■/J^^'rV"*' -i

^:/^<:^

m^-^'^

5i^

/A ■

> ■

<f'

(

>

.^y^

U.X. 1.

THE
LONDON, EDINBURGH, and DUBLIN

PHILOSOPHICAL MAGAZINE

AND

JOURNAL OF SCIENCE.

CONDUCTED BY

SIR DAVID BREWSTER, K.H. LL.D. F.R.S.L.&E. &c.
RICHARD TAYLOR, F.L.S. G.S. Astr.S. Nat.H.Mosc.&c
RICHARD PHILLIPS, F.R.S.L.&E. F.G.S. &c.
SIR ROBERT KANE, M.D. M.R.I.A.

" Nee aranearum sane textus ideo melior quia ex se fila gignunt, nee noster
vilior quia ex alienis libamus ut apes. ' Just. Lirs. Polit. lib. i. cap. 1. Not.

VOL. XXXL

NEW AND UNITED SERIES OF THE PHILOSOPHICAL MAGAZINE,
ANNALS OF PHILOSOPHY, AND JOURNAL OF SCIENCE.

.JULY— DECEMBER, 1847.

LONDON:

RICHARD AND JOHN E. TAYLOR, RED LION COURT, FLEET STREET,

Printers and Publishers to the University of London;

SOLD BY LONGMAN, SHOWN, GREEN, AND LONGMANS J SIMPKIN, MARSHALL
AND CO.; S. HIGHLEY ; WHITTAKER AND CO.; AND SHERWOOD,

GILBERT, AND PIPER, LONDON : BY ADAM AND CHARLES

BLACK, AND THOMAS CLARK, EDINBURGH ; SMITH AND SON,

GLASGOW ; HODGES AND SMITH, DUBLIN ; AND

WILEY AND PUTNAM, NEW YORK.

^K

" Meditationis est perscrutari occulta; contemplationis est admirari

perspicua Admiratio general qusestionenij qusestio investigationem,

investigatio inventionera." — Hugo de S. Victore.

CONTENTS OF VOL. XXXI.

(THIRD SERIES.)

NUMBER CCV.— JULY 1847.

Page

Sir J. Lubbock on the Perturbations of Planets moving in

Eccentric and Inclined Orbits 1

Prof. Schoenbein on the Discovery^of Gun- Cotton 7

The Rev. B. Bronwin on the Inverse Calculus of Definite In-
tegrals 12

Mr. W. R. Grove on certain Phajnomena of Voltaic Ignition
and the Decomposition of Water into its constituent Gases

by Heat 20

Mr. C. R. Weld on the Invention of Fluxions 35

Sir R. Kane's Researches on the Composition and Characters
of certain Soils and Waters belonging to the Flax districts
of Belgium, and on the Chemical Constitution of the Ashes

of the Flax Plant 36

Dr. Schunck on the Colouring Matters of Madder 46

Comparative Analysis of the Urine of the Calf and the Sheep. . 49
Mr. Hind on the expected Reappearance of the celebrated

Comet of 1264 and 1556 50

Messrs. G. Merck and R. Galloway's Analysis of the Water of

the Thermal Spring of Bath (King's Bath) 56

Notices respecting New Books 67

Proceedings of the Royal Society 69

Action of Chlorine on Alcohol. — Formation of Acetal 77

Bisilicate of Iron or Ferruginous Pyroxene 78

Chlorosulphuret of Silicium 78

Meteorological Observations for May 1847 79

Meteorological Observations made by Mr. Thompson at the
Garden of the Horticultural Society at Chiswick, near
London ; by Mr. VeaU at Boston ; by the Rev, W. Dunbar at
Applegarth Manse, Dumfries-shire ; and by the Rev. C.
Clouston at Sand wick Manse, Orkney 80

NUMBER CCVI.— AUGUST.

The Rev. N. J. Callan on a new Voltaic Battery, cheap in its
construction and use, and more powerful than any Battery
yet made ; and on a cheap substitute for the nitric acid of
Grove's Platina Battery 81

Sir J. Lubbock on the Perturbations of Planets moving in Ec-
centric and Inclined Orbits (^concluded) 86

Sir J. Lubbock on the Heat of Vapours 90

a2

IV CONTENTS OF VOL. XXXI. — THIRD SERIES.

Page
Mr. W. R. Grove on certain Phsenomena of Voltaic Ignition and
the Decomposition of Water into its constituent Gases by-
Heat (concluded) 91

Mr. \V. R. Grove's Supplementary Paper on certain Phaeno-
mena of Voltaic Ignition and the Decomposition of Water

into its constituent Gases by Heat . . ' 96

Sir D. Brewster on the Modification of the Doubly Refracting
and Physical Structure of Topaz, by Elastic Forces emanating

from Minute Cavities. (With a Plate) 101

Sir R. Kane's Researches on the Composition and Characters of
certain Soils and Waters belonging to the Flax districts of
Belgium, and on the Chemical Constitution of the Ashes of

the Flax Plant (concluded) . . . . 105

Mr. J. P. Joule on the Theoretical Velocity of Sound 114

Mr. E. C. Nicholson on the Composition of CafFein, and of some

of its Compounds 115

Prof. J. R. Young's Note in reference to the extension of Euler's

Theorem 123

Prof. A. Connell on the Precipitate produced in Spring and

River Waters by Acetate of Lead 124

Mr. J. Mercer on the Action of a mixture of Red Prussiate of

Potash and Caustic Alkali upon Colouring Matters 126

Dr. W. Gregory on the Preparation of Hippuric Acid 127

Proceedings of the Cambridge Philosophical Society 130

Royal Astronomical Society 143

On a new Test for Prussic Acid, and on a simple method of pre-
paring the Sulphocyanide of Ammonium, by Prof. Liebig . . 146

On the Fusion of Iridium and Rhodium, by R. Hare 147

On Testing the Comparative Value of Astringent Substances for

the purposes of Tanning, by Robert Warington, Esq 150

On the two varieties of Arsenious Acid, by M. Bussy 151

On the Preparation of Gun-Cotton 152

On Balsam of Tolu, and some products derived from it 153

On the Equivalent of Titanium, by M. Isidore Pierre 155

On a modification of the Apparatus of Varrentrapp and Will for

the estimation of Nitrogen, by Warren De la Rue, Esq 156

On the Detection of Cotton in Linen, by G. C. Kindt 157

The Planet Hebe 158

Meteorological Observations for June 1847 159

Table 160

NUMBER CCVIL— SEPTEMBER.

Dr. T. Anderson on certain Products of Decomposition of the
Fixed Oils in contact with Sulphur 161

Mr. J. P. Joule on the Mechanical Equivalent of Heat, as de-
termined by the Heat evolved by the Friction of Fluids. ... 173

CONTENTrxW VOL. XXXI. — THIRD SERIES. V

Page
Letter from Prof. Schoenbein to Prof. Faraday, F.R.S., on a

new Test for Ozone 176

Dr. G. Wilson on the Decomposition of Water by Platinum and
the Black Oxide of Iron at a white heat, with some observa-
tions on the theory of Mr. Grove's Experiments 177

Mr. J. J. Sylvester's account of a Discovery in the Theory of

Numbers relative to the Equation Ax^ +By^ + Cz'^=Dxyz . . 189
Experiment made at the Kew Observatory on a new Kite- Appa-
ratus for Meteorological Observations, or other purposes . . 191
Dr. L. Playfair on Transformations produced by Catalytic Bo-

dies 192

Sir W. R. Hamilton on Quaternions ; or on a New System of

Imaginaries in Algebra (continued) 214

Notices respecting New Books 219

Proceedings of the Royal Society 222

Suggestions for the observation of the Annular Eclipse, Oct. 9,
1847, made by the British Association for the Advancement

of Science, Oxford, June 26, 1847 228

On the Preparation and Composition of the Salts of Antimony,

by M. E. Peligot 230

Action of Hydrochloric Acid in the Formation of Oxalic Acid 233

Projection of Aldebaran on the Moon 233

The PufF Parliamentary : — Disinfection 233

A Grant of 200/. to Mr. William Sturgeon 236

Observations on Creatine, by M. Heintz 236

The New Planet Iris 237

Suggestions for Promoting the Science of Meteorology 238

Meteorological Observations for July 1847 239

Table 240

NUMBER CCVIIL— OCTOBER.

Prof. E. Wartmann's Fourth Memoir on Induction. (With a

Plate.) 241

Mr. S. M. Orach on eliminating the Signs in Star- Reductions. . 251

Mr. J. Brown on the Molybdate of Lead. 253

Dr. R. D. Thomson's note on a new Test for Arseniates, &c.. . 258
Mr. E. W. Binney on Fossil Calamites found standing in an
erect position in the Carboniferous Strata near Wigan, Lanca-
shire 259

Mr. E. Frankland and Dr. H. Kolbe upon the Chemical Consti-
tution of Metacetonic Acid, and some other Bodies related

to it 266

Messrs. T. H. Rowney and H. How's Analysis of the Ashes of

the Orange-Tree (Citrus aurantium) 271

Sir W. R. Hamilton on Quaternions ; or on a New System of
Tmaginaries in Algebra (continued) 278

tl CONTENTS OF VOL. XXXI. — THIRD SERIES.

Page
Mr. J. J. Sylvester on the Equation in Numbers Ax^+By^ + Cz^

s^Dxyz, and its associate system of Equations (continued) . . 293
Mr. R. Taylor on the Invention and First Introduction of Mr.

KcEnig's Printing Machine 297

Proceedings of the Cambridge Philosophical Society 301

On the Artificial Production of Minerals, and especially of Pre-
cious Stones 311

Analysis of Kupfemickel 314

On the Dehydration of Monohydrated Sulphuric Acid 314

Observations on Silica, by M. Doveri 315

On Nitric Mannite, by M. Sobrero 316

On the Extraction of Silver, by MM. Malaguti and Durocher 317

Vanadiate of Lead and Copper 319

Meteorological Observations for August 1847 319

Table 320

NUMBER CCIX.— NOVEMBER.

Prof. M. A. De la Rive's Researches on the Voltaic Arc, and
on the influence which Magnetism exerts both on this Arc
and on bodies transmitting interrupted Electric Currents . . 321

Mr. T. Richardson's Analyses of the Ashes of Rough Brown
Sugar and Molasses 336

Letter from Prof. Loomis of the New York University to Lieut.-
Colonel Sabine, Foreign Secretary of the Royal Society, on the
determination of differences of Longitude made in the United
States by means of the Electric Telegraph, and on projected
observations for investigating the Laws of the gi-eat North
American Storms 338

The Rev. B. Bronwin on the Algebraic Equation of the Fifth
Degree 341

Letter from Capt. J. H. Lefroy, R.A., Director of the Mag-
netic Observatory of Toronto in Canada, to Lieut. -Colonel
Sabine, R.A., on a great Magnetic Disturbance on the 24th
of September 1847 346

Dr. H. Kolbe on the Decomposition of Valerianic Acid by the
Voltaic Current 348

Mr. R. Adie's Account of Experiments with Galvanic Couples
immersed in pure water and in oxygenated water 360

Dr. R. Hare on certain Improvements in the Construction and
Supply of the Hydro-Oxygen Blowpipe, by which Platinum
may be fused in the large way , 356

Dr. J. W. Griffith on the Composition of the Bile of the Sheep 366

The Rev. J. Slatter's Notice respecting the Meteor of Septem-
ber 25, 1846 368

Mr. J. Glaisher on the Aurora Borealis, as it was seen on Sun-
day evening, October 24, 1847, at Blackheath 369

CONTENTS OF VOL, XXXI.— THIRD SERIES. Vll

Page

Proceedings of the Royal Society 372

■ Cambridge Philosophical Society 376

Royal Astronomical Society 380

On the Gelatinous Substances of Vegetables 389

Preparation of Protoxide of Tin 392

On the Presence of Arsenic, Copper and Tin, in the Mineral

Waters of Bavaria 392

Solubility of Common Salt in Alcohol 393

On some Improved Forms of Chemical Apparatus, by Thomas

Taylor, Esq 393

Preparation and Composition of Lignin 397

Solubility of Chloride of Silver in Hydrochloric Acid 398

Daubeny on Active and Extinct Volcanos 399

Meteorological Observations for September 1847 399

Table 400

NUMBER CCX.— DECEMBER.

Prof. M. Faraday on the Diamagnetic conditions of Flame and
Gases 401

Prof. Zantedeschi on the Motions presented by Flame when'
under the Electro- Magnetic Influence 421

Mr. T. Weddle on Asymptotic Straight Lines, Planes, Cones
and Cylinders to Algebraical Surfaces 425

Mr. R. A. Couper on the Chemical Composition of the Sub-
stances employed in Pottery 435

Sir D. Brewster on the Polarization of the Atmosphere 444

Mr. A. Smith on the Hydx*ates of Nitric Acid 454

Mr. F. Field on the Products of the Decomposition of Cuminate
of Ammonia by Heat 459

Mr. J. J. Sylvester on the General Solution (in certain cases)
pf the equation j:'^-\-i/^-\-Az^ = Mxyz, &c 467

Mr. W. De la Rue on Cochineal {Coccus Cacti). First Memoir 471

NUMBER CCXI.— SUPPLEMENT TO VOL. XXXI.

Mr. W. De la Rue on Cochineal (Coccus Cacti). First Memoir
(concluded) 481

Sir D. Brewster on the Existence of Crystals with different pri-
mitive forms and physical properties in the Cavities of Mine-
rals ; with additional Observations on the New Fluids in
which they occur. (With a Plate.) 497

Mr. L. Thompson's Observations on Chloric Acid and the
Chlorates 510

Vni CONTENTS OP VOL. XXXI. — THIRD SERIES.

Page
Prof. Sir W. R. Hamilton on Quaternions ; or on a New Sy-
stem of Imaginaries in Algebra (continued) 511

Mr. J. H. Gladstone's Contributions to the Chemical History of

Gun-Cotton and Xyloidine 519

Proceedings of the Royal Astronomical Society 528

On Osmiamic Acid, by MM. J. Fritzsche and H. Struve .... 534
On the Preparation and Properties of some Osmiamates, by

MM. Fritzsche and Struve 535

On Sulphato- chloride of Copper, — a New Mineral, by Arthur

Connell, Esq '. 537

On the Formation of Valerianic Acid, by M. Therault 538

Note on the Measurement of the double Sulphates of Zinc and

Soda, and of Magnesia and Soda, by W. H. Miller 540

Native Carbonate of Nickel 541

An Examination and Analysis of the " Nadelerz," or needle

ore of Bismuth, by E. .T. Chapman, Esq 541

Action of Anhydrous Phosphoric Acid on Ammoniacal Salts, by

M. Dumas 544

Meteorological Observations for October 1847 545

Table 546

Index 547

PLATES,

I. Illustrative of Sir D. Brewster's Paper on the Modification of the
Doubly Refracting and Physical Structure of Topaz, by Elastic
Forces emanating from Minute Cavities.

II. Illustrative of Prof. Wartmann's Fourth Memoir on Induction.

III. Illustrative of Sir D. Brewster's Paper on the Existence of Crystals

with different primitive forms and physical properties in the Cavities

of Minerals.

Erratum in Mr. Sylvester's paper, p. 189
Line \^,for Ti'xyz read D'uviv.

Errata in Sir Graves C. Haugaton's paper, vol. xxx. p. 437.
P. 445, in the thirteenth line from the bottom,/or M024 read 1,1024.
ninth line from the bottom,/o;- 256° read 256.

— 456, in the third line from the bottom,/o;- Hare hair read Horse hair.

— 518, in the thirteenth line from the bottom,/or oxide of hydrogen read

protoxide of hydrogen.

— 522, in the fifteenth line froni the top,/o>" in fault read at fault.

THE
LONDON, EDINBURGH and DUBLIN

PHILOSOPHICAL MAGAZINE

AND

JOURNAL OF SCIENCE.

[THIRD SERIES.]

JULY 1847.

I. On the Perturbations of Planets moving in Eccentric and
Inclined Orbits. By Sir J. Lubbock, Bart.i F.R.S.*

nPHE accuracy of the tables which give for an indefinite
-*■ time the places of the older planets, is at present sufficient
for the purposes of astronomy, and is commensurate with the
accuracy of observation ; or if this statement appears to be
exaggerated, it will at least be admitted that the sensible
errors which remain are owing rather to inadvertence in the
numerical computation than to the imperfection of the method
itself. Such a result is owing to the uninterrupted labours of
the greatest mathematicians from the time of Newton, and is
justly regarded as one of the greatest triumphs of human in-
telligence. But it must be recollected that these methods,
by which the perturbations of the older planets have been
obtained, are applicable only to the case of orbits nearly cir-
cular, and little inclined to each other; so that the general
solution of the problem of the three bodies, as it is called,
remains to the present day imperfect. The methods in use
for the older planets are founded, as is well known, upon the
development of the disturbing function in terms of the mean
anomalies. M. Binet has indeed carried this development to
quantities of the seventh order inclusive ; but such a de,velop-
ment is quite insufficient in the case of comets or planets
moving in highly eccentric and inclined orbits, which problem
presents far greater difficulty ; while the nature of the expres-
sion is such, that it is evidently impracticable to carry further
such a mode of development, even if the expressions were suf-
ficiently convergent when the eccentricity passes a certain
limit.

The only memoir with which I am acquainted which pro-

* Communicated by the Author.
Phil. Mag. S. S. Vol. 31. No. 205. Juli/ 1847. B

2 Sir J. Lubbock on the Perturbations of Planets

fesses to give a general solution of the problem otherwise than
by mechanical quadratures, is clue to M. Hansen. This im-
portant work is translated in the Co7in. des Temps for 1847.
That great mathematician has considered the case when r< r',
that is, when the disturbed body is inferior; and has illus-
trated the question by the numerical calculation of the per-
turbations of the comet of Encke by Saturn. M. Hansen
develops the disturbing function according to multiple angles
of the eccentric anomaly of the disturbed planet literally; and
first, according to multiple angles of the true anomaly of the
disturbing planet; M. Hansen next converts the cosines and
sines of the multiple angles of the true anomaly of the disturb-
ing planet into sines and cosines of multiple angles of the mean
anomaly of that planet; so that finally the disturbing function
is exhibited in terms of the eccentricanomaly of the disturbed
planet and the mean anomaly of the disturbing planet ; but
those series which serve to give the sines and cosines of the
multiples of the true anomaly, in terms of sines and cosines
of the mean anomaly, are not very convergent ; and the pro-
cess becomes extremely laborious, even in the case which M.
Hansen has considered, in which, in consequence of the great
distance of Saturn, the approximation does not require to be
carried nearly so far as in the case of the perturbations of the
same comet by Jupiter, and in many others which may require
consideration. Moreover, in this as in every other mode
which can be devised of developing the disturbing function
literally, all quantities must be retained of a given order ;
although when they are of a different sign, in many instances
they destroy each other ; but such reductions cannot be fore-
seen. The numerical substitutions are also extremely labo-
rious, in consequence of the multitude of terms which have to
be considered.

As the disturbing function, and others which require to be
integrated, are finally exhibited by M. Hansen in terms of
two variables, such that direct integration is impossible, M.
Hansen has recourse to the integration ^ar joar/zW, in which
each term by integration gives rise to a series of other terms,
the nature of which is complicated.

The method which I propose differs from that suggested
by M. Hansen in every particular. Instead of attempting a
literal development, I insert the numerical values of the ellip-
tic constants in the earliest possible stage : by this means the
radical, which expresses the mutual distance of the planets, is
explicitly a function of sines and cosines of various angles
with numerical coefficients. When r < r', I develop in terms
of the eccentric anomaly of m, after having obtained expres-

moving in Eccentric and Inclined Orbits, 3

sions for the co-ordinates of w' in terms of the eccentric ano-
maly of m. Such expressions are very easy to obtain, and are
very convergent. It will be recollected that before I endea-
voured to develop the disturbing function in the lunar theory
in terms of the mean motions of the sun and moon, the inva-
riable practice had been (see MeQunique C6leste, vol. iii. p. 189)
to express the co-ordinates of the sum in terms of the true
longitude of the moon ; but the equation which connects the
eccentric anomalies of two bodies is far simpler than that which
connects the true anomalies, or xf and t;, and therefore the
conversion which I employ is made with greater facility. The
quantity under the radical sign in R may thus be considered
as a function, of which the general term can be represented by

sin /. , ., n' \
^cosr + ^VV'

a being a numerical quantity. The development of this

13
quantity to the power — ^ or — -, may be facilitated by the

use of tables, which give the numerical coefficients in the
development of {l—u4 cos a} ~^, {1— y4cos«}"% &c. Such
tables have been calculated for me by Mr. Farley. By pro-
ceeding in this way, no term is ever introduced which affects
the final result beyond a given place of decimals. For the
development of the radical admits of being exhibited in the
form

>4 + -B+C+Z) + &c.;
such that

5=a^(a, C=(35(a, D=:yC(a,

so that each term is deducible from the one which precedes
it, by the multiplication of that term by a(B, /3®, &c., a, /3,
y, &c. being proper fractions. If therefore the terms in the
two quantities which form those products, such, for instance,

as oiA and <^

which form B, are sorted and arranged in the order of their
numerical magnitude, as soon as any one partial product sinks
below any limit that may be assigned, all the succeeding terms
are necessarily of inferior magnitude ; and the approximation
stops, as it were, of itself, without any exercise of thought on
the part of the computer.

