarist, had no occasion either to conceal or to make known his ac-
quaintance with Watt's letter. There is not the slightest reason for sup-
posing that he requested a perusal of the letter. On the other hand,
Priestley appears to have communicated its contents to whomsoever he
thought proper, and to Cavendish directly or indirectly among others.
Watt's supporters, indeed, speak of the letter as having been in the cus-
tody of Sir Joseph Banks, and Watt himself furnished Blagden with a
note to be added to his paper,]: certifying, as he tells De Lnc, that the
letter was '* left in the possession of the president until it was read."§
Blagden, however, altered this statement, substituting for it that " the
letter was reserved," without mentioning the president's name ; || and he
certainly was justified in making the alteration. Watt's letter may have

* Wait Corr, p. 4C.

t Wait Oorr, p. 39. The passage quoted in the text is Tery ambiguous. A
grammarian wonld find it difficult to decide whether it was Cavendish, Blsgden, or
Lavoisier, who authorised Kirwan to make the communication to Watt. There can be
little doubt, however, that Blagden is the party referred to. It is uncertain, also (as
indeed, Mr. Muirhead and Lord Jeffrey acknowledge), whether the words "your theory*'
signify Watfs theory, or " one identical with yours," that is. Cavendish's. I am by
no means certain, for my own part, that the latter is not the meaning of the words.
The question, however, is not of importance to our present purpose. The letter is quite
explicit as to Blagden having heaid Watt's letter read, and from him> if in no other
yisj, Cavendish would become acquainted with its contents.

X Watt Corr. p. 63. § Ibid, Uid, p. 07.

II Phil. Tram. 1781, p. 330.

416 CAV£NDISU AS A CHEMIST.

been^ in' a oertfdn sense^ in the oustody of the president, ftltbongh seeiw
that it was addressed to Priestley, and not to Sir Joseph Banks, and hai
been desired by its author not to be publicly read, it is difficult to be eer*
tain what control the president had over it. It is beyond question, how*
oyer, that it was not in his possession, or in that of any other office-bearer
of the Royal Society, during a part, at least, of the interval between
its being sent and its being read. On the 4th of April it was in the hands
of De Luc, to whom we find Watt writing, desiring him to make an alto*
ration on it;* and on the 15th of the same month, Sir Joseph Bsoks
writes to Watt, " On the receipt of your favor, I wrote immediately to
M. De Luc, requesting him to deliver to me your letter to Dr. Priest-
ley ;**t so that Sir Joseph's custody of the letter was only nominal, and it
is impossible to discover who took charge of it between April 1788, and
April 1784; but at all events it was read to members of the Royal
Society, or circulated amongst them, with the consent of its writer, its
receiver, and its nominal custodier, so that, in whatever way Cavendish
became acquainted with its contents, he was quite at liberty, and was
probably invited, to study them. Nor did the fact of his being allowed
or requested to peruse the letter, carry with it an obligation to make
public reference to its contents. The only point, therefore, of any im-
portance is, had Cavendish formed his theory before he was acquainted
with Watt's letter? I have already, however, contended that he owed
nothing to it, and that he vras, therefore, under no obligation to refer to
it, even if he had thought himself at liberty to do so. De Luc's mere
assumption, accordingly, that Cavendish derived his theory from Watt*s
letter, requires no further consideration.

The second ground on which De Luc bases his accusation is, the
alleged suspicious identity of the language in Cavendish's paper and
Watt*s letter. It was impossible, however, that two chemists coming to
the same conclusion]: concerning the (imposition of water, could differ
much in their mode of stating that conclusion; and the remark applies
with especial force to chemists of the Phlogiston School, whose nomen-
clature was so limited, that it left them little choice of expression in
expounding the narrow doctrine which was their guiding idea. Never-
theless, it must be acknowledged, that language is so elastic and
expansible a medium of thought, that two independent observers of
the same truth are not likely to exhibit absolute identity in their mode
of stating it. That there is nothing, however, suspicious in the similarity
of Cavendish's language to Watt's, will appear, I think, from the follow-
ing comparison. Watt states his conclusion thus : — '' Are we not then
authorised to conclude, that water is composed of dephlogisticated air
and phlogiston, deprived of part of their latent or elementary heat; that
dephlogisticated or pure air is composed of water deprived of its phlogiston,
and united to elementary heat and light,'' &o., &c.§ Cavendish's
conclusion is as follows : — " I think .we must allow that dephlogisticated
air is in reality nothing but dephlogisticated water, or water deprived of
its phlogiston ; or in other words, that water consists of dephlogisticated

♦ Watt Cknr. pp. 49, 50.

t Wait Corr. p. 53.

% X use the word " same" here, in the qualified sense already referred to. Watfs
conclusion was not identical with Cavendish'si for they used the term phlogiston in
different senses; but their employment of the same term makes the wording of their
conclusions identical.

§ Phil, Tram. 1 784, p. 333.

THE WATER CONTROVERSY, , 411

air aniied to phlogiston ; and that inflammable air is either pnre phlo-
giBtoikf as Dr. rriestley and Mr. Kirwan sappose, or else water united to
phlogiston; since^ according to this supposition, these two substances
united form pure water.'** In comparing these statements, it must be
obeeryed that no one of the phrases used by Watt was peculiar to him or
devised by him. The term " phlogiston/* on the other hand, in the sense
of inflammable air, was introduced by Cavendish in 1 766 as a name for
hydrogen,! '^^ ^ synonymous with inflammable air in a less restricted
signification, was as familiar to him as it was to Watt, as a doctrine of
Priestley's and Kirwan's. Even, therefore, if Cavendish had learned to
interpret his experiments from Watt, he did not require to borrow from
him the terms in which he should state his conclusion; and it was impossible^
on the other hand, holding the opinions that he did, that there could have
been any very great diflerence in the wording of his theory from that of
Watt4 Nevertheless, their terms are not identical, as a comparison of
the passages quoted will show; and whilst Watt attaches great importance
to the function of latent or elementary heat, as concerned in the decom*
position and recomposltion of water. Cavendish attaches no importance to
this, but gives great prominence to the doctrine that phlogiston may con-
tain water as an essential constituent. Nor is this all. In three different
parts of his paper Cavendish gives the conclusion from his experiments
on the production of water from its elements, in terms totally distinct
from any employed in Wattes letter or paper. In the summary of his
experiments with hydrogen and air Cavendish says, " Almost all the in*
flammable air, and about -^th part of the common air, lose their elasticity,
and are condensed into the dew which lines the glass." And again,
*' Almost all the inflammable air, and about \i\i of the common air, are
turned into pure waiei**' In the summary also of the experiments with
hydrogen and oxygen, he says, '^ Almost the whole of the inflammable
and dephlogisticated air is converted into pure water,^^

These passages contain the interpretation of experiments made, and
in all probability as we have seen, interpreted before Watt's letter was
written; and they present not only as good, but in reality a better,
because a less hypothetical statement of Cavendish's conclusions than
the passage quoted from his paper. The substitution of the word
'' phlogiston" for "inflammable air," and the notion that the former
might contain water, which originated in speculations concerning the

• Phil. Trans. 1784, p. 137.

t He does not claim the name a« of his devising, nor had he ever sought to
dimotuiraie that inflammable air was phlogiston, although he inferred that hydrogen
was this entity. Priestley and Kirwan professed to have established this, and he refers
to it accordiogly as their view, but the notion that phlogiston and the inflammable air
of the metals are identical, bad been familiar to him for some seventeen years.

X The reader can test the justice of this remark, by trying to what extent he can
Tary the enunciation of the doctrine that water is a compound of phlogiston and dephlo*
gisticated air, and that the latter is water from which its phlogiston has been withdrawn.
One synonyme may be substituted for another, as. Watt says " Water is composed of '%
Cavendish, '* Water consists of " ; but beyond this there is extremely little room for
change of expression. Cavendish, for example, calls dephlogisticated air '* nothing but
dephlogisticated water.'' Watt does not use this phrase, but employs the exactly
equivalent one, that dephlogisticated air is " water deprived of phlogiston." A proposi-
tion so simple as that published by Cavendish and Watt, in the terms of a common
nomenclature, admitted of no material difference in its wording. An ingenious advocate,
ia truth, might make out as plausible a case against Cavendish on the score of the
suspicious variation of his terms from those of Watt, as De Luc does, on the score of
their suspicious identity.

412 9 CAVENDISU AS A CHEMIST.

vatare of hydrogen, mdded nothing to the eleamess of the orisinal viflv.
It would have been better, in tmtb, if Cavendish had avoidedl tke ieim
^ phlo^pMon,'* and had contented himself with stating, that faydrogeii and
oxygen can be condensed, or converted, or tnmed into pure w^&ter. Thii^
however, is immaterial to oar present inquiry; it is enooj^h that tin
passages quoted prove, that Cavendish has stated his con elusion three
times over in his own terms; and that the words of his fourth statemeat
were as much his as Watt's, nay, were peculiarly his, so hkr as the iden-
tification of phlogiston with inflammable air is oonoemed. De Lnc^s asser-
tion, therefore, that Cavendish used Watt's words was totally unj ost ified,
and was made without any inquiry into Cavendish's independent
researches and opinions.

It may be added, that we have the testimony of De Luc himself (as
the quotations from his letters show), that it was inconsistent with the
known character of Cavendish and Blagden, that the one should have
been guilty of plagiarism, and the other a party to it. It further appears,
that when De Luc requested a perusal of Cavendish s paper, its suithor at
once gave his sanction to its bein^ sent to him by Mr. PJautay Sec R. S.;
and communicated his willingness to allow De Luc to read the manuscript
in a courteous note to him through Blagden, although, as De Luc himself
observes, Blagden was well aware oi the friendship which sohsisted
between him and Watt.*

Such were the baseless grounds upon which De Luc hastened to pre-
possess Watt with the notion that Cavendish had stolen his theory from
him; and such was the origin of the Water Controversy, at least as it
concerns the English rivab.

^ The accusation of plagiarism which has just been considered,
privately preferred against Cavendish before his paper was printed, or
Watt's memoir was publicly read. Cavendish, however, we nave seen^
made some additions to his paper before it was published ; and Watt's
communication, which was originally laid before the Royal Society in the
fonn of two letters — the one to Priestley, the other to De Luc, also
underwent alterations, and received additions before it was printed. In
the course of these modifications of the rival memoirs, certain statements
were made, and certain errors in date committed, which, since the revival
of the Water Controversy, have been made the ground of very grave
accusations against Cavendish and Blagden, who are suspected of unfair
dealing towards Watt. In the discussions to which these charses have
given rise, Watt's jpublished paper has generally been considered as
unexceptionable, so far at least as his alterations are concerned; whilst
Cavendish's memoir has been made the subject of much condemnatory
criticism, on account of the additions made to it between the period of
Its being read and printed. I shall presently show, however, that Watts
additions are obnoxious to the venr same charges as Cavendish's; and I
^alJ subject both memoirs accordingly to a criticism of the same kind.
th^ ^*8<^W88ion will comprehend two points : the one, the interpolations;
«<* other, the erroneous dates in Cavendish's and Watt's papers.

* IVaii Cor}\ p. 45.

THE \^ATER CONTROVERSY. 413

15. Interpolations in Cavendishes and Waifs Papers of date 1784.

Tbe word interpolalion, as employed by the advocates of Watt, to
mark the alterations made on the text of Cavendish's paper, cannot per-
haps be objected to on etymological grounds. Yet, I think, I do not
"Wrong his impugners when I say, that they employ the word in its
secondary sense, as an improper alteration, or falsification, such as would
destroy the validity of a legal document, and throw doubts upon the
authenticity and authority of a sacred codex. It might have been well,
accordingly, if some less equivocal word had been used, such, for exam-
ple, as ' insertion,* ' alteration,' or ^ addition,* as the case mifi;ht be. I
am not anxious, however, to dispute about words, and 1 shall continue
to employ the questionable term in the sense simply of an insertion or
addition, although I hope to be able to show that the so-called interpo-
lations in Cavendishes and Watt's papers are very innocent and harmless
things. First, then, of Cavendish.

The alterations made on Cavendish's paper of 1784, between the
reading and the printing, are three in number; two of them being inser-
tions or interpolations in the body of the memoir, and one of them an
addition at its close. They have been pointed out already in the
abstract of the " Experiments of Air," and are marked by square
brackets in Mr. Muirhead's reprint of the paper.* Nothing appears in
the printed paper to distinguisn these passages from the rest of the text.
They have been discovered by a reference to the MS. in the Royal
Society's archives; and we are indebted to Lord Brougham for bringing
them to light. The supplementary addition or postscript refers solely to
Lavoisier's views on the nature of water, which did not reach this country
in their completed shape till after Cavendish's paper had been read, so
that he could not refer to them previously. This addition is in Caven-
dish's own hand-writing, but it has given rise to no controversy between
his advocates and those of Watt, so that it need not further be referred to.
The two interpolations were doubtless made, in consequence of Watt's
publication of his letter to Priestley of 1783, and his implicit claim of
priority over Cavendish. The first interpolation contains the passage so
frequently referred to, in which Cavendish announces that he had com*
municated certain of his experiments to Priestley, and that Blagden had
made similar communications to Lavoisier. Ihe second interpolation
contains the reference to Watt's paper, and the reasons which influenced
Cavendish in making no allusion to latent or elementary heat, and has
been already discussed. Both insertions are in Blagden's handwriting. f-
Sach are the interpolations so much referred to, consisting, it will be
observed, of two entire paragraphs inserted into the text, which, so far as
1 am aware, was not altered. It is difficult to see what fault can be
found with them, provi<led it was permissible to introduce interpolations
at all — a point to which I shall presently refer. Nor in truth, do the

* PA<7. Tratu. 1784, pp. 134, 140, 150. Mr. Muirhead's reprint. Watt Corr.
pp. 129, 135, 147.

t We are indebted to Lord Brougham for pointing out this; and Mr. Weld has
done those interested in the matter the farther service of printing a facsimile of part of
tbe first interpolation, along with facsimiles of Cavendish's and Blagden's handwriting,
10 that all may determine for themselves by which of these parties the interpolation was
written. {Hi$t, of the Royal Society, vol. ii. p. 174.) Tlie writing is certainly
Blsgden's. It has some similarity to Cavendish's, but is nevertheless easily distinguished
from it* The point, however, is one of very small, if of any, importance.

414 CAVBNDISH AS A CHEMIST*

supporters of Watt openly complain of the contents of these inserted
passages, although natural! j enough thej r^ret them, Beemg that the fini
contains the most cogent proof we possess of Cayendish's priori tj to Watt.
Nevertheless, tbey refer to the interpolations as suspicious, if not nn&ir, j
and attach ^eat importance to the fact of their heing in Blagden's hand- j
writing. Mr. Muirhead, for example, carefully incloses them between
brackets, and marks them as interpolations "hy Dr. Biagden, after the papa
had been read.'* Sir Dayid Brewster, also, speciallj notices that the
two additions are in the handwriting of Dr. Blagden ;* and afterwajds
reproaches him for having "inserted his interpolations in Cayendish^s
memoir; "f adding, '^it is tbe testimony, therefore, of Dr. Blagden alone,
that has disturbed the current of scientific history. It is his testimcmy,
not appealed to by Cavendish, but gratuitously offered by himself, that
contains the allegation that Cavendish mentioned to him and others his
conclusions.''^ It is vain, however, to attempt to separate Blagden from
Cavendish, or to build an3rthing on a matter in reality so trivial as the
handwriting of tbe interpolations. I think that I do not wrong the
supporters of Watt when I say, that a reluctance to impeach Cayendiah
has led tbem to shift tbe supposed blame of making these additions from
him to Blagden. The blame, however, if there be any, cannot so be
shifted. The attempt to do so, if successful to the extent of implicating
Blagden, only gives us two culprits instead of one, and increases Cayen*
dish's guilt, inasmuch as it convicts him of cowardice as well as dishonesty,
and represents him as trying to hide his malpractices by bribing hie
dependent to become his cat's-paw. Sir David Brewster offers no proof
that Blagden's interference was unsolicited and gratuitous; and none
can be given. We have no means of positively deciding why the in-
terpolated passages occur in Blagden's handwriting; but of this we are
quite certain, that Cavendish adopted and homologated them in full, for
he permitted them to appear as integral parts of a paper which went
forth to the world as written " by the Honourable Henry Cayendish,
F.R.S." Much more, however, may be said in defence of Blagden. The
first part of the first interpolation, which refers to the communication to
Priestley, can with no probability be supposed to have been suggested by
him. It refers to statements made by Cavendish to Priestley, concerning
which none could bear direct witnesses but themselves; so that it is infi-
nitely more probable that Cavendish dictated to Blagden what he should
write, than that Blagden proposed to Cavendish to make such a state-
ment. There is nothing, moreover, in the least degree suspicious in
Blagden's pen having been employed in engrossing the interpolations.
It was part of his duty, as Cavendish's assistant, to act as his secretary
and amanuensis; to translate papers for him, and to conduct bis corre-
spondence. The note De Luc received, the extracts from Crell's journal
already given, and various papers among the Cavendish MSS. in the
possession of Lord Burlington, fully establish this. It was probably,
therefore, as part of his customary duty, not as a gratuitous interference,
that Blagden copied out the so-called interpolations for the press. It is
neediest, however, to insist on this. What if Blagden did suggest, nay,
urge the interpolations? What harm was there in this, provided
they contained only truthful declarations? The friends of Watt laud
De Luc for his zealous interference in behalf of Watt, although it was
altogether gratuitous and unsolicited; and why is Blagden to be blamed

♦ North Brit. Rev, 1847, p. 492.
t Ibid. ibid. p. 504. % Ibid. ibid. p. 505.

THE WATER CONTROVERSY. 415

if he interfered to defend his friend's reputation and honour? Hard
accusations have been brought against Blagden, and an eye to his current
salary and expected annuity is said to have been the main motive to his
zeal in Cavendish's cause. But even if this invidious charge were trne^
it matters not if^ in the interpolations, Blagden wrote only the truth. I
put therefore aside, as perfectly irrelevant, the question — whether Blagden
"was the author of the interpolations, or only the clerk who penned them f
Cavendish alone is responsible for their contents.

Thus much settled, the far more important question comes before us :
Were interpolations of any kind, still more of such a kind as Cavendish
inserted in his paper, permissible ) The impugners of his claims do not
assert, in so many words, that they were not, but they constantly argue
as if they were unjustifiable, and represent them as improperly sanctioned
by the office-bearers of the Royal Society, and especially by Blagden,
who became one of its secretaries on May 5th, 1784.* It is quite cer-
tain, however, that it was the practice of the Royal Society, in 1784, to
permit authors to alter their papers after they were read. The chief
judge of the propriety of such alterations appears to have been one of the
seoretaries, who was practically the editor of the Society's Transactions.
Into a minute proof of this practice having been permitted, it is not neces-
sary that I should enter. The complaint of Watt's supporters is, that
liberties were conoeded to Cavendish, which would not have been allowed
had any one but his own assistant been the secretary. If, however, I
oan show that equal and still greater liberties were granted to Watt, the
aocusation will lose all its force. This I shall presently seek to dembn-
strate by a reference to Watt's paper; meanwhile requestiug the reader to
take for granted that interpolations were permitted to be made within
certain limits, by all the contributors to the Philosophical Transactions, I
proceed to inquire whether the character of the interpolations ^hich Ca-
vendish introduced, rendered them unfair or inadmissible. Before doing
so, however, it seems desirable to notice, that the Royal Society has been
severely censured for permitting its MSS. to be altered, on any plea, by
their authors before publication. The censure, however, is undeserved,
and the proof of this lies in the fiict, that some of our best conducted
scientific societies still permit similar alterations to be made. And, in
truth, it is no compliment to our freedom from envy and jealousy, that
much more rigid rules now regulate the majority of our scientific bodies
than were customary in the preceding centurv. The present system
secures many advantages, which were forfeited by the looser practice of
a former period, but it sacrifices some of the benefits which flowed from
the ancient rulcf

The former practice, then, of the Royal Society, is not to be sum-

• Weld't Hist. qfR. Society, vol. ii. p. 561.

t I should not wish to be anderatood as disapproving, in the slightest, of the rigid
rules which are now enforced on the contributors to the TYofuactioru of scientific societies'.
The number of contribators is now so immense, that different observers are often
engaged in exactly similar inquiries, and collisions between rival claimants to discoveries
are much mora liable to occur than they were fifty years ago. Every regulation, accord-
ingly, which can lessen the probability of these occurringi is desirable; for though
controversies, when they do arise, are as difficult to settle now-a-days as they ever were,
many are prevented from occurring by the voluntary subjection of the body of scientific
inqmrers to certain fixed rules, which enable the reality and the date of any discovery
to be sharply defined and readily certified. Such rules, however, were much less
necessary in the days of Cavendish and Watt than they are now, and no one complained
of their non-existence.

416 CAVEKDISII AS A CHEMIST.

marily condemned. The only qnestion, however, of real interest, is^
Were the statutes then in vogue administered impartially f That do
unfair liberty was granted, to Cavendish at least, will appear, I think,
from the following considerations. If the present rules of the Boyal
Society had prevailed in his day, he would have been required to place
his interpolations in the shape of notes or postscripts to the oririnal
paper, and to date them. In that case, they would probably have been
marked, as written some time in April or May, 1784. They would hare
suffered nothing, however, by such an arrangement; on the other hand,
the second would have been more distinct than it is at present, for, as it
stands, it contains an anachronism, inasmuch as it forms part of a paper
dated January, and yet refers to another (by Watt) not made public till
the succeeding April. Otherwise, the interpolations gain nothing by
their insertion in the body of the paper, except that they are read con-
tinuously with that part of the text which they are specially intended to
supplement.*

The first interpolation contains a declaration that certain statements
were made to Priestley and Lavoisier, and a brief criticism of their
researches. No one will affirm that Cavendish did Watt, or any one else,
a wrong, or was guilty of any unfairness in offering the criticism; and aa
for the alleged declarations, either they were or they were not made ; and
if they were, as Cavendish, Priestley, Blagden, and in part Lavoisier,
testify was the case, then their announcement contravened no principle
of justice, and implied no unfairness.

The second interpolation is a criticism of Wattes views on elementary
heat, and a disavowal of participation in them, which could not possibly
have been made till after these views were published; and the concluding
addition, which was a true postscript, though not marked as such, con-
tained a jimilar criticism of Lavoisier's Memoirs on Water, which did
not reach this country till some time after Cavendish's paper had been
read. Cavendish's interpolations were thus of a perfectly legitimate and
admissible character, and although the Royal Society had never permitted
another of its members to interpolate his papers, it would have been
blameless for sanctioning the additions which Cavendish made. I might
well, then, spare any JFurther justification of Cavendish on this point,
but other evidences of his fair dealing can easily be furnished, and I
add them, because Blagden, acting under his sanction, has been so
pertinaciously censured for writing these innocent interpolations.

It may be noticed, then, 1st. That had Cavendish had a fraudulent
intention, of passing off any of the additions made to his paper as having
formed from the first, part of the text, nothing would have been more easy
for him than to have re-written a sheet or two of his MS., so as to incor-
porate the addition with it. The accommodating practice of the Society
rendered this quite feasible, for we have seen that Planta, the secretary,
entrusted Cavendish's MS. to De Luc, and we may be certain that he
would have entrusted it to its author.

2nd. Before the paper was printed. Cavendish's supposed accomplice
was secretary, and could have managed the fraud without the chance of
detection, so far as the appearance of the MS. was concerned ; yet we
find that Cavendish was not at the trouble of writing out the interpola-
tions himself, although neither of them occupies a page in print

* This, however, ia a gain. The present practice often separates, by many pages,
tn innocent addition, which would tell with much more force if inserted in (he text.

THE WATfiR CONTROVERSY. 41/

drd. We learn from Mr. Weld^ that tlie first and most obnoxious
interpolation '^appears in a supplementary form, on a smaller sheet of
paper and of a different quality."* Dr. Davy has most justly referred to
this fkct as irreconcileable with the notion that there was any intention
of palming off the interpolation as part of the original text.f

4th. The anachronism in the second interpolation, is incompatible
with the notion that there was a purpose of antedating it. Cavendish
could easily have disavowed faith in the existence of latent or elementary
heaty without mentioning Watt's name, but his reference to him, and to
his "paper lately read before this society/' enabled every one who chose
to consult the date of Watt's paper, which was published in the same
double volume of Transactionay to observe that the interpolation was of
later date than the body of the paper. I do not, however, mean to
assert that Cavendish was consciously guilty of an anachronism. Like
other men, he wrote chiefly for his contemporaries, to whom the dates of his
paper and of Watt's were familiar, and he did not foresee how paradoxical
his statement would appear to a later generation. His is not the only
paper in the PkU. Trans, containing anachronisms. We shall presently
find that Watt's paper is not free froni them; and I have already referred
to their occurrence in a paper by Priestley. (Ante, p. 384.)

I now proceed to Watt's paper. The interpolations in it have
hitherto passed almost, if not altogether, unregarded, and its supposed
freedom from these has been triumphantly contrasted with their alleged
improper existence in Cavendish's memoir. Mr. Muirhead, for example,
formally notices, that the imperfect reference to Cavendish's experiments
contained in Watt's paper, ''was not in the original draft, nor in the
press copy of the letter as sent to Mr. De Luc, but was afterwards added
in pencil;"^ from which his readers cannot but infer, that the rest of
the paper (no part of which is inclosed in brackets in the reprint) was
all written before it was read on April 29th, 17S4. Nay more, the
paper, which is in the form of a letter to De Luc, is dated Nov. 26th,
1783, except certain portions marked with double commas, which are
extracts from the original letter to Priestley of April 26th, 1783, so that
the latest date of any part of the memoir, as it now stands, is Nov. 26th;
yet it appears, that two most important portions of it were not added
till after it was read, and others not till considerably after the nominal
date of November, which the entire essay bears. Had Mr. Muirhead
applied impartially, his principle of including interpolations between
brackets, the reprint of Watt's paper would have exhibited as many as
that of Cavendish does. I do not regret that he has not done so, for the
interpolations in Watt's paper are of as innocent and permissible a cha-
racter as those in Cavendish's essay, and it was very needless, as we have
seen, to parade those in the latter. Since, however, this was done to the
work of the one author, it should have been done to that of the other,
so that no reader might be misted by the partial application of an invi-
dious rule; and the duty was the more incumbent on Mr. Muirhead, that
the Watt Correspondence, which he edited, supplies the means of indicat-
ing the additions or interpolations in Watt's paper. I proceed to notice
what they are.

This, indeed, might be done without breaking through the rules in fashion, if the inter-
polations were enclosed between brackets^ and marked by foot-notes assigning the dates.

• Weld^t Hitt, of Royal Society, vol. it. p. 173.

t Edinr. Phil. Journal, 1849, p. 45.

$ Walt Corr, p. 80.

2 E

41« GAYENDISH AS A CHEMIST.

Neither the letter to Priestley, nor that to De Luc, appeluB to hUTehid
any heading, and had the first letter, which treated as mnch of the convex
sion of water into atmospheric air, as of the conversion of inflammable air
and oxygen into water, been provided with a title, it most have been
considerably different from the one it now bears. This was not f1l^
nished till May, 1784, on the 25th of which we find Blagden, who was
then Secretary to the Royal Society, and had charge of the printing of
the Transactions, proposing to Watt that his letters to Priestley and De
Luc should be incorporated, and adding, " Be so good as send me what
you think the properest title to be inserted before these papers in the
Transactions/'* Watt replies on the 27th, " I am really at a loss what
title to give the paper, but propose the following, * Thoughts (conjectares)
on the constituent parts of Water and of Dephlogisticated Air; with an
account of some experiments on that subject.'*! This late addition of
the title, is a point of as much importance as the date of any of the
additions made to Cavendish's paper; for, as I have shown already, in
discussing Watt's conclusions, Lord Jeffrey, apparently unaware of the
period at which the title was added, has referred to it, as showing the
scope of Watt's opinions in 1783: " Watt's letter," says his Lor&hip,
'^ professed only to embody his own ' Thoughts on the constituent parts
of Water,'"]: whereas we know that it was as much occupied with
thoughts on the conversion of water into atmospheric air. From thd
same letter of May 27th, we learn that the first note to Watt's paper}
was furnished at the same time as the title. This note is also of
importance, as it contains the implicit claim to priority over Cavendish,
which has been so often referred to. Here, then, are two additionsi
insertions, or interpolations, as unwarrantable as those in Cavendish's
paper, and as much calculated to forward the interests of their author as
those of Cavendish were to serve his. That those interpolations were
sanctioned, nay invited (as the title at least was), by Blagden, constitutes
the very gist of my argument; for here we have the alleged enemy of
Watt, and the unfair favourer of Cavendish, suggesting to the former
the best way of incorporating his two letters, whilst, at the same time,
he tells him that '' it is absolutely at your option to decide upon which*
ever of those methods you shall prefer;"|| and invites him to furnish a
title according to his own discretion. The addition of the note was sug-
gested by Watt; and Blagden at once acceded to it, except that he
altered a statement which it contained, that the letter to Priestley had
been in the custody of the President, in which, as I have already shown, it
had not been during the entire period referred to by Watt. These facts are
of themselves sufficient to show that, in 1784, it was customary to permit
members to alter their papers after they were read, and that Watt was
granted this liberty as amply as Cavendish was. Nor is this all. On 4th
April, 1784, Watt writes to De Luc, begging him to alter a phrase in the
former s first letter, T and, on the 10th, De Luc replies, " I have corrected
the phrase in the letter to Priestley."** So that Watt was permitted to
alter, in 1784, a letter which bore date (including, of course, the alteration),

1783, and which was nominally in the custody of Sir Joseph Banks, but
actually in possession of Watt's friend, De Luc. Further, on April 17th,

1784, Watt writes to De Luc, in reference to the letter addressed to him,

• Watt Corr. p. 62.
t Watt Corr. p. 64. J Edinr. Rev. 1848, p. 99.

§ PAU. Tratu. 1784, p. 330. || Wait Corr. p. 62.

^ Watt Corr. p. 60. *♦ aid. ibid. p. 51.

THB WATER CONTROVERSY. 419

*^I bare not beed able to finish the postscript, but have added some
'notes, and have made some alterations on the first and last page of the
letter^ which I conoeived to be necessary in the present circn instances,
and to make it more suitable to the place where it is now to appear."*
And^ in another part of the same epbtle, he writes, " I shall thank you
to forward the new copy of the letter, which I send by to-morrow's coach,
to Sir Joseph Banks, as soon as you have made the necessary aUerations
and addilwns to the copy you have."t On the same day, Watt wrote to
Sir Joseph Banks, informing him of the alterations, " lest it should be said
by an^ body that the letter was fabricated at a later date than it bears."!
In *spite of this information, however. Sir Joseph did not require the
additions to be dated, and they now form integral parts of a paper
marked Nov. 26th, 1783, although they were not furnished till April
ITtbi, 1784. The proceedings, then, of the office-bearers of the Royal
Society were consistent and impartial, so far as the permission of altera-
tions in Cavendish's and Watt's papers was concerned. Watt was
liberally dealt with, and had every request granted, except one, viz. that
the dates of the experiments mentioned in his letter to De Luc should be
inserted upon the margin of his paper.§ He left it to Bla^den to judge
of the propriety of tbL, and Blagdeu omitted them. Watt would have
gained nothing by the mention of them, bo fax as the theory of the com-
position of water was concerned, for the date of the first is 7th May,
1783, that is later than the letter to Priestley, from which Watt was
recoided as claiming priority.

I need not repeat that 1 have not the slightest intention of imputing
to Watt any sinister purpose in introducing those interpolations, I have
referred to them only to show how idle the objections of his friends to
Cavendish's alterations are. They have unconsciously laid hands on a
double-edged weapon, which smites their client as sorely as it does his
riyal. The interpolations of Watt are, in truth, as numerous, as important,
and as objectionable as those of Cavendish, or, if the reader pleases, they
are as justifiable and as innocent Justice, however, requires that, when
a second edition of the Watt Correspondence appears, the interpolations
in Watt's paper should be marked as distinctly as those in Cavendish's.

16. Erroneous Dates in CavendisJCs and Waifs Papers o/1784.

Two most unfortunate errors in date — the one in Cavendish's paper,
and the other in Watt's — were overlooked during their passage through the
press. For the last of these Blagden was responsible, and probably also
tor the first; though this is not quite certain. Both were speedily
detected, and formally corrected; the one by Cavendish, and the other by
Blagden. Notwithstanding this, the less generous impugners of Caven-
dish's claims have extravagantly magnified the importance of these mis-
takes, and have, not obscurely, denounced them as evidences of most
culpable carelessness, if not wilful fraud, on the part of Blagden, who
allowed them to remain, if he did not contrive them, to serve the inte-
rests of his patron Cavendish. Only those who interpret every doubtful
circumstaoce in the history of the discovery of the composition of water
by the uncharitable and utterly untenable hypothesis that Cavendish and
Blagden conspired to rob Watt of the honour of making it, could have
found proofs of false dealing in these typographical errors. They are

♦ Watt Corr, p. 54. t IHdi ihid. p. 55.

t Ibid. ibid. p. 56. § Ilfid. ibid. p. 64.

2 £2

420 CAVENDISH AS A CHEMIST.

thus referred to by Watt himself, in a letter to De Luc of 27tli Jime,
1786 : " It seems odd, but in the detached memoirs of Mr. Cavendidi
and myself, on the composition of water, they should both be wron^
dated, — Mr. Cavendish's dated 'Read January 1783,' when it was p«w
January 1784, and my letter to Dr. Priestley dated April 1784, when it
was written April 1783."* On this passage Mr. Muirhead furnishes tiie
following comment : " This refers to the copies of Mr. Cavendish's
memoir for private circulation, which were circulated by him before the
publication of the seventy-fourth volume of the Transactions for 1784,
having on their title-page this date : * Read at the Royal Society,
January 15th, 1 783.' The date at the head of the paper itself is rightly
given m the Philosophical Transactions, but omitted in those copiesL
It is not the letter to Dr. Priestley, but that to Mr. De Luc, which is
misdated in the Philosophical Transactions; being there dated '26th
November, 1784,' when the real date was 1783." f

From this account it will be seen that Cavendish's paper bears the
proper date— namely, 15th January, 1784 — in the Phil. Trans.y but that
the separate copies were marked 1783. Watt's paper, on the other hand,
was wrongly dated alike in the Transactions, and in the detached copies,
in both of which it is headed "Read April 29th, 1784," and bears in
addition the date " November 26th, 1784." Mr. James Watt, referring
to the latter date, candidly acknowledges that it " is evidently an error
of the press ;"t and be simply remarks, "that another extraordinary error
of the press was committed in the numerous separate copies of his paper
circulated by Mr. Cavendish. "§ With these statements no fisiult can be
found. The errors in date were singularly unfortunate, and naturally
enough appeared to Watt very extraordinary; but his recent supporters
have gone far beyond him and his son in denouncing these errors. Arago
considers them as most formidable. " To complete the imbroglio,^ says
he, " the foremen, the compositors, and printers of the Philosophical
Transactions also took part in it. Some dates in them [Cavendishes and
Watt's papers] were typographically wrong. In the detached copies of
his paper which Cavendish distributed to various learned men, I observe
a mistake of one whole year. By a sad fatality — ^for it is a real misfor*
tune to give rise unintentionally to annoying and unmerited suspicions —
not one of those numerous errors || of the press was favourable to Watt!
God forbid that I should, by these remarks, intend to cast any imputa-
tions on the literary probity of those illustrious philosophers whose names
I have mentioned; they only prove that, on the subject of discoveries,
the strictest justice is all that can be expected from a rival, or a com-
petitor, however high his reputation may already be."^ Sir David
Brewster takes as unfavotirable a view of the wrong dates. " Mr. Watt's
paper," says he, "with that of Cavendish, was printed under the sole
superintendence of Dr. Blagden, who had been appointed Secretary to the
Royal Society on the 5th May; and in a controversy like this, where
charges of various kinds have been reciprocated by the hostile parties, it
deserves to be seriously noted that Mr. Watt's paper is printed with the
erroneous date of 1784, in place o/17SS, and that the separate copies of
Mr. Cavendish's paper have the erroneous date of 1783 in place of 1784.
The obvious effect of these two errors was to give a priority to the labours

* Wait Corr. p. 70. f Ibid, ibid, p. 70. J Ibid. ibid. p. viii.

§ Ibid. ibid.

il The errors referred to here as numeroas^ were only two.

^ Eloffe ofJamea Watt. Mr. Muirhead's translation, Watt Corr, p. 232.

THE WATER CONTROVERSYi 42 1

of Cavendish over those of Watt; and when we consider that the separate
copies of papers are chiefly circulated ahroad hefore the publication ol
the Transactions, and would not fall to produce their impression in
quarters where no correction of the error could be made, we must repro-
bate the negligence of a functionary — if that be a right name for the deed
— who, in the very Grst act of his official duty, maide so great a mistake
in favour of his friend and patron. We shall have occasion again to
glance at this double contingency, but, in the meantime, we cannot but
express our conviction, that in a court of justice it would shake the testi-
mony of the witness who permitted it, and damage the cause of the party
whom it was intended to benefit.'** In another place also Sir David
repeats his condemnation of Blagden's conduct, declaring that, " in the
performance of his principal duty, viz. in superintending the printing
of the FhUosophical Transactions, the new secretary commits, or allows
to be committed, two gross errors of date, both of which are favourable
to his patron, and unfavourable to Mr. Watt."f Mr. Muirhead, who
always takes the most unfavourable view of Cavendish's and Blagden's
proceedings, affirms that ''it was at least a piece of most singular
negligence, on the part of the secretary to the Royal Society, who
superintended the printing, that those papers should have been circulated
with a double error in their dates; that the tendency, if not the effi^ct, of
both the errors should have been to take the priority from Watt and to
give it to Cavendish; and that, of all the errors which the printer might
have committed, he should have happened to select precisely those which
were best fitted to effect that object.''^

It is refreshing to turn from these uncharitable surmises, to the gene*

rons judgment which the most successful of all the defenders of Watt,

Lord Jeffrey, passes upon both Cavendish and Blagden. '' The higher

elements," says his lordship, ''of our nature are not so discordantly

blended within us, as that the love of honourable fame should lead to the

disregard of truth and honesty. But Cavendish was almost as remarkable

for his indifference to fame, as for the high principle and honour that

belonged to his station and his character; and it would have been strange

indeed, if, for the sake of adding one more to his many intellectual

triumphs, he had stooped, by a deliberate falsehood, to the very lowest

depths of moral degradation. Nor have we any reason to think that his

friend Blagden, who had not even the temptation of a rival claim to

mislead him, would have stooped to such a baseness." § There are few

impartial persons who will find in the erroneous dates anything at variance

with Lord Jeffrey's estimate of the character of Cavendish and his assistant.

^j those who, with myself, believe that Cavendish owed nothing to Watt,

the notion that he could have been a party to the falsification of dates

will be at once discarded as utterly incredible ; and as for Blagden, he

must have been a much less able and accomplished person than he is

universally acknowledged to have been, if he could have expected to serve

his patron by so clumsy and transparent a device as the alteration of a

date in the unauthoritative copies of his paper. In reality, however,

there is no proof that Blagden was responsible for the wrong date. It

formed no part of his official duty as secretary to superintend the

printing of the detached copies of Cavendish's paper. It lay between the

author and the printer to arrange concerning these; and although it

is probable that Blagden, in his capacity of assistant, took charge of them,

Cavendish was as responsible for the contents of the papers as he was,

• North Bnt, Rev, Feb. 1847, p. 493. f Ibid. Ibid, p. 504.

I Watt Corr. p. Ixvi. ( Edinr, Rev. 1848, p. 88.

422 . CAVENDISH AS A CHEMIST.

How the error arose, it is perhaps impossible now to diseoTer; bit
probably an alteration took place in the paging of the TVansactionSy as is
customary at the present day, before or aiter the separate copies were
printed, and dnring the alteration of the types one figure was substituted
for another. I can say nothing positive on this point, however, as I
have not seen any of the detached papers, or any special account of them.
It is not so very rare a thing, nevertheless, for an error to creep into a
revise, that we need wonder very much at it That it was accidental,
we may confidently affirm, for the following reasons : 1. The wrong date
neither did, nor could alter in any respect the relative priority of Caven-
dish, Watt, and Lavoisier. If a passa^ referring solely to the first had
been antedated, whilst the remainder of the paper retained the later date,
there might be some ground for suspicion of fraud. But how does the
case stand ) Cavendish's paper, as it appears in the TranmictioTUi, con-
tains two references to priority; the one is the account of the revelations
made to Priestley concerning the researches of 1781, of which a repeti-
tion is declared to be recorded by the latter '' in the preceding volume of
the TransactionB.^^ The other reference to priority states that an account
of Cavendishes theory was given to Lavoisier '' last summer." A third
passage notices certain opinions of Watt announced by him '' in a paper
lately read before this Society." These references also occur in the
detached copies, so that the relative priority of the three daimants is
represented in the same way in both issues of the paper, with only this
exception, that they are dated backwards from 1784 to 1783 in the
detached copies \ so that if Cavendish antedated his own researches by a
whole year, he also antedated by the same period the revelation to
Lavoisier, and the reading of Watfs paper. Everything, in fiict, was
shifted back twelve months, but this left the relative claims of all parties
exactly as they were before the shift took place. It cannot be imagined
that Cavendish or filagden was so stupid as to expect to gain any advan-
tage by such a useless device as this.

2. Cavendishes contemporaries, whom the separate copies are supposed
to have been Erpecially intended to deceive, were furnished with direct
means of detecting the erroneous date by the references to '' the preceding
volume of the Transactions," and to the "last summer," both of which
allusions applied to 1783, and were therefore incompatible with the paper
containing them bearing the date January, 1 783. And in truth, there is no
reason to imagine that Watt's reputation suffered from the error in date.
It is true that Cuvier, in one essay, gives the date of the reading of
Cavendish's paper as January, 1783, and the friends of Watt make much
of the mistake.* But even here Cuvier has not contrasted Cavendish
with Watt, and is not asserting anything to the disadvantage of the latter
on the authority of the false date; and in his eloge of Cavendishf he gives
the date accurately, ^'le 1 4 de Janvier, 1784." And we may set off,
against what little wrong was done to Watt by Cuvier in his first state-
ment, published " at the distance of four-and-twenty years from the cir-
culation of the erroneous date," when it could do no mischief, the corre-
sponding wrong to Cavendish occasioned by Lord Jeffrey's reference to
the title of Watt's '' Thoughts" as dating from 1783, and consider the two
philosophers as quits.

3. A further proof of the accidental nature of the error in date, will
be furnished to all impartial parties by the consideration, that the utmost
Blagden or Cavendish, or both could hope to achieve by it, if it were

• Rapport HUtorique, p. h7. Quoted by Mr. Muirhead, Watt Corr. p. \xv.

t Bloffe»Hi8toriqueSftome ii. p. 87. •

THE WAtER CONTROVERSY. 423

^i^ilful, waa tbe misleading of those to whom the separate copies were sent,
for a few weeks or months, till the PkiL Trans, were published. And
the onlj parties whom it could have been of any moment to deceive in
this way, viz. the friends of Watt and those of Layoisier, could not
poseiblj have been misled bj it, for the former had had access through
l>e Luc to Cavendish's MS., and the latter knew from Blagden that
Cavendish had read no paper on the composition of water to the Royal
Society in 1783. Whom, then, could the wrong date be intended to lead
astray ? The answer of all unprejudiced persons will be, I think, No one.
A detached copy of a memoir published in the transactions of a society,
is of authority only in so far as it is identical with the text of the Trans-
actiwis, of which it professes to be an isolated portion, so that where there
is any difference between the two, as there is here in a particular date,
tlie Transactions only are authoritative.

4. But all this reasoning might have been spared. Cavendish de-
tected the error, and immediately furnished a correction of it in the form
of a letter to the editor of the Journal de Physique, which I quote in full.

Letter op Cavendish to Mongez.

A Londres, cc 22 Fevrier, 1785.
En lisant, Monsieur, la traduction de mon m6moire sur Tair, publie dans
le Journal de Physique, je fus frappe de le voir datte de Janvier, '83,
comme si la lecture en eut et6 faite ators, devant la Societe Royale. J'eus
reconre aux exemplaires detaches imprimes pour Tusage de mes amis sur
Tun desquels apparemment avoit ete £a.ite votre traduction ; je trouvai k
mon grand ^tonnement que Timprimeur avoit fait cette mSme faute dans
toutes les copies, malgre que Toriginal publie dans les TraTisactions Philo^
sophiques avoit et6 datte, comme il devoit T^tre, de Janvier, '84. Je vous
serai tres oblige, Monsieur, de vouloir bien faire mention de cette meprise
dans le cahier prochain de votre Journal.

Je suis mortifie d'etre dans le cas d'ajoater qu'il s'en faut de beaucoup
uo la traduction soit exacte; on a manque le sens en plusieurs endroits.
'ai Vhonneur d'etre, avec des sentiments distingues,

Monsieur,
Votre tr^s humble et trds obeiss^ scrviteur.
A Monsieur T. A. Mongez, le Jeune, &c. &c. &c.
Au Bureau du Journal de Physique a Paris.*

* Brit. AsBOc, Rep. 1839, pp. 65, 66. I have alluded.to this letter in the personal
narrative, as if it had certainly been published by Mongez, in conformity with the impres-
lion conveyed by Mr. Harcourt's references to the letter (^Brit. Anoc. Report^ 1839^
pp. 41 and 65.) Whilst, however, this sheet is passing through the press, I have discovered
|hat Mongez did not publish Cavendish's letter in his Journal. There is not indeed legal
proof that the letter was sent to Mongez. A draft of it remains among the Cavendish
MSS., from which Mr. Harcourt has printed it. That the letter, however, was actually
despatched to Mongez, and was received by him, admits of every confirmation short of
absolute proof. Cavendish's first series of" Experiments on Air" was published with
the wrong date (1783) in the Journal de Phyeique for December, 1784, p. 417, and
January, 1785, p. 38. The second series of '' Experiments on Air" was published in the
part of the same journal for August 1785, p. 107. To this page a note is attached by
the Redacieun of the Journal de Physique, in which they acknowledge that they have
nodved from Cayendish a more accurate translation into French of his first " Experi-
ments on Air" than the one they had published. For reasons assigned the^ decline to
insert this,* and with Editorial reluctance to confess error, they do not pomt out the
mistake in date, to which Cavendish had drawn their attention in the same letter (tn<f«
eupra)y in which he had complained of their inaccurate translation. A fatality seems to
have attended the dates of Cayendish's papers in 1784, for Kirwan's Remarks on his
" Experiments on Air," and Cavendish's reply to Kirwan, are misdated in the Journal
de Phytique, t. xxvi. pp. 414-425.

424 CAVENDISH AS A CHEMIST,

On this letter Mr. Muirhead remarks "that in one iastaaoey more tiiaa
a year afterwards, (when the error had already been propagated in moei
of the scientific journals of the continent, and when also the Fhilosopkical
TranmdioiUj with the true date of the reading of the paper, had come into
circulation) Mr. Cavendish desired that it might be correctcni."* This
ungenerous comment scarcely requires notice. In how many quartets
Cavendish corrected the error, we do not know ; but it savours of the
ludicrous to suggest, that it was incumbent on him to send letters roiud
Europe, announcing that a solitary typographical error existed in the
private issue of his paper. At the present day, when journals toe so
much more numerous than they were in the days of Cavendish, a writer
who detected an error in the detached copies of a paper from the Trans-
actions of a society, would bo considered to have done enough, if he wrote
to a single influential journal pointing out the mistake; Cavendish,
however, did more. The letter to Mongez, if it had been published, wonld
have informed the French Philosophers of the error. Crell was fnmished
with information which secured the German chemists from mistake, and
in England no correction was needed.

The erroneous date in Watt*s published paper admits, if possible still
more fully than that in Cavendish's, of being proved to have been unde-
signed and accidental. Blagden, certainly, is responsible for it, as Watt
declined to have the proof sheets sent to him to Birmingham, and ex-
pressed his confidence that Blagden would properly incorporate the letters
to Priestley and De Luc, and make the other alterations which were
desired.f Blagden, accordingly, would deserve the severest reprehension,
if it were in any degree probable that he had betrayed the trust Watt
placed in him. He must, moreover, have been a very bold falsifier, if he
could have altered a date, which was to come under the notice of Watt
(who had commissioned fifty separate copies of his paper) soon after the
error was committed, and probably before the volume of the Fraji^actums
in which it appeared was published. The erroneous date, however, bore
error upon its very^face. It might have been overlooked by one who
consulted the paper solely with a view to study its contents, but could
not have been passed over by one who referred to it in order to settle
the chronology of Watt's writings. The paper it will be remembered,
bears two dates, and these as they appear in the Fhil, Trans, are incom*
patible with each other, and thereby betray that one, at least, is erroneous.
The title is as follows :—

*^ Thoughts on the Constituent Parts of Water and of Dephloffisiicated Air;
with an Account of some Experiments on that subject. In a Letter
from Mr. James Wait, Engineer, to Mr. De Luc, F.H.S. Bead
April 29, 1784.

Birmingham, November 26, 1784.
Dear Sir,

In compliance, tkc. <l&c." *

Any one who consulted this heading, with a view to fix the period
when the contents of the paper were written, could not fail to observe
that the dates were irreconcilable, for they represent a letter which was
publicly read in April, 1784, as not having been written till the succeeding
November, that is six months after it was read. A reader who observed
this, would be at a loss to determine which of the dates was wrong ; that
of the private writing of the paper, or that of its public reading. But if

• Wait Corr. p. Uiv. + Ibid. ibid. p. 68.

THE WATER CONTROVERSY. 425

he turned to the errata at the close of the volume^ (which none of the
defenders of Watt or impngners of Blagden, seem to have thou/;ht of
doing), he would find the following correction, '/ Vol. LXXI V. part ii.,
fpage] 329 [line] 7, for 1784, read 1783;" from which he would at once
learn, that the letter was written in November, 1783, and read in April,
1784. Upon this point I might spare any further comment. An error
"vrhich betrays itself, and which was corrected bj him who originally over-
looked it in the same volume in which it occurs, will not be regarded by
any candid inquirer as an artful blunder to serve a sinister purpose.

It is easy, however, to pnsh the defence of Blagden further. Watt,
it will be remembered, sent a note to him to be added to his paper,
referring to the letter to Priestley, but left a blank which Blagden should
fill up. The passage in Watt's letter was as follows, '^ The letter [Priest-

ley's] therefore remained in the custody of the President until ;

"when at the author's request it was read before the society."* The first
clause of this sentence Blagden altered to 'Hhe letter, therefore, was
reserved until the 22nd of April lcut,"f Now the note containing this
passage forms an integral part of a document dated (as corrected m the
erratum) November, 1783, so that the ''April last,*' counting from this,
would bo April, 1783, and thus the passage represents the letter to Priest-
ley, as read to the Royal Society a year before the period when it actually
was conmiunicated to it, and many months before Cavendish's paper was
read.