When r > r', that is, when the planet disturbed is superior
to the disturbing planet, I am not able to suggest any other
course than to develop in terms of the true anomaly of the dis-
turbed planet, and the mean anomaly of the disturbing planet,

B2

4 Sir J. Lubbock on the Perturbations of Vlanets

and to integrate par parties. I have obtained the law of the
coefficients in the series which resuhs in this process, and they
are highly convergent. I am confident that, by the processes
which I have attempted thus so briefly to describe, the per-
turbations of planets moving in orbits, however eccentric and
inclined, may be calculated jvith nearly as great facility as
they are given by existing methods, in orbits nearly circular
and in the same plane, and may be exhibited in tables, giving
their values for an indefinite period, if required. If these me-
thods, which I have described in detail elsewhere, possess the
advantages which I ascribe to them, I hope the time is not
distant when the perturbations of Pallas and of some of the
comets may be reduced to a tabular form; but as the labour
will be very considerable, it will be necessary to limit the in-
quiry in the commencement to the cases of the greatest emer-
gency.

Although my methods are specially adapted to the deter-
mination of the perturbations of bodies moving in eccentric
orbits which cannot be developed in terms of the mean mo-
tions, yet they embrace also the case of a planet moving in an
orbit nearly circular ; and it is easy to show in what manner
the labour is increased by the greater eccentricity. If the
reciprocal of the radical which expresses the mutual distance
of the planets be called

the chief difficulty arises in developing {1+P}~5. If the
numerical values of the elliptic constants are introduced,

1 +P=1— ^1 cos «!— ^2^08 «2 + &c..

Ay, A^ &c. are numerical coefficients, which I here suppose
ranged in the order of their numerical magnitude. I make

{1— -4iCosa,}{l— ^gcosag} • • • • {1— ^;COSa,.} = H-P+Q,

including a limited number of terms in 1-|-P+Q.

{l-^jcosai}"^, {1— .^gcosag}"^ &c.,

can be obtained at once by means of a table. But as the
coefficients given by such a table do not readily furnish, by
interpolation, the values required unless it be considerably
extended, I take for A^, A^, &c. the nearest value given by
the table, and I leave the residue to form part of Q. In this
way it will generally be found sufficient to include not more
than six terms in 1+P+Q, so as to leave Q consisting of

2-p- 72\» ^ ^^ ^^^^ angle included by lines drawn from

moving in Eccentric and Inclined Orbits. 5

terms of which the coefficients are each below •! in numerical
value, and the quantity {1 +P}"i^ can be developed accord-
ing to powers of Q in a rapidly converging series.

„ , 2)]^ a' r so

1+P=1 ^-r— COS^+ljV '

(T ?•' a '^

2^_ 2aa'

«'"2 r^
the sun to m and m\ p= -^ -^ — 1.

2y\^ a! r
The terms contained in r- cos 8 obviously exceed

fj r a •'

greatly in magnitude those contained in )j^^,

- cos 8=2 cosy 4- k siny ;

f being the true anomaly of m', and u the eccentric anomaly ofw,
i= — e^4-^cos 0— 'V^l —e^% sin y,
^=— eC + C cosy + ^1— e^Ssiny;

S[, 93, C and 19 being constants, each necessarily less than
unity, which depend only on the inclination of the orbits and
the position of their line of intersection, and such that when
they are in the same plane

^ = ja and 33 = C.
The process is precisely the same in substance, whether the
orbit of m is highly eccentric and inclined, or circular, and in
the same plane with that of ;«' ; the only difference being that,
while in the former case it may be necessary to detach as many
as six terms to form the quantity 1+P+Q, in order that Q.
may not contain any term of which the numerical coefficient
exceeds •! in magnitude ; in the latter case, supposing e' the
eccentricity of w' to be inconsiderable, 1+P+Qwill only

contain one factor, and therefore {1+P+Q}~^, &c. can be
calculated with greater facility. Thus, for example, in the
perturbations of

Pallas by Saturn, it is convenient that 1+P+Q should
contain three terms.

Pallas by Jupiter, it is convenient that 1+P+Q should
contain four terms.

Encke's comet by Saturn, it is convenient that 1+P+Q
should contain five terms.

Encke's comet by Jupiter, it is convenient that 1+P+Q
should contain six terms.

6 Sir J. Lubbock on the Perturbations of Planets.

And then in each of these examples Q will contain no term of
which the coefficient exceeds -1 in numerical amount.

If ^ -

1+P + Q

{\+P]-^ = A + B-\-C+D,

^=(l + P+Q)-t, 5=|(a^, C=~(^B,

o o

The calculation of {1 +P+ Q}~^ is much facilitated by the
use of a table which gives the values of the coefficients of

{1— ^cosa}"^
-J and 1 +Pcontam terms multiplied by ~j-cosj', -jsmj', -j,

and -j2, and none others require consideration. If the eccen-
tricity of m' is small, they may be developed in terms of ^', the
mean anomaly of m'; and it will be sufficient to consider the
terms depending on

cos ^'5 cos 2^', cos 3^', sin g', sin 2^', sin 3f'.

If/ is the time reckoned from the time of the perihelion
passage of w, r<r',

w/ = y — esin w ;

and if 360° — n'c' is the mean anomaly of m' at the time of the
perihelion passage of w,

^' = n't—n'c';
and if

— v — n'c'=n,
n

. r n! . \

cos?fc' = cos?-< It ^sino >-,

L w J

sin ?^'=sin I < U' e sin 0 > ;

and as — is a fraction, cos /^'and sin i^' can be developed in

a series rapidly converging, and containing explicitly only the
variable quantity u.

[To be continued.]

[ 7 ]

II. On the Discovery of Gun' Cotton.
By Professor Schcenbein*.

THE substance to which I have given in German the name
o'i schiesswolle^ and in English that of gun-cotton, having
excited a lively curiosity, it may be interesting to the scientific
world to become acquainted with some details of the way in
which I was first led to its discovery.

The results of my researches on ozone led me in the course
of the last two years to turn my attention particularly to the
oxides of nitrogen, and principally to nitric acid. The nu-
merous experiments I have made on this subject have led me,
as I have stated in detail in Poggendorff''s Ajinalen, to adopt
a peculiar hypothesis on the so-called hydrates of nitric acid,
sulphuric acid, &c., as well as on the normal nitrates, sulphates,
&c.

For a long time I had entertained doubts as to the exist-
ence of compound bodies of this nature, which cannot be
isolated, and which are stated to be capable of existing only
in combination with certain other substances ; for a long time
also I had come to the notion that the introduction of these
imaginary combinations had only been an apparent progress
in theoretical chemistry, and that it had even impeded its
development.

It is well known that what has most contributed to the
admission of the existence of these compounds has been the
opinion generally received among chemists respecting the
nature of nitric acid. Starting from the existence of the com-
pound of nitrogen NO5, as an undoubted and demonstrated
fact, notwithstanding the impossibility ofisolating it, they always
cite nitric acid to prove the existence of compounds which
cannot exist in an isolated state. In my opinion, there is no
degree of oxidation which is represented by NO5, and what
these chemists designate by the formula NO5+HO must be
considered as being really NO4+HO2; 1 am even inclined
to regard the normal nitrates NO5+ RO, as compounds which
must be expressed by NO4 + RO2. Amongst other motives
which induce me to admit this opinion, I will mention the fact
that we can obtain hydrated nitric acid or a normal nitrate by
the direct mixture of NO4 with HO2 or ROg. Other consi-
derations, which I have had occasion to detail elsewhere, in-
duce me also to consider hydrated sulphuric acid to have the
form SOa-f HO2, and not that of SOg+HO, and a normal
sulphate that of SO2 + RO2. It is sufficient here to observe
that SO2 placed in presence of HOg gives rise to what is

* From the Archives des Sciences Physiques et Naturelles,

8 Prof. Schoenbein on the Discovery of Gun-Cotton.

called hydrated sulphuric acid, and that SO^ placed in pre-
sence of BaOg or PbO^ gives rise to what is called sulphate
of the oxide of barium or of lead. Rose's compound, to which
the formula 2SO3 + NO2 has been assigned, should have, in
my opinion, 2863+ NO4. Admitting this, I considered it
probable that the mixture of 2(S02+H02) ( = 2(803+00))
with NO4+HO2 ( = N05+H0) yields 28O2+NO4, and
that at the same time SHOg is disengaged, or enters into
a loose combination with what is called the bisulphate of
deutoxide of nitrogen. In other words, I conjectured that a
mixture formed with the hydrates of nitric acid and sulphuric
acid would possess a very great power of oxidation, and would
form a kind of aqua regia, in which the combination HOg
would act the part of the chlorine. On this hypothesis, and
abstracting HO2 from the acid mixture by means of a proper
oxidable body, there ought to remain Rose's compound.

Guided by these suppositions, which, I admit, may be as
little founded as they are contrary to the ideas received among
chemists, I commenced in December 184-5 a series of experi-
ments with a view to put my hypothesis to the proof: it will
be seen in the sequel whether the results at which I arrived
tend to confirm it.

I mixed some flowers of sulphur and a certain quantity
of the acid mixture of which I have spoken : immediately,
even at the temperature of 32° F., a lively disengagement of
sulphurous acid gas took place without the production of
deutoxide of nitrogen. After the reaction, which was accom-
panied by a development of heat, there remained a colourless
liquid, which, mixed with water, disengaged a considerable
quantity of deutoxide of nitrogen, and acted generally as a
solution of Rose's compound in hydrated sulphuric acid would
have done.

I should add here, that a mixture of four ounces of hydrated
sulphuric acid with a single drop of nitric acid, on the addition
of flowers of sulphur, disengages a sensible quantity of sulphu-
rous acid. To assure himself of the presence of the latter, the
operator has only to hold over the liquid a strip of paper which
has been covered with iodide of potassium paste, and tinged
slightly blue by exposure to chlorine. The liberated sulphu-
rous acid will soon dissipate this blue colour.

Selenium and phosphorus are oxidized in the same manner
at low temperatures in the acid mixture in question ; and this
latter is modified to such an extent, that, on the addition of
water, an abundant disengagement of deutoxide of nitrogen gas
takes place.

Iodine even, in the state of powder and shaken up with the

Prof. Sclioenbein on the Discoveiy ()f Gun-CoUon. 9

acid mixture, rapidly absorbs oxygen, when exposed to a low
temperature; and there is formed, besides iodic acid, the
compounds to which Millon has lately drawn attention. After
the reaction a li(|uid remains, which, diluted with water, gives
an abundant disengagement of deutoxide of nitrogen and
liberates iodine.

My experiments on ozone having shown that this body,
whicfi I consider to be a distinct peroxide of hydrogen, forms,
as well as chlorine, at the ordinary temperature, a peculiar
compound with olefiant gas, without apparently oxidizing in
the least either the hydrogen or the carbon of this gas, I had
the idea that it would not be impossible that certain organic
matters, exposed to a low temperature, would likewise form
compounds, either with the peroxide of hydrogen alone, which,
on my hypothesis, occurs in a state of combination or of mix-
ture in the acid mixture, or with NO4. It was this conjecture,
doubtless very singular in the eyes of chemists, which princi-
pally led me to commence experiments with common sugar.

I made a mixture of one part (volume) of nitric acid, of 1*5
spec, grav., and two parts of sulphuric acid of 1'85, at the
temperature of 36° F. ; I then added some finely powdered
sugar, so as to form a very fluid paste. I stirred the whole, and,
at the end of a few minutes, the saccharine substance formed
itself into a viscous mass entirely separated from the acid
liquid, without any disengagement of gas. This pasty mass
was washed with boiling water, until this last no longer exer-
cised any acid reaction ; after which I deprived it, as much as
possible, at a low temperature, of the water it still contained.
The substance now possessed the following properties: — Ex-
posed to a low temperature, it is compact and brittle ; at a
moderate temperature, it may be moulded like jalap resin,
which gives it a beautiful silky lustre. It is semi-fluid at the
temperature of boiling water ; at a higher temperature, it
gives off red vapours ; heated still more, it suddenly defla-
grates with violence, without leaving any perceptible residue.
It is almost insipid and colourless, transparent like the resins,
almost insoluble in water, but easily soluble in the essential
oils, in aether and concentrated nitric acid, and in most cases
it acts in general like the resins in a chemical and physical
point of view : thus friction renders it very electro-negative.
I will add, that the acid mixture, by means of which this resi-
nous body was obtained, has an extremely marked bitter taste.

I wished to make experiments also with other organic sub-
stances ; and I soon discovered, one after another, all those
about which there has been so much said of late, especially in
the Academy of Paris. All this passed in December 1845,

10 Prof, Schoenbein on the Discovery of Gun-^Cotton.

and the first few months in 1846. In March, I sent speci-
mens of my new compounds to some of my friends, in parti-
cular to Messrs. Faraday, Herschel and Grove. It is neces-
sary to note expressly that the gun-cotton formed part of these
products; but I must add, that hardly was it discovered when
1 employed it in experiments of shooting, the success of which
encouraged me to continue them. Accepting the obliging
invitation which I received, I went in the middle of April to
Wurtemburg, and made experiments with gun-cotton both in
the arsenal of Ludwigsburg, in the presence of artillery offi-
cers, and in Stuttgard, before the king himself. In the course
of May, June and July, with the kind cooperation of the
Commandant de Mechel, of M. Burkhardt, captain of artil-
lery, and other officers, I subsequently made in this city (Bale)
numerous experiments with arms of small calibre, such as
pistols, carbines, &c., and afterwards with mortars and can-
non,— experiments at which Baron de Kriidener, the Russian
ambassador, was several times present. I may be allowed to
mention, that I was the person who fired the first cannon
loaded with gun-cotton and shot, on the 28th of July, if I
remember aright, after we had previously ascertained, by ex-
periments with mortars, that the substance in question was
capable of being used with pieces of large calibre.

About the same time, and indeed previously, I employed
gun-cotton to blast some rocks at Istein in the Grand Duchy
of Baden, and to blow up some old walls at Bale ; and in
both cases I had opportunities of convincing myself in the
most satisfactory manner, of the superiority of this new explo-
sive substance over common gunpowder*.

Experiments of this kind, which took place frequently and
in the presence of a great number of persons, could not long
remain unknown ; and the public journals soon gave, without
participation on my part, descriptions, more or less accurate,
of the results which I had obtained. This circumstance, joined
to the short notice which I inserted in the May number of Pog-
gendorff's Annalen, could not fail to attract the attention of
German chemists : in the middle of August I received from
M. Boettger, Professor at Frankfort, the news that he had
succeeded in preparing gun-cotton and other substances. Our
two names thus became associated in the discovery of the sub-
stance in question. To M. Boettger the gun-cotton must have
been particularly interesting, as he had previously discovered
an organic acid which deflagrates readily.

In the month of August I went to England, where, assisted

* In tlie month of June I made also the first capsules, and employed
them with success for muskets, in the presence of the above-named officers.

Prof. Schoenbein o« the Discovery of Gun-CoUon, 11

by the able engineer, Mr. Richard Taylor of Falmouth, I
made numerous experiments in the mines of Cornwall, which
were entirely successful, in the opinion of all competent wit-
nesses. Experiments on the action of gun-cotton were also
made in several parts of England, under my direction, both
with small fire-arms and with pieces of artillery, and the re-
sults obtained were very satisfactory.

Until that time there had been little or nothing said of
gun-cotton in France ; and it will appear that the short notices
which Mr. Grove gave at Southampton at the meeting of the
British Association, and the experiments with which he ac-
companied them, served first to attract the attention of French
chemists to this substance. At Paris, the thing was at first
considered hardly credible, and jokes even were passed upon
it ; but when there could no longer remain any doubt as to
the reality of the discovery, and when several chemists in
Germany and other countries had published the processes
which they employed to prepare the gun-cotton, then a lively
interest was manitiested in a subject which had just before ex-
cited derision, and it was soon pretended that the new explo-
sive substance was an old French discovery. It was declared
to be nothing more than the xyloidine first discovered by M.
Braconnot, and afterwards investigated anew by M. Pelouze,
and the only merit left me was to have conceived the happy
idea of putting this substance into a gun-barrel. The know-
ledge of the composition of xyloidine ought to have sufficed
to convince those who put forward that opinion, that it is not
suited for fire-arms, on account of its containing too much
carbon and too little oxygen for the chief part to be converted
into gaseous matters during the combustion. It was moreover
very easy to discover the essential differences which exist be-
tween the xyloidine of Braconnot and gun-cotton. Never-
theless the error was kept up for some months.

Matters stood thus, when, on the 4th of last November, a
Scotch chemist, Mr. Walter Crum of Glasgow, published a
memoir, in which he showed that gun-cotton is not the same
product as xyloidine, but that it presents an essentially differ-
ent composition ; and towards the end of the same month, the
French Academy received a communication of the same na-
ture. The gun-cotton was then no longer xyloidine, it was
called pyroxyloidine, and the first was admitted to be unsuit-
able for fire-arms.

If, therefore, it is proved that from the commencement of
1846 I prepared gun-cotton, and applied it to the discharge
of fire-arms, and that M. Bcetlger did the same in the month
of August, — if it be admitted that xyloidine cannot serve the

12 The Rev. B. Bronwin on the Inverse Calculus

same purposes as this cotton, and if it be notoriously known
that what is now called pyroxyloidine was not brought before
the French Academy and the scientific world until towards
the middle of last November, the idea of attributing to France
the discovery of gun-cotton cannot be seriously entertained,
or of assigning to me merely a practical application of that
which another would have discovered.

I appeal to the justice of Frenchmen, to decide the point to
whom belongs the honour of not only being the first to apply
the new substance in question, but also of having first pre-
pared it — to MM. Braconnot and Pelouze, or myself. I
must, moreover, add expressly, that it was not xyloidine even
which led to my discovery, however intimate may be its rela-
tion with gun-cotton ; it was theoretical ideas, possibly very
erroneous ones, but which are peculiarly my own, as well as
some facts which I was also the first to discover. Suum cuique
is a principle of morality on which society at large rests ;
why should it not be strictly respected in the republic of sci-
ence? M. Pelouze is a distinguished chemist, and already
possesses a sufficiently high reputation not to require to ele-
vate his pretensions on the merits of others ; and I am fully
persuaded that this estimable chemist, of well-known truth
of character, will, appreciating with impartiality the circum-
stances which have occurred, freely render me the justice to
which I consider myself entitled.

Bale, Dec 28,1846.

III. On the Inverse Calculus of Definite Integrals.
^ By the Rev. Biiice Bronwin*.

"T^HIS paper contains several very simple and easy methods
■■ in the inverse calculus of definite integrals; and they
show that the function under the sign of integration may have
more than one form. The exponents n and p are always
positive, and n+p = i an integer.

First, let (p{x)='ZAmX"^i an ascending series. Then

=2A;„a'"+'* /" x'^-^dx{\^x)'^

= r(w)2;A^«-+« r^^'^^^^j) =4'(«) suppose.
* Communicated by the Author.

of Definite Integrals. 13

Tlien also

and

or

Operate with ( y j on both members, and we have
by making a— ;r = «t;. Therefore

A

Next, let (p (a?) = 2-^, a descending series. Then

<p(a + a;) = 2 ^"^

(1-)

and r"^ n-M ^f , N VA r"^ x^-'dx

- ^ -^;^J^ (TT^ ~ ^''''^;^^^ r(m) -'^^''^

suppose. Therefore

^, >^ A;;, Tlm—n) ,, ,

and
or

(2.)

14 The Rev. B. Bronwin on the Lwerse Calculus

Operate with (— j on both members; then

{-iyT{n)T{p)^ ^ = {£)J^ ^v-d,^a + a^) ;
or

^ 0
We may put ip [a) under a different form by making

a+d?= -. The forms of f [a) obtained in (1.) and (2.) differ

from those given by Mr. Boole in the Cambridge Mathema-
tical Journal, No. 20 ; but by varying the process a little, we
might obtain his results. We may observe that the least
valueof 7win(l.) must be greater than (—1), and in (2.) greater
than n+p or /.

In (p(^) = e^^(p(0), which is Taylor's theorem (D standing for

-r-), change (^[x) into <p (s-^), and then x into log x\ we have

(p(a:) + .r»<p(eO) (a.)

Tlierefore, also,

<^{a—x) = {a — x)^<^{s%
and

r''x'^-^dx<^[a — x)=i< J x'^-'^dx{a—x)^ \<^{^^)
Consequently

and

rwimi)(„_^)D.„^(,o)=^(,_^),

or

= / xP-'^dx'\>{a—x)',

nn)np) Y{S^rU) "''^'■^^^°) ^fo""'" ^-^^(^-^^

Operating with ( y-) on both members, we find

of Definite Integrals. [5

the same result as in (1.).

In (a.) change <p (x) into <p (-), and then x into—; we

have

• <^{x)=x-^<^{i-')\ {h.)

and therefore

(p(a4-^) = (a + ^)-"f(6-*'),
and

r(?z)r(D-w) „ n / m •/ X

Hence
and

or

r(«)r(;.) ^]5z^) ai-D^(,-o) =^ °° ^p-1 ,/^^(« + ^^).