There can be no doubt that Blagden intended the '' April last," to date
from the period of the reading, not from that of the writing of the letter
to De Luc. The date of the reading was April 29th, 1784, and the letter
to Priestley was read on the 22nd of the same month and year. There is
nothing, however, in the note to point this out. Bat for the Watt Cor-
respondence, every reader would imagine Watt's paper to have been
written in November, 1783, so that the April preceding it was of neces-
sity the April of the same year» 1 have pointed this out to show the
exaggeration of the statements already quoted, in which the advocates
of Watt declare, that of all the typographical errors that could have been
made, those only were suflered, or committed, which gave Cavendish an
advantage over Watt; whereas Blagden thus appears to bave filled in a
blank in Watt's note, in such a manner as to represent the public reading
of the first version of his paper, as having occurred many montbs before
the reading of Cavendish's. This may suffice as a reply to the declaration,
that none of Blagden's errors were favourable to Watt.

The worst, then, that can be charged against Blagden, is negligence
In the discharge of his duties as Secretary to the Royal Society, nor am I
required to defend him against this accusation ; yet I will say a word in
defence of one who has been so unjustly and severely blamed, because he
overlooked typographical errors. It should be remembered by those who
wish to do him justice, that he was new to the duties of the secretaryship
when the mistakes we have been considering occurred, and that he ofiered
to send the proofs of his paper to Watt, which the latter declined. It should
further be noticed, that Watt's paper was not a continuous single document
which could at once be put in the hands of the printers, but consisted of
at least three papers ; the first. Watt's letter to Priestley; the second, his
letter to De Luc ; the third, his letter to Blagden, containing the title and
the note claiming priority. There were thus three documents which had

* Wali Corr. p. 64. t PhU, TVans. 1784, p. 330.

426 CAVBNDISH AS A CHEMIST.

in wliole or in part io be doretailed together, 00 as to oonfititate a
tinnous memoiri and the risk of error was mach greater than if the whole
had been written out by Watt^ and despatched to Blagden ready for tiie
press. It is probable, moreover, that Blagden pnt into the printers'
hands the letter to De Luc with the proper date affixed to it, as be re-
oeiyed it from the latter, and that they were guilty of the blunder aboot
which so much has been said; but whether or not this was the case, it
does not demand a great stretch of charity towards Blagden to suppoee
that the error, which he afterwards corrected, was an oversight. Walt
liimself, was so confused with the many alterations which he had made
on his paper, that he blundered still more grossly than Blagden, in re-
ference to the error in date, which he stated occurred in the letter to
Priestley, instead of in the letter to De Luc*

And as if to show how charitable authors should be in judging typo-
^praphical errors, and what need there is for caution in imputing motives
to those who have suffered misprints, Mr. Muirhead, who judges Blagden
80 harshly, has imitated him in introducing an error into his edition of
Watt*s paper. I have already referred to Mr. Muirhead*s care as an
editor, in the publication of the Watt Correspondence, and the error I
have to notice I fell upon accidentally, for I have not busied myself
seeking for mistakes. It occurs in the note already referred to, as altered
by Blagden, containing the explanation of the reason for withdrawing the
letter to Priestley. In the FhiL Trans, it runs thus: '^The author^
having heard of Dr. Priestley^s new experiments, begged that the reading
might be delayed."! Whereas according to Mr. Muirhead's reprint,
Watt " begged that the meeting might be delayed.*'! What would Mr.
Muirhead say, if some of the extreme supporters of Cavendish were to
insinuate that there was something veiy suspicious in this innocent
misprint, and that he had a sinister purpose of serving his client by
representing him as delaying only the meeting, so quickly did he expect
to be able to obviate the difficulties which Priestley's new experiments
threw in the way of his theory ? Mr. Muirhead would justly be indig-
nant at a charge so unfounded as this, yet it is as well founded as that
preferred against Blagden. If Mr. Muirhead, guided by a standard of
editorial accuracy much more precise than any author set before him in
Blagden's days, with abundant leisure to execute his task, and printed
papers to copy from, could not, with all his solicitude to avoia error,
escape making a blunder, we may well be lenient towards the new Secre-
tary of 1784, who, in the days of careless editing, had to construct a
continuous paper out of three unconnected documents; and we may further
assign to him a merit which we cannot impute to Mr. Muirhead, viz. that
he at least detected his error, and pointed it out in an erratum. There
are few literary or scientific men who have not had the mortification of
discovering, when it was too late, some overlooked error in their printed
productions; and I know not who among them can venture to cast the
first stone at the erring Blagden. I, at least, will request that my typo-
graphical errors may be leniently judged.

17. Alleged Plagiarism of Lavoisier .

Four French philosophers, La Place, Lavoisier, Meusnier, and Monge,

are connected with the discovery of the composition of water^ but only

Lavoisier has been excepted against by his English rivals. Monge, as

* Wait Corr. p. 70. t Phil. Trans. 1784, p. 330.

t Watt Corr, p. 78.

•THE WATER CONTROVERSY. 42ir

hare already Been, disclaimed priority to Cavendisli, and did not even
establish a right to be considered an independent discoverer of the true
natnre of water^ so that I shall saj nothing concerning him in this section;
bis integrity having been conspicuous and never called in question. La
Place also^ is without reproach, so that he will be passed over without
notice; although in another section, where I shall have the more pleasant
duty of contrastiug the merits of the rivals in the Water Controversy, he
'vrilf receive honourable mention. Meusnier, likewise, appears to have
been simply an experimental assistant or colleague of Lavoisier, and need
not be specially referred to, so that the last named chemist alone will
come before us as accused of plagiarism. I should have been glad to
bave been spared the consideration of this charge, encouraging as I do the
bope that among the unpublished papers of Lavoisier, which Dumas is
understood to be editing, will appear something fitted to lessen the force
of the accusations which have so long remained unanswered and unan-
swerable. But it would be affectation to conceal that at present, so far
as published documents are concerned, Lavoisier stands forth as a detected
plagiarist. His distinguished countrymen, Arago and Dumas, have in
the meanwhile abandoned his defence ; for Arago styles him in his 61oge
of Watt " a pretender" to the discovery of the composition of water,*
and Dumas <' adopts completely and in all its parts, the history which
M. Arago has written of the discovery of the composition of water."f
Thus much, however, is said in the Eloge in vindication of Lavoisier, that
with the exception of Watt, all those whose names figure in Arago*s
narrative were to a greater or less degree guilty of falsehood, at least to
the extent of concealment of truth. { I am slow, however, to credit that
Arago and Dumas can sanction so hateful a doctrine, as that disregard of
truth by one man can in the slightest degree justify its disregard by
another; but if they do, I need only say that the multiplication of
Lavoisier's fellow-transgressors will not lessen the guilt of his transgres-
sion. Though it were true indeed, which it is not, that Lavoisier's rivals
as well as himself, were plagiarists, he could not, on the lowest ground,
plead justification on that score. None of them borrowed anything from
him, so that he could not say that he only committed reprisals; nor did
he know anything concerning their alleged obligations to each other
when he claimed the disputed discovery, so that he could not even plead
that he had only followed a bad example. He must stand therefore
alone, and be judged as to his fair dealing, without any reference to the
shoH-comings or delinquencies of others; and I fear that he stands self-
condemned. It appears from the Watt Correspondence§ that he found
in private a zealous defender in De Luc, who tried hard to convince
Watt that though Cavendish had shamefully wronged him, Lavoisier was
altogether innocent, and owed nothing to the account which had been
given him of the English experiments. Watt, however, was deaf to all
De Luc's protestations. " You see," says he, in a letter to the latter,
''that it is possible for a philosopher to be disingenuous. For Mr.
Lavoisier haa heard of my theory before he formed his, or before he tried
the experiment of burning dephlogisticated and inflammable air together,
and saw the product was water."|| De Luc laboured hard in reply to
shake this opinion;ir but Watt returned him for answer, '^ I must still

* Mr. Muirhead's Translation, Wait Corr. p. 228.

t Compies Rendus de VAcad, de$ Scieneet, 20 Janvier, 1840, p. IIL

t Eloge of Watt, Mr. Muirhead's Translation, Watt Corr. p. 230.

I P. 41. ' II Watt Corr. p. 40. f Ibid, ibid, p. 41.

428 CAVENDISH AS A CHEMIST.

differ from you in regard to Mr. Laroisicr's knowledge of my ibeoiy before
lie even made bis experiments.'**

Watt was justified in his suspicions of Lavoisier. Reference lus
already frequently been made to the visit which Blagden paid to Paris in
June, 1783, and to the account which he gave to Lavoisier of Caven-
dish's experiments, and the conclusions drawn from them. Lavoisier, as
we have also seen, acknowledged that a communication had been made
to him by Blagden, but represented it as much more limited than the
latter asserted it was. The problem, therefore, of Lavoisier's innocence
turns chiefly upon the question. Is his account of matters, or Blagden*^
the more credible ? The difference, it will be remembered, lay chiefly in
this, that Lavoisier represented Blagden as having only told him that
water could be obtained by the combustion of inflammable air in close
vessels (ante, p. 336); whilst Blagden asserted that he had further
informed Lavoisier, that " the water was equal to the weight of the two
aira added together;" that he had likewise made known to him the con-
clusion drawn both by Cavendish and Watt as to the composition of
water; and that both Lavoisier and his friends were incredulous as to the
equality of weights, and as to the validity of the conclusion. Sir David
Brewster is the only party, so &r as 1 am aware, who has called in
question the accuracy of Blagden's statement. He seeks to convict him
of a failure of memory in 1786, when he impeached Lavoisier*s veracity
in his letter to Crell, as to the statements which he had made to the
French chemists in 1783. " The assertion of Lavoisier," says Sir David,
" that Blagden mentioned to him only the experiments of Cavendish, and
the fact that, in the account ^iven of the French experiment of the
Academy of Sciences on the 25th June, Lavoisier states that the conclu-
sion as to the compound nature of water was drawn by La Place and
himself, may be fairly held as a proof that Dr. Blagden had forgotten in
1786 the extent of the communication which he made to the French
chemists in 1783, and may have made a second mistake also in his state-
ment that Cavendish communicated to him, and his particular friends in
the Royal Society, in the spring of 1783, the conclusions which he drew
from his experiments ;"f and again: ''Are we not entitled to suppose
that in his [Blagden's] mind, the year 1783 took the^ place of 1784, and
that the communication of his conclusions, alleged to have been made by
Mr. Cavendish in 1783, were actually made in the beginning of 1784,
just before his piiper was read to theKoyal Society, and that he did not
communicate these conclusions to the Academicians in 1783, because
they had not then been communicated to himself. This seems to be the
only supposition upon which we can reconcile the declarations of Lavoisier
and La Place with the declarations of Dr. Blagden : and it relieves both
parties from the mutual recrimination of their mends that neither of them
had told the whole truth."J

Sir David's objections would be very serious if they could be main-
tained, but they are so completely answered by Lord Jeffrey, that I need
only quote his observations on the subject. "As it has been surmised,
iQ palliation of the disingenuousness which it appears to impute to most
ominent and meritorious individuals, that Blagden, when thus writing at
inf ^*®^°^ ®^ ^^^ years, may have misrecoUected the extent of the
fort *^^^^^^^° he had given verbally to the Parisians so long before, it is

u«»ate that we can now show, from documents recently brought to

•*. j^ ,^ * TTfl/r Corr. p. 42.

-'vorM Brit, Rev, Feb. 1847, p. 487. J Uid. ibid, p. 488.

THE WATEH CONTROVEKSY. 429

ligbtj that be had openly given the same account vmMediaUly on hia
return to this country; and desired it to be communicated on his autho-
rity, to those whom it most concerned. In a letter, accordingly, addressed
to Watt by Kirwan, in December, 1783, immediately after the former had
first heard of the French experiments, and expressed his suspicions of
their origin, he says, ^^ Lavoisier certainly learned your theory (not expe-
riments merely, but theory) from Dr. Blagden^ who first had it from Mr.
Cavendish; and afterwards from your letter to Dr. Priestley, which he
heard read; and explained the whole mimUely to Lavoisier , last July
(mistake for June). This he anthorised me to tell you,^** And Cavendish
himself, in his paper read in January, 1784, put openly on the record of
the Kqyal Society, that '' during the last summer a friend of mine gave
an account of these experiments to M. Lavoisier, as well as of the conclvr
sion drawn fromi them, that dephlogisticated air is only water deprived of
its phlogiston; but at that time M. Lavoisier was so far from thinking any
8uch opinion warranted, that till he was prevailed upon to repeat the expe-
riment^ he had difficulty in believing that nearly the whole of the two
airs could be converted into water."

"These we think/* continues Lord Jeffrey, "were public enough
challenges to the advocates for the French discovery; and we are yet to
learn that any champion ever appeared to take them up.*'t

It appears that none did. Yet it is incredible that the French
chemists can have been ignorant of the charges preferred against Lavoisier.
Cavendish's paper was translated in CrelFs Chemische Annalen, and was
thus made accessible to readers of German. It appeared also in Mongez'
Journal de Physique, which was published in Paris; and it is not at all
improbable that De Luc, who corresponded with La Place about
Lavoisier s synthetical experiments on water, gave the former some hint
as to Watt's estimate of the originality and fair dealing of Lavoisier,
Blagden's letter to Crell also, was very unhesitating and minute in its
statements, and implicated, to a certam extent, other members of the
French Academy besides Lavoisier. One of these was La Place, and
among others were in all probability MM. Le Roi and De Yandermonde,
who, Lavoisier states, were witnesses along with several other Acadc«
micians^ of his repetition of Cavendish's experiments. Berthollet also
took great interest in the experiments made in England and in France on
the production of water, and corresponded with BWden on the subject.
A letter which passed between them is ^iven elsewhere. At all events
some half-a-dozen members of the Academy heard the account which
Blagden gave Lavoisier " of theso new experiments and of the opinions
founded upon them;" and Blagden taxes them all with having been
sceptical, both as to the experiments and the conclusions drawn from
them. " They did not doubt," says Blagden, in his letter to Crell, '^ that
in such manner a considerable quantity of water might be obtained; but
they felt convinced that it did not come near to the weight of the two
species of air employed; on which account it was not tooo regarded as
water formed or produced out of the two kinds of air, but was already
contained in and united with the airs, and deposited in their combustion.
This opinion was held by Mr. Lavoisier, as well as by the rest of the gentle-
men who conferred on the subject; but as the experiment itself appeared to
them very remarbable in all points of view, they unanimously requested

* Wait Corr, p. 39.
t Edinr. Ilev. Jan. 1848, pp. 70, 71. t ^atl Corr. p. 170.

430 CAVENDISH AS A" CHEMIST.

Mr. Layoisiery who possessed all the necessary preparatioils, to repeat the
experiment on a somewhat larger scale, as early as possible. This desire
he complied with on the 24th JonOi 1783, (as he relates in the latest
Tolume of the Paris memoirs)/ *

It mast be considered in the highest degree improbable that BIagdea*8
letter shoald have escaped the attention of everj one of the Academiciami
referred to in it ; yet if they were acquainted with it, we may be oertaia
that they would haye called Bktfden to account, had he imputed to them
opinions which they did not hold! Against Lavoisier himself, moreover,
Blagden reiterated with great pertinacity a formidable list of charges, which
I abridge from the letter to Crell. He affirmed — Ist. That at the period
of his communication to him, Lavoisier had not forme<l the opinion that
water was composed of hydrogen and oxygen, but believed that an acid
would result from their union. 2nd. That he was informed of Cavendish's
experiments same days before he made his own ; not on the very day that ho
tried them. Srd. That his experiments were made in consequence of what
Blagden had told him. 4th. That he was not told that Mr. Cavendish
had obtaiued ^'une quantite d'eau tres sensible,*' but that the water waa
equal to the weiffht of the two airs added together. 5th. That he was
made acquainted with Messrs. Cavendish and Watt's condusions. 6tfa.
That he was not led to the alleged discovery, as he represented he was,
by following up bis own experiments commenced in 1777, bat waa in-
duced to make the experiments which he reported to the Academy, solely
by the account which he received from Blagden of the English experi-
ments. And the indictment is wound up by the unequivocal declaration
that Lavoisier discovered nothing but what had before been pointed oot
to him, to have been previously made out and demonstrated in England, f

That such an impeachment of Lavoisier's originality, veracity, and
ffood faith, would have been allowed to pass unnoticed, if it could have
been answered, is incredible. It may be suggested, that it did not attract
the attention of those whom it concerned at the period of its publication ;
but this is exceedingly improbable. Lavoisier survived its appearance
some eight years, and many of his fellow-academicians. La Place among
the rest, loug outlived him, so that it must be considered extremely un-
likely that the contents of Crell's letter to Blagden escaped the attention
of the party accused, and of all bis friends and contemporaries. The
force of this argument is increased by the consideration that Blagden,
who often visited Paris, and was in constant correspondence with the
philosophers there, is very likely to have put his letter in the way of
those whom it concerned, or to have drawn attention to it. The first
reference to it in the Water Controversy was made by Arago, so that,
sooner or later it came to be known in France. It is not necessary, how-
ever, to insist upon Lavoisier's silence as implying inability to answer
any or all of Blagden's statements. We have no right to affirm this, and
I do not wish even to assert that Blagden was justified in all he said,
although I feel assured that he thought he was. I have pointed out in
another place (ante, p. 345) that he was wrong in affirming that Monge's
experiments were certainly of later date than Lavoisier's, and were con-
fessedly repetitions of them. It must, however, be considered npon any
view very inexplicable, that no notice should have been taken in France
of Blagden's letter. The only point, nevertheless, of very great import*

* Mr. Muirhead's Translation, Watt Corr, p. 72.

t Mr. Muirhead's Translation, Watt Corr. pp. 72 — 74.

THK WATER CONTROVERSY; 431

ance i^ Were Blagden's accusations subdtantially true; and that they
were, will, I think, appear from the following considerations. Lavoi*
sier acknowledges that tome account of Cavendish's experiments was
given him by Dlagden, and La Place bears independent testimony to the
fact. It is quite certain, moreover, that all that Blagden professes to
hare told the French chemists, he was in a condition to tell them, and no
reason can be assigned why he should have concealed anything when he
spontaneously made a communication eonceming the experiments and
conclnsiona of Cavendish and Watt. On the other hand, everything
favours the belief that having once entered upon the subject, he would
claim for one or other, or both of the English chenusts, all that he could
claim for them.

It is quite certain, however, that before June, 1783, Cavendish had not
only ascertained that when inflammable air and oxygen are burned together
in proper proportions, they yield their own weight of water, but that
Priestley had also publicly declared that he had confirmed this result.

It is also (Certain that both Cavendish and Watt, whether indepen-
dently or not, had inferred from the experiments of the former, and their
repetition by Priestley, that water was composed of inflammable air and
oxygen ; and that Blagden knew that they had drawn these conclusions.
He is fully entitled to credit therefore, when he declares that he an-^
noonced aO this to Lavoisier and his colleagues. But if he did, Lavoisier
cannot be acquitted of disingenuousness and plagiarism. Again ; the
only party, except Blagden and Lavoisier, who nas given any account of
the conference between them is La Place. And his account does not
tally with Lavoisier's, for he tells De Luc that he and Lavoisier had been
occupied in repeating before Blagden and several other persons, " Mr.
Cavendish's experiment" on the conversion of inflammable and dephlo-
gisticated air into water by their combustion.* Such an experiment
was necessarily quantitative, and La Place, accordingly, refers, as if
it were a matter of course, to the question of equality of weights be-
tween the burned gases and the produced water; and although he
writes three days after Lavoisier had announced to the Academy a
conclusion which was not justifiable unless the weights were equal,
he declares that neither he nor Lavoisier yet knew whether they were*
His statement, accordingly, so far as it goes, is in accordance with Blag-
den's, and at variance with Lavoisier's account of matters. Further;
Lavoisier acknowledges that his original expectation was, that inflammable
air in burning would yield sulphuric or sulphurous acid;t and he tried, on
two occasions, to demonstrate the production of an acid, whilst he does
not profess to have had the slightest expectation that water would be
generated. Blagden's declaration, therefore, that he and his colleagues
were incredulous as to the production of water, is highly credible, and
quite in keeping with Lavoisier's own statements. It seems probable
also, although it is impossible to be certain on the point, that before
Blagden communicated with him, he had arranged an apparatus resembling
that for the oxyhydrogen blowpipe, with a view to t^ on a large scale
what was the product of the oxidation of hydrogen; and although Blagden
had never visited Paris, Lavoisier would certainly have sooner or ukter
employed this apparatus, and might have made the discovery which he
missed. As it was, however, before he had put it in action, he was

♦ Watt Corr. p. 41.

t Mimoire par Lavoisier, Watt Corr. p. 173; or Mim, de VAcad, paw 1781,
p. 473.

432 CAVBNDISH AS A CHBMIST.

informed that tlie problem which interested him was already solved. He
seems to have thought that his pre-arrangements jastified him in proceeding
as if he had not been anticipated, and entitled htm to publish his conda-
sion as if it were new to the world. A conscious rediscovery, however,
is one of the strangest contradictions in terms. Finally; Blagden was
on very friendly terms with the scientific men in Paris, which he often
visited. When he appealed, therefore, to Lavoisier's fellow-academicians,
he appealed to those with whom he would soon be confronted, and by
whom he would be called to account for any misrepresentation be should
give of their views. He was, moreover, a cautious and somewhat form^
person, who had no interest in ofiending the French philosophers : yet he
risked their good-will by the version which he gave of his conference
with Lavoisier. Nor was it at all necessary for Cavendishes vindication,
that he should have entered into so minute an account of his interviews
with the French philosophers as he did. It may be added, as pointed
out already in the personal narrative, that Blagden did not forfeit the
good-will of the French philosophers by his letter to Crell. ' The terms of
his will sufficiently demonstrate this. When all these things are consi-
dered, the credibility of Blagden's uncontradicted accusation of Lavoisier
must be considered as greatly enhanced. It seems impossible, therefore,
to acquit Lavoisier of the charge preferred against him alike by Cavendish,
Blagden, and Watt, which, in the cutting words of the last, amounted to
this, that after Lavoisier had had the theory of the composition of water
explained to him, " he invented it himself."*

GENERAL SUMMARY.

18. Relative Merits of Cavendish, Watt, and Lavoisier,

I have stated frequently throughout the long discussion which is now
brought to a close, that to each of the three great claimants of the dis-
covery of the composition of water, a large though unequal share of merit
must be assigned. I proceed now, accordingly, to state in what propor-
tion each appears to me entitled to honour. Merit, however, is not a
thing which in any case admits of precise definition or accurate estima-
tion, as the question of the reality, the nature, or the date of a discovery
does. Those who dissent from the conclusions advocated in the preceding
sections of this discussion, will of course apportion the merit due to
Cavendish, Watt, and Lavoisier, otherwise than I do; and against this
I have nothing to say. Some of those, also, who entirely or substan-
tially accept as valid the conclusions already urged, may nevertheless
differ widely from me, as to the way in which the honours of the disputed
discovery should be divided amongst the three great claimants. I do not
regret tne probability of this, but rather welcome it, as more consistent
with an intelligent acquaintance with the facts of the case, than a uniform
estimate of the relative merits of the rivals would bo. The questions
already considered are the only ones on which it is worth while to take a
side strongly. It would be a more waste of time to discuss the question

• Wait Corr, p. 61.

THE WATER CONTROVEtlSY. 433

of merit with those who rank themselves on the opposite side from me, for
there can be no agreement between ns. It would be as unprofitable to
fieek to compel all who are on the same side as myself, to use the same
balance and weights as I do, in apportioning the merit due to the dis-
coverer of the composition of water, between Cavendish, Watt, and
Lavoisier. It will be enough, therefore, if I briefly state the grounds on
vhich I urge the following conclusions. First, then, of Cavendish. He
only of the three is entitled to be called tiie discoverer of the com-
position OF WATER, if that title is to be given undivided to any one of
them. He first consciously converted hydrogen and oxygen into water,
and first announced the possibility and reality of such a conversion.
Whatever else is doubtful in the Water Controversy, this at least is certain,
that Cavendish was the first who formed water out of its elements. He
alone is entitled to the undivided merit of having first observed that
(nearly) two measures of hydrogen and one of oxygen may be burned
into their own weight of water; and he supplied the data to which,
whether conscious of the fact or not, Watt and Lavoisier were indebted
for the foundations of their conclusions. Even, therefore, if we should
concede to tho supporters of Watt all that they can claim for him, viz.
that he first drew the just inference from experiments on the synthesis of
hvdrogen and oxygen, we should still be entitled to claim for Caven-
dish one-half of the merit of the discovery; inasmuch as be supplied the
data, whilst Watt supplied the conclusion. Such a liberal division of merit,
however, is by no means satisfactory to some of the advocates of Watt,
Trho, in their extreme and confident partisanship, would exclude Caven-
dish from any share of merit whatever. They have sought, accordingly,
OS Watt can only claim the conclusion, to impute to certain of Cavendish s

Predecessors the observation of the production of water from its elements,
'he claimants against Cavendish are Macquer, Warltire, and Priestley.
Their pretensions, which are urged by others, not by themselves, may be
dismissed in a word. All that Macquer did, was to observe that inflam-
mable air, when burning, deposited a liquid, to appearance water. On
the source of this fluid he ofiered no opinion, and he only assumed, he did
not ascertain, that the liquid was water. Arago indeed states, that
Macquer analysed it; ''qui apres verification se trouva etro de I'eau
pure. '* No authority, however, for this statement is given; and to the
best of my knowledge, Macquer's results were published only in his
" Dictionnaire de Chymie." From this work, Mr. Muirhead has trans-
lated Macquer's account of what and he and M. Sigaud de Lafond saw,
when they held a saucer over burning inflammable gas, "it was only
moistened by small drops of a liquor as clear as water, and which in fact
appeared to tu to be only pure water. "f Nicholson, in his dictionary,
comments thus on the statement of Macquer: — " He does not say whether
any test was applied to ascertain this purity."^ The French chemists, in
short, only guessed that the liquid was pure water, and possibly guessed
wrongly; for it may have contained a trace of sulphuric acid transferred
from the materials employed to yield the gas, which appears to have
been hydrogen. It is strange that Mr. Muirhead, after quoting from
Macquer the passage which 1 have borrowed fi*om him, should thrice in

* Anuuaire du Bureau des Longitudes, for 1839, p. 349. In Mr. Muirhead's
Translation of the Eloge of Watt, the quotation is rendered, " a liquid like water, and
which on analjfri$ proved to be pure water." {Watt Corr. p. 224.)

+ Dictionnaire de Chymie, tome ii. p. 314. Quoted from JVatt Corr. p. xxviii.

i Article, Water, vol. ii. p. 1018.

2 F

434 CAVENDISH AS A CHEMIST.

effect state 'Ubat the French chemists aseertained*' the liquid ''to be
pure water.***

Warltire's obsenrations on the appearance of water aie thns reported
by Priestlej : — " I mnst add, that the moment he [Warltire] saw the
moisture on the inside of the close glass ressel, in which I afterwards
fired the inflammable air, he said that it confirmed an opinion he had
long entertained, viz, that common air deposits its moisture when it is
phlogisticated. With me [Priestleyl it was a mere random experiment,
made to entertain a few philosophical friends.'*!

Priestley and Warltire, it thus appears, assumed without examination
that the liquid was water. The former drew no conclusion as to its
origin; and the latter did not regard it as produced, but as simply
deposited from the common air in which it had pre-existed as ready
formed vapour.

From this account it will be seen, that although Warltire and
Priestley's experiments confessedly suggested Cavendish's researches, it
was left to him to demonstrate that the liquid was pure water, and Uiai
it was equal in weight to the gases burn^. Till both those facts were
ascertained, no one was entitled to infer, or did infer, that water was
prodttced during the combustion of hydrogen, and Cavendish was the first
to ascertain them. This cannot, in truth, be denied; and the impngners
of his claims have thought it well to lessen if possible his merits by
declaring that the water he procured in his experiments was not quite
free from foreign matter, and was not sufficiently analysed. Sir David
Brewster affirms that in one of the experiments the water ''was not
absolutely pure ;" and dwells upon the nict, that in another the liquid
yielded on evaporation a trace of sediment. | Mr. Muirhead states that
Cavendish applied "some tests to the liquor condensed ;"§ and Lord Jeffrey
declares that Cavendish's processes were conducted with far less care
and accuracy than those afterwards instituted by Lavoisier, Monge,
Berzelius, or Dumas.H These objections are easily disposed of. We
may concede to Sir David Brewster that the water was not abto-
ItUely pure, but in this respect his rivals had no advantage over
him. Absolutely pure water, indeed, is a thing which very few per-
sons, even chemists, have seen. We may also concede to Lord
Jeffrey, that Berzelius and Dumas might have improved on Caven-
dish's process; but not that Lavoisier and Monge did, for their pro-
cesses were still more improvable than Cavendish's ; and Berzelius and
Dumas would have been the first to protest against experiments of 1781
and 1782 being tried by the more exacting rules, which the united
labours of a host of chemists, spread over more than half a century, have
rendered imperative, by the improvements which they have introduced.
As for Mr. Muirhead, I do not know on what principle of justice, he
three times over attributes to Macquer the observation that hydn^en
yields, when burned, pure water, although he only made a guess, and
did not try a single test ; and yet denies that Cavendish's careful
analysis entitled him to call the liquid he analysed pure water. It must
be a substance much more difficult to recognise than people generally
imagine, if the ''some tests" which Cavendish applied, did not entitle
him to affirm what it was.

* VTatt Corr, pp. xxviii, Ixxxix, and cxxiii.

t Barpt9. Sf Obg. on Air, 1781, p. 398.

t North Bnt. Revieto, Feb. 1847, p. 494.

i Wait Corr. p. xxxv. || Edinr, Rev, Jan. 1848, p. 134.

THE WATER CONTROVERSY. 435

The friends of Watt are using a very dangerous argument in
reasoning thus; for if Cayendish is to be condemned because he only
tried " some tests/' what is to be said of Priestley who tried none, and
has left on record no proof that either he or Watt was entitled to infer
that pure water had been produced? In reality, as I have shown
already, Gayendish's analysis was more complete than Layoisier s, and
more accurate than Mouge's or Priestley^s; and the fact that they oyer-
looked the occasional production of nitric acid is the best proof that
Cayendish more rigidly investigated the product of the combustion of
hydrogen and oxygen, than any or all of them did. It is mere hyper-
criticism, then, to cavil at Cavendish's experiments; for if they were not
such as to entitle him to his conclusion, still less were those of Priestley,
Lavoisier, and Mouge; and the result we must arrive at is, that not one
of the rivals discovered the composition of water, and that it has not yet
been discovered. I shall assume it, therefore, as a settled point, that
Cavendish was the first who couverted a given weight of hydrogen and
oxygen into the same weight of pure water.

Monge's original and independent observations were misinterpreted,
so that they do not demand consideration here; but the other con-
clusions, viz. those of Priestley, Watt, and Lavoisier, go back to Caven-
dish's researches, on which at first or second hand they were based.
His observations thus stand first in chronological order; and first also in
what I may call organic sequence. They were the root out of which his
own conclusions immediately sprang, and those of his rivals mediately
branched.* He is thus doubly entitled to at least half the merit attach-
ing to the discovery of the composition of water, inasmuch as he furnished
the data on which it was based by all its claimants.

I claim, however, for Cavendish the remaining half of the merit due
also. And this on two distinct grounds already discussed at length.
The first, that though direct proof is wanting, indirect evidence of the most
cogent kind can be adduced to show that he had formed a just theory of
the composition of water, before Priestley had repeated his experiments,
or supplied Watt with the data from which he drew his conclusions.
The proofs of this are, as we have seen,

1st. That Cavendish, a highly honourable, upright, and modest man,
claimed the conclusion as his own; acknowledged no obligation to Watt;
and accused Lavoisier of plagiarism.

2nd. That the dates of his journal strongly favour the belief, that he
had formed a just theory of the composition of water as early as January,
1783.

3rd. And, most important of all, that he announced his experiments
and their results to Priestley, in a statement to the effect, that hydrogen
and oxygen can be converted into water, which has its origin in these
gases.

This announcement contains a just theory of the composition of water,
although it does not make use of the word phlogUton, which Cavendish
employed in the full exposition of his theory. The substitution, however,
of that word for hydrogen (inflammable air from the metals) was not
occasioned by any alteration in opinion as to the convertibility of hy-
drogen and oxygen into water, but arose solely out of speculations on

* Lord Jeffrey concedes this so far as Watt is conoemed, and it is still more
evident in reference to Priestley and Lavoisier.

2f 2

436 CAVENDISH AS A CHEMfST.

the nature of hydrogen. Cavendish's theory, indeed, was altered for ilie
worse by the cbange in temis, and rendered needlessly hypothetical. So
soon, however, as he affirmed that water could be generated out of two
gases, in which it had its origin, he had demonstrated his disooveiy of
the nature of that liquid, and announced the true, though not the com*
plete theory of its composition. Cavendish, therefore, in so far as he tc^
Priestley that hydrogen and oxygen can be entirely converted into wat^,
of which they are the factors or constituents, was prior to him, to Watt,
to Lavoisier, and Monge, and was the first to interpret his own experi-
ments.

Cavendish has a second daim of a lower, but, nevertheless, im-
portant kind, to be called the discoverer of the composition of water;
for, even if it were certain that Watt drew a conclusion from Priestley's
repetition of Cavendish's experiments, before Cavendish drew one from
the original trials, we have seen that Priestley did not employ hydrogen,
and that Watt did not affirm that water consisted of hydrogen and oxy-
gen, but spoke of the combustible element of water as inflammable air,
without special reference to hydrogen, and apparently with special refer-
ence to the inflammable gas from charcoal ; so that whatever Cavendish
borrowed from Watt (if he borrowed anything), he did not, and coold
not, borrow from him the doctrine that water consists of hydrogen and
oxygen; and even if it were conceded that Cavendish's conclusion was
not arrived at till after he had read Watt's letter, it would still be true
that his theory, though later in date than Watt's, was the first true
theory of the composition of water. It has been urged against this
view, that Cavendish accepted Watt's theory as identical with his own,
and that it is, therefore, idle for any one else to attach importance to
differences he may detect between the two theories. The argument is
specious, but not the less hollow. Cavendish did not accept Watt's con-
clusion as of equal value with his own, without a protest in reference
to an apparent difference between them. The wording of their theories
closely corresponded, and so far their conclusions appeared identical, but
the term phlogiston, or inflammable air, stood speciaUy for hydrogen in
the one theory, and for charcoal gas (if for any single elastic fluid) in the
other, and to this fact Cavendish drew attention. He did not affirm that
charcoal gas would not serve as well as hydrogen for the production of
water, but he would not make himself responsible for the declaration
that it would ; he only professed to be certain that the inflammable air
from metals (hydrogen) was the combustible constituent of water. Watt
and Priestley together were sponsors for the inflammable air from char-
coal being a true hydrogen, or water producer, and on their authority be
accepted Watt's conclusion as equivalent to his own. He would have
acted unjustly if, without repeating Priestley's experiments, he had done
otherwise. Nevertheless, his approval was only conditional; for he
speaks of Priestley's experiments as " of the same kind," not as abso-
lutely identical, and he draws attention to three points of difference, vis.
that Priestley obtained no nitric acid; that he " used a different kind
of inflammable air, namely, that from charcoal," and that he perhaps
used a greater proportion of it.* And, at a later period, we know from
his MS. Journal, that he justified his caution by discovering that the
charcoal-gas and hydrogen are not identical. Cavendish's conditional

* Phil TVojM. 1784, p. 135.

THE WATER CONTROVERSY. 437

approiva), then, of Wait's conclusion^ does not make him responsible for
the errors it contained; still less does it affect the fact^ that Cavendish's
own theory was the earliest true theory of the composition of water.

As for Watt^ the merit which can be assigned to him cannot^ accord-
ing to the Tiew taken in the preceding pages, be great, at least in one
respect. If his conclusions, as I have contended, were drawn from expe-
riments in which hydrogen was not used, and if his theory did not affirm
that hydrogen was one of the elements of water, then he was at no time
a discoverer, much less the discoverer, of the composition of water. His
theory, however, at the period of its announcement, passed for as good as
Cavendish's and Lavoisier's, and had as much effect for a time as their
juster conclusions, in inducing belief in the compound nature of water.
On this account, even the extreme advocates of Cavendish's merits
shonld, if impartial, assign Watt a place of honour in connexion with
the discovery of the true nature of water; and, if they do not, the histo-
rian of the progress of science must be his eulogist. His theory, though
inaccurate, unquestionably made a close approximation to the true
theory, and was very clearly and perspicuously stated. In particular,
lie urged with ereat fulness and emphasis, in his private letters, that
water is not an element or simple substance, that it is composed of ingre-
dients, and that he knew what they were. His quaint recipe ''to make
water" shows how fully he realised, that it might be compounded or
manufactnred out of substances which he specified. There can be no
doubt that Watt's theory increased the faith of Cavendish and Lavoisier
in their own views, and won the approval of the great majority of their
scientific contemporaries.

It did service thus, at a certain epoch in the progress of discovery,
and has a place in the history of science, whether it pleases us that it
should have such a place or not. When we consider, indeed, that Watt
expounded very fully that water is a compound, and came so near its
true composition as to assert that one ingredient of it is oxygen, and the
other combustible gas, or a combustible gas, we must acknowledge that,
had his been the earliest interpretation of the earliest experiments on the
subject, his merit would be very great. But when we call to mind that
the experiments which he interpreted were shaped to his hand by Caven-
dish, and presented to his consideration in their simplest and most
demonstrative form, with their validity as establishing the convertibility
of a combustible gas and oxygen into water, already announced by
Cavendish and confirmed by Priestley, we can assign him only a seconda^
and subordinate place as a discoverer of the composition of water, ife
was but a follower in the path of Cavendish. Had the latter never
experimented, or had he never reported his results to Priestley, there is
no reason to suppose that Watt would have conjectured, even remotely,
that water is a compound of oxygen and inflammable air. He was not
on the track of such a discovery. His speculations on the convertibility
of steam into a permanent gas by the change of all its latent into sensible
heat, did not point in that, but in exactly the opposite direction. He
was following Priestley in all his devious wanderings, and going astray
along with him, into the belief that water was transmu table into atmo*
spheric air, when Priestley's repetition of Cavendish's experiments (in
which all that was true and significant was Cavendish's) arrested him in
his mistaken course, and enabled him to approximate to the true theory
of the composition of water. He taught nothing, however, to Cavendish,
who, in due time, when his inquiry was completed, made it public to all.

438 CAVENDISH AS A CHEMIST.

Watt most rank below Carendish and Lavoisier, though how far below
them I will not attempt to decide. To the consideration of the Freneii
philosopher's merits I now tnrn.

Layoisier has not the slightest claim to be considered the discoTerer
of the true nature of water; nor did he demonstrate its synthetical pro-
duction more fullj than Cavendish did. He added absolutely nothing to
the results which the latter had obtained, and his method of procedure
was less accurate and ingenious, and differed for the worse, in so far as
the experiments were tried on too large a scale.

Berzelius, in apportioning the merit of the discovery under discus-
sion between its three great claimants, assigns to Lavoisier the first
announcement that water is a compound, and that when produced by the
combustion of hydrogen and oxygen, it " weighs exactly as mvch as the
combined gases weigh together.*'* In a previous section, however, it
has been shown so fully that he was anticipated in both those annoanoe-
ments by Caveudish, and in the first by Watt, and that he was informed
of their conclusions by Blagden before he formed his own theory, that I
need not recur to the subject. What was especially his, was the precise
definition of the combustible element of water, as a peculiar gas having
the physical properties of the other gases, and the ordinary attribntes of
matter, so that the halves of water were physically similar, though
chemically unlike. This seemed to Cavendish, who commented on it,
in the close of his own paper, a slight alteration and a very doubtful im-
provement on his own view. To us it appears in a very diflerent light
Lavoisier's merit was assuredly very great. Cavendish encumbered hia
theory by needless and unwarranted speculations on the nature of phlo-
giston. His declaration that it consisted of water and hydrogen involved
the awkward and embarrassing doctrine that water was an element of
water. He further contended, that every substance when oxidized

Jrielded water, and that the addition of oxygen to any body was eqnivan
ent to the removal of water from it. His view was thus, to a great
extent, only a modification and extension of the doctrines of Stahl and
the other theorists of the Phlogiston School. Cavendish, like them,
believed that every combustible and oxidable body contained phlogiston,
but, unlike them, he identified it with hydrogen. The main point, how-
ever, of difference between their views was, that the older chemists
were content to affirm, as their most general proposition, that burning
bodies evolved or emitted phlogiston, which, by its addition to atmospheric
air, converted it into phlogisticated airjf whereas Cavendish contended
that the phlogiston or hydrogen always united with the oxygeo of the
air, and produced water. Such a view, however unfounded, implied
no confusion or hesitation as to the nature of water. An erroneous
opinion was entertained as to the nature of hydrogen, and likewise as to the
extent of its distribution among combustible bodies, but no error was
thereby introduced into the definition of water, as a compound of hydro-
gen and oxygen. The clearness, indeed, of Cavendish's conception of
this, is shown by the very excess of his ^neralisation, for he showed the
importance which he attached to his discovery, by representing it as
fitted to explain every combustion and every oxidation; and with con-

* Lehrbueh, 1843, p. 371.

+ This term was restricted by Carendish and his contemporaries to nitrogen, but
by the older chemists it was applied to atmospheric air so altered as to extinguish flame,
whether this resulted from the loss of oxygen or the addition of carbonic add (or other
irrespirable gas) ; or the simultaneous loss of the one and gain of the other.

THE WATER CONTROVERSY. 439

Busient and rigorous fidelity to his theorji he contended that wherever
hydrogen and oxygen met and combined^ water moat be produced. It is
quite clear, howeyer, that although Gayendish announced the true nature
of water with so great ezplicitness that it was not left for any later expo-
sitor to depriye him of the honour of being the discoyerer that it is a com-
pound, nevertheless, the erroneous views he connected with it could not
but have led both himself and his contemporaries far astray; and, had
they not been immediately neutralised by the expositions of Lavoisier,
they would have delayed indefinitely the overthrow of the mystical and
delusive doctrine of phlogiston.

Similar remarks apply to Watt's theory apart from its errors. He,
like Cavendish, believed that phlogiston contained water : and he further
held, which Cavendish did not, that nitrogen and carbonic acid, as well
as water, were, what we should now call oxides of phlogiston, which entity
he identified with inflammable air as procured from various sources, not
"With hydrogen. Had Wattes theory, accordingly, been published first and
alone, it would have diverted chemistry out of its legitimate course, and
delayed its rational progress still more than Cavendish's erroneous doc-
trines would have done. Phlogiston and oxygen, according to Watt's
▼iew, might be converted into carbonic acid or nitrogen, or atmospheric
air, as well as into water, so that substances consisting of very different
ingredients were held to consist of but slightly varying proportions of
the same elements. This view, apart from its not defining phlogiston
as hydrogen, could not but have set the chemistry of the gases on the
wrong track for at least many years. It was the publication of Lavoisier's
memoirs, contemporaneously, or nearly so, with Cavendish's and Watt's
papers, which prevented their erroneous speculations from doing any
mischief. And however strong may be our feeling and expression of
indignation against Lavoisier's plagiarisms, we must guard against allow-
ing our disapprobation of his concealment of obligation to his English
rivals, to make us unjust towards him, in reference to what was all his
own. He has, in truth, paid a heavy penalty for his disingenuousness ;
and for some seventy years has been held up to the scorn of the world as
a detected plagiarist. Yet never, perhaps, was the policy of honesty more
apparent than in the case of Lavoisier, for he might safely have conceded
to Cavendish and Watt all that was theirs, and yet have left to himself
what would have secured, and does secare to him, the highest honour.
Had he done so, the English chemists would have been foremost in
praising him, and his name would now be revered by us in a way it is
not likely soon to be. The wrongs, however, which he did to others
cannot justify us in wronging him ; and I feel peculiar pleasure in insisting
upon his merits in connexion with the discovery of the true nature of
water. He assuredly did chemistry an unspeakable service by abolishing
the vague and mischievous phlogiston, which proved an ipnii fcUuus even
to men like Cavendish and Watt, and substituting for it, as a name for
the combustible element of water, the simple term, aqueous inflammable
air, and afterwards hydrogen. He thus inade the one element of water
as manifest and tangible a reality as the other, and deprived it of all
those mysterious and imaginary attributes which had so long invested
phlogiston with a phantom existence. The believers in that entity,
unwisely for themselves, had identified it with inflammable air, or located
it in that body. From that moment it had to forswear any physical
difference from other forms of gaseous matter. It needed Lavoisier,
however, to perceive and to demonstrate this, and to urge it by multiplied

440 CAV£ND1SH AS A CHEMIST.

experiments and arguments. And it was most forianftte for us tiiat his
views were published before those of the English chemists had time to
spread, so that the discovery of the nature of water^ which seemed in
their hands destined to prolong the existenC'e of phlogiston, was Gonverted
by Lavoisier into an instrument by which he for ever effected its orer-
throw. It would be difficult, accordingly, to over-estimate Lavoisier's
merits in this respect. At all events, for my own part, looking back,
almost with regret, on the many hours I have spent in wading through
the wearisome and often unprofitable writings of the Phlogiston School, I
will be foremost in acknowledging obligation to Lavoisier for having put
so decisive and final a stop to the speculations of its disciples. Modern
chemists cannot be too thankful that they have not grown up under tlie
fablse system which led to so much waste of time and capacity by their
predecessors, and all their gratitude is due to Lavoisier.

Further, he was the first who consciously analysed water into its
elements. He appears to have been indebted — as he tells us himself — to
La Place* for the suggestion that the hydrogen which is obtained by the
solution of metals in diluted acids results from the decomposition of Mrater,t
so that, to some extent, he must divide the honour with Lavoisier. Into
this question, however, I do not enter further than to remark, that it adds
another honour to the many which render La Place illustrious. The pro-
cesses by which water was first decomposed, appear all to have been of
Lavoisier's own devising, although he had the assistance of Meusnier in
performing the necessary experiments; so that he is justly entitled to he
called the first analyst of water. And we may affirm with great con-
fidence,' that the faith of mankind in' the modern theory of the natnre of
water, was increased twofold by the discovery of the French chemist that
water can not only be compounded out of hydrogen and oxygen, but can
also be decomposed into these gases. Analysis — i,e, the resolution of any
composite or compound whole into its constituent parts — is a method of
investigation more familiar to us as a means of ascertaining truth, and
more easily and more frequently employed by the great majority of
thinkers in all their investigations, than synthesis, i.€, the production or
construction of a compound whole out of its components. Inductive
reasoning, accordingly, is much more practised by all but the highest
intellects than the deductive method of research. I will venture to
affirm, indeed, that no teacher of chemistry at the present day who wished
to demonstrate to his pupils the compound nature of water, would select
Cavendish's synthetical process as the means of proving it, but wonld
exhibit in preference the methods by which it may be analysed. The
universal practice of our lecturers, and of the authors of our text-booksj
sufficiently proves this. Lavoisier's method is not the best analytical one^
for it exhibits only one of the elements of water in a state of isolation; so
that we should now prefer Mr. Grove's application of a high temperature,
which sets both elements of water free; or Nicholson and Carlisle's appli-
cation of the voltaic current, the action of which is very rapid, and may
be exhibited on a scale of great magnitude, which Cavendish's experi-
ment, even if equally demonstrative, cannot be. Lavoisier's process^
however, was the only, and therefore the best analytical method of the
time, and was as new to mankind, and as unexpected as Cavendish's. It
could not but increase the belief of those who had already accepted as

♦ Mim. par Lavoisier. Watt Corr. p. 181; or MSm, de VAead, pour 17B1,
p. 476.

t Levant ntr la PhihsopJde Chimiqve, par M. Dumas, p. 158.

THE WATER CONTROVERSY. 441

true the latter's eonclufiions ; and it probabl j carried conviction to the
minds of manj who, like Monge, hesitated as to the significance of the
Bjnthetical experiments. No solicitade, accordingly, for Cavendish's repu-
tation can justify us in refusing Lavoisier a very liigh place in the history
of the discovery we are considering. When we remember that he gave
the simplest and most accurate exposition of the phenomena attending
the synthesis of hydrogen and oxygen ; that he first efiected and most
lucidly expounded the analysis of water ; and that, by the use which he
made of the fact that water is a specific compound in his interpretation
of other phenomena, he did more than any of his contemporaries to im->
press mankind with a sense of the importance of the discovery which had
been made, we cannot hesitate to place him side by side with Cavendish,
with whom he would probably, from the first, have been ranked in hon-
ourable brotherhood, but for his own lack of generosity to his English
rival.

Some other points, however, require to be noticed, before we can settle
Gompletelv the relative merits of Cavendish, Watt, and Lavoisier. The
friends of Watt are fond of affirming that Cavendish did not appreciate
the importance of his discovery, as the very title of his paper, '' Experi-
ments on Air," shows. The very opposite, however, is the truth ; for it
was Cavendish's fault, as we have seen, to over-estimate the value of his
discovery, which he affirmed supplied the rationale of every combustion
and oxidation, besides explaining the most important phenomenon of
vegetation and the action of sunlight in efifecting chemical changes.
The title, too, was quite appropriate, for the main object of the paper, from
first to last, was to record an experimental inquiry into all thepheuomena
which attended the phlogistication, or deoxidation of atmospheric air by
various phlogisticating or oxidable bodies besides hydrogen. His paper
entered largely, moreover, into the consideration of the nature of nitrogen
and oxygen as constituents of the atmosphere, and furnished a new pro-
cess by which their relative amount, or its quantitative composition, could
bo determined. Its title, accordingly, was most befitting, and we might
as well contend that its emplo3rment a second time, as the heading of the
next paper Cavendish published,* implied an indifierence to the greatness
of the discovery of the composition of nitric acid, which it recorded, as
affirm that its earlier adoption as the title for the paper of 1784, is at
variance with the idea that Cavendish appreciated his discovery of the
composition of water. Such reasoning, if valid, would tell still more
against Watt than against Cavendish, for the former put no title ori-
ginally to hia paper, so that its readers were left to discover for themselves
what the importance of its contents was, and Cavendish and Lavoisier
had drawn the attention of all to the importance of the new views con-
cerning water, before Watt did the same by his title.