And, as before,

( - 1)' r(w)r(p)a-»^(e-«) = ( - 1 )T(«)r(;;)(p(a)

\daj J Q

X

a;''~*(/a;\|/(a + ^)j

the same result as in (2.).

We might by this method derive the forms of (p {a) given
by Mr. Boole; but my object is merely to show one use out
of many which may be made of the formulae (a.) and {b.)

If Ar=l, and E=l+A; 'EJ'r=r-\-k,W'x'' =x'-x^. Giving
to k an infinity of different values, multiplying the results by
any constants, and taking the sum, we have

x^(p{x) = (^{^)x'' (c.)

It is plain that we may give to /^, not only integer values,
but fractional ones also, and any values whatever, and nega-
tive as well as positive ones; for the operation E''^ performed
on r, or on a*", merely changes them into r + k, and ^''^^ re-
spectively. The function (^{x) is therefore very general.

16 'The Rev. B. Bronwin on the Inverse Calculus

Change x into a—x, and we have

{a — xY-<p {a-x) = f (E) {a - xf.
Therefore

/* V-i dx (a ^xYfia-x) = <p (E) /* V-^ dx{a - xY

*y 0 <^ Q

= f{E)ar-^-/^\--.dx{l -^)'' = r(«)f (E)a''+«j,p^^^ =Ha)

suppose. Change a into «— ^, and we have

T{n)f(,E)ia-xY'--^^^:^^ =Ha^x),
and

rO^)'P(E)_y'"^^\. rxP-hlx[a-xY^''=r''xP-'dx'\;{a-x)',
or

and

r(w)r(p)f(E)a'-=r(«)r(;))fl'-<p(a)=^^)^"^^-'^7^4'(«-^).

Change ^'"(p (.r) into <^ {x)^ and transform the second member ;
then

as before.

Resuming the equation

T{n)<^{E)[a-wY^^ ^^}f) =^(«-^)'
we have

or

Multiply by a}-P=ia'^, and operate with (-7- j ; there results
r(«)r(;3)f (E)a'-/^= Cj-Yf v-^dv{l -v)P-'^av).

of Definite Integrals. 17

If therefore we change x^<^{fc) into <^{x\ we now have

(3.)

0

If we put D for ^, then

dr

and, as in (c), we have

6''f^(^)=(p(D)£'-S s-'-'^^(^)=(p(-D)s-'-^ . . (r7.)
Or, if we put p* for the operation which converts e'"' intoa^e'*'',
and ^^ for that which changes s~'* into a^^s"'"'^, then

and A: may be positive or negative, integer or fractional, or
any quantity whatever. I believe these forniulae are new,
and they admit of many uses.
Changing x into a + ^, we have

and

= r(«)f(S) ^^''""^Ha) suppose.
Changing a into a+x, this gives

and

' ^ 0 '

0

or

(-i)'r(w)r(;7)^(fl)s— = (-i)'r(w)r(p>— f(a)

Change the function e-''*'(p(^) into <p(a;), and we have

/ xP-^dx^{a + x)

as heretofore.

Phil, Mag. S. 3 . Vol. S 1 . No. 205. Jidy 1 84-7 . - C

18 The Rev. B. Bronwin on the Inverse Calculus

The equation

r(nM9)^e--=4;(«),

found in this investigation, gives

and

T{n)f{&)\r v-Pdv{l-v)P-^s-^CO=J v-Pdv

Differentiate this for a, then
Make 1 =.27, and

V

r v-P-^ dv{ 1 - v) p- ^ s-'v«y

*y 0

will be transformed into

0

.—ra

J Q rP

and the preceding will become

Multiply by aP, and operate with ( ;t- ) j there results
After changing the function (p, as before, we now have

(ra)'"«4/>-'*(>-^)'-<l)-J

The formulae (5.) and (4 ) are the same as those given by
Mr, Boole in the paper before referred to. From the last

of Definite Integrals. 19

method of investigation, it appears that the functions <p and \p
may be any whatever, consistent with the required relation
between them. But if we are obliged to integrate by series,
they will in general be subject to the restrictions mentioned in
(1.) and (2.); I say, in general, for infinite quantities may

vanish by the operations ( ^ ) •

To give an example in each of the theorems : in (1.) let

We find

w=2>P=2»4'(^)='/^.

{K0}V)=|(f).

and , V 1

then 4/(a)='/a,

as it should be.
In (2.) let

We find

In the last example n and p are not conformed to the re-
strictions, but the infinite quantity goes out by differentiation.
The theorems (3.) and (4.) are likewise satisfied by these ex-
amples. It must not be supposed that the values of <p (a),
given in (1.) and (3.), or in (2.) and (4.), are necessarily equal;
for they will not reduce the one to the other. Yet we may
have

^ 0 ^ 0

in both cases ; since we know from examples that the inte-
grals of different functions may be of the same form.

Gunthwaite Hall, near Barnsley,
May 24, 1847.

C2

[ 20 ]

IV. On certain Phenomena of Voltaic Ignition and the De-
composition of Water into its constituent Gases by Heat.
% W. R. Grove, Esq., M.A., F.R.S.'^

IN the Philosophical Magazine for August 1841, I recom-
mended for eudiometrical purposes, the use of a platinum
wire ignited by a voltaic battery. In fig. 1 is re- Fig. 1.
presented a form of apparatus for this purpose ; it .
consists of a tube of Bohemian glass, vi'ith a loop of ^
platinum wire |j\jth of an inch diameter sealed into
its upper end ; the size of the glass tube may be
adapted to the quantity of gas sought to be analysed,
and may when necessary be reduced to extremely
small dimensions, one-eighth of an inch being ample:
into this the gas may readily be made to ascend, by
the insertion of a wire of copper, platinum, or glass,
as may be suitable to the gas : two cells of the nitric-
acid battery are sufficient fully to ignite the wire,
and the same battery supplies, by electrolysis, pure
oxygen and hydrogen for the analysis. Since the -— '^^—
period when I first proposed this, I have seldom used any
other apparatus for such gaseous analyses as are performed
by combining the gas to be examined with oxygen or hydrogen.
This eudiometer possesses the advantage of enabling the ope-
rator either to detonate or slowly to combine the gases, by
using different powers of battery, by interposing resisting
wires, or by manipulation alone, — a practised hand being able
by changing the intervals of contact to combine or detonate
the gas at will. My general practice has been to produce a
gentle heat in the wire until the gases contract, and then gra-
dually to increase the heat until a full ignition takes place, by
which means all the objects of the eudiometer of Volta are
fulfilled, without detonation, without dependence on the fickle
electric spark, and without thick tubes, any danger of explo-
sion, or of the gases being projected from the eudiometer.

I have commenced with a description of this eudiometer,
as it has been indirectly the means of my undertaking the
experiments detailed in this lecture; and as its very great
convenience has never been generally understood, I think that
in strongly recommending it, I shall be of service to chemists.

In a paper honoured by insertion in the Philosophical
Transactions for 1845, p. 358, I have shown another method
of eudiometry also performed by voltaic ignition ; in that ex-
periment the vapour of camphor was decomposed into car-
bonic oxide and carburetted hydrogen ; it was an application

* From the Philosophical Transactions for 1847, part i. ; having been
received by the Royal Society September 3, and read November 19, 1846.

Mr. Grove on the Decomposition of Water by Heat. 21

of voltaic ignition to effects analogous to those produced by
Priestley and others, by passing compound gases through
ignited tubes of porcelain.

But the voltaic process has this immense advantage, that
the heat can be rendered incomparably more intense ; that
the quantity of vapour or gas to be operated on may be inde-
finitely small ; that there are no joints, stop-cocks or ligatures;
and that there is no chance of ehdosmose, which takes place
through all porcelain vessels. I therefore determined to
examine by these means several gases, both with a view of
verifying, under different circumstances, known results, and
seeking for new effects by this new and advantageous appli-
cation. I used an eudiometer (fig. 1) of 8 inches long and
0'4' inch internal diameter, exposing the gases to intense heat,
and subsequently analysed the residues in one of the same
length, but 0*2 inch diameter.

I will first consider the physical effects of different gases on
the ignition of the wire itself.

In a paper on the Application of Voltaic Ignition to lighting
Mines*, I have mentioned the striking effects of hydrogen in
reducing the intensity of ignition of a platinum wire, so much so
that a wire voltaically ignited to incandescence in atmospheric
air, is apparently extinguished by inverting over it a jar of
hydrogen ; with other gases the effects are not so striking, and
with them these differences are best shown by including a volta-
meter in the circuit. Davy found that the conducting power of
a wire diminished in proportion to the degree to which it was
heated : assuming the accuracy of this position, the amount of
gas in the voltameter would be inverse to the intensity of ig-
nition in the wire. The following is the result I obtained with
different gases, employing the same battery (the nitric-acid
combination at its most constant period), the same wire, and
the same vessel : —

Cubic inches of gas evolved in
Gases surrounding the wire. the voltameter, per minute.

Hydrogen ....... 7*7

defiant gas 7*0

Carbonic oxide ...... Q'Q

Carbonic acid 6'6

Oxygen 6'5

Compressed air, 2 atmospheres 6'5

Nitrogen 6'4

Atmospheric air 6*4

Rarefied air Q'3

Chlorine 6-1

* Phil. Mag. Dec. 1845.

22 Mr. Grove on the Decomposition of Water by Heat.

To ascertain the relation between the amount of radiant
heat generated by the same battery and wire in gases which
presented striking differences as to the luminous effects of the
platinum wire, an apparatus was prepared in which the bulb
of a thermometer was retained at a certain distance from the
coil of wire ignited by a battery of four cells, and exposed,
first, to an atmosphere of hydrogen, and then to one of atmo-
spheric air, at the same temperature and pressure ; the ther-
mometer rose 1^° in five minutes in the hydrogen, and 15°
in the air in the same time. Both the heating and luminous
effects appear therefore to be greater in atmospheric air than
in hydrogen. I cannot satisfactorily account for the differ-
ences shown in the above table ; there appears a general ten-
dency to greater ignition in the electro-negative than in the
combustible gases, but the facts are far too few to found a
generalization. I was at first inclined to regard the difference
of effect in hydrogen as analogous to the peculiarity mentioned
by Leslie* respecting its convection of sound, but the pa-
rallel does not hold ; sound is transmitted imperfectly through
rarefied air, and also through hydrogen; on the contrary, the
heat of the ignited wire is most intense in the former, and
least so in the latter ; the heat is also very much reduced in
intensity in the compounds of hydrogen, ammonia and olefiant
gas, or even by a small admixture of hydrogen with another
gas, such as nitrogen ; hydrogen, therefore, appears to have
a peculiar and specific action in this respect.

I now pass to the consideration of the effects of the ignited
wire on different gases. The ignition was in every case raised
to the fullest extent, and the gases after exposure to it were
carefully cooled down to their original temperature.

When the experiments were
made over water, the whole
eudiometer was immersed in a
vessel of distilled water, occa-
sionally having an inch depth of
oil on the surface (see fig. 2t);
when over mercury, and a long-
continued exposure was required,
a bent tube was employed, as at
fig. 3, the closed end being im-
mersed in water or oil, to prevent
the fusion of the glass which
would otherwise have ensued.

* Transactions of the Cambridge Philosophical Society, vol. i. p.267'
t In this and in figs. 3 and 5, the lines leading from the platinum loop
to tfie mercury cups represent copper wires.

Fig. 2.

Fig 4.

J'x

Mr. Grove on the Decomposition of Water by Heat, 23

The tubes are much more easily preserved from cracking,
and the ignition better kept up with oil on the exterior than
witli water, but as in many of these experiments I might have
been considerably mis- Fig. 3.

led by a crack in the
glass, or a bad seal-
ing of the wire, al-
lowing a portion of
oil to enter the tube,
I used water in the
greater number of
them until I was as-
sured of the phaenomena.

The apparatus, fig. 3, is superior in one respect
to fig. 2, even for experiments over water, as the
wire being situate below the volume of gas, the circu-
lation is more rapid. This object may also be effected
by employing the form of eudiometer, fig. 4, in which
the loop of wii*e is near the centre of the tube, so as
to be just above the surface of water in the tube;
there are, however, some difficulties of manipulation
with this form, which render it practically of less
value than fig. 1.

Binoxide of nitrogen o\ev distilled water contracted
differently in proportion to the heat of the wire ; in
the best experiment it contracted to one-third of its
original volume; the residual gas was nitrogen.
Nitric acid was found in solution in the water.

Over mercury the effects were nearly the same ; the mer-
cury was attacked, and the orange fumes of nitrous acid were
visible.

Protoxide of nitrogen was decomposed into nitrogen and
oxygen ; the volume increased by 0*35 of the original volume ;
I could not get the full equivalent proportion, or 0*5 of oxygen.
Carbonic acid underwent no perceptible alteration.
Ammonia increased to double its original volume ; it was
now no longer absorbable by water, and gave three volumes
of hydrogen, plus 1 nitrogen.

Olejiant gas contracted slightly, deposited carbon, the residue
being hydrogen and olefiant gas, more of the former in pro-
portion to the heat, but I could not succeed in entirely de-
composing it.

Nitrogen suffered no change.

Oxygen gave a very slight contraction, amounting to jV^h
of its volume ; the oxygen employed was very pure, obtained
from chlorate of potash and manganese, and also from water

uL

24 Mr. Grove on the Decomposition of Water by Heat.

by electrolysis : no change in properties was perceptible in
the oxygen after its exposure to the ignited wire. This con-
traction I incline to attribute to a slight portion of hydrogen
present, which view will, I think, be considered as strengthened
by the effect of the ignited wire on hydrogen, to be presently
detailed. I at one time thought that the contraction might be
due to a slight oxidation of the wire, but it never went beyond
a very limited point; nor was the wire altered in size or weight,
though it was kept ignited for many hours.

Chlorine over water gave dense white fumes; a grayish-
yellow insoluble powder accumulated on the sides of the tube
near the platinum wire, which appeared of the same nature
as the vapours ; the deposit was insoluble in cold nitric, sul-
phuric, or muriatic acid, but dissolved by the last when boiled.
The fumes did not, as far as I could judge, affect litmus paper ;
a barely perceptible tinge of red was indeed communicated to
it, but this, I had every reason to believe, was attributable to
a slight portion of muriatic acid not absorbed by the water.
I have not yet woi'ked out this result, as it is probable, con-
sidering the number of experiments that have been made on
heated chlorine, that it is a known product, though I cannot
find, in several books to which I have referred, any substance
answering to it in description, and the field opened by voltaic
ignition is so new that each result demands a separate and
prolonged examination ; if I find that this is an unknown
compound I shall probably resume its investigation*.

Cyanogen gave, though in very minute quantities, a some-
what similar deposit, but at its then very high temperature it
began to act rapidly on the mercury, and I was obliged to
give up the experiment after an hour's ignition. Both these
gases require peculiar and novel apparatus for examination
by voltaic ignition. It will presently be seen that my whole
attention and disposable time were necessarily occupied with
certain phaenomena to which this class of experiments ulti-
mately led me.

Hydrogen gave a very notable contraction, amounting in
some cases to one-tenth of its volume. This was an unex-
pected result, and I examined it with care. It took place
both over water and over mercury; rather more with the
former than with the latter. It obtained equally with hydrogen
procured by electrolysis from carefully distilled water and pure
sulphuric acid ; with that procured from common zinc and
pure sulphuric acid diluted with distilled water; and with
that obtained from distilled zinc and pure diluted sulphuric
acid. The contraction was less when the water from which
* See Supplemental paper.

Mr. Grove on the Decomposition of Water by Heat. 25

the hydrogen was obtained was carefully purged of air by
boiling and the air-pump, but yet there was a notable con-
traction-even when the water had been freed from air to the
utmost practicable extent. In the numerous experiments
which I made on this subject, the contraction varied from the
y\jth to the 7\jth of the whole volume.

After many fruitless experiments I traced it to a small
quantity of oxygen which I found hydrogen to contain under
all circumstances in which I examined it. Phosphorus placed
in hydrogen, obtained with the utmost care, gives fumes of
phosphorous acid, shines in the dark and produces a slight
contraction, but there is after this a further contraction by the
use of the ignited wire.

I may cite the following as an easy experiment and simple
illustration of the rapidity with which hydrogen appropriates
oxygen. Let hydrogen be collected over water well-purged
of air; let a piece of phosphorus remain in it until all com-
bustion has ceased, the hydrogen will then be full of phos-
phoric vapour ; fill another tube with water, and pass the
hydrogen rapidly into it, the second tube will instantly be
filled with a dense white fume of phosphorous acid ; the hy-
drogen having instantly carried with it oxygen from the stra-
tum of waier it has passed.

A very careful experiment was made as follows : — distilled
water was boiled for several hours, to this was added one-
fortieth part by measure of pure sulphuric acid, and it was
cooled under the receiver of an air-pump ; it was now placed
in two test-glasses, connected by a narrow inverted tube, full
of the same liquid: platinum electrodes were placed in each
glass, and the hydrogen caused to ascend immediately into the
eudiometer tubes ; the whole was completed within two or
three minutes after the water had been removed from the air-
pump. Here the ordinary sources of impurity in hydrogen
were avoided ; no zinc was used, the sulphuric acid was pure,
and the quantity was so small, that, had it not been pure, the
error could have been but very trifling. The hydrogen so
obtained, contracted in volume ^-^ih ; hydrogen prepared in
the same way, and exposed to phosphorus, gave dense white
fumes; the phosphorus was luminous in the dark for more
than an hour, and the contraction (temperature and pressure
being carefully examined) was g\jth ; the amount of contrac-
tion by the wire would of course equal three times the volume
of oxygen mixed with the hydrogen, consequently the oxygen
would be y^jth of the whole volume; the platinum wire induces
therefore a greater absorption of oxygen than the phosphorus,
unless the volume of hydrogen is increased by the phosphoric

26 Mr. Grove oti the Decomposition of Water hy Heat.

vapour; the sequel of this paper will render it probable that
even the ignited wire does not and cannot induce combination
of all the oxygen existing in the hydrogen. •

I have looked into the papers of MM. Berzelius and Du-
long, and of M. Dumas on the equivalent weight of hydrogen.
The latter contains a most careful experimental investigation,
and is by far the best determination we have ; although it is
not there mentioned that hydrogen contains oxygen, yet a
correction is made for the air contained in the sulphuric acid
employed. M. Dumas does not state how the quantity of that
air is calculated. There can be no question that nothing ap-
proaching in elaborate care to these experiments has been yet
performed on the subject; but with the fullest consciousness of
M. Dumas' skill, I have, in all my experiments, perceived
such an inveterate tendency of hydrogen to possess itself of
oxygen, that I cannot help entertaining some doubts whether
we have yet the real weight of hydrogen within the assigned
limits of error.

It is difficult to see how hydrogen can be absolutely deprived
of oxygen which has once existed in it; neither an oxidable
metal as zinc, or an ignited inoxidable metal as platinum, get-
ting rid of all the oxygen, and phosphorus, if it does so, re-
places it by its own vapour. The near approach, however, of
the equivalent of hydrogen, as determined by M. Dumas, to
the ratio of whole numbers, renders it probable that it is a
very close approximation to the truth.

1 have not been able to detect nitrogen in the hydrogen,
but the probability is that a slight quantity also exists in it.
Whether the oxygen proceeds from portions of air still re-
maining in solution in the liquid from which the air is ex-
hausted, or whether it is a part of the water actually decom-
posed, but of which the oxygen is not absorbed by the zinc,
is a question to resolve which further experiments are neces-
sary.

Hydrogen and carbonic acid mixed in equal volumes were
readily acted on by the ignited wire; they contracted to O'^S
of the original volume ; the residue was carbonic oxide ; one
equivalent of oxygen had therefore united with the hydrogen;
and the slight additional contraction was probably due to the
further combination of hydrogen with oxygen, as above stated.

Carbonic oxide exhibited a remarkable effect, and one which,
coupled with the last experiment, gave rise to considerations
which mainly led to the results to be detailed in the body of
this paper. Carbonic oxide, very pure and carefully freed from
carbonic acid, was exposed to the ignited wire over distilled
water ; the gas increased in volume in one experiment to one-

Mr. Grove on the Decomposition of Water by Heat. 27

third of its original volume, in the greater number of instances
to one-fifth : this increase depended upon the intensity of ig-
nition, which it was very difficult to maintain at its maximum
on account of the frequent fusions of the platinum wires.

Here again I had a long research and many erroneous
guesses, which I need not detail. The effect did not take
place with perfectly dry gas over mercury, and I thence was
led to attribute it to some combination with aqueous vapour;
the increase turned out to be occasioned by the formation of
carbonic acid. By agitation with caustic potash or lime water
the gas was reduced to exactly its original bulk, but it was
now found to be mixed with a volume of hydrogen equal to
the volume of carbonic acid by which it had been increased ;
it was thus perfectly clear that half a volume or one equivalent
of oxygen derived from the vapour of the water, had combined
with one volume or equivalent of carbonic oxide, and formed
one volume or equivalent of carbonic acid, leaving in place of
the carbonic oxide with which it had combined, the one vo-
lume or equivalent of hydrogen with which it had been origi-
nally associated.