Much reference has also been made in depreciation of Cavendish's
merit, to the delay which attended the publication in full of his views.
In reality, however, it is much more easy to account for Cavendish's delay
than for Watt's. The latter withheld from publication a declaration that
water consisted of infiammable air and oxygen, because a totally difierent
proposition, viz. that water can be converted into atmospheric air, proved
untenable. His friends refer triumphantly to the Watt Correspondence
as demonstrating that he uniformly maintained in private, the theory
which nevertheless he would not publish. This fact, however, only reu-

• Phil. Traw. 1785, p. 372.

442 CAVENDISH AS A CHEMIST.

den his publio silence the more inexplicable ; and in spite of his alleged
nnhesitating confidence in the troth of his theory, he told Sir Joseph
Banks that one of his reasons for withdrawing his letter was, his ''being
informed that that theory was considered too bold, and not suficietUly
supported by fads. These reasons made me think it prudent to delay the
publication until I should have considered it more maturely, and have
made some experiments to determine the truth or falsehood of it.*" This
passage proves how deliberate Watt's delay was, and what amount of
confidence he had in his theory when he withdrew it. He was convineed,
I believe, all throughout, that his conclusion was just, provided the data
on which it was founded were correct, but after Priestley had so grievouaiy
misled him, as to the conversion of water into air, he very naturally began
to doubt whether his ingenious, but inaccurate friend might not have
equally misled him as to the conversion of air or gases into water, aad
he hesitated to publish his theory till he should be certain that Priestley's
experiments were accurate. How well founded this suspicion of Priestley^s
accuracy was. I need not say, and how much it weighed with Watt in
causing his delay, is shown by the fact that when he drew up the new
version of his views, he was careful to fortify Priestley's observations by
a minute reference to those of Lavoisier, which De Luc had reported to
him.t We may confidently affirm, accordingly, that Watt's faith in the
convertibility of inflammable air and oxygen into water, was shaken by
Priestley's mistake so much that he woidd not assert its certainty, but
waited for many months, seeking for the means of establishing its truth.
It is idle, therefore, for the advocates of Watt to reproach Cavendish with
his delay.

From this needful digression, I return to Cavendish. His merit is
not lessened in any respect ; on the other hand, it is positively increased
by the delay which attended the publication of his first series of "Experi-
ments on Air." Let it be noticed in the first place, that it is not the
fact that Cavendish kept back a completed inquiry, and thereby gave
grounds for the suspicion that he distrusted his researches, or did not
appreciate them. I have shown that his MS. journal records experi-
ments referred to in the paper of 1784, which were made so late as
Christmas, 1783, whilst the paper was laid before the Royal Society on
the 15th of the succeeding January, so that no delay attended the publi-
cation of the inquiry after it was completed. It is quite true that the
experiments on the production of pure water from the combustion of
hydrogen and atmospheric air, were made as far back as 1781, and might
have been reported earlier than 1784. A very sufficient reason, however,
can be given why Cavendish did not immediately publish these researchesy
viz. the occasional appearance of nitric acid, when oxygen was substituted
for atmospheric air. Whilst this unexpected and perplexing phenomenon
remained unexplained, it would have been premature to have published a
declaration, that hydrogen and oxygen yield by their combustion pare
water. Had Lavoisier, who expected the oxide of hydrogen to be an
acid, encountered nitric acid (in his earlier researches),! in the circum-
stances in which Cavendish found it, (viz. where the oxygen was slightly
in excess of the hydrogen, whilst a little nitrogen was present as an im-

* Watt Corr. p. 52. t Phil. Traru. 1784, p. 332.

X The letter from BerthoUet to Blagden, quoted in another place, shows that
Layoisier, in his later trials, confirmed Cavendish's obsenration of the prodaction of
nitric add, and that La Place was led completely astray as to the nature of the acid, by
the circumstances of its origin.

THE WATER CONTROVERSY. 443

purity in tlie gases) we may be certain that this phenomenon would have
appeared to him more important than even the production of pure water^
and that he would have thoroughly investigated the source of this acid,
before he made public his interpretation of the phenomena which he had
witnessed. Watt also wrote to Priestley, " I maintain my hypothesis until it
shall be shown that the water found aiter the explosion of pure and inflam-
mable air^ has some other origin ; nor shall I believe that air is a child of
cbcidSf or rather a modification of them, until stich acids can be found after
the decomposition of it"* He tells De Luc also that '' the water remaining
after inflammation is not in the least acid, which must be the case if the air
was formed of the acid part of the substances."f From these references^
it is manifest that Watt would have attached ajs much importance as
Cavendish did, to the appearance of nitric acid^ and would have altered,
or perhaps abandoned, his hypothesis, bad he been aware of the produc-
tion of acid, during the combustion of apparently pure oxygen and hydrogen.
Cavendish's rivals, therefore, may be regarded as both testifying to the
wisdom of his delay, till he should have discovered the origin of the
nitric acid. He did. not, however, complete his investigation into this
subject, till January, 1783, and he announced his results to Priestley,
not later than the succeeding March, and to the French chemists in the
succeeding June. He stands, therefore, altogether acquitted of suspicious
or blameable delay. His whole paper he did not publish till a year later,
but the records of his journal completely demonstrate that this apparent
procrastination did not result from any loss of confidence in the accuracy
of his conclusions concerning the composition of water, or any change of
opinion on this subject, but was occasioned solely by the prosecution of
collateral inquiries, which he deemed essential to the completeness of his
memoir. One chief cause of delay was the publication of his account of the
new eudiometer, which formed an essential part of his elaborate inquiry
into the phenomena which attend the phlogistication of air. It was
essential that this paper should be published before the record of his
views concerning water, for the interpretation which he gave of the pro-
duction of this liquid when hydrogen and atmospheric air were burned
together, would have been unintelligible to his readers, had he not pre-
viously announced his discovery of the constancy in composition of air^
and stated what was the relative amount of oxygen and nitrogen con-
tained in it.

Another cause of delay was the publication, in 1783, of his commentary
on Mr. Hntchins*s experiments on the freezing point of mercury. These
experiments were made in 1781 and 1782, at Hudson's Bay, but the
record of them did not reach this country till late in the latter year, and
it was not read to the Royal Society till 1783. As Cavendish, however,
had furnished Mr. Hutchins with the apparatus he employed, and with
directions how to use it, it devolved upon the former to furnish a commen-
tary which should accompany Mr. Hutchins's paper. An abstract of this
commentary is given in the analysis of the papers on heat, from which it
will appear that the experiments which it called for, and the study of Mr.
Hutchins's numerous results, must have occupied a considerable period.

Again : it was impossible to adopt the conclusion that water is a com-
pound of hydrogen and oxyeen, and not be led to speculate in a way
which had not been done be^re, concerning the nature of these gases.
Cavendish, accordingly, and Watt also, made many experiments to

♦ Watt Corr. p. 27. t Ibid. ibid. p. 30.

444 CAV£NJ)1SH AS A CIIBMIST.

determine the constitution of the elements of water, and dkcossed at
great length the theory of the different processes bj which oxygen can be
prepared. Cavendish, further, was directly led by his discovery that
nitrogen and oxygen can be burned together into nitric acid, to investi*
gate minutely the nature of the former gas as well ae of this acid. la
addition : Cavendish saw at once that the doctrine of the compound
nature of water, could not but throw much light on the phenomena td
vegetation, and he explains at length his view of the function of water in
maintaining the growth of plants, and discusses the deoxidising power of
sunlight, which he refers to its power of causing the separation of
hydrogen from the chemical substances it decomposes. In this way ha
accounts for the evolution of oxygen by the green parts of living plants,
when exposed to the sun.

It was natural — it was, I may almost say, inevitable — ^that such col-
lateral speculations should have been entertained by a believer in the
compound nature of water. Similar views, more or less fully expounded,
occur in the papers of Watt, Lavoisier, and Monge, and to have omitted
them would both have lessened the value of their separate papers, and
have left other writers to claim the credit of making these applications of
the new doctrine. Add to all this, that Cavendish was exceedingly
cautious, never published in haste, but on the other hand, left behind
him many finished inquiries which he had not given to the world, and
nothing remarkable, or blameable, or fitted in any way to lessen his
merit, will be found in his so-called procrastination.*

If ever, indeed, delay was justified by its fruits, it was in Cavendish's
case. Had he hastened to publish, he might have won a single laurel
wreath, but as it was he gained a triple crown. He claimed in the end,
not only the discovery of the composition of water, but that also of the
constant quantitative composition of the atmosphere, and he had in his
hands the rapidly expanding germ of the discovery of the nature of
nitric acid.

Once more. Cavendish's merit as discoverer of the composition of
water, is much greater than that of Watt and Lavoisier, even if it
were conceded that they also had made the disputed discovery. The
French chemist, before he made an experiment on the subject, had been
informed that at least three persons, viz. Cavendish, Watt, and Blagden,
were satisfied that water was composed of inflammable air and oxygen.
The path, accordingly, along which he should travel, and the objects
which he should encounter, were pointed out to him, so that he could
not fail to observe all the essential phenomena, and could foresee the
inference which he should draw, in the light of the foregone triple
conclusion which had been made known to him. It was easy, therefore,
for Lavoisier to appear before the Academy, with a single insufficient
experiment, and profess to found his theory of the composition of water
on it alone.

Watt, likewise, had much simpler data supplied to him by Priestley,

* It was in truth with Cavendish the exception, not the role, to publish an inquiry.
The large number of papers, several in a state fit for publication exactly as they stand,
which remain among the Cavendish MSS., show both that he was in the practice of
delaying indefinitely the publication of finished researches, and that he has left us in
entire ignorance of his reasons for so doing. Any endeavour, therefore, like that of Lord
Brougham, to say when Cavendish should have published a paper, seems to roe a hope-
less attempt; still more, any endeavour to make delay in publishing an argument in
favour of want of belief in what was delayed.

TIIE WATER CONTROVERSY. 445

than Gavendish'fi experiments furnished to himself. The problem the
first had to consider was : ^'^iyen inflammable air and oxygen, changing
into their combined weight of pare water, what conclusion is warranted]*'
Cavendish had the much more difficult enigma to solve: ''given those
gases changing sometimes into water, sometimes into water and nitric
acid, what is the just inference?*' Again: Priestley and Watt con-
sidered only what happened when inflammable air and oxygen were
burned together, but Cavendish's inquiry embraced also the action of
burning hydrogen and air on each other; and as a necessary step towards
this, he spent months of labour on the analysis of air, before he could
proceed to deduce from his researches into its composition the proportion
in which hydrogen and oxygen should be burned together, so as to ensure
the combination of the entire volume of each gas taken.

Further: Cavendish had no predecessor in his researches, and was
embarrassed by the assertion that heat was ponderable. Priestley, on
the other hand, had Cavendish's experiments before him as an example
and guide: and Watt was assisted towards his conclusion by being fami-
liarised with the doctrine of the convertibility of gajses into water, which
Priestley undertook to test, and professed to have verified. Priestley
thus comes between Cavendish and Watt, and transferred from the former
to the latter, not only his experiments, but at least an outline also of his
oonclusions.

Finally : I know of no explanation of Priestley's extraordinary char-
coal-gas experiments so plausible as this, viz. that aware that Cavendish
ha«i obtained pure water equal in weight to the gases burned, he took
for granted that the water he himself procured, was also pure, and set
down as due to his imperfect balance the deficiency in weight which
arose from the charcoal-gas being employed instead of hydrogen ; so that
Priestley^s experiments appeared valid, only because they were supple-
mented by those of Cavendish, on which they rested. It was to the
latter^ therefore, and not to Priestley, that Watt was really indebted for
the foundation of his claim to be one of his rivals.

When thns it appears that Cavendish was first in the field, and that
he famished his rivals with the grounds of their conclusions ; whilst the
one also was fully informed of his theory through Blagden, and the other
received an account of it through Priestley; when we further consider
that the problems he had to investigate were more numerous and more
difficult than those they undertook to solve; and when, lastly, we learn that
besides first observing all that they claimed to have observed concerning the
combination of hydrogen and pure oxygen, he also investigated the action
of hydrogen on atmospheric air, which led him to a new method of
analysing air; and also the action of hydrogen on artificial mixtures of
nitrogen and oxygen, which led to the discovery of the true nature of
nitric acid, — we shall not hesitate to affirm that so far as the discovery of
the composition of water by synthesis is concerned. Cavendish must rank
above Lavoisier^ and far above Watt, however liberally their merit be
estimated.

CAVENDISH AS A NATURAL PHILOSOPHER.

<»^M»#»^^^»>i^<W^M»<WS»<»

PAPERS ON HEAT.

Gatsndish devoted much attention to the snbjeet of Heat, and appeara
to hare discovered for himself the great laws, which in the langoa^ of
the material hypothesis of heat, determine the relation of bodies to latent
and specific caloric. Among his unpublished papers remains an extensive
series of experiments on the heat of liquidity and gaseity, and on specific
heat, besides an inquiry into the change of temperature which accom-
panies chemical combination when that is attended with the evolution of
a gas. The original record of these researches occupies 120 small octavo
pages, which are not dated up to the 89th. It bears date February 5th,
1 765, and as the experiments registered on the preceding pages must have
occupied many weeks, it is certain that Cavendish's investigation into the
chemical relations of heat, must have commenced in 1764, and occupied a
considerable portion of that year. He was thus, probably, contempo-
raneous with Black in many of his discoveries on heat, though not in his
earliest, which go back to 1758. From the way, however, in which
Black made his observations public, not through the press, but in his
University Lectures at Glasgow and Edinburgh, it is difficult to determine
the exact date of many of his discoveries; and still more difiicult
to ascertain to what extent those who were not privileged to hear his
prelections, and were not in direct communication with his pupils, were
familiar with his views.

How far accordingly Cavendish was conversant with Black's specula-
tions cannot be precisely determined. Cavendish refers to some of them as
known to him, but of others he seems to have been quite ignorant; and it
is incredible that he should have made the elaborate researches he records,
had he been aware of all that now appears in the posthumous works of
Black on heat. The inquiry seems to have been undertaken solely for
his own instruction, for he published no part of it till nineteen years after
most of the experiments were completed, and then only a trifling frag-
ment of it appeared incidentally in a paper which will be presently
referred to, on the freezing of Mercury, read to the Royal Society in
1783.* Perhaps a reluctance to enter into even the appearance of
rivalry with Black, prevented him from publishing researches which
might be thought to trespass upon ground which the latter had marked
ofiT for himself and preoccupied. In truth, however, with Cavendish
publication was the exception, not the rule, and he has left so many
completed researches unpublished, that no special hypothesis is needed to
account for those upon heat having remained in manuscript.

* Among the Cavendish MSS. are both the original notes on heat, and a treatise
drawn up from them, which, as Mr. Harcoort has pointed out. and, as the personal
references show, "was written for the use of some individual, it does not appear whom."
(Brit. Assoc. Rep. 1839, p. 45.) Cavendish's views were thus reduced to a shape in
wliich they might at once have been printed, had he cared to publish thenu

VALUE OF THE THERMOMETER. 447

Mr. Haroourt has given an interesting sketch of the contents
of the Cavendish MSS. on heat, to which I would refer those who
are curious to know what our philosopher observed for himself on
this subject.* A brief notice, however, of his observations will make his
printed papers on heat more intelligible, and will justify what was said
of Cavendish as an investigator of the laws of heat in the personal
memoir. I shall quote entirely from what Mr. Haroourt calls ''the
paper of results and deductions." It occupies 50 pages quarto, and has
evidently been drawn up from the notes, already referred to, which
extend over 120 pages octavo.

The first experiments commented on, were similar in principle to those
made by Dr. Brooke Taylor and others, to determine whether or not the
mercurial thermometer is an accurate and uniform measurer of tempera-
ture; and consisted in mixing a given weight of cold water with a given
weight of hot water, and vice versa, with a view to ascertain whether the
temperature of the mixture is the arithmetical mean of the temperatures of
the hot and cold water. The experiments were made with all the precision
and accuracy which might be expected from the known character of the
observer; and the immediate results obtained were carefully corrected, so
as to eliminate the loss of heat occasioned by the cooling of the vessel in
which the mixture was made, and by the abstraction of that which was
spent in raising the temperature of its walls, and of the stirrer used in
effecting the complete mixture of the two liquids. The results were so
accurate that, to take one example, the temperature of the mixture came
out 149*2^, whilst "the heat of the mixture by computation should be
149'1°, therefore the heat of the mixture is iV^h of a degree greater than
it should be by computation." (MS. p. 9.)t By these trials Cavendish
justified the truth of the proposition which determined him to make them,
namely, " that on mixing hot and cold water the quantity of heat in the
liquors taken together should be the same after the mixing as before, or
that the hot water should communicate as much heat to the cold water, as
it lost itself."! Thus much settled. Cavendish proceeded to try the effect
of mixing unlike liquids at different temperatures with each other. He
introduces these experiments with the following significant notice: " One
would naturally imagine that if cold mercury or any other substance is
added to hot water, the heat of the mixture would be the same as if an
equal quantity of water of the same degree of heat had been added,
or in other words that all bodies heat and cool each other when
mixed together equally in proportion to their weights. The following
experiments, however, will show that this is very far from being the
case." (MS. p. 12.)

It is plain from this statement that Cavendish was not aware of the
experiments which Black and his pupils had tried, or about that period
were trying; and the fruit of his researches shows that had he published
his results when he obtained them, he would have deprived Irvine,
Crawford, and Wilcke of much, if not all the praise thev have gained,
^nd would have anticipated or rivalled Black in some of his most im-
portant observations.

* Brii. A$$oc, Rep, 1839, pp. 45^50.

t This quotation, and those which follow, are from the Quarto Manuscript referred
to in the text, which, along with Cavendish's other unpublished papers, Lord Burlington
has kindly placed at my disposal.

page

X The whole passage will be found in Brit, Amoc, Rep, 18.39, p. 48. It occurs on
8 of the MS., but is directed to be substituted for p. 1, which is deleted.

448 CAVEKDISH AS A NATUILIL PHILOSOPHER.

His first expertiuents were made with quicksilver and water, at
different temperatures, and the general result he arrived at was, that
"hot water is cooled near as much by the addition of 1 part of cold
water as by that of 30 parts of mercury of the same heat.'^ (MS. p. 14.)
These numbers refer to parts by weight, as all those in Cavendish's
similar experiments do, and his determination of the specific heat of
water as 30 times greater than that of mercury, as inferred from a com-
parison of equal weights, is exactly in accordance with later observations^
The results thus obtained by mixing cold mercury with hot water, were
checked by others, where the mercury was the hotter of the two liquids.
These trials led to the conclusion, that " water cools hot mercnry as much
as 30*42 its weight of mercury can do, or the effect of water in cooling
hot mercury is 30'42 times greater than that of mercury.'* (MS. p. 18.)
This result is confirmed in various ways, and the statement of it is
followed by the account of experiments with other liquids. When spirits
of wine and mercury at different temperatures were mixed, it was foand,
that " by a mean of all the experiments the effect of spirits of wine in
cooling hot mercury or spirits seems to be 22*7 times greater than that of
mercury, and consequently 1*326 times less than that of water." (MS.
p. 23.)

From those experiments and others which I have not mentioned,
another and important conclusion was deduced, which is thus stated by
Cavendish: " It should seem, therefore, to be a constant rule that when
the effects of any two bodies in cooling one substance are found to bear a
certain proportion to each other, that their effects in heating or cooling
any other substance will bear the same proportion to each other." (MS.
p. 26.) After a brief commentary on this law of reciprocal proportior,
the following genci*alisation of all the researches detailed is given. " The
true explanation of these phenomena seems to be, that it requires a
greater quantity of heat to mise the heat of some bodies a given number
of degrees by the thermometer, than it does to raise other bodies the same
number of degrees/' (MS. p. 27.)

This clear and complete definition of the fundamental law of specific
heat, is followed by the account of experiments of the same kind as those
previously described, made with hot mercury and solutions of sea-salt and
pearl ashes, as well as diluted oil of vitriol; and that is succeeded by a
table of similar trials made upon solids. In these observations water
was taken at an average temperature of about 1 50°, whilst iron filings,
lead shot, tin in fragments, sand, white glass, white marble, brimstone,
Newcastle coal, and charcoal, at an average temperature of 50% were
respectively added to the hot water.

An endeavour was made to determine the specific heat of air by blow-
ing it from a smith's bellows, through the worm of a still surrounded by
hot water. This method was identical in principle with that followed
at a later period by Delaroche, Berard, and others, except that in
Cavendish's trials the air was colder than the water, whilst in later
researches the gas has been raised in temperature, and made to heat, not
to cool, the water. " The quantity of air blown through the worm was
216^ oz., and was heated 63° in passing through; therefore it requires as
great a quantity of heat to raise 216^ oz. of air, 63 degrees by the ther-
mometer, as to raise 2573 oz. of water, 0°*96, and therefore the effect of
air in heating and cooling other bodies is 5'5 times less than that of
water. By another experiment made in the same manner, its effect
seemed D-2 times less than that of water, but the quantity which the

VIEWS ON SPECIFIC AND LATENT HEAT. 449

water was cooled by blowing the air tbrohgli the worm, was so small ia
both these experiments, that one can give bat a very imperfect gaess at
how much its effect is." (MS. p. 32.) Many years had to pass before
more accurate results were obtained in reference to the specific heat of
any of the gases, including air.

The passage last quoted, is the concluding one of Cavendish's obser-
vations on specific heat. Page 33 of the quarto MS. is headed Part 2nd,
and is devoted to latent heat.^ This subject is introduced by the following
statement : ^^ As far as I can perceive, it seems a constant rule in nature,
that all bodies in changing from a solid state to a fluid state, or from a
non-elastic state to the state of an elastic fluid, generate cold, and by the
contrary change they generate heat.'* — MS. p. 33. Cavendish then proceeds
to explain that Dr, Cullen has sufficiently proved that most, if not all,
liquids generate cold, when evaporated at a less heat than that which is
snflScient to make them boil;f whilst ''there is also a circumst-ance daily
before our eyes, which shows that water generates cold/' when heated till
it passes into ebullition. The circumstance to which Cavendish refers, is
the same which fixed Black's attention, and led him to experiment on
what he termed the conversion \)i sensible into latent heat during evapo-
ration; namely, ''that water as soon as it begins to boil, continues
exactly at the same heat, till the whole is boiled away, which takes up a
very considerable time. No reason, however, can be assigned why the
fire should not continually communicate as much, or nearly as much heat
to it after it begins to boil, as it did when it wanted not many degrees of
boiling; and yet during all this time it does not grow at all hotter. This
I think shews that there is as much heat lost, or, in other words, as much
cold generated by the evaporation, as there is heat communicated to it by

the fire If no cold was produced by the evaporation, the

water should cither grow hotter and hotter the longer it boiled, or else
it should be entirely converted into steam immediately after it arrived at
the boiling point." — MS. p. 34. From this consideration the conclusion
is generalised, that all volatile solids and liquids must generate cold when
distilled, for they all require a long time after they reach their point of
evaporation or ebullition, before they are completely volatilised.

These views pointed to a method by which it might be ascertained
what " was the quantity of cold produced," or, as Black would have said,
what was the quantity of heat rendered latent during the ebullition of
water. To determine this, it was only necessary to ascertain at what
rate a given weight of water rose in temperature when exposed to a
(hypothetically) uniform source of heat, from a given thermometric de-
gree lower than 212% till it began to boil; and thereafter, to estimate
the addition of heat to the boiling water, as proceeding at the same rate
as it did before the liquid passed into ebullition. The source of heat
which Cavendish employed, was a compound spirit lamp, so arranged that
he could kindle seven or fewer wicks at pleasure. It was placed below a

* Cavendish does not xue the term * latent heat/ bat prefers, as will be seen in
the sequel, to speak of the ' generation' of heat and cold. His views on this point will
be considered more ftdly in the notice of his pap«r on the freezing of mercury, where he
states them at length.

f Cullen's paper first appeared in the Edinr, Physical and Literary Ettayt, vol. ii.
p. 159. This was published in 1770, and determines the date of Cavendish's observa-
tions, discussed in the text, as later than that year. It was afterwards reprinted along
with Black's ** Experiments upon Magnesia Alba, &c.,'* with the tide, '* An Essay on
the cold produced by evaporating fluids, and of some other means of producing cold."

2 O

450 CAVENDISH AS A NATURAL PHILOSOPHER.

tin vessel containing a weighed qaantitj of water, and enclosed in a me-
tallic case to prevent loss of heat. A thermometer in the liquid marked
its temperature at the heginning of the experiment; and the time which
the water took to rise a given number of degrees after the lamp was
lighted, was carefully watched, up to the point of ebullition. The liquid
was allowed to boil some twenty minutes, and the vessel being uien
weighed, the loss in weight gave the amount of water vaporised. The
result in the first recorded trial, with four wicks of the lamp lighted, was,
that " there is as much heat lost by converting any quantity of water
into steam, as is sufficient to raise that quantity of water 982% or in other
words, there are 982° of cold generated by converting water into steam.^*
—(MS. p. 39).