Comparing the last experiment, viz. that of mixed carbonic
acid and hydrogen with this, I was naturally struck with the
curious reversal of affinities under circumstances so nearly
similar ; in the one case, hydrogen taking oxygen from car-
bonic acid to form water and leaving carbonic oxide ; in the
other, carbonic oxide taking oxygen from water to form car-
bonic acid and leaving hydrogen.

1 thought much upon this experiment; it appeared to me
ultimately that the ignited platinum had no specific effect in
producing either composition or decomposition of water, but
that it simply rendered the chemical equilibrium unstable, and
that the gases then restored themselves to a stable equilibrium
according to the circumstances in which they were placed with
regard to surrounding affinities ; that if the state of mixed
oxygen and hydrogen gas were, at a certain temperature,
more stable than that of water, ignited platinum would de-
compose water as it does ammonia.

This is a very crude expression of my ideas, but we have
no language for such anticipatory notions, and I must adapt
existing terms as well as I am able.

It now appeared to me that it was possible to effect the
decomposition of water by ignited platinum ; that, supposing
the atmosphere of steam in the immediate vicinity of ignited
platinum were decomposed, or the affinities of its constituents
loosened, if there were any means of suddenly removing this
atmosphere I might get the mixed gases ; or secondly, if, as

28 Mr. Grove on the Decomposition of Water hy Heat.

appeared by the last two experiments, quantity had any influ-
ence, that it might be possible so to divide the mixed gases
by a quantity of a neutral ingredient as to obtain them by
subsequent separation (or as it were filtration) from the neu-
tral substance. Both these ideas were realized.

To effect the first object, after, as usual in such circum-
stances, much groping in the dark, I cemented a loop of pla-
tinum wire in the end of a tube retort similar to fig. 3, and
covered it with asbestos, ramming this down so as to form a
plug at the closed extremity of the tube, the platinum wire
being in the centre. My object was, by igniting the platinum
wire, to drain the water out of the asbestos, and the ignited
wire being then in an atmosphere of steam, I hoped the water
would by capillary attraction keep constantly oozing down to
the platinum wire as the steam or decomposed water ascended.
The experiment did not succeed ; the water established a
current through the asbestos by washing away fine particles,
and the phaenomena of ordinary ebullition took place, unless
the intensity of the battery was very much exalted, when a
very slight decomposition was perceptible, which I attributed
to electrolysis. This experiment, however, suggested another
which did succeed. In one or two cases the asbestos plug
became compressed above the platinum and so choked up the
tube that the wire suddenly fused. It now occurred to me
that by narrowing the glass tube above the platinum wire I
had the result at my command, as the narrow neck might be
made of any diameter and length, so as just to allow the water
to drip or run down as the steam forced its way up; a tube
was so formed, and is shown with its accompaniments at fig. 5.

Fig. 5.

The result of this experiment was very striking: when two
cells of the nitric-acid battery were applied the air was first
expanded and expelled, the water then soon boiled, and at a
certain period the wire became ignited in the steam. At this
instant a tremulous motion was perceptible, and separate
bubbles of permanent gas of the size of pin-heads ascended,

Mr. Grove on the Decompositioji of Water hy Heat. 29

and formed a volume in the bend of the tube. It was not a
continuous discharge of gas as in electrolysis, but appeared to
be a series of rapid jerks ; the water, returning through the
narrow neck, formed a natural valve which cut off by an in-
termitting action portions of the atmosphere surrounding the
wire; the experiment presented a novel and indescribably
curious effect. The gas was oxyhydrogen. It will occur at
the first to many of those who hear this paper read, that this
effect might be derived from electrolysis. No one seeing it
would think so for a moment; and although I shall by my
subsequent experiments, I trust, abundantly negative this sup-
position, yet as this was my first successful experiment on this
subject, and is 'per se an interesting and striking method of
showing the phaenomenon of decomposition by heat, I will
mention a few points to prove that the phaenomenon could
not be occasioned by electrolysis.

To account for it by electrolysis, it must be supposed that
the wire offered such a resistance to the current that this di-
vided itself, and the excess of voltaic power passed by the
small portion of water which trickled down, instead of by the
wire.

In the first place, the experiment was perfornved with di-
stilled water, and only two cells of the battery employed, which
will not perceptibly decompose distilled water.

2ndly. No decomposition took place until the instant of
ignition of the wire, though there was a greater surface of
boiling water exposed to the wire before than after the period
of ignition.

Srdly. A similar experiment was made, but with the wire
divided in the centre so as to form two electrodes, and the
water boiled by a spirit-lamp ; here the current had no wire
to conduct any part of it away, but the whole was obliged to
pass across the liquid, and yet no decomposition took place,
or if there were any it was microscopic.

4thly. When, instead of oil, distilled water was used in the
outer vessel*, even the copper wires, oneof which would form
an oxidable anode, gave no decomposition across the boiling
water outside, while the ignited wire inside was freely yielding
mixed gases.

5thly. To prevent the water from being the shortest line
for the current, I repeated the experiment with a perfectly
straight wire (fig. 6). The result was precisely the same,
but the experiment is more difficult ; as a certain length of

* January 8.— I have since found that the exterior tube of oil or water
may be dispensed with in this experiment, as the water which trickles
down prevents the fusion of the glass.

so Mr. Grove 07i the Decomposition of Water hy Heat.

wire is necessary, the sealing is more troublesome, and the
size of the bulb is much more difficult to adapt to the produc-
tion of steam in exactly the requi- Fig. 6.
site quantity ; the straight wire
being more suddenly extinguish-
ed and more easily fused : with
careful manipulation however it
succeeds equally well with the
former experiment.

I might add other experiments
and arguments, but I believe when
the remainder of this paper has been read, that the above will
be thought scarcely necessary.

I now directed all my efforts to produce the effects by heat
alone without the battery. I will mention a few of my unsuc-
cessful attempts, as it will save trouble to future experimenters.
I sealed a platinum wire into the extremity of a curved tube,
filled the latter with water, and applied a strong heat by the
blowpipe to the projecting end of the wire, hoping that the
conducting power of the platinum, although inferior to that of
most other metals, was sufficiently superior to that of glass to
enable me to ignite the portion of the wire within the tube,
and thus surround it with an atmosphere of steam ; the water
however all boiled off from the glass ; nor could I succeed in
igniting the platinum by heat from without. A similar failure
occurred when, on account of its superior conducting power,
a gold wire was substituted for that of platinum.

I sealed spongy platinum and bundles of platinum wire into
the ends of Bohemian glass tubes, closing the glass over them,
and then filling the tubes with water and heating the whole
extremity; but the water boiled off" from the glass, and the
platinum could not be made to attain a full incandescence.

After many similar trials I returned to the battery, and
sought to apply it in a manner in which electrolysis could not
possibly take place. I had hoped, as I have above stated, to
obtain a residual decomposition of water by masking or dilu-
ting the gases by a neutral substance. I therefore tried the
following experiment: a tube Fig. 7.

similar to fig. 1 was filled with
water which had been carefully
freed from air by long boiling
and the air-pump ; it was then
inverted in a vessel of the same
water, and a spirit-lamp applied
to its closed extremity, until the
upper half was filled with va-

Mr. Grove o?i the Decomposition of Water by Hetit. 31

pour (fig. 7). The wire was brought to a full ignition by
the battery, and kept ignited for a few seconds ; connexion
was then broken and the lamp removed, so that the water
gradually ascended. A bubble of the size of a large mustard-
seed was left in the extremity of the tube, and I was much
gratified at finding that when this was caught by a lighted
match at the surface of the water-trough it detonated. The
experiment was then repeated, continuing the ignition for a
longer time, but the gas could not be increased beyond a
very limited quantity ; indeed it was not to have been expected,
as supposing it to be mixed gas, recombination of the excess
would have taken place, and the fact of any uncombined gas
existing when exposed to incandescent platinum, will doubt-
less surprise those who hear it for the first time.

I'he experiment was repeated as at first and the bubble
transferred to another tube ; the wire was then again ignited
in vapour, another bubble was instantly formed and trans-
ferred, and so on, until after about ten hours' work sufficient
gas was collected for analysis ; this gas was now placed in an
eudiometer, it detonated and contracted to 0*35 of its original
volume ; the residue being nitrogen. The experiment was
repeated several times with the same general results, the re-
sidue sometimes containing a trace of oxygen.

Here electrolysis was out of the question ; the wire was
ignited in (if I may use the expression) dry steam, the upper
part of the tube being far above the boiling-point, and of
course perfectly transparent; if not an effiict of heat, it must
have been a new function of the electric current, at least one
hitherto unknown.

As the voltaic arc and electric spark afford heat of the
greatest intensity, I tried a succession of electric sparks from
platinum wires through steam in the apparatus fig. 8, the
water, as in all my experiments,
having been previously purged of Fig. 8.

air (to save circumlocution I will
in future call it prepared water).
The sparks were taken from the
hydro-electric machine of the
London Institution ; they had in
the steam a beautiful crimson ap-
pearance; on cooling the tube a
bubble was perceptible, which
detonated by the match.

As in the previous experiments, a whole day's work did not
increase the bubble, but when it was transferred another
instantly formed ; the gas was similarly collected ; it detonated

32 Mr. Grove on the Decompositioji of Water by Heat.

and contracted to 0*4 of its original volume; the residue was
nitrogen with a trace of oxygen.

This experiment will again surprise by its novelty ; the
very means used in every laboratory to combine the mixed
gases and form water, here decompose water*. From a vast
number of experiments which I have made on the voltaic and
electric disruptive discharges (which are I believe similar phae-
nomena, differing only in quantity and intensity), I believe the
decompositions produced by them are the effects of heat alone,
and this experiment was therefore to my mind a repetition of the
last under different circumstances; others however may think
differently. This experiment also I several times repeated.

By counting the globules given off, and comparing a cer-
tain number of them with the average volume of steam in the
last two experiments, an attempt was made to ascertain what
proportion of water could be decomposed by an ignited pla-
tinum wire in aqueous vapour, or, which amounts to a corol-
lary from this proposition, what degree of dilution would
enable mixed gas to exist without combustion in an atmo-
sphere of steam exposed to an ignited platinum wire. The
proportion in an experiment in which the globules were so
counted, was 1 to 2400 ; the probability is however that dif-
ferent temperatures of the platinum wire would give diffei*ent
volumes of gas so decomposed, the volume being greater as
the wire is more intensely ignited.

Although there was no known effect of electricity which
could produce the phaenomenon exhibited by the last two
experiments, and it was in any event new, still, firmly con-
vinced that it was an effect of heat, I again determined to
attempt its production by heat alone, and without the use of
the battery. I procured a tube of silver 9 inches long and
0*4 inch diameter; at the extremity of this was a platinum cap
to which a smaller tube, also Fig. 9.

of platinum, was soldered.
This platinum tube was
closed at the end and sol-
dered with gold solder. The
apparatus was filled with
prepared water ; the water
was boiled in the tube to ex-
pel the air from the narrow
tube and any which might
have adhered to the vessel ;

• I need scarcely point out the distinction, in fact, between this experi-
ment and those in which liquid water has been decomposed by the electric
spark. See Supplemental Paper.

Mr. Grove on the Decomposition of Water by Heat. 33

the tube was then, when full of hot water, inverted into water»
and the flame of a common blowpipe made to play against the
platinum tube (fig. 9) until a white heat was obtained. Upon
inverting it under water, a bubble of the size of a mustard-
seed rose to the surface, which gave a very feeble detonation
with the match. Similar bubbles were collected as before,
and the gas in an eudiometer contracted to 0'7. On re-
petition the experiment did not succeed so well, and upon
several repetitions it sometimes succeeded and sometimes
failed, and I should not mention it but that it was the first
experiment which gave me, although not very satisfactorily,
the effect of decomposition by heat alone. The reason of
its uncertainty I believe to have been the want of a suffi-
ciently intense heat, as I dared not venture on account of
the gold solder to push the ignition very far ; in fact, I sub-
sequently fused the extremity and spoiled the apparatus by
applying the oxyhydrogen flame to it; had the platinum tube
been welded instead of gold-soldered, it would doubtless have
succeeded better. I should state that the object of the silver
tube was to prevent the chance of recomposition by the cata-
lytic effect of a large platinum surface; to have, in short, a
small portion of platinum exposed to the steam, and that at a
high temperature: oeconomy was also no indifferent consi-
deration. This experiment, although, coupled with the pre-
vious ones, tolerably conclusive, did not satisfy me, and I
attacked the difficulty in another manner. The experiment
(fig. 5) induced me to believe that if I could get platinum
ignited under water so as to be in an atmosphere of steam,
decomposition would take place ; and M. Boutigny's experi-
ments on the spheroidal state of water led me to hope I might
keep platinum for some time under conditions suitable for my
purpose.

After a few failures I succeeded perfectly by the following
experiment. The extremity of a stout platinum wire was
fused into a globule of the size of a peppercorn, by a nitric-
acid battery of thirty cells ; prepared water was kept simmer-
ing by a spirit-lamp, with a tube filled with water inverted in
it ; charcoal being the negative terminal, the voltaic arc was
taken between that and the platinum globule until the latter
was at the point of fusion ; the circuit was now broken, and
the highly incandescent platinum plunged into the prepared
water : separate pearly bubbles of gas rose into the tube, pre-
senting a somewhat similar effect to experiment (fig. 5). The
process was repeated, the globule being frequently plunged
into the water in a state of actual fusion ; and when a sufficient
quantity of gas was collected it was examined, it detonated,

Phil. Mag. S. 3. Vol. 3 1 . No. 205. July 1847. D

34- Mr, Grove on the Decomposition of Water by Heat.

leaving 0"4< residue ; this was as usual nitrogen vi^ith a trace
of oxygen. A second experiment gave a still better result,
the gas contracting to 0*25 of its original volume.

On making the platinum negative and the charcoal positive,
a very different result followed : the carbon was, as is known
to electricians, projected upon the platinum; and the gas in
this case was mixed with carburetted hydrogen and carbonic
oxide. I know no experiment which shows so strikingly the
different effects at the disruptive terminals as this ; when the
platinum is negative it gives much carbonic gas, when it is po-
sitive, not a trace (the gas was delicately and carefully tested
for it); nay, more, by changing the platinum from negative to
positive the carbon is instantly removed, and in a single ex-
periment the platinum becomes perfectly clean.

Here then I produced very satisfactorily decomposition by
heat ; it is true, the battery was used, but used only as a means
effusing the platinum, as this was, as soon as fused, entirely
separated from the circuit and could have no possible voltaic
action. Wishing however altogether to avoid the use of the
battery, I repeated this experiment, employing as my means
effusing the platinum the oxyhydrogen blowpipe; the expe-
riment was equally successful, perhaps more so, as the mani-
pulation was more easy.

I could readily by this means collect half a cubic inch or
more of the gas ; when detonated, the residue of nitrogen
averaged 0*35 of the original volume.

In carefully watching this experiment, 1 observed that at
first a rapid succession of bubbles ascended into the tube from
the incandescent platinum, it then became quiescent; the
spheroidal state was assumed by the water and no gas
ascended; on losing the spheroidal state a sudden hiss was
heard, and a single bubble ascended into the tube. I deter-
mined to examine separately the gas from the platinum before
and after the quiescent state; to effect this I placed two in-
verted tubes in the capsule with the orifices near each other ;
the platinum at the point of fusion was immersed under one
tube, say tube A, and as soon as the ascent of bubbles ceased,
it was removed across to tube B, and the last bubble then
entered that tube ; the gases from each tube were separately
analysed, and tube A gave nearly all detonating gas, the re-
sidue being only 0*2 ; tube B gave none ; the gas collected in
it was nitrogen, with a trace of oxygen.

In order to examine the effect of an oxidable metal under
similar circumstances, I fused by the oxyhydrogen blowpipe
the end of a stout iron wire, plunged it into prepared water
and collected the globules of gas ; no oxygen was given off,

Mr. C. R. Weld on the Invention of Fluxions, 35

or at least no more than I have always found to accompany
hydrogen, which with a small residue of nitrogen was the gas
given off in this experiment.

[To be continued.]

v. Invention of Fluxions. By Charles Richard Weld, Esq.
To the Editors of the Philosophical Magazine and Journal.

Gentlemen,

IN the course of my researches in the archives of the Royal
Society, with reference to a history of the Society which I
am compiling, I have been much struck with a very remark-
able discrepancy on a most important point, connected with
the celebrated dispute of the invention of fluxions, between the
original Minutes of the Society and the statements of writers
on this subject.

Sir David Brewster and Professor De Morgan, following
others, state that at a meeting of the Society held on the 20th
of May 1714, a resolution was inserted in the Minutes, that
" it was never intended that the report of the committee was
to pass for a decision of the Society *." This alludes to the
report presented by the committee appointed by the Society
to determine the question of the invention of fluxions. Now
the exact words of the minute are these: —

" It was not judged proper (since this letter was not directed
to themf) for the Society to concern themselves therewith,
nor were they desired so to do ; but that if any person had
any material objection against the Commercium or the report
of the committee, it might be reconsidered at any time J."

There is nothing here to show that the Society resolved (and
this is the word Mr. De Morgan uses) upon repudiating the
report of their committee ; so far from this, the opposite con-
clusion is at once obvious, which is in keeping with the ori-
ginal resolution of the Society adopting the report of their
Committee, nemine contradicente. The point is of great mo-
ment ; for had the Society come to the resolution as repre-
sented, a strong case would be made out against Newton. I
have examined the Minutes of the meeting in question with
the greatest care, and confidently assert that there is no other
allusion to the dispute between Leibnitz and Newton. In
conclusion, I wish to state that it is at the request of some of

* See Life of Newton by Brewster, p. 211,and Life by De Morgan, p. 93.
f Alluding to a letter of Leibnitz to Cliamberlayne, complaining of the
report of the committee.
|. Journ. Book, vol. xi. p. 431.

D2

36 Sir Robert Kane on the Composition and Characters

our most eminent philosophers that I send you this " correc-
tion," which they conceive ought to be made public through
the medium of the Philosophical Journal.
I am, Gentlemen,

Your humble Servant,
Royal Society, Somerset House, ChaRLES RichaRD Weld.

June 11,1847.

VI. Researches on the Composition and Characters of certain
Soils and Waters belofiging to the Flax districts of Belgi^miy
and on the Chemical Constitution of the Ashes of the Flax
Plant. By Sir Robert Kane, M.D., M.R.I.A.^

A BOUT two years since, I had the honour to submit to
-^■^ the Royal Irish Academy the results of some inquiries
into the chemical composition of the flax and hemp plants,
and into the chemical phsenomena of the treatment which they
undergo in the preparation of the ligneous fibre for the pur-
poses of the arts. The main object of that memoir was to
point out that, whilst the plant, as a whole, was rich in alka-
lies, earths, sulphuric and phosphoric acids, &c., the fibre, as
ultimately purchased in the market, was practically destitute
of all these materials, which therefore remained amongst the
substances removed from the plants during their preparation,
and hitherto rejected as of no use. Those results being pub-
lished in the Proceedings of the Royal Irish Academy, and
copied thence into various agricultural books and journals,
have in some degree led to the oeconomising of those valuable
residues ; and it is to be hoped that, according as the atten-
tion of farmers becomes more definitely fixed upon the real
and philosophical principles of the growth and composition of
various crops, the utilization of the different parts of plants
will be still more carefully attended to.

The researches to which I have referred, involved the de-
termination of the elementary composition of the plants, only
so far as it was necessary to prove the presence and propor-
tional quantity of certain materials in the plant as it grows,
and their absence in the fibre as prepared ; but it was not my
design therein at all to discuss the very important questions,
so fundamental to vegetable chemistry and physiology, of the
degree within which the composition of the ashes of a plant
may vary; or whether there is any general expression within
which the constitution of the mineral elements of a plant is
necessarily contained ; or finally, whether there can be traced

* Read at theAgricultural Evening Meeting of the Royal Dublin Society,
held on the 6th of April 1847.

of certain Soils and Waters hi Belgium. 37

any positive relation between the composition of the plant and
the composition of the soil upon which it grows. To answer
these questions even approximatively, will require investiga-
tions frequently repeated, and the concurrent labours of many
different investigators ; and although my present inquiries may
serve to furnish certain grounds for arriving at an opinion upon
these points, I would not in any way be understood as putting
them forward with that view.

My main object, in the inquiry which forms the subject of
the present paper, was to ascertain, if possible, whether there
existed any difference between the composition of the ashes of
the ordinary flax of Ireland and the flax grown in those loca-
lities in Belgium, where that plant is known to yield a fibre of
so much commercial value. Further, to ascertain the compo-
sition of the soils of those districts, in order to compare them
with the soils of the localities in Ireland where flax is, or may
be, successfully cultivated. Finally, as it is known that in the
preparation of the fibre the most important stage consists in
the steeping or retting of the plant, I considered it of the
greatest interest to trace, if possible, whether the superior qua-
lities of some rivers or ponds in Belgium could be connected
with any peculiarity of chemical constitution. For the mate-
rials and specimens necessary for these investigations, I am
indebted to the kindness and liberality of Mr. Marshall of
Leeds ; who was anxious also to connect therewith the dis-
cussion of some most important points of special technical
application, for which, however, the pressure of other avoca-
tions did not allow me time. I therefore publish the results
contained in the present paper, solely under their scientific
relations to agricultural chemistry and physiology, and shall
not enter upon any considerations belonging to manufacturing
practice.

Before entering into the description of the numerical re-
sults of the analyses, I think it better to premise a succinct
notice of the modes of analyses adopted for the different classes
of substances, as I shall thereby be enabled to avoid a great
deal of repetition.

I. Of the Modes of Analyses used for the Ashes, Soils and.
Waters.

The preparation of the flax-ashes was effected by chopping
up the plant stems into moderately small bits, and then car-
bonizing them gently in a Hessian crucible. The material so
obtained was further incinerated by very gentle ignition in a
platinum capsule over a gas flame; but it. was not in any way
sought to burn off all charcoal, or to obtain the ash perfectly

38 Sir Robert Kane on the Composition and Characters

white, as such would require a temperature capable of mate-
rially altering the constitution of the ash, a fact of which I
have been long aware, and which has latterly fixed the atten-
tion of several chemists. The ash so prepared was carefully
dried in a stove, and then treated in the following manner: —

Dilute muriatic acid having been poured over the quantity
of ash selected for analysis, the whole was heated in a water-
bath until it dried completely down ; water was then added,
and when the soluble materials had been completely taken up,
the whole was thrown upon a weighed filter and the liquor
separated ; there remained upon the filter such particles of
sand or soil as had been adherent to the plants, the unburned
charcoal of the ash, and the silica which had existed in the ash,
either free or in combination with alkaline or earthy bases.

The weight of this insoluble residue having been properly
determined, it was boiled in a strong solution of caustic potash,
by which all the proper silica of the ash was taken up, and
the residue then remaining being weighed, gave the sand and
charcoal, the silica being thus determined by difference.

The muriatic solution was then divided into three parts for
the determination —

1. Of the alkaline constituents.

2. Of the phosphoric acid, manganese, alumina, magnesia,
and lime.

3. Of the sulphuric acid and oxide of iron.

The first portion of solution was rendered slightly alkaline
by carbonate of ammonia, and then mixed with solution of
caustic barytes in excess, and allowed to stand for some hours.
By this means the sulphuric and phosphoric acid were perfectly
removed, as well as the earthy constituents, except a small
quantity of lime, which remained dissolved in a caustic state,
and which was then perfectly removed by the addition in ex-
cess of a mixture of caustic and carbonated ammonia. The
liquor, after filtration, was evaporated to dryness, and the
residue gently ignited, when the ammoniacal salts were per-
fectly expelled : there remained the alkalies of the ash as
chlorides. This residue was weighed, then dissolved in water,
and a solution of bichloride of platinum added. The liquor
and precipitate were then evaporated nearly to dryness, the
potash platinum salt washed by a mixture of alcohol and aether,
and the amount of platinum determined in the usual way.
The soda was ascertained by subtracting the weight of the
chloride of potassium from the weight of the mixed chlorides,
as given in the first instance.

To the second portion of the liquor was added so much
ammonia as nearly neutralized it without producing any per-

of certain Soils and Waters in Belgium. 39

manent precipitate. A quantity ofperchloride of iron was then
added, and acetate of potash, until a deep wine-red colour was
produced ; the liquor was then boiled until all odour of acetic
acid ceased, and a copious brown precipitate formed, which
was separated by the filter. This precipitate was then re-
dissolved in muriatic acid, boiled until all odour of acetic acid
ceased, and the liquor then precipitated by ammonia. The
precipitate, collected on a filter, was dried, ignited, and weighed,
then redissolved by muriatic acid. A quantity of tartaric acid
was added to the liquor, and ammonia then added in such
excess as to redissolve the precipitate which first forms. To
the solution thus got, hydrosulphuret of ammonia was added
in excess, the sulphuret of iron collected on a filter, and, when
washed, redissolved in aqua regia. The peroxide of iron,
precipitated from the liquor by ammonia, collected, dried,
ignited and weighed, and its weight subtracted from the weight
of the basic phosphate previously given, determines in an ab-
solute manner the quantity of phosphoric acid.

To the liquor from which the phosphoric acid had been
separated by the means described above, hydrosulphuret of
ammonia was added, by which a precipitate was formed, which
was collected, and, while moist, boiled with caustic potash
liquor; the undissolved matter was dissolved in muriatic acid
nearly neutralized, and treated with benzoate of ammonia ; by
this a trace of iron, generally remaining from the preceding
process, was removed, and the manganese was then precipi-
tated by carbonate of ammonia, collected, ignited and weighed.
The potash liquor was then acidulated by muriatic acid, and
the alumina which it had dissolved was precipitated, and its
quantity detei'mined in the usual way.

The solution, from which the iron, alumina, and manga-
nese had thus been separated by hydrosulphuret of ammonia,
was next boiled until all odour of sulphuretted hydrogen
ceased, and then treated with oxalate of ammonia, the oxalate
of lime collected was gently ignited with carbonate of ammonia,
and the quantity of lime determined. The liquor was then
very much concentrated by evaporation, and treated with
phosphate of soda and ammonia, set aside until the ammoniaco-
magnesian phosphate had perfectly deposited, and the quan-
tity of magnesia determined from that of the latter salt.

The third portion of the ash liquor was treated with nitric
acid, so as perfectly to peroxidize the iron ; it was then de-
composed by chloride of barium, by which all the sulphuric
acid was separated as sulphate of barytes, collected and weighed.
To the filtered liquor there was added a great excess of phos-
phate of soda and ammonia, and then an excess of acetic acid.

40 Sir Robert Kane on the. Composition and Characters

By boiling, the whole of the iron separated as perphosphate,
and was collected, ignited, and the quantity of oxide of iron
calculated from its weight.

For the determination of the chlorine, a totally distinct
portion of ash was taken, digested with water, acidulated with
nitric acid, and then precipitated with nitrate of silver in the
usual way.

It will be observed that, in all its main features, this plan
of examination of the ash coincides with that employed by
Will and Fresenius, and proposed by them in iheir memoir
on the Composition of certain Ashes. It is, however, that
which I had followed in all my former ash analyses, except
in regard to the determination of the phosphoric acid, for
which I had previously made use of the method proposed by
Schulze, but now replaced with so much advantage by that
invented by Will.

It is necessar}', however, to remark, that the composition
of the perphosphate of iron, given by Will, and upon which
he founds his mode of determining the quantity of oxide of
iron, has been contested very recently by Wittstein, who has
not succeeded in preparing that salt with the composition
assigned to it by Will. According to the latter chemist, it
consists, in its anhydrous form, of 2Fe2 Og + SPOg, that is,

SFegOg ... 160 4289

3PO5 .... 213 57-11

373 100-00

whilst the salt uniformly obtained by the other chemists was

Fe^Og+PO.y or

Fcg O3 . . . . 80 52-98

PO5 .... 71 47-02

151 100-00

But the circumstances of preparation of the different salts do
not appear to have been quite identical ; and I do not, there-
fore, reject Will's numbers, which have been, moreover,
verified by some trials made in my own laboratory. I have
consequently employed his formula in calculating the amount
of iron in the different materials; but it is easy to calculate,
for each analysis, the change (in most cases trifling) which the
employment of Wittstein's formula for the perphosphate should
introduce.

In the examination of the soils, the process consisted, first,
in mechanically separating the sandy and gravelly materials
from the finely-divided portion, by careful elutrialion with the
smallest possible quantity of pure water. This having been

of certain Soils and Waters in Belgium. 41

clone, and the quantity of sand determined by direct weighing,
the finely-divided earthy material was carefully dried at the
highest temperature it would bear without its organic consti-
tuents being injured, and then weighed. It was then carefully
but gently ignited in a current of air, until the organic mate-
rials were burned out, and was then again weighed. The loss
of weight gave the quantity of organic substance, together,
however, with some traces of water, from which the soil could
not be previously perfectly freed.

The soil was then subjected, for the determination of its
chemical constituents, to precisely the same general plan of
treatment which I have described in the case of the ash. The
matter, insoluble in muriatic acid, was however found to be
(the sand and organic matters having been previously sepa-
rated) ferruginous clay, which it was not necessary further to
examine, as all the materials of importance, in studying the
chemical nature of the soil, had been taken up by the different
solvents used.

In the case of the waters, the quantity employed for ana-
lysis was, with one exception, about two gallons ; in that case,
owing to a vessel having leaked (No. 3), but one gallon was
employed. The waters were, in the first instance, very care-
fully filtered ; and where any sensible quantity of sediment
was found upon the filter, its nature and quantity observed.

The water was then evaporated, at first upon the sand-bath,
but finally upon a water-bath, to perfect dryness, and the re-
sidue having been collected and dried at 212°, was weighed.
It was then incinerated ; the residue, moistened with carbonate
of ammonia, again gently ignited and weighed. By the dif-
ference of weight, the quantity of organic matter present was
ascertained in the state in which it exists when dried at 212°
Fahrenheit.

The solid material thus obtained was treated with water,
until all soluble salts were taken up, and the alkalies, lime,
magnesia, with sulphuric and muriatic acids, therein deter-
mined. The undissolved residue was next treated with muri-
atic acid, and the amount and nature of the earthy substances
taken up, as well as oxides of iron, &c., and phosphoric acid,
if any, ascertained. The material insoluble in muriatic acid,
when present, was of course determined.

The detailed modes of analyses pursued in these cases were
precisely the same as in those of the ashes and soils.

In carrying out the greater part of the practical details of
these analytical methods, I derived valuable aid from Mr.
William Sullivan, then my private assistant, but now first
chemical assistant in the Museum of Irish Industry, founded
by Her Majesty's Government in Dublin.

42 Sir Robert Kane oti the Composition and Characters

2. Results of the Analyses of the Soils.
The general character of all the soils submitted to exami-
nation was that of light, sandy loams, in some cases almost
purely sandy ; excessively loose in texture, non-coherent and
permeable : usually rich in organic matters containing nitro-
gen. These soils all coloured water boiled upon them, and
gave to it a sensible, though very small, quantity of alkaline
and earthy salts.

A. Soil from Heestert, in the Courtrai district : —

Composition per cent.

Potash 0-160 .

Soda 0-298

Peroxide of iron . . « . 3*298
Oxide of manganese . . . a trace

Alumina 2-102

Lime 0-357

Magnesia 0-202

Sulphuric acid 0*025

Phosphoric acid. . . . . 0*121

Chloride of sodium . . . 0*017

Organic matter and water"! qioq

not driven off at 212° J -^ ^^-^

Clay 14*920

Sand 75-080

99*703

Loss 0*297

100*000

B. Soil from Escamaffles, some of the very best flax lands
of the Courtrai district: —

Composition per cent.

Potash 0*123

Soda 0*14.6

Peroxide of iron .... 1*663
Oxide of manganese ... a trace

Alumina 1*383

Lime 0-227

Magnesia 0*153

Sulphuric acid 0'017

Phosphoric acid . . . . 0-152

Chloride of sodium . . . 0*030

Organic matter and water"! „ „.

not driven off" at 212° J ^'"^^^

Clay 9*280

Sand 84*065

99-600

Loss '400

100-000

of certain Soils and Waters in Belgium.

43

C. Soil from Hnmme Zog, the best flax land in the Ant-
werp district : —

Composition per cent.

Potash ....

Soda ....

Peroxide of iron
Oxide of manganese
Alumina . . .
Lime ....

Magnesia . . .
Sulphuric acid .
Phosphoric acid
Chloride of sodium
Organic matter and water \
not expelled at 212° J
Clay ........

Sand

Loss

0-068

0 110

1-202

a trace

1-125

0-481

0-140

0-013

0-064

0-067

4-209

5-760

. 86-797

99-975
. -025

100-000

D. Soil from a district producing coarse flax and poor
crops generally : —

Composition per cent.

Potash 0-151

Soda 0-206

Peroxide of iron .... 1-543

Oxide of manganese ... no trace

Alumina 0-988

Lime ........ 0-366

Magnesia . 0-142

Sulphuric acid 0*026

Phosphoric acid . . . , 0-193

Chloride of sodium . . . 0*009

Organic matter and water"\ „ „^^

not expelled at 212° . J "^'^'"^

Clay 4-400

Sand 88-385

100-081

44- Sir Robert Kane on the Composition and Characters

E. Soil from a district in Holland, where flax is well-
grown : —

Composition per cent.

Potash 0-583

Soda 0-306

Peroxide of iron .... 6-047
Oxide of manganese ... a trace

Alumina 5-626

Lime 3-043

Magnesia 0-105

Sulphuric acid 0-023

Phosphoric acid . . . . 0*159

Chloride of sodium . . . . 0-023

Organic matter and water") c.qa,\

not expelled at 212° . J ^ "*'

Clay 17-080

Sand 60-947

99-783
Loss 0-217

100-000

Mr. Marshall was also kind enough to forward to me a speci-
men of the kind of soil which is found deposited in the Humber,
and the gradual silting up of which has formed the extensive
flat districts reclaimed along that eastern coast. This speci-
men of soil, or iioarp^ as it is termed, from the operation by
which the ground becomes permanently gained from the sea,
had not yet borne any crop. It is from Crowie, in Lincoln-
shire.

Its composition per cent, was found to be as follows : —

Potash 0-534

Soda 0-083

Peroxide of iron .... 4-500

Oxide of manganese . . a considerable trace

Alumina 3-065

Lime 5-538

Magnesia 0-052

Sulphuric acid 0-113

Phosphoric acid .... 0*222

Chloride of sodium . . . 0*067

Organic matter and water ^ K'^xqh

not expelled at 212° . J ^ "^^^

Sand 80*702

100*204

of certain Soils and Waters in Belgium. 45

By these analytical results, it is abundantly evident how
completely due to artificial means is the fertility of those dif-
ferent Belgian soils; the large quantity of azotized organic
matter, the proportionally large quantities of phosphoric acid
and magnesia, and of the alkalies, being evidently the result
of the copious treatment with animal manures, to which, as all
persons conversant with Flemish agriculture are aware, the
soil in Belgium is subjected. This will become still more
evident, when hereafter I have to notice the course of cultiva-
tion which those soils are made to undergo. The duty, so
important in the preparation of our Irish soils for flax, of di-
viding the soil to the finest possible state, and rendering it
perfectly friable and porous, is seen by the above results to be
naturally effected in the Belgian soils, of which a well-manured,
incoherent sand, might be more correctly the title; none of
them containing, except that marked A, and that from Hol-
land, E, as much clay as would even justify the title of a light
loam. There is, therefore, no doubt but that the soils most
adapted for the successful growth of flax are of this very light
and porous character; and that, in the selection of districts in
this country into which the flax culture may be extended, this
quality of lightness and permeability of soil is of the first im-
portance.

The quantity of lime contained in the Belgian soils will be
observed to be extremely small ; but in that from Holland
and from the Lincolnshire warped land it is much larger, in-
deed so as to constitute the most dominant earthy material.
This has evidently had its cause in the source from whence
these soils were derived, the silt deposited in shallow, quiescent
waters by the sea, and which contains, mixed with sand, a
proportion of comminuted shells or chalk. There is no posi-
tive evidence that this amount of lime is connected with any
decided inferiority in the flax ; but it is still worthy of atten-
tion, that the soil of the districts which have been longest and
best known for the production of good flax have but a mere
trace of lime in their constitution.

The comparatively large quantity of magnesia which the
Belgian soils contain, and which is so remarkably contrasted
with its inferior proportion in the warp soil, is, in my opinion,
produced by the artificial manuring by animal liquids ; and
to this source also I attribute the great richness of these soils
in phosphoric acid.

[To be continued.]

[ ^Q ]

VII. On the Colouring Matters of Madder.
By Dr. Schunck*.

THE organic colouring matters present such a wide field for in-
quiry, that it would require the labour of years to enable one
person fully to elucidate their properties, or even to bring this depart-
ment of organic chemistry into a state of development proportionate
to the present condition of the science. The substances included under
the name of colouring matters by no means agree in their chemical
characteristics ; they merely coincide in being possessed of certain
vivid colours, or in giving rise to coloured compounds. Strictly con-
sidered, some of them ought to be classed among the resins and others
among the extractive matters ; and on the other hand, if we attempt
a definition of the class according to their chemical characteristics,
we shall find it impossible to exclude a large number of bodies, which,
like tannin and catechin, are capable of giving rise under peculiar
circumstances to brown substances, which in nowise differ in their
general properties from the bright red colouring matters of archil,
logwood, &c. Some colouring matters are presented to us ready
formed in the different parts of plants and animals ; others are pro-
duced artificially from colourless substances, which undergo very
complex changes during the process ; others arise spontaneously
during the first stages of oxidation or putrefaction following the ex-
tinction of organic life. In the investigation of substances thus
widely differing in properties and formation, it would be vain to
expect at present anything approaching to general results in regard
to the class as a whole. I must therefore content myself on this
occasion with giving a short account of the results of some ex-
periments which I have made on one branch of the subject, at the
same time apologising for their present vague and undefined nature.

I have directed my attention in the first instance to madder, partly
because the colouring matters contained in it are almost unknown,
or rather worse than unknown, viz. known in such a manner as
merely to mislead those who wish to inform themselves by the ac-
counts given of them, and partly because madder is an article of such
an immense importance in the art of dyeing that every discovery in
relation to it acquires immediately a practical bearing.

It will be unnecessary for me to allude to the former numerous
investigations of madder, except so far as to mention that Robiquet
discovered in it a crystallized volatile colouring matter, which he
called Alizarin, and that Runge described five colouring matters
which he obtained from it, viz. madder purple, madder red, madder
orange, madder yellow and madder broivn. I may here state as one
result of my investigation, that I agree with Runge in thinking that
there is more than one colouring matter in madder, though I am of
opinion that the substances which he enumerates and describes are
not pure. Before however entering on this part of the subject, I
shall first give the results at which I have arrived in regard to ali-

* From Report of British Association for 1846,

Dr. Schunck on the Colouring Matters of Madder. 47

zarin. Alizarin is doubtless the most interesting and the most
definite in its nature of all the substances contained in madder. It
also presents itself the most easily to the observer even on the most
superficial examination. If we heat madder spread out in a thin
layer on a metal plate without carrying the heat far enough to char
the woody parts of the root, we shall in the course of a few hours
find its surface covered with small red or orange-coloured crystals,
which consist of alizarin. In the same way any extract of madder,
whether with water, alcohol or alkalies, evaporated to dryness and
gently heated, gives a crystalline sublimate of alizarin, which is va-
riously coloured from a light yellow to a dark red or brown. Now
one of the first points to be ascertained in regard to this body was
whether it exists as such in the root, or whether it is formed by the
process of sublimation. Robiquet, the discoverer, states that it pre-
exists in the plant. He considered alizarin as the colouring principle
of madder, and merely subjected it to sublimation for the purpose
of purifying it. But his investigation presents us with no convincing
proof of this opinion, for the extract of madder with water, alcohol,
&c., from which he prepares his alizarin by sublimation, shovvs no
trace of anything crystalline; and many chemists have asserted in
consequence that it is u product of decomposition, being formed by
the action of heat in the same way as pyrogallic, pyrotartaric acitl,
and many other bodies. I have however no hesitation in affirming
that it exists in the plant as such, having in more than one way ob-
tained it in a crystallized state without the intervention of heat.
If we make an extract of madder with cold water, we obtain a brown
fluid which produces no reaction on test-paper. After being ex-
posed however to the action of the atmosphere for some hours, it
acquires a distinctly acid reaction ; and if it be now examined care-
fully, there will be found floating about in it a number of long hair-
like shining crystals : these crystals are alizarin. If the fluid be still
further exposed to the influence of the atmosphere, a yellow amor-
phous substance begins to separate, which I shall mention afterwards.
This is succeeded by a gelatinous substance, and after some days a
complete state of putrefaction ensues. It seems as if the alizarin in
madder, or at all events that part which dissolves in the water, exists
in combination with lime. On exposure to the atmosphere, there
is formed, from some constituent of the root dissolved in the fluid
through the instrumentality of the oxygen, some acid, which seizes
hold of the lime in the solution and separates the bodies which are
combined with the lime. Now the alizarin, being a body of very
slightly acid properties, is separated first, and the other substances
follow in succession. The fresher the madder is, the purer will be
the alizarin, which separates on exposure to the atmosphere ; in some
instances it forms on the surface of the fluid a thick light yellow
scum ; but in most cases it is mixed with brown or red substances,
from which it is separated with difficulty. It is therefore most
advisable to separate the crystals which are deposited after twelve
hours' standing, by filtration. These crystals are then washed from
the filter and boiled -with very dilute nitric acid until they have be-

48 Dr. Schunck on the Colouring Matters of Madder.

come of a bright yellow colour. They are then dissolved in boiling
alcohol, from which they separate on cooling in yellow transparent
plates and needles having a strong lustre. Alizarin prepared in
this way has the following properties : — It has a pure yellow colour
without any admixture of red. It may be volatilized without leaving
any residue. The vapour crystallizes on cooling in beautiful yellow
plates and needles. It suffers hardly any change if exposed to the
action of the most powerful reagents. It dissolves without change
in cold concentrated sulphuric acid. Concentrated nitric acid hardly
affects it even on boiling. It is not changed by chlorine. It is in-
soluble in water, but soluble in alcohol with a yellow colour. It
dissolves in alkalies with a beautiful purple colour. Its compounds
with the alkaline earths are red and slightly soluble in water. Its
compounds with the earths and metallic oxides are insoluble in water
and exhibit different shades of red. It imparts no colour to cloth
mordanted with acetate of alumina or oxide of iron, on account of
its insolubility in M'ater. Very little alizarin is obtained in this way ;
perhaps one 1 gr. from 1 lb. of madder, though there is more of it
contained in the root.