Dr. Black had observed, as Cavendish was informed, ^Hhat in distil-
ling water, the water in the worm tub is heated thereby much more than
it would be by mixing with it a quantity of boiling water equal to that
which passes through the worm. Upon this principle I made some expe-
riments to determine how much heat is generated by converting water
from the state of an elastic to that of a non-elastic fluid." — (MS. p. 40).
" The heat produced by the condensation of the vapours of boiling water,
by a mean of several experiments tried in the foregoing manner was about
920% so that it seems likely that there is just as much heat prodnoed
by the condensation of steam into water, as there is cold by the changing
of water into steam." — (MS. p. 41.)*

The account of these trials is followed by that of an entirely original
series of experiments undertaken with a view " to find whether any cold
was generated by the emission of fixed air in dissolving alkaline sub-
stances in acids. The way,'* coutinues Cavendish, *' I tried it, was by
finding how much more heat was produced by saturating soap-leys, spirits
of sal ammoniac made with lime, and lime slaked with water (all which
substances contain no fixed air), with spirit of salt, than by saturating
the same substances saturated with fixed air, that is, a solution of pearl
ashes, the mild spirits of sal ammoniac and whiting mixed with water,
with the same acid« By a comparison of the experiments, it seemed
that the cold generated by the emission of the fixed air, was sufficient to
heat a quantity of water equal in weight to the fixed air emitted, about
1000 or 1700 degrees.''— (MS. p. 42.)

Cavendish now passes from the heat of gaseity to that of liquidity,
and records a series of ''experiments to show that bodies, in changing
from a solid state to a fluid state, produce cold ; and in changing from a
fluid to a solid state, produce heat." — (MS. p. 43.)

These are introduced by a statement that the propositions in question
may be proved by a reference to the long time required to thaw ice, or
to freeze water, and also by the phenomena which attend the congelation
of this liquid, when cooled some degrees below 32° without freezing, and
then agitated. In these circumstances it suddenly begins to freeze, and
its temperature immediately rises to the freezing point. A direct trial

* From the OctaTO Notes, which fonned the basis of the Quarto MS., we learn some
of the numbers from which the mean given in the text was calculated. At page 68
(8vo), occurs the entry, " Therefore heat gen. by condensation of vapours = 923*."
At page 70 there is a similar entry, *' = 941°"; and at page 71 another, ** = 942*."
Despretz made the latent heat of water-vapour 955^.8; Dulong, 977°.4; and one of
the latest and most accurate observers, Brix of Berlin, makes it 972°. [GraJkam'e
Chemittry, p. 57.) Cavendish, as already mentioned, obtained 982* as the result of
his observation on the rate at which boiling water acquired heat.

EXPERIMENTS ON LATENT HEAT. 451

]«( tben described, in wbich "the quantity of cold produced bj the
changing of snow into water" was determined, " by dissolving a given
quantity of snow in warm water. The cold produced seemed to be
about 170 degrees. There seemed no difference between the cold pro-
duced by snow, and by the same quantity of ice." — (MS. 44.)*

Reference is then made to the cold produced by mixing salt and snow.
'* There can be no doubt," Cavendish observes, " but this increase of cold
is owing to the melting of the snow," (MS. p. 45.); and then he alludes
very briefly to some trials made to determine the quantity of cold pro-
duced by mixing snow with different liquids, which did not lead to any
very precise result. He was more successful with other liquids. " I find
also,** says he, '* that cold is generated by the melting, and heat by the
hardening of spermaceti. The cold produced by the melting of spermaceti
is sufiicient to cool a quantity of water equal to it in weight, about 70
degrees, and nearly the same degree of heat is produced by the hardening
of spermaceti." (MS. p. 45.)

The last series of recorded experiments has the most direct bearing
on his published papers on heat. They were made with the more fusible
metals, taken singly and in combination. "Some tin and lead were
melted separately in a crucible, and a thermometer put into them and
suffered to remain there till they were cold. The thermometer cooled
pretty fietst till the metal began to harden round the edges of the pot ; it
then remained perfectly stationary till it was all congealed, which took up
a considerable time. It then began to sink again. On heating the metal
with the thermometer in it, as soon as the metal began to melt round the
sides, the thermometer became stationary as near as I could tell, at the
same point that it did in cooling, and remained so till it was entirely
melted. On putting a thermometer into melted bismath, the phenomena
were the same, except that the thermometer did not become stationary
till a good deal of the metal was hardened, unless I took care to keep
the thermometer constantly stirring about. It then remained stationary
till it was almost all hardened.

" I do not know what this difference between bismuth and the two other
metals [lead and tin] should be owing to, except to its not transmitting
heat so fast as them. I forbear to use the word conducting, as I know
you have an aversion to the word, but perhaps you will say the word I
use, is as bad as that I forbear." (MS. pp. 46, 47.)t

Similar observations were made with mixtures in various proportions, of
lead, tin, and bismuth. The mixtures were found to "begin to abate of their

* In the Octavo Notes, different experiments made by melting ice and snow with hot
water are recorded. In these the cold generated, in other words, the latent heat of ice (or
snow) IS 154^ 151°, 142°, and again 154^ (MS, Svo. pp. 86—88.) In his published
views on heat, Cayendish preferred the mean of those numbers to 170°> which he
^ves in the Quarto Manuscript, apparently from a single experiment recorded in the
notes as yielding 171°.

+ The party thus addressed is not known, and cannot be guessed at Cavendish
admitted very few to his intimacy; but even among those few, it would be impossible to
assign a satisfactory reason for selecting one rather than another, as the probable object
of his confidence, and even deference, on the point referred to in the text. It is quite
possible that the paper, though intended speciislly for some one, was never sent to him.
It is difficult to account for its existence among Cavendish's papers, if it had been
despatched to some friend. A copy, however, may have been made, and the original
retained, and perhaps this is rendered probable by the incomplete condition in which
the Quarto Manuscript remains, so far as its conclusion is concerned. This is referred
to more particularly a little further on.

2o2

432 CAVENDISH AS A NATURAL PHILOSOPHER.

fluidity in a heat considerably greater tban that in which they grow hard,
whereas in the simple metals, I could not perceive any difference between
the heat in which they ceased to be perfectly fluid and that in which
they hardened

''As soon as the metal began to abate of its fluidity the thermometer
began to sink, extremely slow in comparison of what it did before, and
continued to do so till it was taken out, so that I think there can be very
little (MS. p. 48,) doubt but what these metallic substances generate heat
in hardening, as well as the simple metals." (MS. p. 49.) Cavendish then
mentions his belief that the reason why the thermometer never became
stationary in the melted alloys, was the separation of their constituent
metals from each other, when solidification commenced. He demonstrated
the truth of this view by showing that when an alloy of lead and tin was
allowed to cool slowly, and cut into two pieces, the lower half had a
higher specific gravity, and evidently contained more lead than the upper.
(MS. p. 49.) This concludes the quarto MS. which ends abruptly thus:
"To understand this yoU must read the following proposition." (MS. p. 50.)*

Such is a short account of Cavendish's unpublished researches on heat
That he was unacquainted with BlacVs contemporaneous researches, ex-
cept to a small extent, is evident from the tenor of his remarks. On this
subject, Mr. Harcourt, after stating the number which Cavendish assigned
to the latent heat of steam, offers the following commentary. "At what
date Black and Watt arrived at a similar result, I know not. Nor do I
know the precise year in which Black first taught the doctrine of specific
heat. Dr. Thomson says, ' that the specific caloric of bodies is different,
was first pointed out by Dr. Black in his lectures at Glasgow between
1760 and 1765. Dr. Irvine afterwards investigated the subject between
1765 and 1770 (Black's Lectures, i. 504), and Dr. Crawford published a
great number of experiments on it, in his treatise on heat (1779), but
Professor Wilcke, who published the first set of experiments on the sub-
ject {Stockholm Transactions, 1781), introduced the term, specific caloric'
'I have been informed,* he adds, *by the late Professor Robison, that
Wilcke's information was got from a Swedish gentleman who attended
Dr. Black's lectures, about 1770.' It appears probable, from what I have
stated, that this unpublished series of experiments by Cavendish is tfie
first made upon this subject. After these and immediately preceding
those of the date Feb. 1765, is Cavendish's determination of the number
of degrees of 'cold gen. by thawing ice or snow,' which he found on an
average to be 150°. In the account which Black gives in his lectures,
of his determination of the quantity of heat absor^d in the melting of
ice, he says, 'These two experiments, and the reasoning which accompanies
them were read by me in the Philosophical Club or Society of Professors
in Glasgow in the year 1762.' "f This jpassage has given offence to some
of the advocates of Watt ; for it is one of the unfortunate secondary resalts
of the Water Controversy, that in the anxiety on the one side to exalt, and
on the other to depreciate Cavendish's merits, he has been contrasted and

* In the parcel which oontains the Quarto Manuscript, are three fragments on heat,
two of which Mr. Harcourt has printed. {Brii. Aswc, Rep, 1839, p. 47.) They are only
repetitions, however, with slight variations of part of the contents of the Quarto Manuscript,
and I have not thought it necessary to reprint them. The third fragment is entitled, " A
compleat proof that the quantity of heat in different bodies, at a given temperature, is
not in proportion to their specific heats." From the employment of the term apecifie
heat, which Cavendish never employs in his other MSS., I think this fragment must
be of much later date than the Octavo Notes or Quarto Manuscript.

+ Brit. Atioe. Rep. 1839, p. 46.

OBSERVATIONS ON SPBCIFIC HEAT. 453

compared with other natural philosophers, hut especially with Black, for
whom Lord Brougham has claimed discoveries which the opponents of
Watt attribute to Cavendish, whilst Mr. Harcourt, in the passage just
quoted, assigns Cavendish some share in a discovery which others regard
as due only to Black. In defence of the latter, Mr. Muirhead quotes a
passage from an unpublished letter of Black to Wait, of 15th May, 1780,
which remains among the papers of the latter. " I began," says Dr. Black,
'' to give the doctrine of latent heat in my lectures at Glasgow, in the
winter 1757-58, which I believe was the first winter of my lecturing
there, or, if I did not give it that winter, I certainly gave it m 1758-50,
and I have delivered it every year since that time in my winter
lectures, which I continued to give at Glasgow until winter 1766-67, when
I began to lecture in Edinburgh."* After adducing this passage, Mr.
Muirhead adds in a note: — " Preposterous pretensions have also been, by
insinuation, set up for Cavendish to the discovery of the same theory;
pretensions which are quite unfounded. See p. 30 of the Birmingham
address of the Rev. W. V. Harcourt.**f Mr. Harcourt^ however, neither
states, nor insinuates that Cavendish had anticipated Black in recognising
the existence of latent heat in fluids. He has drawn attention, three
times over, in the context of the passage quoted, to Cavendish's acquain-
tance with Black s observation " that in distilling water and other liquors,
the water in the worm -tub is heated thereby much more than it would be
by mixing it with a quantity of boiling water equal to that which passes
through the worm." j: What Mr. Harcourt suggests is, that Cavendish may
have been the first to determine, not the presence, but the amount of latent
heat in steam, and further that he may have been earlier than Black, or
his pupils, in prosecuting experiments on the specific heat of bodies. The
letter quoted by Mr. Muirhead refers only to the general doctrine of latent
heat, and is silent on the subject of specific heat, so that it brings to light
nothing incompatible with Mr. Harcourt's cautious and temperate sugges-
tions. A peculiar difficulty, indeed, attends the determination of all the
dates of Black's discoveries, for his posthumous lectures, edited by Pro-
fessor Robison, are frequently in error as to dates (doubtless accidentally),
and this to the disadvantage of Black himself. It is often, accordingly,
impossible to determine within several years, when a particular series of
researches was made, so that we cannot do justice either to Black or
to bis rivals. From a statement in Professor Robison*s preface to Black's
Lectures, it appears that *' it was late before he [Black] had made such
experiments as satisfied him in respect to the precise quantity of the heat
latent in steam, not till the summer of 1764," (page XLlL); so that
Cavendish's determination of that point cannot be considered as earlier
than his. It would appear, however, to have been contemporaneous, nor
is it in any degree probable that Cavendish was aware of Black's observa-
tions, for he so frankly refers to him in various of his papers, both pub-
lished and unpublished, that we may be certain that he would have alluded
to his determination of the latent heat of steam, had he been acquainted
with it. That he does not, will be sufficient to convince all candid judges
of his ignorance, and if Black's ipse dixit is to be accepted as settling the
nature and date of his observations, (which is Mr. Muirhead's argument) then
Cavendish's ip^Jm^ should also be accepted as deciding the nature and date
of his. It may be added, as already incidentally mentioned, that it is im-

* Wait Corr. Introd. p. xxiii. t Wati Corr. p. xiiv.

X Brit. As90C, Rep, 1839, p. 48.

454 CAVENDISH AS A NATURAL PHILOSOPHER.

poflsible to imagine that Cayendisli should have taken the trouble he did,
to make observations which he would not and did not publish, could he
hare obtained the information which they taught him, accredited bj one
whom he respected so much as he did Black. Further^ it is certain that,
except the latter s students in the University of Glasgow, very few pre-
vious to 1766 had the means of learning what discoveries Black had
made; and how imperfect a medium of publication university lectures are,
we see proved at the present day, when news of all kinds travel so much
more rapidly than they did eighty years ago, by the difficulty which pro-
fessors, who confine the publication of their views to their unprinted lectures,
have had in vindicating their priority against rivals at no great distance
from them, to whom these prelections were unknown. This remark applies
with peculiar force to Black's earlier lectures, when he was little known out
of Scotland as an authority on heat, and it is specially pertinent in reference
to one who led so retired a life as Cavendish did. We shall presently find,
also, that Cavendish ultimately published some of his observations on latent
heat, with an incidental object in view, and that though pointedly referring
to Black, whom he had plainly no thought of rivalling, yet he spoke of the
researches as original on his own part.* So far as regards specific heat,
there exists no room for doubt, that had Cavendish published, early in
1765, his views on it, he would now be reputed one of its independent
discoverers.

I have made these remarks chiefly because it might otherwise seem
as if Cavendish actually wronged Black; not with any purpose of putting
him forward as a rival of that great chemist. Cavendish voluntarily
suppressed all, or nearly all publication of his experiments on heat; and
it would be idle now to put forward claims for him as a discoverer in
that department of science. But it is only justice to him, and it is no
injustice to Black, to refer to the beautiful and independent researches
of the former, as evidences of his ingenuity, capacity, perseverance, and
accuracy, when seeking to form an estimate of his intellectual character.
But for the statements of Mr. Muirhead, I should have referred to his
unpublished experiments on heat in no other light than in this.

Thus much premised, I now proceed to the consideration of Caven*
dish's published papers on heat, and first, of that published in 1776,
entitled "An Account of the Meteorological Instruments used at the
Royal Society's House."t

This paper is occupied with an account of the thermometers, baro-
meter, rain-guage, hygrometer, variation-compass and dipping-needle,
then in use at the Royal Society's apartments. As all those instruments
have long been replaced by others constructed with the improvements
which later discoveries have suggested, I have thought it needless to
make a formal analysis of the entire paper.}: The only part of it which
concerns us in this section, is that treating of the Society's thermometers;
and this only in so far as it refers to the best method of constructing

* Phil. Trans. 1783, p. 312.

+ Read to the Royal Society, March 14th, 1776. Phil. Trans, vol. Ixvi. p. 375.

X It may also be added, that the Royal Society has, since 1843, ceased to prosecote
systetaatic meteorological observations. They were commenced in 1773, and carried on
without interraption till 1843, when they were discontinued, in consequence of the
Government, on the recommendation of the President and Council of the Royal
Society, having established at the Royal Observatory at Greenwich, under the super*
intendence of the Astronomer Royal, a magnetical and meteorological obtervatory,
where observations are made on an extended scale, which are regularly published.
Phil. Trans, for 1844, and WelcTs Hist, of Royal Society, vol. ii. p. 76.

GRADUATION OF THERMOMETERS. 45$

such instruments. To this subject Cayendish devoted much attention,
and he was the means of introducing methods of accurate construotiouy
which have since been universally followed. The accuracy of all
Cavendish's observations on heat was no doubt in great part owing to
the care which he evinced in selecting or constructing the thermometers
he employed. Before his time, even so-called standard instmments, as
they were made in England, were frequently in error two or more degrees*
This he showed to be mainly owing to the inaccurate way in which the
degree of boiling water was ascertained. Two points in particular were
neglected even by the best artists. To the first he refers in the follow-
ing terms: — " It has been too common a custom, both in making experi-
ments with thermometers and in adjusting their fixed points, to pay no
regard to the heat of that part of the quicksilver which is contained in
the tube, though this is a circumstance which ought by no means to be
disregarded; for a thermometer dipped into a liquor of the heat of
boiling water, will stand at least 2° higher, if it is immersed to such a
depth that the quicksilver in the tube is heated to the same degree as that
in the ball, than if it is immersed no lower than the freezing point, and
the rest of the tube is not much warmer than the air. The only accurate
method is to take care that all parts of the quicksilver should be heated
equally. For this reason, in trying the heat of liquors much hotter or
colder than the air, the thermometer ought if possible to be immersed
almost as far as to the top of the column of quicksilver in the tube."*
As the entire immersion, however, of the thermometer into boiling liquids
would often have been extremely inconvenient, and indeed impossible, a
table is subjoined by means of which the difiepence between the indicated
and the real temperature may be ascertained and allowed for ; the table
being constructed on the datum, ''that quicksilver expands 11,500th
part of its bulk by each degree of heat.**

The other circumstance " which ought to be attended to, in adjusting
the boiliug point of a thermometer, is, that the ball should not be im-
mersed deep in the water ; for, if it is, the fluid which surrounds it will be
compressed by considerably more than the weight of the atmosphere, and
will therefore acquire a sensibly greater heat than it would otherwise
do.*'t An apparatus is then described, by means of which both the
sources of inaccuracy pointed out may be avoided. It consists essentially
of a tall metallic boiler filled to a small depth with water, and having
two apertures in its lid; one for the insertion of the thermometer, which is
made steam-tight by corks or tow after the tube has been introduced; and
a second and wider perforation, elongated into a canal, and acting as a
chimney, to carry off the steam, which is allowed to escape by a small
aperture. With this apparatus matters could easily be arranged, so that the
bulb should be but a slight depth below the surface of the boiling water,
whilst the greater part of the stem of the thermometer was surrounded by
hot steam, so that bulb and stem were at the same temperature. In such an
arrangement, the thermometer is found to stand not sensibly higher when
the water boils vehemently, than when it boils gently; and the method
admits of great accuracy in the fixing of the boiling point of water.

These directions were afterwards generaUy sanctioned by the Royal
Society,]; and are still in practice in the construction of standard ther-
mometers; but there is as much reason now as there was in the days of

♦ PML TroM. 1776, p. 376.
t nwU ibid. p. 379. t IM. ibid. 1777, p. 816.

456 CAVfiNDISlI AS A NATURAL PHILOSOPHER.

Cavendieb, to complain of the want of uniformity between professedly
accurate instruments.

The remaining papers of Cavendish on heat will enable us to discoyer
what use he made of his acquaintance with a method for accurately con-
structing thermometers; and how he applied his theory of the nature of
heat, and his observations on its generation and destruction, to the
explanation of remarkable natural phenomena. The first paper I have
to notice is entitled, *^ Observations on Mr. Hutchins*s Experiments for
determining the degree of Cold at which Quicksilver freezes."* To
understand this paper, two communications of Mr. Hutchins's to whicli it
refers, and the preliminary observations which led to his trials, must be
considered. This may be done most conveniently by adducing a few
details from a very elaborate paper which follows Cavendish's in the
73rd vol. of the FhiL Trans., entitled " History of the Congelation of
Quicksilver, by Charles Blagden, M.D., F.R.S., Physician to the Army."

From the time of the alchemists downwards, chemists had been in the
habit of regarding quicksilver as a quite peculiar substance which differed
from all the other metals, in being liquid at ordinary temperatures, and
appeared possessed of an essential principle of fluidity. These views
were first overturned by Professor Braun of Petersburgh, who, after his
removal to that city, prosecuted an inquiry which had already been
commenced in Germany, on the action of freezing mixtures, and the
lowest temperature which they could secure. Availing himself of
the cold which characterises a Russian winter, Professor Braun selected
a day (14th December, o.s., 1759) when the thermometer in the open
air stood so low as —34° F., and made a freezing mixture with nitric
acid and snow, into which he plunged his thermometer. He was aston-
ished to perceive it rapidly sink to —69^, which was lower by almost
thirty degrees than it had fallen in any preceding experiment then on
record. It did not stop however here; for, by using fresh supplies of the
freezing mixture, the mercury descended to —100®, —244°, and —352**.
At this stage in the inquiry, Braun removed the thermometer, and
was astonished to find the mercury quite fixed and immovable. He
repeated the experiment some days later, and as soon as the quicksilver
became immovable, broke the bulb of his thermometer, and " obtained a
solid, shining, metallic mass, which extended under the strokes of a
pestle; in hardness rather inferior to lead, and yielding a dull dead sound
like that metaL"f The conclusions which these observations justified,
viz. that mercury can be solidified, and that it is a true metal which
melts with a very small degree of heat, were clearly perceived by their
first observer. He failed, however, to see the significance of another
phenomenon which he observed, viz. that mercury contracts and becomes
denser when congealed, so that the solid metal sinks in the liquid.
Making no allowance for this, Braun supposed the descent of the mercury
in the stem of the thermometer to be solely due to the contraction occa-
sioned in the metal whilst still liquid by its reduction in temperature;
and, in consequence, he set down the freezing point of mercury as some
hundred degrees below Fahrenheit's zero. He was unable, however, to
determine it exactly, " owing to the various impediments that occurred
from adhesion of the quicksilver in the themiometrical tube, hollows left
in the bulb as it froze, portions of the mercury remaining uncongealed^
and many other causes. }:

• Read to the Royal Society, May 1st, 1783. PhiL Trans. 1783, p. 303.
t Phil. Trans. 1783, p. 332. % Ibid. ibid. p. 335.

CONGELATION OF MERCURY. 457

Braun's experiments excited universal interest throughout Europe,
and, in compliance with his suggestion or request, that his observations
might be repeated at Hudson's Bay, the Royal Society procured the co-
operation of Mr. Hutch ins, who tried the necessary experiments at
Albany Fort, Hudson's Bay, in January and February, 1775. He found
no difficulty in freezing mercury, by placing it in a mixture of nitric
acid and snow, when the temperature of the air was — 37° F. The mercury
descended to -— 480°, and, when the tube was broken, appeared in greater
part frozen, and admitted of being hammered.* This was the fullest
confirmation which Braun's experiments had received, but Hutch ins
found it impossible to determine, even within wide limits, the freezing
point of mercury. In 1781, however, he repeated his experiments with
a different apparatus; and this brings us to Cavendish's paper. The
result of Braun and Hutchius's observations had been to show, that it was
vain to attempt to determine the freezing point of mercury by enclosing
it in a graduated thermometer tube, seeing that it was impossible to
determine how much of the diminution in volume of the mercury as it
cooled was owing to contraction before, how much to contraction after,
congelation. Cavendish, accordingly, suggested a very simple apparatus,
which enabled the one phenomenon to be observed without interference
from the other. No attempt was made to freeze the mercury in the bulb
of the thermometer, which was employed solely to indicate the tempera-
ture. It ''was enclosed in a glass cylinder swelled at bottom into a
ball, which, when used, was filled with quicksilver, so that the bulb of
the thermometer was entirely surrounded with it."t The object of this
arrangement was to freeze the quicksilver in the outer cylinder, without
freezing that in the thermometer, and to observe the point at which the
latter stood when the greater part, but not the whoUy of the mercury
surrounding it was frozen. The principle of this method was the fact,
very familiar in our own day, but clearly understood by few at the period
when Cavendish wrote, viz. that when a liquid is undergoing congelation,
its temperature remains stationary, after it begins to freeze, till it is
entirely frozen. This fact, of which Black was the great expounder,
Cavendish, as we have seen, had examined minutely for himself, some
nineteen years before the paper we are considering was written, and
he expounds it in this paper as a thing still requiring exposition, in
the following terms : " If a glass of water, with a thermometer in it, is
exposed to the cold, the thermometer will remain perfectly stationary
from the time the water begins to freeze, till it is entirely congealed, and
will then begin to sink again. In like manner, if a thermometer is
dipped into melted tin or lead, it will remain perfectly stationary, as I
know by experience, from the time the metal begins to harden round the
edges of the pot till it is all become solid, when it will again begin to
descend; and there was no reason to doubt that the same thing would
obtain in quicksilver.

'' From what has been iust said it was concluded, that if this apparatus
was put into a freezing mixture of a sufficient coldness, the thermometer
would immediately sink till the quicksilver in the cylinder began to
freeze, and would then continue stationary, supposing the mixture still to
keep cold enough, till it was entirely congealed. This stationary height
of the thermometer is the point at which mercury freezes, though, in
order to make the experiment convincing, it was necessary to continue

♦ Phil, Trans, 1776, p. 170. t Ibid, ibid, 1783, p. 303.

458 CAVENDISH AS A NATURAL PHILOSOPHER.

the procees till so much of the quicksilyer in the cylinder was frosen as
to pat the fact out of donbt/'*

The apparatus already described was sent by Cavendish to Mr.
Hatchins, who employed it at Albany Fort, in Hudson's Bay, in 1781.
It completely answered the expectations of its deviser, as appears from
Mr. Hutchins's paper, entitled ^' Experiments for ascertaining the point
of Mercurial Congelation, read [to the Royal Society] April 1 0th, IJBS^f
He had every assistance in prosecuting his observations, for, curiously
enough, Dr. Black, without being aware of the process which Cavendish
had proposed in 1776, suggested an exactly similar mode of procedure.
His letter containiug this is dated 1779, and seems to have reached Mr.
Hutchins before Cavendish's communication did.}: It thus appears, that
Hutchins had the advantage of Cavendish and Black's suggestions, and
that the apparatus he employed was furnished by the former, and was
in conformity with the instructions of both the philosophers.