I shall now shortly describe two other colouring matters which I
have obtained from madder. If an extract of madder be made with
hot or cold water, and a strong acid, such as muriatic or sulphuric
acid, be added to the fluid, a dark reddish-brown flocculent preci-
pitate is produced. This precipitate was separated by filtration and
washed until the acid was removed. On being treated with boiling
water, a part of it dissolves with a brown colour. On adding a few
drops of acid to the filtered solution a dark brown precipitate is
produced, which seems to me to be a peculiar colouring matter
similar in its properties to orcein, hematin and other soluble colour-
ing matters. It dissolves in alkalies with a red colour, and is capable
of imparting very lively colours to mordanted cloth. As far as I
am aware it has not been described in the former investigations of
this subject, though it seems to be the principal substance concerned
in the production of the colours for which madder is used in the
arts. I have however only examined it very slightly as yet. The
residue left behind by the boiling water was treated with dilute
boiling nitric acid, by which every trace of the preceding substance is
destroyed, and the residue itself acquires a bright yellow colour
and a more powdery consistence. This yellow powder contains
alizarin, as is shown by its giving crystals of that substance on
being gently heated; in fact it contains all the alizarin of the root,
but mixed with another substance of an amorphous natui'e but very
similar properties, from which it is difficult to separate it. By crys-
tallising from alcohol no separation can be effected, as they are both
about equally soluble in that menstruum. They also behave in a
similar manner towards the alkalies, the earths and most of the me-
tallic oxides. I have hitherto only succeeded in discovering one
method of separating them, which is as follows : — The mixture of the
two is dissolved in a little caustic potash. To the solution is added
perchloride of iron, which produces a dark reddish-brown precipi-

Analjjsis of the Uihie of the Calf and Sheep. 49

tate consisting of peroxide of iron in combination witli the two sub-
stances. Now on boiling this precipitate with an excess of perchlo-
ride of iron, the alizarate of iron dissolves, forming a dark brown so-
lution, while the iron compound of the other substance remains
behind. On adding muriatic acid to the filtered solution, the alizarin
separates in yellow flocks and may be purified by crystallization from
alcohol. The other substance, to which I have not yet given a
name, is obtained by decomposing its iron compound, which remains
behind on treating with perchloride of iron, with muriatic acid, and
washing till all the oxide of iron is removed. It seems also to be a
colouring matter, as it dissolves with a red colour in alkalies and
gives red compounds with the earths and metallic oxides. It is
insoluble in water, but soluble in alcohol with a yellow colour. It
therefore resembles the resins in its general properties. It cannot
be obtained in a crystallized state. From a hot concentrated solu*
tion in alcohol it separates on cooling as a yellow powder. It im-
parts no colour to mordanted cloth.

VIII. Comparative Analysis of the Urine of the Calf and the

Sheep^.

MBRACONNOT finds that the urine of the calf,
• nourished by tlie milk of the mother, consists of —

grs.
Ammoniaco-magnesian phosphate .... 0"18

Chloride of potassium 3*22

Sulphate of potash . , 0'44<

Urinarv animal matter") „ _ „

Urea / ^'^^

Phosphate of iron "

Phosphate of lime

Phosphate of potash

Combustible acid combined with potash ^ . . traces

Silica

Mucus

Chloride of sodium ?

Water 993*80

looaob

A litre of the urine of the sheep yielded — grs.

Chloride of potassium 6*13

Sulphate of potash 3*74f

Carbonate of magnesia . r40

1

> quantities undetermined.

Urea

Urinary animal matter
Hippurate of potash
Bicarbonate of potash
Carbonate of lime

Mucus

Oxide of iron . .

* From the Annates de Chimie et de Physique, Juin 1847.
Phil. Mag. S, 3. Vol. 3 1 . No. 205. My 1 847. E

50 Mr. Hind on the expected Reappeaj-ance of

IX. On the expected Reappearance of the celebrated Comet of
1264. and 1556. By Mr. Hind*.

'^I'^HE time is now near at hand when the return of the comet
-*- of 1264) and 1556, signalised by Mr. Dunthorne and M.
Pingre, may be expected to take place. It is therefore de-
sirable that observers should be in possession of everything
that may tend to facilitate their search for the comet ; and 1
venture to communicate to the Society the results of some re-
cent calculations of my own on the subject, preceded by a very
brief view of the principal circumstances connected with former
appearances of the comet, and a short notice of calculations
already published.

"The great and celebrated comet" of 1264, as Pingre
terms it, is mentioned by nearly all the European historians
of the time, and was observed by the astronomers of the dy-
nasties then reigning in the north and south of China. It is
described as presenting a most imposing appearance, with a
tail 100° in length, stretching from the east part of the " mid-
heaven." The comet was of "surprising magnitude," far
exceeding any remembered by those who beheld it. Contem-
porary writers generally considered it the precursor of the
death of Pope Urban IV., and many of them relate that it
disappeared on the same night that the pope died, or on Oc-
tober 2 ; thus, in the words of Thierri de Vaucouleurs,
" Quo (Urbano) moriente, velut mortem cognosceret ejus,
Apparens minime Stella comata fuit."

In 1556 the appearance of the comet was not on the same
scale of splendour as in 1264, but still was sufficiently imposing
to call forth from historians the epithets "ingens et lucidum
sidus." It was observed by Paul Fabricius, a mathematician
and physician at the court of the emperor Charles V. of Au-
stria. M. Pingre, the celebrated cometographer, sought in vain
for the original observations; the only information he could find
on the subject was contained in a small rough chart found in
Lycosthenes and other authors. I have before f suggested the
probability that these observations were given by Fabricius in
his work upon the comet, published at Niirnberg in 1556, and
mentioned by Lalande in his Bibliographic ; but, as far as I
am aware, tfiis work has not been discovered in any library.
M. Pingre would have at his command the splendid collec-
tions of St. Genevieve and the Royal Library at Paris ; and his
ineffectual search for the observations in these libraries makes
it at least doubtful whether they are now in existence. The
chart just mentioned enables us to form a tolerably definite

* From the Proceedings of the Royal Astronomical Society, No. 14.
f Ast. Nach. 493.

the celebrated Comet o/1264- and 1556. 51

idea of the path followed by the comet, and we have ample
information for a rough determination of the elements.

When Halley published his Synopsis of Cometary Astro-
nomy, he gave a set of parabolic elements for the comet of
J 556, founded upon the observations made by Paul Fabricius;
but he remarks that these elements are not so certain as those
of other comets he had computed, the observations being made
"neither with sufficient instruments nor due care," and by no
means to be reconciled with any regular calculation.

The elements of the comet of 1264 were first computed by
Mr. Dunthorne. His discussion of the observations and cir-
cumstances relating to the comet's apparition are published in
vol. xlvii. of the Philosophical Transactions. The elements
are chiefly founded on the £Puthority of a manuscript preserved
in the library of Pembroke Hall College, Cambridge, entitled
Tractatusfratris JEgidii de Cometis. But it must be observed
there are manifest contradictions in this account not easily set
right. The other authorities consulted were the Chronicon
Sampettinum Erphiirtense and the Chronicle of John Vitodu-
ranus. The orbit deduced by Mr. Dunthorne much resem-
bles that calculated by Halley for the comet of 1556.

In the Memoirs of the Royal Academy of Sciences at Paris
for 1760, appears a valuable memoir by M. Pingre on the
comet of 1264. After collecting together a great number of
accounts from different chronicles and histories of the day,
he proceeds to the discussion of the elements. The contra-
diction in the Cambridge manuscript which relates to the
comet's motion in longitude is pointed out; and since this
manuscript was Mr. Dunthorne's chief authority, it might be
supposed that his orbit would differ entirely from M. Pingre's.
This, however, was not the case ; for although there are dif^
ferences of some moment in one or two of the elements, there
is still a striking similarity between the two orbits taken as a
whole, and M. Pingre's approaches much nearer than Mr.
Dunthorne's to the orbit of the comet of 1556. A closer
agreement might have been produced if he had not wished to
preserve the path laid down by Thierri de Vaucouleurs with
as little alteration as possible. M. Pingre concludes from his
researches that there is little doubt of the identity of the comets
of 1264 and 1556, and, therefore, that the return to perihelion
may be expected to take place in the year 1848. In No. 493
of the Astronomische Nachrichten will be found the results of
my first calculations relating to this comet. I have there de-
duced elements from the observations by Fabricius in 1556,
and computed an ephemeris for comparison with the comet's
observed path. The agreement, though not so close as could
be wished, was the best that could be obtained from the data

E2

52 Mr. Hind on the expected Reappearance of

given by M. Pingre in his Cometography. I then reduced the
elements to the year 1264, and with the assistance of a passage
in Thierri's poem, I fixed the time of perihelion for July 9'9
(old style). The passage alluded to is as follows : —
" Undecimumque gradum Phcebo superante Leonis,
Ter deno Cancri restitit ilia loco."

With M. Pingre, I have understood by " Ter deno Cancri,''
the 120th degree of longitude; but I am not quite sure that
this is the true interpretation.

With perihelion and node reduced as before stated, and the
other elements as for 1556, an ephemeris of the comet's geo-
centric path in 1264 was computed. During the month of
July, calculation and observation agree pretty well ; but after
the beginning of August the theoretical places entirely differ
from the positions of the comet, as deduced from the accounts.
Instead of traversing Orion towards the end of its appearance,
as some historians relate, it would take a higher declination,
passing through Auriga and Taurus.

Since the publication of this paper in the AstronomiscJie
Nachrichten, I have made some further investigations on the
subject, and with more success than in my first calculations.
A closer comparison of data showed pretty clearly that the
observation of March 5, on which I had chiefly relied, must
be erroneous as it is given by M. Pingre. In tome i. of his
Cometography, p. 503, we learn that on March 5 the comet
was almost in the right line joining the stars y and fl Virginis,
and was equidistant from the stars. A trigonometrical cal-
culation from these data gives the place of the comet in lon-
gitude 188'^ l', and latitude +2° 19', and this position was
employed in my earlier investigations. But I have recently
satisfied myseltj that the observation as given above cannot be
reconciled with those of March 3 and 4, and on subsequent
days, by any set of elements. The cause of this anomaly is,
I believe, an error in the name of the star. If instead of y
and fl Virginis we read 8 and i9, then the place of the comet
would be in longitude 188° 41', and latitude +5° 13', which
agrees very well with the track which the comet ought to have
followed, according to the other observations.

A recalculation of the elements from an interpolated posi-
tion for March 5, and from those of March 9 and 14, gives
the following values: —

Passage through perihelion, 1556, April 220233, G. M.T. [Old style.]

o '

Longitude of perihelion 274 14"9 1 ,:, . ncK/j

Ascending node 1^5 ^^.g j Equinox of 1556.

Inclination 30 12-2

Log. least distance 970323

Motion direct.

the celebrated Comet of 19,6^ and 1556.

S3

The following ephemeris of the comet for the appearance
in 1556, Greenwich mean midnight, old style, is deduced from
these elements: —

1836.
Old style.

Geo. long.

Geoc. lat.

Log. r.

A-

March 3

188 13

0 /

+ 1 9

00732

0-193

4

188 0

3 40

00670

0-175

5

187 44

6 45

00606

0-157

6

187 22

10 36

0-0541

0-140

7

186 54

15 29

0-0476

0-124

8

186 14

21 43

00409

0-109

9

185 18

29 49

00341

0-096

10

183 49

40 12

0-0272

0085

11

181 11

52 50

00201

0-078

12

175 21

67 5

00130

0075

13

153 35

80 29

00057

0-078

14

55 19

82 30

9-9983

0085

15

27 16

73 26

99908

0095

16

20 37

65 30

9-9831

0-108

17

17 44

59 16

9 9753

0-122

27

12 19

34 58

9-8903

0302

April 6

12 7

27 1

9-7959

0-505

16

14 13

20 30

9-7178

0733

26

19 12

+ 13 52

97130

0-974

If this ephemeris be compared with the descriptions of the
comet's apparent path in the heavens, we shall find the agree-
ment as close as could be expectetl, considering the uncer-
tainty and irregularity of the data.

With the above elements reduced to 1264, the time of pe-
rihelion was found to be July 13*42, /. e. assuming with Pingre,
that the comet was in longitude 120° when the sun had reached
the 11th degree of Leo, according to the narration ofThierri
de Vaucouleurs. The geocentric places of the comet, Green-
wich mean midnight, old style, would then be as follows : —

1264.
Old style.

Geo. long.

Geoc. lat.

r.

A.

July 7

138° 10

+ 18° 14

0-53

0-82

17

132 36

22 9

0-51

0-62

22

126 29

21 54

0-55

0-55

27

118 36

20 14

0-61

0-48

Aug. 6

101 14

+ 10 17

075

0-41

16

85 23

— 3 47

0-92

0-39

26

70 47

17 10

109

0-42

Sept. 5

56 39

27 8

1-26

0-48

15

43 11

33 4

1-43

0-57

25

31 35

35 26

1-59

0-69

Oct. 5

22 47

-35 30

1-75

084

If we are to depend solely on the European accounts of
this comet's path, the above is liable to two objections : first,
too high a declination in August; and secondly, that the posi-

54 Mr. Hind on the expected Reappearance of

tions are in Eridanus during the latter part of the comet's
apparition; historians generally contenting themselves with
stating that the comet " finally traversed Orion." M. Pingre's
elements, which are not open to these objections, do not agree
so well as mine with the more circumstantial details left us in
the Chinese annals. The two orbits differ chiefly in the lon-
gitude of the node and perihelion distance, but the discord-
ances are by no means great.

The results of my calculations have satisfied me that the
comet of 1264< was, in all probability, the same as that of 1556,
and consequently, that its return to perihelion must be very
near at hand. The nodes of the comet's orbit lie very close
to the earth's path. The ascending node is passed fifty days
before perihelion, the radius vector being 1*193, and conse-
quently the distance outside the earth's orbit about 0*197.
The passage through descending node occurs 31| days after
perihelion, and the distance of the point from the earth's orbit
inside is 0*126. However, the nearest approach of the comet
to the earth will not happen at the nodes, but soon after its
passage through them; thus in 1556 the least distance between
the two bodies was O'OT'i, nine days after the transit through
ascending node. The effect of this close proximity to our
globe on the period of revolution of the comet has been inves-
tigated by Professor Madler, of the Dorpat Observatory, as
detailed in No. 501 of the Astronomische Nachrichten\ it
amounted to 14| days only, and the return of the comet to
perihelion was fixed for the end of February 1 848.

The following table contains the heliocentric co-ordinates
referred to the equator and the log. radii vectores of the comet
in my last orbit, reduced to 1848, for every tenth day, from
ninety days before to 90 days after perihelion.

Time from

X.

y.

«.

Log. r.

perihelion pass.

Days.

-90

-1*7430

+ 0-5750

-0-0603

0-2640

80

1-6231

0-4370

0-0445

0-2257

70

14931

0-2963

0-0284

0-1826

60

1-3504

0-1533

-0-0122

0-1333

50

1*1917

+0-0084

+ 0-0041

0-0762

40

1-0120

-0-1363

0-0206

00092

30

0-8039

0-2770

0-0363

9*9300

20

0-5570

0-4031

0-0501

9-8385

-10

— 02611

0-4907

0-0592

9-7474

0

+ 00738

0-4961

0*0583

9-7032

+ 10

0-3929

0*3951

0-0450

9-7474

20

0-6507

0-2266

0-0239

9-8385

30

0-8503

—0-0352

+ 0-0004

99300

40

1-0086

+ 0-1590

-00233

0-0092

50

1-1385

0-3490

0-0463

0-0762

60

1-2484

0-5331

0-0685

0-1333

70

1-3433

0-7109

00900

0-1826

80

1-4268

0-8828

0*1107

0-2257

+90

+ 1-5015

+ 1-0492

-0-1307

0-2640

the celebrated Comet qfl264i and 1556. 55

With the above values for x, y and ;?, and those of X, Y,
Z, taken from the Nautical Almanac, the position of the comet
for different suppositions as to the time of passage through
perihelion may be readily obtained. If vi^e suppose March 0,
which is about the epoch fixed by Professor Madler, we shall
have the following ephemeris for facilitating the discovery of
the comet, mean noon at Greenwich : —

1847—8.

R.A.

Decl.

A.

Dec. 1

187 16

o /

-11 22

2-16

11

193 .55

12 56

1-92

21

201 52

14 29

1-68

31

211 43

15 52

1-46

Jan. 10

224 16

16 50

1-26

20

240 18

16 47

Ml

30

259 53

15 3

1-02

Feb. 9

281 23

11 24

1-03

19

302 15

7 1

1-13

29

321 18

3 23

1-29

Mar. 10

338 5

— 0 54

1-48

20

352 11

+ 0 50

1-66

30

3 50

2 9

1-84

April 9

13 32

3 11

201

19

21 46

3 59

2-17

29

28 52

4 35

2-32

May 9

35 5

4 58

2-46

19

40 36

5 11

2-59

29

45 31

+ 5 12

2-69

It appears from this ephemeris, that according to the most
probable supposition we can make respecting the time of pe-
rihelion without actual calculation of the perturbations, the
position of the comet in the heavens during the approaching
reappearance will be extremely unfavourable for observation ;
and it is therefore the more desirable that those who look
out for comets should be on the alert. Nearly the whole of
the vast trajectory of this comet lies below the plane of the
ecliptic, and Jar Jt'om the paths of the larger planet s^ but it ex-
tends into space more than twice the distance of Neptune ;
and surely we are not yet able to say what causes may operate,
at this immense distance from the sun, to affect the time of the
next return to perihelion. If however the comet can be de-
tected and observed, we shall then have the means of ascer-
taining something more on these points

[ 56 ]

X. Analysis of the Water of the Thermal Spring of Bath
{Kinfs Bath). By Messrs. George Merck and Robert
Galloway*.

THE water of this celebrated spring, the efficacy of which
was known in the time of the Romans, has been analysed
repeatedly by various chemists at different periods. Richard
Phillipsf, Scudamorejj Walker§, and more recently Noad||,
have occupied themselves in the investigation of this water.
In their several analyses, the whole amount of the fixed ingre-
dients of the water agrees very closely ; but in regard to the
composition of these substances there are considerable dis-
crepancies, as may be seen in a table which we have annexed
at the end of this paper.

Besides great differences in the quantitative analysis, we
find discrepancies even in regard to the presence and absence
of certain constituents. Among the chemists that have been
mentioned, Walker is the only one who has recognised the
presence of potash. The same chemist corroborated Scuda-
more's statement as to the presence of magnesia, overlooked
by their predecessors ; but he states also that he detected
alumina, which none of the others found. In all these ana-
lyses iodine has been omitted. Mr. Cuff ^ however has in-
dicated the presence of this element in the spring.

These discrepancies made another investigation of the mi-
neral water of Bath very desirable ; the following analysis was
performed at the suggestion of Dr. A. W. Hofmann.

To obtain the water genuine, and especially for the pur-
pose of ascertaining the amount of free carbonic acid it con-
tained, we collected the water ourselves, an operation in which
we were kindly assisted by Messrs. Green and Simms, lessees
of the establishment.

The water was taken from the principal w^ell, which sup-
plies the King's and Queen's baths, which are the most
esteemed and valued in the city. Of the two other w ells, one
supplies the Hot Bath and the other the Cross Bath, ^vhich
are in the neighbourhood of those first mentioned.

* Communicated by the Chemical Society j having been read Nov. 16,
1846.

t An Analysis of the Bath "Water, by Richard Phillips. London, 1806.

X A Chemical and Medical Report of the properties of the Mineral Waters
of Buxton, Matlock, &c., by Ch. Scudamore, M.D. 1820.

§ Quarterly Journal of Science, Literature and Arts, vol. xxvii. 78. 1829,

II Pharmaceutical Journal, vol. iii. 526.

^ Memoir on the occurrence of Iodine and Bromine in certain Mineral
Waters of South Britain, by Charles Daubeny; Transactions of the Royal
Society of London, 1830, ii. p. 223.

Analysis of the Water of the Thermal Spring of Bath. 57

The King's Bath is an oblong cistern, 65 feet long and 40
feet broad, in which the water stands at the height of 46
inches. It is supplied from the bottom by means of twelve
large and about twenty smaller apertures. By far the largest
amount of water rises however from an opening made in the
centre of the bath, 18 inches in diameter. Although the
water flows under the influence of a very small pressure, the
quantity is such, that the two reservoirs, the King's and the
Queen's bath, are entirely filled in about nine hours. The
quantity of water entering each minute is 126 gallons, upon
the authority of Dr. Daubeny*.

I. Qualitative Analysis.

The water as it issues from the well has a temperature of
46° C. (115° Fahr.), the temperature of the air being 20° C.
(68° Fahr.) ; it is clear and without odour, and has no effect
upon vegetable colours ; it has a saline and slight iron taste ;
the iron is deposited as sesquioxide in rather large quantities
in the pipes leading from the well.

The following experiments gave the qualitative composition
of the mineral water ; on boiling for some time a white cry-
stalline precipitate formed. The qualitative analysis was
therefore divided into two parts.

a. The analysis of the precipitate formed on boiling.

b. The analysis of the substances remaining dissolved.

a. Analysis of the Precipitate formed on boiling.

1 . The precipitate was treated with hydrochloric acid ; a
small portion of it dissolved with effervescence, indicating the
pi-esence of carbonic acid. The portion insoluble in hydro-
chloric acid dissolved on the addition of a large quantity of
water : — Indicating sulphate of lime.

Another portion of the water was boiled some time, with
the precaution of replacing the evaporated water, in order
that all the sulphate of lime should remain in solution ; in
this case only a very small precipitate was formed, which was
entirely soluble in hydrochloric acid.

2. On heating this solution and adding ammonia, a very
slight flocculcnt precipitate of a yellowish-white colour was
produced after some time: — Indicating oxide of iron.

3. In the filtrate from the sesquioxide of iron (2.), on the
addition of oxalate of ammonia, a white precipitate was
formed: — Indicating salts of lime.

* On the Quantity and Quality of the Gases disengaged from the Thermal
Spring which supplies the King's Bath in the City of Bath, by Charles Dau-
beny ; Transactions of the Royal Society of London, 1834, 1. i.

58 Messrs. Merck and Galloway's Analysis of

4. In the liquid filtered off from the oxalate of lime (3.),
phosphate of soda produced an exceedingly slight crystalline
precipitate : — Shoiving the presence of magnesia.

Note. — ^This precipitate could only be distinctly seen in
testing a large quantity of the water.

b. Analysis of the substances remainiiig dissolved.

The liquid which was filtered from the precipitate (a.)
formed on boiling had no alkaline reaction ; a portion of it
was evaporated nearly to dryness and treated with hydro-
chloric acid ; no carbonic acid was evolved, from which com-
portment the absence of alkaline carbonates could with safety
be concluded.

1. A portion of the liquid gave on addition of chloride of
barium a copious white precipitate, insoluble in hydrochloric
acid : — Indicating sulphuric acid.

2. In another portion of the liquid nitrate of silver pro-
duced a copious white precipitate, easily soluble in ammonia :
— Evidencing the presence of chlorine.

3. The entire solubility of the silver precipitate seemed to
indicate the absence of iodides. To make ourselves perfectly
certain of the absence of these salts, 30 or 40 pounds of the
water were evaporated to 2 or 3 pounds, and the liquid filtered
off from the precipitate which had been formed ; a part of this
fluid was evaporated with precaution to dryness, the residue
was mixed with some starch paste, and a few drops of nitric
acid being added, feeble but distinct violet spots were ob-
served : this experiment was repeated several times with the
same success : — Indicating the presence of iodine.

4. Another portion of the liquid {b.) was treated with hy-
drochloric acid, evaporated to dryness, and gently ignited : on
treating the residue with a large quantity of water an insolu-
ble portion remained: — Showing the presence of silicic acid.

5. Another portion of the liquid {b.) gave, on addition of
chloride of ammonium and oxalate of ammonia, a white pre-
cipitate : — Indicating lime.

6. On adding, to a portion of the filtrate, ammonia and
phosphate of soda, a slight crystalline precipitate was formed :
— Indicating magnesia.

7. For the discovery of the alkalies, the remaining portion
of the filtrate from the lime precipitate Avas evaporated to
dryness, and the residue ignited until the ammoniacal salts
had been expelled. The ignited residue was then dissolved in
water, the sulphuric acid and magnesia precipitated by baryta
water, and after separation of the excess of baryta by means
of carbonate of ammonia, the filtrate evaporated to dryness

the Water of the Thermal Spring of Bath. 59

and ignited. The residue imparted a yellow colour to the
blowpipe flame : — Evidencing the presence of soda.

An alcoholic solution of the residue gave with a concen-
trated solution of bichloride of platinum a yellow crystalline
precipitate : — Indicating potassa.

The precipitate which had formed on evaporating for the
iodine determination, was treated with hydrochloric acid, the
filtrate saturated with ammonia and precipitated by sulphide
of ammonium ; this precipitate was re-dissolved in nitro-hy-
drochloric acid mixed with chloride of ammonium, and the
sesquioxide of iron separated by ammonia. The filtrate, eva-
porated and fused with nitrate of potash and carbonate of
soda, gave a green mass: — Showing traces of manganese.

Lithia, alumina, bromine and phosphoric acid were found
to be absent.

In regard to the presence of gases in the water, it was
scarcely necessary to test for the presence of free carbonic
acid. On mixing a solution of lime with the mineral water a
precipitate was formed, which dissolved in an excess of the
mineral water. The quantity of free carbonic acid however
is not very large ; the water has no reaction on blue vegeta-
ble colours ; hydrosulphuric acid is not contained in the water.
Acetate of lead gave only a white precipitate of sulphate of
lead free from all trace of brown colour, which might indicate
the presence of sulphur.

A large quantity of gas is continually disengaged from the
chief spring as well as from the secondary ones. Dr. Dau-
beny* paid particular attention to the composition of this gas.
He found that it consists principally of nitrogen, together
with small quantities of carbonic acid and oxygen.

He employed a peculiar apparatus, constructed on purpose
for these experiments, by which he was enabled to collect the
whole of the gases from the principal well, as well as from
those adjoining it. The experiments of Daubeny are so nu-
merous and accurate as to preclude any other researches on
the subject.

II, Quantitative Analysis,
Determination of the Specific Gravity.

A small bottle, which contained at the temperature of
16°'5 C. (60° Fahr.) 10 grms. of distilled water, contained at
the same temperature 10*025 grms. of the mineral water;
from this the specific gravity of the water is calculated
as 1-0025.

* Vide Memoir mentioned.

60 Messrs. Merck and Galloway's Analysis of

1. Estimation of Sulphuric Acid.

The mineral water was heated with a little hydrochloric
acid and chloride of barium added.

I. 534*199 grms. of water gave l"340grm. of sulphate of
baryta = 0'4605 grm., or 0*08620 percent, of sulphuric acid.

II. 475*003 grms. of water gave 1*1791 grm. of sulphate
of baryta = 0*4050 grm., or 0'08526 per cent, of sulphuric
acid.

Mean of the results, 0*08573 per cent.

2. Estimation of Chlorine.

The water was treated with nitric acid and precipitated by
nitrate of silver ; the precipitated chloride of silver was washed
by decantatioUj fused and weighed.

I. 101 grms. of water gave 0*1137 grm. of chloride of silver
= 0*2811 grm., or 0*02778 per cent, of chlorine.

II. 100*006 grms. of water gave 0*1093 grms. of chloride
of silver =0*02702 grms., or 0*02701 per cent, of chlorine.

Mean of the results, 0*02739 per cent.

3. Estimation of Silicic Acid.

To the water was added nitric acid in excess ; it was then
evaporated to dryness and the residue for some time heated
on the sand-bath. On treating this residue with water and
hydrochloric acid, the silicic acid remained behind; it was
collected, washed and weighed.

I. 765*325 grms. of the water gave 0*0342 grm., or
0*00446 per cent, of silicic acid.

II. 732*015 grms. of water gave 0*0289 grm., or 0*00407
per cent, of silicic acid.

Mean of the results, 0*00426 per cent.

4. Estimation of Iron.

The iron was estimated, —

A. In the precipitate formed on boiling the mineral water.

B. In the water which had not been boiled.
Both estimations gave the same results.
A.^Estimation of the iron in the precipitate : —

A certain quantity of the water was boiled for some time ;
the precipitate which had formed was washed, dissolved in
hydrochloric acid and precipitated by an excess of ammonia.

I. 777*215 grms. of water gave 0*0079 grm., or 0*00101
per cent, of sesquioxide of iron.

B. Estimation of the iron in the water which had not been
boiled : —

the Water of the Thermal Spring of Baih. 61

The liquid filtered off from the silicic acid (3.) was concen-
trated and precipitated by an excess of ammonia.

II. 765-325 grms. of Mater gave 0-0078 grm., or O'OOlOl
per cent, of sesquioxide of iron.

III. 732-015 grms. of water gave 0*0086 grm., or 0-00116
per cent, of sesquioxide of iron.

Mean of the results, 0-00106 per cent., corresponding to
0*00153 per cent, of carbonate of oxide of iron.

5. Estimation of Lime.

The estimation of the lime was divided into —

A. Estimation of the lime contained in the water in the
state of carbonate.

B. Estimation of the lime contained in the water in the
state of sulphate.

C. Estimation of the total amount of lime for control.

A. Estimation of the lime combined with carbonic acid : —
The ammoniacal liquor filtered off from the precipitate of

sesquioxide of iron was precipitated by oxalate of ammonia ;
the oxalate of lime was converted in the known way into
carbonate.

I. 712-747 grms. of water gave, on boiling, a precipitate
containing 0*0904 grm. of carbonate of lime = 0-05062 grm.,
or 0'00712 per cent, of lime.

II. 623-881 grms. of water gave, on boiling, a precipitate
containing 0*0782 grm. of carbonate of lime = 0*0437 grm.j
or 0*00700 per cent, of lime.

Mean of the results, O-OO7O6 per cent.

B. Estimation of the lime combined with sulphuric acid:
The mineral water was kept boiling for one or two hours,

replacing the water w^hich evaporated ; the precipitate formed
was filtered off, washed, and to the filtrate was added chlo-
ride of ammonium, ammonia, and oxalate of ammonia; the
oxalate of lime was converted into carbonate.

I. 710747 grms. of water gave in this way 0*6072 grm. of
carbonate of lime = 0-3400 grm., or 0-04783 per cent, of lime.

II. 623-881 grms. of water gave 0-5165 grm. of carbonate
of lime = 0-2892 grm., or 0*04635 per cent, of lime.

Mean of the results, 0*04709 per cent.

C. Estimation of the total amount of lime for control : —
The ammoniacal liquid which was filtered off from the pre-
cipitate of sesquioxide of iron was precipitated after the addi-
tion of chloride of ammonium by oxalate of ammonia, and the
oxalate of lime converted into carbonate.

62 Messrs. Merck and Galloway's Analysis of

I. 765*325 grms. of water gave 0*7211 grra. of carbonate
of lime = 0*4038I6 grm., or 0*05276 per cent, of lime.

II. 732*015 grms. of water gave 0*6981 grm. of carbonate
of lime = 0*3909 grm., or 0*05340 per cent, of lime.

Mean of the results, 0*05308 per cent.
Mean of the lime combined with carbonic acid 0*00706
Mean of the lime combined with sulphuric acid 0'04709
Total amount found by addition 0*05415
Mean of the total amount found by direct ^sti-l ^ „^^^q
mation j

6. Estimation of Magnesia.

The estimation of the magnesia was divided in the same
manner as the estimation of lime into —

A. Estimation of the magnesia combined with carbonic
acid.

B. Estimation of the magnesium combined with chlorine.

C. Estimation of the total amount of magnesia for control.

A. Estimation of magnesia contained in the water as car-
bonate : —

To the liquid filtered off from the oxalate of lime was
added phosphate of soda ; on stirring, after some time a pre-
cipitate of phosphate of magnesia and ammonia was formed,
which was converted by ignition into pyrophosphate of mag-
nesia.

I. 777*215 grms. of water gave, on boiling, a precipitate
which contained 0*0046 grm. of pyrophosphate of magnesia
= 0*001685 grm., or 0*00021 per cent, of magnesia.

II. 623*881 grms. of water gave, on boiling, a precipitate
which contained 0*0044 grm. of pyrophosphate of magnesia
= 0*00016 grm., or 0*00025 per cent, of magnesia.

Mean of the results, 0*00023 per cent.

B. Estimation of the magnesia contained in the water as
chloride of magnesium.

The liquid filtered off from the oxalate of lime was con-
centrated by evaporation, ammonia added filtered off from a
small portion of silicic acid which separated, and the mag-
nesia precipitated by phosphate of soda.

I. 414*279 grms. of water gave in this way 0*1007 grm.
of p3'rophosphate of magnesia = 0*03689 grm., or 0*008906
per cent, of magnesia.

II. 427*1 grms. of water gave 0*1050 grm. of pyrophos-
phate of magnesia = 0*03846 grm., or 0*009004 per cent, of
magnesia.

Mean of the results, 0*008955 per cent.

the Water of the Thermal Spring of Bath. 63

C. Estimation of the total amount of magnesia for con-
trol :—

The liquid filtered off from the precipitate of oxalate cf
lime was concentrated, ammonia and phosphate of soda added.

I. 7G5*325 grms. of water gave 0-1936 grm. of pyrophos-
phate of magnesia = 0-070929 grm., or 0*00926 per cent, of
magnesia.

II. 732-015 grms. of water gave 0*1837 grm. of pyrophos-
phate of magnesia = 0*0673 grm., or 0*00919 per cent, of
magnesia.

Mean of the results, 0*00922 per cent.

Mean of the magnesia combined with carbonic acid 0*00023
Mean of the magnesia contained in the water asl o*00895

chloride of magnesium j _^_

Total amount found by addition 0*00918
Mean of the total amount found by direct estimation 0*00922

7. Estimation of the Alkalies.

For the estimation of the alkalies the mineral water was
evaporated to one-third of its volume and baryta water added
in excess, the precipitates of sulphates of baryta, lime, mag-
nesia and sesquioxide of iron were filtered off, and the excess
of baryta precipitated by means of carbonate of ammonia.
To get rid of the silicic acid the filtrate was evaporated to
dryness with hydrochloric acid, gently ignited, dissolved in
water, again filtered and evaporated to dryness ; the mixed
chlorides obtained in this manner were weighed.

I. 632*481 grms. of the mineral water gave 0*2937 grm. of
chloride of sodium and chloride of potassium = 0*04643 per
cent, of the mixed chlorides.

II. 546*032 grms. of water gave 0*2538 grm. of chlorides
of sodium and potassium = 0*04648 per cent, of the mixed
chlorides.

Mean of the results, 0*04645 per cent.

8. Estimation of the Potassa.

The chlorides of potassium and sodium were dissolved in
a small quantity of water and an excess of bichloride of
platinum added ; the liquid was then evaporated to dryness
in the water-bath, the residue digested with alcohol, the in-
soluble chloride of platinum and potassium filtered off from
the soluble sodium salt and washed with alcohol ; the preci-
pitate was dried in the water-bath and weighed.

I. 632*481 grms. of the mineral water, or 0*2987 grm. of
the mixed chlorides, gave 0*124 grm. of chloride of platinum

64 Messrs. Merck and Galloway's Analysis of

and potassium = 0*0378 grm. of chloride of potassium
= 0*00597 per cent, of chloride of potassium, which equals
0*00377 per cent, of potassa.

II. 546*032 grms. of water, or 0*2538 grm. of the mixed
chlorides, gave 0*0975 grm. of chloride of platinum and po-
tassium = 0*02977 grm. of chloride of potassium = 0*00545
per cent, of chloride of potassium, which equals 0*00342 per
cent, of potassa.

Mean of the results, 0*00571 per cent, of chloride of potas-
sium and 0*0359 per cent, of potassa.

9. Estimation of the Soda.

The quantity of soda was found simply by the difference
of the mixed chlorides and the quantity of chloride of potas-
sium found by direct estimation.

Mean of the mixed chlorides . . . 0*04645
Mean of the chloride of potassium . 0*00571

Chloride of sodium . . 0*04074

corresponding to 0*02168 per cent, of soda.

10. Estimation of Carbonic Acid.

To find the quantity of free carbonic acid contained in the
water at the moment it was taken from the well, a siphon of
exactly known capacity was immersed in the well, and the
water obtained in this way put in bottles, containing a mix-
ture of ammonia and chloride of calcium. In this way the
free carbonic acid as well as the carbonic acid in combination
was precipitated in the form of carbonates. Four bottles were
filled with mineral water by this method. The capacity of
the siphon was exactly 533 cubic centimetres, therefore
533 X 4 X 1*0025 = 2137 grms. of water were taken.

The precipitate from the water contained in these four bot-
tles was collected, washed, dried and weighed; it yielded
1*4748 grm. of carbonate mixed with some alumina from im-
purity in the solution of chloride of calcium.

To estimate the quantity of carbonic acid in this precipitate,
tw^o portions of it were taken and estimated separately after
the method proposed by Drs. Fresenius and Will.

I. 0*66 grm. of the carbonate, &c. gave in this way 0*22
grm. of carbonic acid, therefore 1*4748 grm. of the carbonate,
or 2137'0 grms. of water, gave 0*4916 grm. of carbonic acid.

II. 0*718 grm. of the carbonate, &c. gave 0*23 grm. of car-
bonic acid, therefore 1*4-748 grm. of the carbonate, &c., or
2137*0 grms. of water, gave 0*4718 grm. of carbonic acid.

tJie Water of the Thermal Spring of Bath.

65

Mean of the results.
0*48l7grm. of carbonic acid,which equals 0*02254 per cent.

Total amount of carbonic acid . . 0-02254
Carbonic acid existing in combination —

With oxide of iron . . . 0-00057

With lime . . . . . 0-00554

With magnesia .... 0-00024

Sum total ...... 0-00635

Free carbonic acid remaining 0*01619

From the details contained in the preceding pages, it fol-
lows that the thermal spring in the King's Bath contains the
following constituents in 100 parts : —

Carbonate of lime . . .

. 0-01260

Carbonate of magnesia
Carbonate of oxide of iron

. 0-00047
. 0-00153

Sulphate of lime . . .
Sulphate of potassa . .
Sulphate of soda . . .
Chloride of sodium . .

. 0-11436
. 0-00663
. 0-02747
. 0-01806

Chloride of magnesium .
Silicic acid

. 0-02083
. 0-00426

0-20620
Traces of manganese and iodine.

Estimation of the total amount of the fixed ingredients in the
water for control.

The water was concentrated in a porcelain dish, and after-
wards evaporated to dryness in a platinum basin. The resi-
due was heated in an air-bath until the weight was constant.
Two estimates were made.

I. 217*058 grms. of water gave 0*4540 grm., or 0*20916
per cent, of residue.

II. 319*57 grms. of water gave 0*6726 grm., or 0*21040
per cent, of residue.

Mean of the results, 0*20978 per cent.

But in this experiment the iron was obtained in the state
of sesquioxide, whilst in the preceding calculation it is taken
as the carbonate of the oxide, in which form it exists in the
water.

On calculating the absolute weights from the above, we
obtain the following numbers : —

Phil. Mag. S. 3. Vol. 3 1 . No. 205. Jidij 1 847. F

66 Analysis of the Water of the Thermal Spring of Bath.

Carbonate of lime . .
Carbonate of magnesia .
Carbonate of oxide of iron
Sulphate of lime . . .
Sulphate of potassa . .
Sulphate of soda . . .
Chloride of sodium . .
Chloride of magnesium .
Silicic acid

In a litre

0-1260 grm.

0-0047

0-0153

1-1436

0-0663

0-2747

0'1806

0-2083

0-0426

0-20621..

In an imperial gallon

(70,000 grs.).

8-82000 grs.

0-32900 ...

1-07100 ...
80-05200 ...

4-64100 ...
19-22900 ...
12-64200 ...
14-58100 ...

2-98200 ...

144-01800

According to our experiments, 1 litre of the water contains
95-64 cubic centimetres of free carbonic acid at the tempera-
ture of 46° C. (115° F.) and normal atmospheric pressure.

One imperial gallon contains therefore 26-45 cubic inches
of free carbonic acid of 46° C, being more than double the
quantity which has been determined by former experiments.

This however is not surprising, as the estimations previ-
ously made had been effected by the expulsion of the carbonic
acid from the water. Besides the difficulty of avoiding a loss
of carbonic acid before the operation, it is scarcely possible,
as Mr. Philli])s justly notices in his paper, to expel all carbonic
acid by simple ebullition. Besides, we see from the experi-
ments of Daubeny, that the gas which escapes from the well
contains at different periods highly varying amounts of car-
bonic acid. He found by several experiments that the King's
Bath evolves on an average 267 cubic inches of gas per minute,
or 223 cubic feet in twenty-four hours. He further ascer-
tained that this gas consists nearly entirely of nitrogen, mixed
with a small amount of oxygen and carbonic acid, and that
these gases were generally in the following proportion : —

Nitrogen . . =91 '9
Oxygen . . . = 3-8
Carbonic acid = 4*3

In many instances, however, he observed as much as 7 '4
to 8-2, and even once 11*5 parts of carbonic acid.

From these observations there is no doubt that the quan-
tity of carbonic acid dissolved in the water is very variable.

In the following Table we give the analyses of former ex-
perimenters, calculated in an imperial gallon (70,000 grs.).

Notices respecting New Booh.

67

Phillips.

Scudamore*.

Walker.

Noad.

7-680
0-274

86-460

14-460
31-680

1-960

5-280
0-200

98-320

i-520
12-240
15-360

i-920

10-667
0-243

81-624

2-927

19-371

15-122

13-339

0-150

3-233

0-521

5-760

96-240

27-456
7-142

3-360

Carbonate of oxide of iron
Carbonate of soda

Sulphate of lime

Sulphate of potassa

Sulphate of soda

Chloride of sodium

Chloride of magnesium ...
Alumina

Silicic acid ..,.•

Quantity directly observed
Carbonic acid

142-394
144125

134-840

146-676
147-622

140-479
149-72

11-52 cub. in.