The results of Hutchins's observations are stated succinctly in the
following passage from Blagden's history of the congelation of quick-
silver. " They have not only confirmed the preceding observations
relative to the solid state into which quicksilver can be brought by cold,
its metalline splendour and polish when smooth, its roughness and
crystallisation where the surface was unconfined, its malleability, softness,
and dull souud when struck; but have also clearly demonstrated, that its
point of congelation is no lower than — 40% or rather— 39°, of Fahrenheit's
scale; that it will bear, however, to be cooled a few degrees below that
point, to which it jumps up again on beginning to congeal; and that its
rapid descent in a thermometer through many hundreds of degrees, when
it has once past the above mentioned limits, proceeds merely from its
great contraction in the act of freezing."§ The results thus referred to,
were obtained, both by the artificial congelation of mercury, and also,
and very satisfactorily, by its natural freezing. On 26th January, 1782,
the temperature at Albany Fort was so low, that mercury standing in
the open air was frozen through a considerable part of its mass. ''As
this," observes Mr. Hutchins, " was a certain method to find the point of
congelation, I introduced the mercurial thermometer and the spirit ther-
mometer into the fluid part, after breaking off the top of the phial, and
they rose directly and became stationary, the former at 40° or 40^% the
latter at 29f °, both below the cvpher."||

I now return to Cavendish's comments on ihese observations. A
large part of the former is occupied with the account of an examination
of the accuracy of Mr. Hutchins's thermometers, which were sent home,
and carefully examined as to their graduation by Cavendish. The most
accurate thermometer showed the freezing point of mercury to be — 40°,
and "as it appeared," observes Cavendish, "from the examination of
this thermometer after it came home, that — 40° thereon answers to
— 38f°, on a thermometer adjusted in the manner recommended by the
Committee of the Royal Society, it follows, that all the experiments
agree in shewing that the true point at which quicksilver freezes is d8-|°,

* Phil. Trans, 1783, p. 305.

t This paper will be found at the end of the 73rd vol. of the Phil, TVam, after
the index, where it appears out of its proper place, in consequence of the MS. having
been mislaid. The paging is the same as that of Cavendish's paper in the same volume,
but is distinguished from it by an asterisk at each figure.

t Phil, Traru. 1783, p. *305.

§ Ibid, ibid, p. 346. (I Ibid. ibid. p. *368.

THEORY OF HEAT. 459

or, in whole numbers, 39° below nothing.*'* From later observations it
appears, that the corrected temperature is about half a degree, too high,
and that we may call the freezing point of mercury 39°'5.

The most interesting portions, however, of Cavendish's paper, are
those which contain his speculations on the relation of heat to fluidity.
The first of these refers to some remarkable appearances observed by
Mr. Hutchins in his experimeuts. One of these was, that the thermo-
meter, when plunged into very cold mercury, was occasionally observed
to sink three or four degrees below —40°, the reputed freezing point of
the metal, and then suddenly to rise again to the point from which it fell,
and remain there stationary for many minutes. This phenomenon,
Cavendish infers to depend upon mercury admitting, like water, of
being cooled in certain circumstances, below its freezing point, without
freezing, whilst, if agitated, as by the insertion of a thermometer, it im-
mediately rises to the point of congelation, and remains stationary there
whilst the liquid is freezing. " The cause of the rise of the thermometer,"
he observes, '^ when the water begins to freeze, is the circumstance, now
pretty well known to philosophers, that all, or almost all, bodies, by
changing from a fluid to a solid state, or from the state of an elastic to
that of an unelastic fluid, generate heat; and that cold is produced by
the contrary process, "f

In continuation of these remarks. Cavendish observes, a little farther
on, " that if it was not for this generation of heat by the act of freezing,
whenever a vessel of water exposed to the cold was arrived at the freez-
ing point, and began to freeze, the whole would instantly be turned into
solid ice."t He pursues this idea somewhat further, and adds, '4n like
manner, it is the cold generated by the melting of ice which is the cause
of the long time required to thaw ice or snow. It is this also which is
the cause of the cold produced by freezing mixtures ; for no cold is pro-
duced by mixing snow with any substance, unless part of the snow is
dissolved.'*§ He then refers to the unpublished experiments already
noticed, which led him to the conclusion that the latent heat of ice or
snow, is 150°. In explaining this, as in all his papers on heat, he uses
the terms, production, or generation of heat and cold, in reference to which
he adds the following important note : " I am informed that Dr. Black
explains the above mentioned phenomena in the same manner; only
instead of using the expression, heat is generated or produced, he says,
latent heat is evolved or set free. But as this expression relates to an
hypothesis depending on the supposition, that the heat of bodies is owing
to their containing more or less of a substance called the matter of heat,
and as I think Sir Isaac Newton's opinion, that heat consists in the
internal motion of the particles of bodies, much the most probable, I
chose to use the expression, heat is generated. Mr. Wilke, also, in the
TraTisactions of the Stockholm Academy of Sciences, explains the pheno-
mena in the same way, and makes use of an hypothesis nearly similar
to that of Dr. Black. Dr. Black, as I have been informed, makes the
cold produced by the thawing of snow 140°; Mr. Wilke, 130°."|| This
passage, it will be observed, is closely accordant in its general bearing
with the second interpolation in the '^ Experiments on Air," (First Series,
1784), in which Cavendish declines to adopt Watt's phrase of elementary
heat.

• PhU. Tram. 1783, p. 321. f Ibid. ibid. p. 311.

t Ibid. ibid. p. 312. § Op. et he. cit.

il Ibid, ibid. Foot note, pp. 312, 313.

460 CAVENDISH AS A NATURAL PHILOSOPHER.

Tbe reference to tbe generation of cold by the liqnefaction of snow,
is followed bj an allusion to the unpublished experiments alreadj consi-
dered, on the solidification of melted tin and lead, and the coDclnsion Li
drawn '' that those metals, as well as water and quicksilver, may bear
being cooled a little below the freezing or hardening point (for the harden-
ing of melted metals and freezing of water seems exactly the same process)
without beginning to lose their fluidity."*

The succeeding part of the paper contains a minute and careful criti-
cism of all Mr. Hntchins*s experiments, which does not call for special
reference, as the result at which he arrived has long been UDiversally
credited by men of science.

Another point, however, required examination. The mercury; as we
have seen, appeared to sink four or five hundred degrees below zero,
although in reality the temperature was little below — 40^ Cavendish
has added a sequel to his paper referring to this. It is entitled, '' On the
contraction of quicksilver in freezing," and the general purport of it may
be gathered from the first sentence, which is as follows. ''All tbe^e ex-
periments prove that quicksilver contracts or diminishes in bulk by
freezing, and that the very low degrees to which the thermometers have
been made to sink, is owing to this contraction, and not to the cold having
been in any degree equal to that shewn by the thermometer. ''f The ex-
tent to which mercury contracts. Cavendish inferred to be "not much less
than its expansion by 500° or 510° of heat, that is, almost one twenty-
third of its whole bulk.*' He thought it likely, however, that congealed
mercury might vary in density like other metals solidified from a state of
fusion, especially when frozen in the stem of a thermometer, and he had
no means of ascertaining the amount of this variation.

The concluding part of this paper discusses the cold of the freezing
mixtures employed to congeal mercury. This subject Cavendish pnrsued
in other papers, where his views are displayed more fully. His chief object
is to shew that the cold produced by a mixture of nitric acid with snow,
which was the freezing mixture made use of, is owing to the melting of the
snow; but that in all probability there is a certain degree of cold in which
the nitric acid, so far from dissolving snow, will yield part of its own water,
and suffer that to freeze, as is the case with solutions of common salt.
And as the veneration of cold, or reduction of temperature can only be
determined by the liquefaction of the snow, the cooling must be arrested,
if the opposite process, that, viz. of the freezing of veater, is going on. It
should thus happen that the warmer, within certain limits, the materials
are, the greater will be the additional cold produced, and such was the
case in Mr. Hutchins's observations, and also in experiments made by
Cavendish himself.

He then notices that, paradoxical though it may appear, nitric acid,
somewhat diluted, produces a greater cold when mixed with snow, than
stronger acid does. He points out very lucidly the cause of this
anomaly, viz. that there is a point, up to which nitric acid generates heat
by combining with water, and that it is not till this degree of dilution has
been attained, that cold is produced. In other words, as we should say
at the present day, strong nitric acid combines with water in several pro-
portions to form definite hydrates, and during this chemical combina-
tion, heat is evolved ; but when these hydrates are added to snow, such
combinations are not formed, but a simple solution of the snow occurs,

* PML Trent. 1783, p. 313. f Ibid, ibid. p. 322.

THEORY OF FREEZING MIXTURrrS. 461

and it renders much heat latent, or generates much cold during its
liquefaction. In conformity with these views^ Cayendish recommended
that in using nitric acid and snow as a freezing mixture, the snow should
be gradually added to the strong acid, till the mixture ceased to rise in
temperature, hefore any hody to he cooled was placed in it ; aud that
then a large addition of snow should he quickly made. By following
this method he succeeded, for the first time in this country, in freezing
quicksilver. This was effected in February, 1783, when the temperature
of the air was as high as 20° or 25° F.

The publication of Hutchins's and Cavendish's observations, besides
establishing incontestably the freezing point of mercury, and the limit
downwards to its use in thermometers, put an end to many extravagant
speculations which had been afloat concerning the enormously low tem-
perature of the colder regions of the globe. These were not only repre-
sented as colder by many hundred degrees than the more temperate
countries, but the vicissitudes of their climates were declared to be such,
that the thermometer might fall in an hour some four or five hundred
degrees. Such opinions, founded on a misrepresentation of the cause of
the contraction which accompanies the solidification of mercury, neces-
sarily carried error into every branch of physical science, but especially
into physiology, and natural history. It now appeared that the enor-
mous natural and artificial colds which had been believed in, had no
existence; and Black and Cavendish were both able to predict, even
before Mr. Hutch ins*s experiments were made, that what the one called
the " evolution of sensible heat,'' and the other simply the " generation of
heat," rendered it impossible that there could be such rapid and enormous
reductions of temperature as Braun thought he had observed.

The remaining papers by Cavendish on heat are two upon freezing
mixtures, in continuation of the views contained in the concluding portion
of the paper which has just been noticed. The first is en titled " An
account of Experiments made by Mr. John McNab at Henley House,
Hudson's Bay, relating to Freezing Mixtures. Read [to the Royal
Society] February 23rd, 1786."* The experiments recounted in this
paper were made at Cavendish's request, and according to his directions,
with a view chiefiy to test the truth of the opinion he had published
concerning the cause of the cold produced by mixing snow with different
liquors. It has already been mentioned that he thought it probable, that
at certain temperatures nitric acid, instead of dissolving snow, would part
with a portion of its water of combination, which would become frozen,
80 that if the temperature of the materials was equal to this, no additional
cold could be produced. There seemed an objection, however, to the
truth of his views in a fact observed both by Fahrenheit and Braun, viz.
that even frozen nitric acid will produce cold if mixed with snow.
Cavendish accordingly thought it desirable to ascertain whether it was
possible to freeze the entire mass of the acid, or at least the more concen-
trated part, without separation of water from it, in the expectation that
such congealed acid might dissolve snow aud generate cold in so doing.
To determine this, and some other questions, he sent to Hudson s
Bay bottles of strong and of diluted nitric acid, of oil of vitriol, and of
diluted alcohol. The liquids were ordered to be exposed to the cold, and
if they froze the temperature was to be ascertained; the more fluid portion
which might appear was to be decanted into another bottle, and each

♦ Phil. Trans, 1786, p. 241.

462 CAVBUDISH AS A NATURAL PHILOSOPHER.

portion sent home so that it might be ascertained by their examination, .
whether the original liquid had frozen as a homogeneous bodj, or had
separated daring congelation into dissimilar parts.

CaveDdish also, it will be remembered, drew attention to the fact that
snow in certain circumstances produces heat, not cold, when added to
nitric acid, and that when employing those materials for freezing mixtures,
the snow should be added gradually to the acid till it begins to fall in
temperature, and then rapidly so as to secure its maximnm effect in
producing cold. By the extension of this method he thought a greater
cold might be produced than had been observed before; and for this
purpose he furnished Mr. McNab with bottles of nitric acid and oil of
yitriol, wHich had been diluted nearly to the point at which they began
to generate cold, when mixed with snow.

Another question sought to be determined was, the lowest point to
which thermometers filled with other liquids than mercury, could be made
to sink. According to Professor Braun, nitric acid and snow caused
thermometers filled with oil of sassafras, and other essential oils, to fall to
— 100^ or — 124°, whilst a thermometer filled with the most rectified
spirits of wine sank to — 1 48°. To test the truth of these statements, a
thermometer filled with oil of sassafras, and two others with spirits of
wine were furnished to Mr. McNab. The first of those instruments was
early deranged by the appearance of a large air-bubble in its ball, so that
no further observations could be made with it. The spirit of wine
thermometers could by no arrangement of matters be made to sink " by a
mixture of snow and spirit of nitre, to a degree near approaching to that
mentioned by Professor Braun."

I now notice briefly the results of these different inquiries. Nitric
acid was found capable of a kind of congelation in which the whole, and
not merely the watery part, freezes. Its freezing point was found to differ
greatly according to its strength, and this variation was found to follow a
very unexpected law. The acid was also found to admit of being cooled
very much below its freezing point without congelation occurring, but
when that did happen the temperature immediately rose to the freezing
point. During congelation great contraction occurred, so that the crystals
which first formed sank rapidly to the bottom, and when the whole liquid
was frozen solid, the surface was depressed and full of cracks.

On mixing snow with nitric acid, results were obtained considerably at
variance with those anticipated. In two experiments, the addition of snow
produced heat, until the mixture arrived pretty exactly at what was found
to be the freezing point of the diluted acid; but as soon as it reached that
point the addition of more snow began to produce cold. The cause of
this, Cavendish inferred to be the following: The freezing point of nitric
acid, when diluted to a certain extent, is much less cold than when it is
considerably more diluted, and also than when it is not diluted at all; so
that there must be a certain degree of strength, not very different from
that to which these acids were reduced by dilution, at which they freeze
with a less degree of cold than when they are either stronger or weaker.
The acids had thus what Cavendish called " a point of easiest freezing;"
and he shows that in the experiments referred to, "when they were
diluted to the strength of easiest freezing, they would also be at the heat
of easiest freezing." For if the liquid was below that point, so much of
the acid would immediately freeze as would raise it to it, and if the
liquid was above that point, so much of the congealed acid would dissolve
as would sink it down to it.

CONGELATION OP OIL OF VITRIOL. 463

Id farther trials the nitric acid was largely diluted with water, and
"when exposed to a low temperature was found to congeal in part, so as
to present the appearance of flakes or spiculfe of ice, floating through a
syrupy liquid. This liquid was decanted from the crystals, and its
strength determined by saturating it with marble; and the crystals, after
liquefaction, were treated in the same way. They were found to consist,
in greater part, of pure water, whilst the unfrozen liquid was strongly
acid. From these experiments Cavendish drew the conclusion that nitric
acid, or as he called it spirit of nitre, *' is subject to two kinds of conge-
lation, which we may call the aqv>eoiu and the spirituous; as in the first
it is chiefly if not entirely the watery part which freezes, and in the latter
the spirit [i. €, the acid] itself. Accordingly, when the spirit is cooled to
the point of aqueous congelation, it has no tendency to dissolve snow and
produce cold thereby, but on the contrary is disposed to part with its own
water; whereas its tendency to dissolve snow and produce cold, is by no
means destroyed by being cooled to the point of spirituous congelation or
even by being actually congealed.'*

Experiments of a similar nature were made with oil of vitriol, which
was frozen at a temperature of —15^. It was allowed to melt partially
in a warm room, and the fluid part was then decanted, but was not found
to differ sensibly in strength from the undecanted portion of the acid. In
other experiments snow was added to the diluted oil of vitriol, ajs in the
preceding experiments to the nitric acid, till it began to fall in tempera-
ture, which it did speedily. When snow was then rapidly added to it,
the temperature of the air being —39'', the mixture sunk to —5S\^, In
another trial the thermometer descended to — 68^°, and in a third to
— 78i^, a greater cold than had certainly been produced up to this time.
Oil of vitriol was also shown to be capable of the spirituous congelation,
and to freeze with a less degree of cold when strong than when much
diluted; but Cavendish could not at this time make out that it had any
point of easiest freezing. During the congelation, however, of oil of
vitriol, he thought, both from his own experiments and from those of
others, some separation of its parts took place, so that the congealed
portion difiered in some respect from the rest, and in consequence froze
with a less degree of cold ; and as he could detect no difference between
the strength of the congealed and uncongealed portions of the oil of vitriol,
he thought it must be owing to the presence of some peculiar substanc^-^
probably that which makes glacial sulphuric acid difier from common oil
of vitriol. He then refers to the separability from the glacial acid, by
the gentlest heat, of '' a peculiar concrete substance in the form of saline
crystals;" but that this was anhydrous sulphuric acid, and that glacial
and common oil of vitriol difiered from each other only in the proportion
of this substance which they contained, he was not aware. A reference
is then made to the action of a mixture of oil of vitriol and spirit of nitre
on snow, which does not call for special notice, and the paper concludes
with an account of experiments on the congelation of spirits of wine.-
Little cold was found to be produced when the spirits were mixed with
snow. When diluted spirit was exposed to the natural cold of the atmo-
sphere, it froze in part. The congealed portion consisted in greater part
of water, what had not frozen retaining the spirit.

Not satisfied in some respects with the results we have been consider-
ing. Cavendish, with that extraordinary love of accuracy which he carried
into every inquiry, had a second series of trials made by Mr. MoNab.
The report of these is entitled, '' An Account of Experiments made by

464 CAVENDISH AS A NATURAL PHILOSOPHER.

Mr. John McNab, at Albany Fort, Hudson's Bay, relative to the Freeiing
of Nitrous and Vitriolic Acids. Read [to the Royal Society] Febmajy
28th, 1788.*** Cavendish begins this paper by statin? that, from Mr.
McNaVs previous experiments, he had drawn the conclusion, that when
nitric acid is of such strength as not to dissolve ya^ of its weight of
marble, or when its strength is less than '243, as he calls it for shortness-
sake, it is liable to the aqueous congelation solely; in other words, the water
in it freezes, but not the acid. But if of greater strength, it undergoes
the spirituous congelation, or the acid itself freezes. The latter congela-
tion occurred at the highest temperature, when the strength was '411, in
which case the freezing point is at— 1^^. If the acid is either stronger
or weaker, it requires a greater degree of cold, so that, for example, if its
strength be such as *54, the freezing point goes down to — 31^®; and if it
be so much diluted as to have the strength -243, it goes down to 44^''.
Cavendish, however, as he tells us, thought some of these conclusions
" were deduced from reasoning notsufficiently easy to strike the generality
of readers with much conviction;*' and therefore he desired Mr. McNab to
try some more experiments to ascertain their truth. For this purpose
specimens of nitric acid of different strength were sent to Hudson's Bay,
and exposed to the cold till they froze. They were then warmed till the
greater part was melted and thereafter exposed a second time to the cold,
till a considerable portion of the acid had frozen into a more hard and solid
ice than could be obtained by a single congelation. The temperature of
freezing having been ascertained both times, the uncongealed liquid was
separated from the congealed, and each sent back to England for exami-
nation. Bottles of oil of vitriol of different strengths were also sent to be
treated in the same way, with a view especially to discover whether oil of
vitriol has a strength of easiest freezing.

The general result drawn by Cavendish from these new trials is stated
as follows : '* These experiments confirm the truth of the conclusions I
drew from Mr. McNab's former experiments ; for, 1 st, there is a certain
degree of strength at which spirit of nitre freezes with a less degree of
cold than when it is either stronger or weaker ; and when spirit of nitre
of a different strength from that is made to congeal, the frozen part
approaches nearer to the foregoing degree of strength than the unfrozen."
He proceeds, 2nd, to notice that the numerical results obtained in the
second series of experiments came out very nearly the same as in the
earlier trials.

Oil of vitriol presented greater anomalies in its behaviour with freezing
mixtures than nitric acid had done, and Cavendish thought he was thereby
strengthened in the view we have already seen he entertained, concerning
the presence of some peculiar substance in the congealed part of oil of
vitriol.

The general conclusion to which these experiments led, is thus stated:
" From these experiments it should seem, that the freezing point of oil of
vitriol answering to different strengths, is nearly as follows :

Strength. Freezing Point.

,977 + V

,918 — 26®

,846 + 42**

,758 — 45**

"From hence we may conclude that oil of vitriol has not only a strength
of easiest freezing, as Mr. Keir has shown ; but that at a strength superior

• Phil. Trans. 1788, p. 166.

VIEWS OF CHEMICAL EQUIVALENTS. 465

to this, it has another point of contrary flexure, beyond which, if the
strength be increased, the cold necessary to freeze it again begins to
diminish/**

The most interesting part, perhaps, of this paper, is an incidental
passage in which Cayendish states how he estimated the strength of sul-
phuric acid. The numbers in the table representing strength, state the
weight of marble in 1 OOOths of the weight of acid which could dissolve it.
Cavendish, however, as he tells us, "did not find their strength by actually
tryinir how much marble they would dissolve, as that method is too un-
eertain, on account of the selenite (sulphate of lime) formed in the opera-
tion, and which in good measure defenas the marble from the action of the
acid." He then proceeds to state, " The method I used was, to find the
weight of the plumbum vitriolatum formed by the addition of sugar of
lead, and from thence to compute the strength, on the supposition that a
quantity of oil of vitriol, sufficient to produce 100 parts of plumbum
vitriolatum, will dissolve 33 of marble; as I found by experiment that so
much oil of vitriol would saturate as much fixed alxali, as a quantity of
oitrous acid sufficient to dissolve 33 of marble. It may be observed that
the quantity of alkali necessary to saturate a given quantity of acid, can
hajdly be aetermined with much accuracy, for which reason the foregoing
less direct method was adopted ; especiaUy as the precipitation of plum-
bum vitriolatum shows the proportional strengths, which is the thing
principally wanted, with as great accuracy as any method I know.'^f
The interest of this passage lies in the fact pointed out elsewhere, that
Cavendish had detected, not only the law of constant proportion which
regulates the quantitative combinations of chemical substances, but had
also, at least in one case, perceived and applied the law of reciprocal pro-
portion. It will be seen from the preceding quotation, that he goes upon
the principle, that being aware of the quantity of sulphuric acid which
will neutralise as much fixed alkali as the quantity of nitric acid able to
saturate 33 parts of marble can neutralise, he is thereby made acquainted,
without further experiment, with the amount of sulphuric acid requisite
to saturate 33 parts of marble, inasmuch as it will be identical with that
needed to neutralise the fixed alkali.

* Phil. Tran». 1788, pp. 180, 181.
t PhU. Tratu. 1788, p. 178.

466 CAVENDISH AS A NATURAL PHILOSOPHER.

An Attempt to explain some of the principal Phenomena of Electricity,
by Means of an Elastic Fluid, Read to the Royal Soddy,
Dec. 19, 1771, and Jan. 9, 1772 *

It will suffice io gire the title of this weU-known and elaborate paper,
in which Cavendish upholds a theory of electricity similar to that of
Epinus, whose views, however, were entirely unknown to his English
contemporary.

An account of some Attempts to imitate the Effects of the Torpedo
by Electricity. Read to the Royal Society , Jan. 18, 1775.t

Cavendish begins this paper by pointing out that Walsh's experiments
leave little room for doubt '' that the phenomena of the torpedo are pro-
duced by electricity;" but that, nevertheless, there are difficulties in the
way of this conclusion. One of the principal difficulties is " that a shock
may be perceived when the fish is held under water," although the elee-
tricity hajs a readier passage through the liquid, than through the body of
the person receiving the shock. This difficulty, however, Cavendish shows
is not insurmountable; for it is a mistake to imagine, as some electricians
do, that electricity always takes the shortest and readiest circuit;
whereas, in reality, it takes all the circuits provided for it, only a greater
quantity passes through the bodies which oppose least resistance to its
passage. This law is then applied to the demonstration of the truth,
that though the electricity discharged by the torpedo must pass chiefly
through the water by which it is surrounded, a certain quantity will pass
through the body of a person who lays his hands on opposite parts of the
surface of the animal; or who merely dips his fingers into the water close
to its body.

The second great difficulty in the way of the identification of the
torpedinal power with electricity, to which Cavendish draws attention, is,
that '^ no one hath ever perceived the shock to be accompanied with any
spark or light, or with the least degree of attraction or repulsion." Those
phenomena, however, can only be observed when the circuit along which
the electricity passes is interrupted; but the shock of the torpedo ^' would
never pass through the least sensible space of air, or even through a small
brass chain." The torpedo, accordingly, in this respect, does not differ
from a Leyden battery, which, if of large size, will give a sharp shook,
although "so weakly charged that the electricity will nardly pass through
any sensible space of air; and the larger the battery is, the lees wUl this

• PhU. Tratu. 1771, pp. 584—677.
t PML Tram. 1776, pp. 196—225.

PAPERS ON ELECTRICITY. 467

space be/' Into the proof of this doctrine, Cayendlsh enters at length,
and for the first time lays down that distinction between iiUenmty and
quantiUy as afiecting electrical phenomena, which has since proved so im-
portant a guide to Uie explication of electrical problems. He does not use
the word 'ifUensUy,* but the equivalent terms * force* or ^degree.* These
he employs in describing experiments by which it appeared that when

Leyden jars of different sises are eleclrified " in a given degree,"

'' the distance to which the spark will fly is not sensibly affected by the
number or size of the jars, but depends only on the force with which
they are electrified." In other words, the phenomena referred to, are de-
termined by the itUensUy of the electricity, not by its quantity. The
peculiar and very ingenious electrometer made use of, is then described ;
and thereafter experiments are recorded made with jars of the same size
une<iually charged, from which it appeared that a given quantity of elec-
tricity of a certain intensity produces a rather less shock, than twice that
quantity with half that intensity. It is further shown, that the rapidity
with wnich the electricity of the torpedo is discharged, renders it impos-
sible that it should move pith-balls suspended near the fish, or otherwise
exhibit attraction and repulsion. A description is then given of the arti-
ficial torpedo which Cavendish constructed in order to test the truth of his
sagacious theories. It was made of wood, with plates of pewter on either
side to represent the electrical organs, and covered with sheepskin. A
wire attached to each metallic plate, passed through an insulating glass
tube and terminated in a brass knob. When this instrument, or 'artificial
torpedo,' as Cavendish calls it, was employed, it was first soaked in salt
water. One of the knobs was then placed in connexfon with the negative
side of a large Leyden battery charged with electricity, and the other
knob with the positive side. Whilst the battery was thus discharged
through the artificial torpedo by an assistant. Cavendish placed his hands
in different positions upon it, and observed the result. In these trials,
the battery employed was always charged by transferring to it the elec-
tricity of a certain number of jars, which were electrified till the baUs of
an electrometer stood at a given distance, so that the intensity of the elec*
tricity was determined with great exactness. On trial it was found that
the artificial torpedo, when charged so as to give a shock in air, equal in
intensity to that of the living fish, gave a shock ' iust perceptible' under
water; whilst, if charged so as to give a shock under water, equal to that
of the real torpedo, it gave too strong a shock in air. To remedy the dis-
proportion thus observed between the strength of the shock in water, and
in air, a second artificial torpedo was constructed of sole-leather soaked
in salt water, but otherwise like the wooden one, from which it differed
only by being a better conductor of electricity. The event, as Cavendish
teUs us, answered his expectation; for there was a much smaller difference
between the shock in air and the shock in water of the leather torpedo,
than between the shock in air and in water of the wooden one. Experi-
ments are then described, which proved that the shock received by dipping
the hands into water dose to the charged torpedo, is occasioned by elec-
tricity which passes from hand to hand throuffh the water, not from arm
to arm through the body. Various trials are then recorded which showed
that the artificial torpedo closely imitated the natural animal in its electri-
oj deportment. Cavendish then proceeds to investigate the cause of the
shock of the torpedo not passing through a sensible space of air. The
wooden model was used in these trials. It was found that the shock

2h2

468 CAVENDISH AS A NATURAL PHILOSOPHER.

passed freely along tinfoil laid upon sealing wax, but tbat if the foil was
cut across witb a penknife, without being otherwise disturbed, the shock
could not traverse the minute space in the divided sheet of metal, a pheno-
menon exactly in accordance with what Mr. Walsh observed in the case
of the living torpedo. The electricity of the wooden model traversed a
short chain, especially if stretched, and the phenomena then exhibited,
were found to be in conformity with those observed with the torpedo and
the gymnotus. A difficulty, however, occurred in reference to the action
of the living torpedo, as contrasted with the artificial one, inasmuch as
the '' real torpedo was never known to force his [shock! through a single
interval'* or link of a metallic chain; whereas the wooden imitation gave
a shock through several links of chain. Cavendish refers the difference
in this respect to his not having used a sufficiently large battery with the
model fish, and shows by direct experiment '' that the greater the batteiy
is, the less space of air, or the fewer links of a chain, will a shock of a
given strength pass across." The general conclusion to which these ob-
servations conducted Cavendish, was, that the peculiar physiological sensa-
tion, termed emphatically, *^ the shock," depends somewhat more upon the
quantity than upon the intensity of the electricity passing through the
animal ' shockea ;' or perhaps, the conclusion arrived at, may be better
expressed by saying, that a large quantity of electricity possessing a low
intensity, will produce as severe a shock, as a small quantity of electricity
possessing a high intensity. In illustration of this proposition it is shown
that if a large and a small Leyden battery be so charged as to give shocks
of equal severity, the electricity of the smaller battery will be found able
to travel by a longer^route across air or along a chain. The electricity of
the smaller battery has thus a higher intensity, than that of the larger
battery, but it does not on that account give a shock of greater severity.
Accordingly, as the living torpedo can give a severe shock, but cannot
make this shock pass *^ through any sensible space of air," the qtuxntily
of electricity which it develops " must be extremely great," for when
electricity possesses a feeble intensity, it must be sent through the body
of the animal ' shocked ' in large quantity, otherwise it will not produce
a shock. Yet if so large a quantity of electricity as the torpedo certainly
develops were suddenly transferred from one side of its body to the other,
it could not but possess an intensity ^' sufficient to make it dart through air
to a great distance, unless there was something within it [the torpedo]
analogous to a very large battery." Cavendish then contends that there
is '^ room in the fish for a battery of a sufficient size," according to the
observations of John Hunter, and concludes with reiterating his statement^
that even those phenomena exhibited by the artificial torpedo, which
differed most from the actions of the living animal, are by no means re-
pugnant to the supposition that the shock is produced by electricity.

It would be difficult to exaggerate the merits of this beautiful essay,
on which, however, my limited space does not allow me to dwell at
length. Singularly enough, its aim has been entirely misapprehended by
several of its critics. Dr. Charles Hutton and others have referred to
Cavendish as having pointed out that animal electricity is peculiar in its
nature, and different from that evolved by inorganic bodies; whereas his
aim from first to last is to insist upon the identity of the electricity of the
torpedo and the gymnotus with that of the Leyden jar. Faraday has
done full justice to Cavendish's merit in this respect. In truth the Third
and Fifteenth series of Faraday's Electrical Researches, (January, 1833

PAPERS ON BLECTRICITY. 469

and NoYember, 1838,) form a complete commentary on Cavendish's
labours, and on tliose of his able saccessors who have experimented on the
electrical fishes.

I quote a single passage from the Electrical researches in reference to the
torpedo; bat the reader who wishes to do justice to Cavendish will consult
Faraday's entire papers: — " In concluding this summary of the powers of
torpedinal electnci^^ I cannot refrain from pointing out the enormous
absolute quantity of electricitv which the animal must put in circulation
at each enort. It is doubtful whether any common electrical machine
has as yet been able to supply electricity sufficient in a reasonable time to
cause true electro-chemical decomposition of water; yet the current from
the torpedo has done it. The same high proportion is shown by the
magnetic effects. These circumstances inaicate that the torpedo has
power (in the way probably that Cavendish describes) to continue the
evolution for a sensible time, so that its successive discharges rather
resemble those of a voltaic arrangement, intermitting in its action, than
those of a Leyden apparatus, charged and discharged many times in
succession. In reality, however, there is 9U> phUosophicaZ difference
between these two cases."*

I add the remark that since the date of Faradav's experiments on the
living gymnotus (1838), an interesting addition has been made to our
knowledge concerning the electrical fishes. In December, 1844, Dr.
James Stark of Edinburgh discovered an electrical organ like that of the
torpedo and gymnotus in the tail of the common skate and other Rays.f
Professor Qoodsir took up the investigation after Dr. Stark| and has
deposited in the Anatomical Museum of the University of Edinburgh, a
beautiful series of dissections of the electrical organ of the skate. In
1846, Dr. C. Robin, apparently unaware of the observations made in
Edinburgh in the previous year, announced his discovery of an electrical
organ in the Rays.§ It is still doubtful whether the organ in these fishes
is rudimentary or a fully developed and active electrical apparatus. I
would direct the attention of our experimental electricians to the problem
as one worth their investigation.

Besides his two published papers on Electricity, Cavendish has left
behind him some twenty packets of manuscript essays, more or less
complete, on Mathematical and Experimental Electricity. These papers
are at present in the hands of Sir William Snow Harris, who most Kindly
sent me an abstract of them, with a commentary of gr^t value on their
contents. I regret that I cannot do more in this volume than allude to
Sir William Harris's communication. It will, I trust, be made public.
Meanwhile, I will only mention that Sir William states that " Cavendish
had really anticipated all those great facts in common electricity, which
were subsequently made known to the scientific world through the
investigations ana writings of the celebrated Coulomb and other philo-
sophers, and had also obtained the more immediate results of experiments
of a more refined kind instituted in our own day."

Professor William Thomson also,|| who saw Cavendish's Electrical

* Bjcperimental Retearehew in Sleetriciiy, by Michael Faraday, vol. i. p. 101,
par. 359.

t Proceedingt of the Boyal Soeitty of Edinburgh, December 2, 1844.
X ProeeedingM q/zAe Royal Society of Edinburgh, Janaory 6, 1845.
$ Ann, deu Sciences Natur, 1816, tome vii. pp. 193—302.

II Profeuor of Natural Philosophy in the University of Glasgow.

470 CAVENDISH AS A NATURAL PHILOSOPHER.

MSS. whilst in the possession of Sir William Harris, writes to me con-
cerning them, that althongh he was not able to do more than glance ai
them, thej appeared to him ^'to contain descriptions of excessirelj
ingenious experiments leading to important qnantitative results^ with
reference to electricity in equilibrinm on bodies of various forms and
dimensions."

From the concurrent testimony of two such high authorities as Sir
William Snow Harris and Professor W. Thomson, it cannot be doubted
that the Electrical MSS. of Cayendish would amply repay publication*

DETERMINATION OF DENSITY OF EARTH. 4?!

Experiments to determine the Density of the Earth. Read to the

Royal Society June 21 st, 1798 *

The object of ibis paper will appear from tbe following brief account of
tbe principles whicb cbaracterised the method saggested by the Rev.
Jobn Michell for tbe determination of the density of the Earthy and
'were adopted by Cayendish.f

Michell's apparatus was a delicate torsion balance, consisting of a
light wooden arm suspended in a horizontal position by a slender wire 40
inches long, and having a leaden ball about 2 inches m diameter hung at
either extremity. Two heavy spherical masses of metal were then
l>rought near to the balls, so that their attractions conspired in drawing
the arm aside. The deviation of the arm was observed ; and the force
necessary to produce a given deviation of the arm being calculated from
its time of vibration it was found what portion of the weight of either
ball was equal to the attraction of tbe mass of metal placed near it.
From the known weight of the mass of metal; the distance of the centres
of the mass, and of the ball; and the ascertained attraction, it is easy to
determine the attraction of an equal spherical mass of water upon a
particle as heavy as the ball placed on its surface; and from this can be
found the attraction of a sphere of water of the same diameter as the
earth, upon the ball placed on its surface. Now the attraction of this
sphere will have to that of the earth, the same ratio as their densities;
and as the attraction of the earth is equal to the weight of the ball, it
follows that as the calculated attraction is to the weight of the ball, so is
the density of water to the earth's density, which is thus determined.

It is unnecessary to criticise this paper at any length, as the late
Francis Baily, Esq. has devoted a large quarto volume to a discussion of
all the researches which have been made concerning the density of the
earth, including his own careful repetition of Cavendish's observations,
recorded in the paper under notice.]: The following extracts will show
what Mr. Daily's estimate of Cavendish's experiments is : — " Mr. Caven-
dish proceeds to describe the apparatus which he had erected, and

to explain his mode of operation, which appears to have been conducted
with great judgment and accuracy. Yet, notwithstanding the precautions
wbich he had taken, he still met with some anomalies for which he could
not satisfactorily account, and which appear to have affected the results
rather more than he had anticipated. He made several attempts to eluci-
date this difficulty; yet, although he had evidently hit upon the probable
sonrce of the principal anomalies, he does not appear at that time to hare
taken any effectual steps to remove it, but deferred his intention of pur-
suing this subject, as well as some other improvements in his apparatus,
to a future period.

" The number of his experiments is very few; yet, with one exception,
they are very accordant^ and show the diligence and care with which they

* PhU. Tram. 1798, p. 469.

f In drawing up this abstract I have been kindly assisted by my friend W. Swan,
Esq., F.R.S.E., Teacher of Mathematics, Edinburgh.

" X Exporiments with the Torsion Rod for determining the Mean Density of the
Earth, forming VoL XIV. of the Memoirs of the Royal Astronomical Society, by Francis
Baily, Esq. Vice-President of the Society. 1843.

472 CAVENDISH AS A NATURAL PHILOSOPHER.

have been made. From 17 sets of experiments (being all that are on
record) he deduced 23 results; from the mean of which he compntea the
Density of the Earth to be e^ual to 5*45.* Some objection, indeed, may
be made to the paucity of his experiments, and to certain parts of his

mode of proceeding : it is but just, howeyer, here to remark, that

CayendisVs object in drawing up the Memoir appears to faaye been more
for the purpose of exhibiting a specimen of what he considered to be an
excellent method of determining this important inquiry, than of deducing
a result that should lay claim to the full confidence of the scientific

world."t

From allusions in Cayendish's paper it appears that he intended to

repeat his experiments with a yiew to discoyer the cause of the anomalies
which appeared in his earlier researches. As he neycr, howeyer, put this
intention in practice, Mr. Baily, at the request of the Astronomical Society,
undertook the repetition of Cayendish's obseryations. A preyious repeti-
tion had been made by F. Reich, Professor of Natural Philoeophy in the
Academy of Mines at Freiberg in Saxony, an account of which was read
before the German Scientific Association, which met at Prague in Sep-
tember, 1837. Mr. Baily remarks on this repetition: —

" ReicVs experiments also were (like Cayendish's) too few in number;
57 only haying been made, from which 14 results haye been deduced;
the mean of which makes the Density of the Earth equal to 5*44, almost
identical with that of Cayendish.'*^

Baily*s own repetition of Cayendish's experiments commenced on
October 19, 1838, with an apparatus constructed generally like that used
by Cayendisb, but including many ingenious modifications, calculated to
render the apparatus more delicate, and, aboye all, more accurate in its
indications.

In spite, howeyer, of all the care and ingenuity bestowed upon the
reconstruction of the apparatus of Mr. Baily, the same anomalous motions
of the torsion rod showed themselyes which had perplexed Cayendisb; and
so difficult was the detection of the nature of the disturbing force, that
eighteen months of labour (daring which nearly 1300 experiments were
made) were thrown away, before the chief cause of the anomaly was
detected. § The results of these 1300 experiments, as Mr. Baily obseryes,
" althouffn in many cases yery consistent amongst themselyes, were, upon
the whole, so discordant and unsatbfactory, that no confidence could be
placed on the general result, as a correct yalue of the true object of
inquiry. "II The whole experiments accordingly were rejected.

In January, 1841, a new series of trials was commenced with a modi-
fication of the apparatus. Professor J. D. Forbes, who agreed " with
Cayendisb in opinion that one source, at least, of the anomalies might
arise from the radiation of heat from the masses, suggested the pro-
priety of having the masses gilt, and also of procuring a gilt case as a cover
to the torsion box, for the purpose of preventing the effect of radiation,
from whatever source it might arise."^

This alteration was followed with the happiest effects, '' for the results
soon convinced me," says Mr. Baily, *' that the proper mode had been
taken for the removal of the principal source of discordance."** The new

* " Cavendish says 5*48 ; but there is a singular error in his computation."
t Memoirs of thi Royal Astronomical Society, vol. ziv. pp. 7 and 8.
% Op, cit, p. 10.

% Mem, Attr, Soc. vol. ziv. p. 41. || Op. cit, p. 41.

\ Op, et loc, cit, ♦* Ibid, p. 42.

MISCELLANEOUS PAPERS. 473

experiments weie eontinned till May, 1842, and the final resnlt was, that
the Mean Density of the Earth is 5*6604.

It only remains to notice the papers on the Density of the Earth,
whioh have appeared since the publication of Mr. Baily s volume in 1849.
In 1847, G. W. Hearn, Esq. read a paper to the Ro^l Society ''On
the Cause of the Discrepancies observed by Mr. Baily with the Cavendbh
Apparatus for determining the Mean Density of the Earth.*' The object
of this communication was to draw attention to the probability of the
anomalous vibrations of the torsion-rod, being in part occasioned by the
maffnetio or diamagnetic condition of the masses.* In 1849, Prof. J. D.
Forbes re-directed attention to a method of determining the density of
the earth suggested by Prof. Robison, by taking advantage of the high
tide which rises in the Bay of Fnndy, Nova Scotia. ** The object was to
determine the earth's density by the attraction of the tide-wave on a
plummet or spirit-level, on the same principle as Maskelyne's experiment
on Schiehallien, but with the superior advantages arising from the per-
fect homogeneity of the attracting mass, and £rom the circumstance that
all the observations might be made at a single station. The experiment
might, in short, appear to unite the advantages both of Maskelyne's and
Cavendish's methods of determining the earth's density." Prof. Forbes
has made the calculation approximately for an assumed height of the
tide-wave. Robison reports the water in Fundy Bay to rise 100 feet at
spring tide. Professor Forbes accordingly has " calculated tJie horizontal
attraction of a semicylinder of water 1 00 feet thick, and of about two,
four, and eight miles radius upon a point at the extremity of the axis of
such a semicylinder." The influence, however, of a tide- wave 100 feet
thick, with a radius of 40,000 feet, upon a plumb-line, would produce a
deviation of only 0'*'5S (fifty-three hundredths of a second). '' Even the
greatest of these calculatea deviations afibrds no ground for hoping that
the method of Robison could be applied with any success to determine
the earth's density."f

It is not a little curious that Robison had been anticipated by Caven-
dish in suggesting this methoil of procedure. Among the Caven^sh MSS.
is a parcel in Cavendishes handwriting, entitled *' Attraction," and con-
taining a variety of packets of papers on various matters connected with
the estimation of the earth's density. One of these packets is entitled^
^' Paper given to Maskelyne relating to Attraction and Form of Earth.
No. 6." In this paper Cavendish calculates the deviation which the tide
in the Bristol Channel would occasion on a plumb-line, on the supposition
that the Channel is ten miles broad, and that the tide rises fifty feet; and
comes to the conclusion that the mass of water ** would make the plumb-
line deviate If seconds, if the mean density of the earth is the same as
that of the surface." (P. 9.) It further appears that Boscovich has pre-
ceded Cavendish; for the latter, in another part of the paper, says, '^ Since
I saw you [Maskelyne], I have looked again into Boscovich's book (' De

Littoraria Expeditione, &c.'), and find that he supposes the arm of

the sea to be 100 miles broad, in which case he says the plumb-line will
deviate 2" 38'''." (P. 10.) From these passages it will be seen that
Boscovich, Cavendish, and Maskelyne were aware of the method gene-
rally believed to have been suggested by Robison. I do not know the
exact date of the communication of Cavendish to Maskelyne, but a letter

* See Report in AtheruBum, March 27th, 1847.

f The extracts in the text are taken from Prof. Forhes's interesting Note regarding an
experiment tuggeeted by Pro/eesor Robieon, (Proceed, of R.S.E., toI. il. (1849) p. 244.)

474 CAVENDISH AS A NATURAL PHILOSOPHBIU

irom tiie latter to the fonner lemains among the Cayendish MSS., dated
January 5, 1783. It refers to Cavendiah's " Bales and directioiis for the
choice of hills haying a considerahle attraction,*' which he had famiahed
to Maflkeljne. Cayendish was a memher of the *' Committee of Attrac-
tion/' appointed bj the Royal Society to assist Madcelyne in his search
for a moantain snitable for his experiment, and the acoonnt quoted most
haye been giyen before Jnne^ 1774, when he b^gan his obeerya-
tions on Schehallien.* Robison does not explicitly claim the method
described by him as his own deyice. Perhaps, howeyer, he deyised the
process for himself ; at all eyents, the sa^gestion of Fnndy Bay as apeoiiJly
suitable for trying the sapposed experiment was his, and so was the
application of a syphon to indicaite the attraction.

The remaining papera haye been referred to in the Personal Ni
tiye. I will only mrther remark concerning them, that they are all
important. That on the Height of a Luminous Arck\ has been commAnted
on by Daltoa.t The paper on the CivU Year of the Ilindooi% should be
read in connexion with a work of high authority, with the loan of whidi
I haye been £ftyoared by James Daunahoy, Esq. It is entitled *^Kala
Sankaliia, a collection of memoirs on the yarious modes according to
which the nations of the southern parts of India diyide time, &c By
Lieutenant-Colonel John Warren. Madras, 1825."

Cayendish's latest published paper, that, namely, on the Division of
AsiroTiomical Instruments, \\ is commented on in the '^ Encyclopsadia
Brittanica," art Gr<iduaium, Finally, I m^ notice that among the
Cayendish MSS. I haye found a pa^r on the Density of the Atmosphere
of the Earth and of Jupiter. In this Cayendish supposes the air to con-
sist of " particles disposed as in the angular points of cubes." He then
shows, (that if the density of the air be diminished beyond a certain
limit, depending on the weight of particles of air, its elasticity will not
be sufficient to support the weight of the next row of partides. The
necessary consequence of this hypothesis is, that the atmosphere has a
definite limit; and this conclusion seems eyidently to anticipate WoUas^
ton's speculations on the same subject. IT

* Weld's History qfihe Royal Society, vol. ii. pp. 79 — 80.

t Phil. Tram. 1790, pp. 101 and 105.

% Meteorological B$8ay9, 2nd. ed., p. 146.

$ Phil. Tram. 1792, pp. 383—399. II Ibid. 1809» pp. 221—231.

f Ibid, 1822^ p. 89, and Tram. R.8.E., vol. xn. part 1, p. 79.

475

CAVENDISH'S APPARATUS.

Catbndish left behind him an immense amount of apparatus, which
was inherited by Lord George Cavendish. From him it passed into the
possession of various parties, bat the Earl of Burlington, Sir Humphry
Davy, and Mr. Newman of Regent Street, obtained the greater part of it.
By them or their heirs it has to a great extent been dispersed among those
liKely to value it. Mr. Tomlinson has had drawings made of two curious
specimens of Cavendishes apparatus, which have been engraved for this
volume. The descriptions of the instruments given in the succeeding
paffes have been kindly communicated by Mr. Tomlinson, who has care-
fully examined them.

I have been unable to discover any reference to the brass Eudio-
meter in Cavendishes published or unpublished writings, but it seems
probable that it was employed in the course of the investigations which
led to the cUscovery of the composition of water and of nitric acid. I
Iiave applied to all the parties known to me who possess Cavendish's
apparatus, with a view to learn whether the glass globe-eudiometers
which he describes in his paper of 1784 (ante, pp. 42, 43), are still in ex-
istence. As yet, I have discovered no traces of these vessels. The Earl
of Burlington, however, writes to me, that he thinks it possible that relics
of them may bo found among a collection of Cavendish's apparatus, which
18 at present, from circumstances, inaccessilile. Lord Burlington has
kindly engaged to send the relics of the Water- Eudiometers, should he dis-
cover them, to the Museum of the Royal Society, in which they would
doubtless find an appropriate place beside Newton's Telescope and Davy's
Safety-lamp.

Cavendish seems to have taken greater interest in thermometers than
in any other instruments applicable to the extension of physical science.
His house at Clapham Common was crowded with them. The late Pro-
fessor Daniell possessed two, which were provided with wooden scales, on
which the degrees had been marked by Cavendish himself. In the col-
lection of apparatus belonging to the Natural Philosophy Chair of Edin-
burgh, is a singular instrument of Cavendish's, presented to Professor
J. D. Forbes by Dr. Davy. It may be called a Balance Thermometer.
Its construction is such that a glass thermometer turns upon an axis,
and moves an index, according as the expansion or contraction of the
included quicksilver renders heavier either extremity of the balanced
tube.

I
I

;^^' <§f§if§(§rBi$ii>$rHiu>sopaEii.

HessH

■ft. -di' .ft.

" lE^i^fiuuent belonging to the coUec-
Ifl&t^SXin of Great Biitsin. It goes
i^w^' and wu presented to the

""i^P^^lMaiii length, &nd about 2 inchea

ii^»3^f>bT means of two iron straps.

„ JtWrgEic^ left of the figure hj a plate,

^^&^S*^:eption of a stop-cock. The

^jAiKli*liV(iS^6>' through this stop-cook, and

ia^^^l«J£«r gasometer for the pnrpoee

M*^^'^^ ^ partial Tacnum within it.

a bent piece of metal, added

ge, so as to give the operator

„ extremity of the cylinder to

t^bSBihallow cap, in the centre of

' — ^wn-out jet. ThiB Btop-<»ck

^lesnlting from the combustion

■;£l>,een added

for the purpose of

There is, however, a special

By the side of the opening

rhicb is screwed a perforated

r tube, the bore of which is

■«-:§-»-s

is>?'i'--i— ;:;"S

iltl^-tliitfi

S* e'fwf pis^"^ ^&V of Caveudieh's

tube oontuniog

>fft, the atmoephere,

a. The snr&oe

which prooeeda

iheiy of a wheel

[7, with a email

The axis of the
hand: thia hand
luated circle; hnt
piece carrying a

pointing at 50°.
acoompanies the
e. In the figute

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "The Life of the Hon. Henry Cavendish: Including Abstracts of His More ..."

See other formats