7-60 cub. in.

Our analysis agrees, as may be seen, best with that of
Walker. According to Professor Liebig'sf arrangement of
mineral M'aters, the thermal spring of Bath would belong to
the saline waters containing carbonic acid.

XI. Notices respecting New Books.

On the Correlation of Physical Forces : being the substance of a Course
of Lectures delivered in the London Institution, in the year 1843.
ByW. R. Grove, Esq., M.A., F.R.S., Barrister-at-Law. Printed
at the request of the Proprietors of the London Institution. London ;
Samuel Highley, 32 Fleet Street.

''r^HIS publication treats of subjects which might have been advan-
-*• tageously considered at much greater length ; but it must be
acknowledged that in the brief space to which the author has con-
fined the announcement of his views and speculations, he has done
them no small degree of justice ; it may indeed be questioned
whether the opinions broached are not of such a nature as to defy
the test of experiment to realise or to refute them. This is certainly
the case as far as experiment has yet been carried ; but although we
discover great reason for doubting whether the difficulties which
beset the subjects may ever be overcome, we discover no cause for
despair, seeing that new modes of research and new instruments for
carrying them out are of almost daily occurrence. As a proof of
this we may cite the author's excellent invention of his well-known
and justly- appreciated voltaic battery ; and his still more recent
discovery, that water may be decomposed by heat so as to exhibit
both its elements in the gaseous form.

Mr. Grove states tliat "the position which he seeks to establish
in this Essay is, that the various imponderable agencies, or the affec-
tions of matter which constitute the main objects of experimental

* Recalculated according to a more correct principle by Walker,
t Handworterhuch der Cliemie, Art. ' Analyse der Mineral wasser,'

F2

68 Notices respecting New Books,

physics, viz. heat, light, electricity, magnetism, chemical affinity and
motion, are all correlative, or have a reciprocal dependence ; that
neither, taken abstractedly, can be said to be the essential or proxi-
mate cause of the others, but that either may, as a force, produce
or be convertible into the other ; thus heat may mediately or imme-
diately produce electricity, electricity may produce heat ; and so of
the rest."

In further illustration of the author's views, we may quote what
he states to be the sense that he has attached to the word correlation,
which is, that " of a reciprocal production or convertibility ; in other
words, that any force capable of producing or being convertible into
another, may, in its turn, be produced by it, — nay, more, can be
itself resisted by the force it produces, in proportion to the energy
of such production, as action is ever accompanied and resisted by
reaction ; thus, the action of an electro- magnetic machine is reacted
upon by the magneto-electricity developed by its action."

In order to support his speculations by facts, the author appeals
in the first place to the agency of electricity. " To commence, then,
with electricity as an initiating force, we get motion directly pro-
duced by it in various forms ; for instance in the attraction and re-
pulsion of bodies, evidenced by mobile electrometers, such as that of
Cuthbertson, where large masses are acted on ; the rotation of the
fly wheel, another form of electrical repulsion, and the deflection of
the galvanometer needle, are also modes of j)alpable, visible motion.
Electricity directly produces heat, as shown in the ignited wire, the
electric spark, and the voltaic arc, in the latter the most intense heat
with which we are acquainted, so intense, indeed, that it cannot be
measured, every sort of matter being dissipated by it. Electricity
directly produces light in the same pha?nomena. It directly produces
magnetism in all ferruginous bodies placed at right angles to its line
of direction, and, indeed, in the substances, of whatever nature,
traversed by the electrical current, in a direction at right angles to
that of the current ; in this case giving us a new character of force,
viz., a force acting, not in direct straight lines, but in a tangential or
rather rectangular direction.

" Lastly, electricity directly produces chemical affinity, and by its
agency we are enabled to obtain efi^ects of analysis or synthesis, with
which ordinary chemistry does not furnish us. Of these effects we
have examples in the brilliant discoveries by Davy of the alkaline
metals, and in the peculiar crystalline compounds made known by
Crosse and Becquerel."

Having stated thus much respecting electricity in support of his
peculiar views, Mr. Grove adduces additional confirmation of them
from considering the action of light, in a passage which we shall
quote at length. He observes that "light is, perhaps, that mode of
force the reciprocal relations of which with the others has been the
least traced out. Until the discoveries of Daguerre and Talbot, very
little could be definitely predicated of the action of light in produ-
cing other modes of f^rce ; and, even, since these discoveries, it is
doubted by many competent investigators, whether the phsenomena

Royal Society^ 69

of photography are not mainly dependent upon a separate agent
accompanying light, rather than upon light itself. It is, indeed, dif-
ficult not to believe that a picture, taken in the focus of the camera
obscura, and which represents to the eye all the gradations of light
and shade shown by the original luminous image, is not an effect of
light ; certain it is, however, that the different coloured rays exer-
cise different actions upon various chemical compounds, and that
the effects on many, perhaps on most of them, are not proportionate
in intensity to the effects upon the visual organs ; those effects,
however, appear to be more of degree than of specific difference,
and without pronouncing myself positively upon the question, hitherto
so little examined, I think it will be safer to regard the action on
photographic compounds as resulting from a function of light : so
viewing it, we get light as an initiating force, capable of producing,
mediately or immediately, the other modes of force. Thus, it imme-
diately produces chemical action ; and having this, we at once ac-
quire a means of producing the others."

Mr. Grove then relates the following beautiful experiment, by
which he conceives th.it he showed the production of all the other
modes of force by light : — " A prepared Daguerreotype plate is in-
closed in a box filled with water, having a glass front, with a shutter
over it ; between this glass and the plate, is a gridiron of silver wire ;
the plate is connected with one extremity of a galvanometer coil,
and the gridiron of wire with one extremity of a Breguet's helix ;
the other extremities of the galvanometer and helix are connected
by a wire, and the needles brought to zero. As soon as a beam of
either daylight or the oxyhydrogen-light is, by raising the shutter,
permitted to impinge upon the plate, the needles are deflected : thus
light being the initiating force, we get chemical action on the plate,
electricity circulating through the wires, magnetism in the coil, heat
in the helix, and motion in the needles."

We have had some difficulty in selecting passages for quotation
from this publication, on account of the profusion of interesting
matter which it contains, though in so small a space ; we believe,
however, that the selections which we have given are such as will
well and sufficiently illustrate the interesting views of their author.

XII. Proceedings of Learned Societies.

ROYAL SOCIETY.

[Continued from vol. xxx. p. 207.]
Feb. 11, "/^N the Amount of the Radiation of Heat, at night,
184'7. ^^ from the Earth, and from various Bodies placed on,
or near the surface of the Earth." By James Glaisher, Esq. Com-
municated by G. B. Airy, Esq., F.R.S., Astronomer Royal, &c.

The author enters into a very detailed description of the construc-
tion of the thermometers he employed in these observations, and
the precautions he took to ensure their accuracy ; and gives tabular
records of an extensive series of observations, amounting to a num-»

70 Royal Society,

ber considerably above ten thousand, with thermometers placed on
nearly a hundred different substances, exposed to the open air, under
different circumstances, and in various states of the sky, at the Royal
Observatory at Greenwich.

Feb. 18. — " On the Diurnal Variation of the Magnetic Declina-
tion of St. Helena." By Lieut.-Colonel Edward Sabine, R.A., For.
Sec. R.S.

It has long been known that the diurnal variation of the magnetic
needle is in an opposite direction in the southern, to what it is in
the northern hemisphere ; and it was therefore proposed as a pro-
blem by Arago, Humboldt and others, to determine whether there
exists any intermediate line oi' stations on the earth where those
diurnal variations disappear. The results recorded in the present
paper are founded on observations made at St. Helena during the
five consecutive years, from 1841 to 1845 inclusive; and also on
similar observations made at Singapore, in the years 1841 and 1842;
and show that at these stations, which are intermediate between the
northern and southern magnetic hemispheres, the diurnal variations
still take place ; but those peculiar to each hemisphere prevail at
opposite seasons of the year, apparently in accordance with the
position of the sun with relation to the earth's equator.

Feb. 25. — " On certain Properties of Prime Numbers." By the
Right Hon. Sir Frederick Pollock, M.A., F.R.S., Lord Chief Baron
of the Exchequer, &c.

The author of this paper, after noticing Wilson's Theorem, (pub-
lished by Waring about the year 1770, without any proof), which
theorem is that, if A be a prime number, 1. 2. 3. . . . (A — l)-f-l is
divisible by A ; refers to Lagrange's and Euler's demonstrations,
and mentions Gauss's extension of the theorem, to any number, not
prime ; provided that instead of 1, 2, 3, &c. (A — 1), those numbers
only be taken which are prime to A, and 1 be either added or sub-
tracted. This theorem was published by Gauss without a proof in
1801, with a rule as to the cases in which 1 is to be added or sub-
tracted, the correctness of which is questioned by the author, who
proceeds to propound the following theorem, which he had previ-
ously, for distinctness, divided into three.

If any number, prime or not, be taken, and the numbers prime to
it, and less than one half of it be ascertained, and those be rejected
whose squares +1 are equal to the prime number, or some multiple
of it (which may be more than one), then the product of the re-
maining primes (if any), + 1 shall be divisible by the prime number.

He gives as examples, 14, the primes to which, and less than one
half, are 1, 3, 5, and 1.3. 5=15; therefore 1.3.5 — 1 = 14; also
15, the primes to Avhich and less, are 1, 2, 4, 7; but 4x4 = 16
= 15 + 1 ; therefore 4 is to be rejected, and 1. 2. 7 + 1 = 15. The
author adds another theorem, that if A be a prime number, all the
odd numbers less than it (rejecting as before) ; also, all the even
numbers (making the same rejection except A — 1) will, multiplied
together, be equal to A+1.

The author then proceeds to prove Gauss's extension of Wilson's

Rmjal Society, 71

theorem, and to give the cases in which 1 is to be added or sub-
tracted ; and in the course of the proof, he mentions that the num-
bers prime to any number not only are found in pairs, one greater
and one less than one-half of the number, but that they associate
themselves in sets of four, with an odd pair in certain cases. Thus,
the primes to 7 are 1, 2, 3, 4, 5, 6, —

2x4=8=7 + 1.

Put the complemental numbers underneath crosswise, thus, —

2x4

\/
/ \

y \

3X5
80 that 2 + 5 and 4+3 may equal 7 ; and then

3x5=15=2x7 + 1

2x3= 6=7-1 4x5=20=3x7-1

Multiplied together one way the product exceeds 7, or a multiple
of it, by 1 ; multiplied the other way, the product is less than 7, or
some multiple of it, by 1. By assuming the prime number to be A,
and the two primes to it to be jo, q, and that p-\-q he not equal to
A, but j05'=wA+l, it is shown that the complemental primes
{A—q) and (A—/?) will have a product=w'A+l, and that, in-
stead of 1, the number may be any other prime to A. Upon this
foundation the author proceeds to show that Wilson's theorem, and
also Gauss's, may be made much more general ; that if A be a prime
number, as 7, the numbers less than it may be arranged in pairs,
not only with reference to 1, but to any number less than 7. Take
4 as an example : —

1 X 3=7-4

4 X 6=4x7—4
2 X 5=2x7—4

therefore 1.2.3.4.5 . 6=7w— 4' ;

therefore ] .2.3.4.5.6+43=7^; that is, is divisible by 7.

The same is then shown as to numbers not prime, provided those
numbers alone are taken which are prime to it, and the number of
pairs will be half the number of primes. The general theorem
therefore is this : — If A be any number, prime or not, and m be the
number of primes to it, which are l,p, q, r, &c. ; then 1 .p.q.r, &c.,

m

+ Z2 will be divisible by A, provided Z be prime to A, whether it
be greater or less.

It follows from this that z^+ 1 must be divisible by A, and there-
fore that z"*- 1 must be divisible by A. If A be a prime number

74 Royal Society.

and z a number prime to it (which every number not divisible by it
is), this is Fermat's theorem, and the author has given a new proof
of it. But the theorem is true though A be not a prime number,
provided z be prime to A and m be the number of primes to A,
and less than it ; and instead of 1 , any other number prime to A
raised to the mi\\ power may be substituted : and z'^—if"^ will be di-
visible by A, provided z and y be primes to A, and m be the number
of primes to A and less than it.

The author has therefore in this paper offered a proof of Gauss's
theorem, and proved that it applies in certain cases to one half of the
primes, and in all cases, with certain modifications, has shown that
a similar property belongs to the product of the odd numbers, and
also of the even numbers which precede any prime number ; and
lastly, has shown the intimate connexion between Wilson's theorem
and Fermat's, and shown that each is but a part of a much more
general proposition, which, he observes, may itself turn out to be
part only of a still more universal one.

In a postscript, the author has shown that the well-known law of
reciprocity of prime numbers is an immediate corollary from his
theorem ; and that it may be extended thus : if A and B be any
two numbers (not prime numbers but) prime to each other, and the
primes to A, and less than it, are (in) in number, and the similar
primes to B are (?^), then (A"—!) is divisible by B, and (B""— 1) is
divisible by A.

" On the reabsorption of the Mixed Gases in a Voltameter." By
Professor M. H. Jacobi, in a letter to Michael Faraday, Esq., F.R.S.
Communicated by Dr. Faraday.

The author found that if the mixed gases developed from the
decomposition of water by a voltaic current, be allowed to remain
in the voltameter in which they were collected, in contact with the
fluid which produced them, they by degrees diminish in volume,
and ultimately disappear by being absorbed by the fluid. He has
not yet fully determined the precise conditions on which this phe-
nomenon depends ; but he is inclined to think that it is owing to a
portion of the mixed gases, diffused throughout the whole liquid,
comipg into contact with the platinum plates, and being recombined
on the surface of those plates ; and this process being renewed with
every fresh portion of the gases which takes the place of the former,
the whole of the gases are thus reconverted into water.

March 4. — " Researches into the effects of certain Physical and
Chemical Agents on the Nervous System." By Marshall Hall, M.D.,
F.R.S., &c.

The professed object of the author, in the present paper, is " to
detail the results of an investigation of the phenomena and the laws
of production and action of certain secondary or induced conditions
of the nervous system, which are effected by a voltaic, and proba-
bly by any other electric current, but persistent after the influence
of that current is withdrawn." This condition he designates by the
new term electrogenic, as describing at once the origin and the inde-
pendence of that condition. On the present occasion he confines

Royal Society. 73

himself to the subject of the electrogenic condition of the muscular
nerves, postponing to future inquiries that of the incident nerves
and of the spinal marrow ; and also the modes of action of other
physical and chemical agents, such as mechanical injury, heat and
cold, strychnine, and the hydrocyanic acid.

The bones and muscles of the brachial lumbar and pelvic regions
of a frog, being isolated from all the other parts of the body, except-
ing only by means of their respective brachial and lumbar nerves,
which were perfectly denuded on all sides, and raised from the glass
on which the limbs were laid, a voltaic current from a pair of the
" couronne de tasses" was passed downwards through the nerves, in
a direction from their origin in the spinal marrow towards their ter-
minations in the muscles. Energetic muscular movements were at
first excited ; and the current was thus continued during the space
of five, ten, or fifteen minutes, and at the end of this period was
withdrawn. No sooner was the current discontinued than the mus-
cles were affected with spasmodic contractions, and with a tetanoid
rigidity, constituting the secondary, or what the author denominates
the electrogenic condition ; an effect, which as instantly subsides on
the restoration of the voltaic current.

' The author proceeds to state the precautions which must be taken
to ensure the success of experiments on this subject; and traces the
effects of desiccation of the nerves from spontaneous evaporation,
and 6f the application of external moisture, on the phenomena ; and
also the modifications introduced by varying the extent of voltaic
contact. Various experiments are then described, which the author
instituted with a view to ascertain the nature of the electrogenic
condition of the nerves, and the circumstances under which it is in-
duced ; and he is led to the conclusion that the phenomena involve
some voltaic principle which has not hitherto been fully investigated.

March II. — "On the cause of the discrepancies observed by
Mr. Baily with the Cavendish Apparatus for determining the Mean
Density of the Earth." By George Whitehurst Hearn, Esq., of the
Royal Military College, Sandhurst. Communicated by Sir John
F. W. Herschel, Bart., F.R.S.

After taking a summary review of the methods employed by Mr.
Baily for determining, on the plan devised by Mr. Cavendish, the
mean density of the earth, and of the anomalies, hitherto unac-
counted for, which had introduced jierplexity in the results obtained,
the author, suspecting that these anomalies had their source in the
variable magnetic states of the masses which were the subject of
experiment, traces the effects which such an influence might be
supposed to have on those results. He finds that, the attraction
arising from gravitation between a mass and one of the balls being
exceedingly minute, an almost inconceivably feeble magnetic state
may be the cause of great perturbations. He then proceeds to in-
vestigate the subject by the application of mathematical analysis ;
from which he is led to the conclusion that the masses and balls do
actually exert on one another influences which are independent of
the action of gravitation. He finds that such influences are of a

74 Royal Society.

very fluctuating nature ; the action arising from them being either
positive or negative, and its sign also changing in each revolution
as the masses are turned round a vertical axis ; and he observes that
such action may either fall short of that arising from gravitation or
exceed it many times. Such disturbing force he conceives can be no
other than a magnetic influence ; not however one of the ordinary
kind, but that which Faraday has recently discovered as affecting
all diamagnetic bodies.

The author concludes by proposing methods by which the inquiry
should in future be conducted, so as to obviate or eliminate this
source of error. Such an inquiry, he remarks, would, by exhibit-
ing the magnetic and diamagnetic powers under new aspects, lead,
in all probability, to important consequences.

March 18.—" Researches to determine the Number of Species
and the Mode of Development of the British Triton." By J. Hig-
ginbottom, Esq., F.R.C.S. Communicated by Thomas Bell, Esq.,
F.R.S.

The observations of the author, of which he gives a detailed ac-
count in the present memoir, have led him to the following con-
clusions : —

Two species only of the genus Triton are met with in England ;
namely, the Triton verrucosus and the Lisso-triton punctatus. It is
three years before the animal is capable of propagating its species,
and four years before it attains its full growth. In its tadpole state,
it remains in the water till its legs acquire suflficient strength to
qualify it for progressive motion on land. While a land animal, it
is in an active state during the summer, and passes the winter in a
state of hybernation ; but does not then, as has been erroneously
supposed, remain at the bottom of pools. Very dry, or very wet
situations are incompatible with the preservation of life during the
period of hybernation. At the expiration of the third year, the
triton revisits the water, in the spring season, for the purposes of
reproduction, and again leaves it at the commencement of autumn.
Impregnation is accomplished through the medium of water, and
not by actual contact. The growth and development of the triton
are materially influenced by temperature, and but little by the action
of light. The triton possesses the power of reproducing its lost
limbs, provided the temperature be within the limits of 58° and 75°
Fahrenheit ; but at lower temperatures, and during the winter, it
has no such power.

April 15. — " On the Proper Motion of the Solar System." By
Thomas Galloway, Esq., A.M., F.R.S.

The object of this paper is to communicate the results of a calcu-
lation for determining the direction of the proper motion of the
solar system from the apparent proper motions of stars in the
southern hemisphere, deduced mostly from a comparison of the
observations made by Lacaille at the Cape, about the middle of the
last century, with the recent observations of Mr. Johnson and the
late Professor Henderson at St. Helena and the Cape respectively.
After adverting to the papers of Sir William Herschel in the Philo-

Royal Society. 75

sophical Transactions for 1783 and 1805, and some other investi-
gations of the same subject, the author remarks that up to a
recent period astronomers seem generally to have entertained the
opinion that our knowledge of the proper motions of the stars is
not sufficiently advanced to enable us to pronounce positively either
on the fact or the direction of the motion of our own system. This
opinion was grounded on the discrepancies which present them-
selves when it is attempted to explain the observed displacements
of individual stars by referring them to the motion of the sun in an
opposite direction ; it being always found that whatever direction
is assigned to the sun's motion, there are many stars whose proper
motions cannot thereby be accounted for. But if the sun be in
motion it is very improbable that any star is absolutely at rest;
hence the proper motions deduced from a comparison of catalogues
must be regarded as the effect partly of the true proper motions of
the stars, and partly of the apparent systematic or parallactic mo-
tion caused by the displacement of the point of view ; and as we
have no reason for supposing the true proper motion of a star to be
more probable in one direction than in another, it may be expected,
a priori, that the observed directions will form angles of all different
values with the direction of the sun's motion, or any other fixed
line. The observed discrepancies are therefore not incompatible
with a general drifting of the stars towards a particular region of
the heavens ; but in order to deduce the direction of the systematic
motion, it becomes necessary to take account of a very considerable
number of proper motions, and to represent them by equations,
involving the unknown quantities required for determining the
direction of the sun's motion, and to solve the equations so as to
obtain the most probable values of those quantities. The first person
who investigated the subject under this point of view was Professor
Argelander of Bonn, in a paper published in the Petersburg Me-
moirs for 1837. From the proper motions of 390 stars deduced
from a comparison of Bessel's catalogue of Bradley's observations
with his own catalogue of stars observed at Abo, Argelander found
the direction of the sun's motion, for 1792*5, to be towards the point
of the sphere whose right ascension is 259° 47'*6 and declination
-1-32° 29'*5. Lundahl, subsequently, from a comparison of the places
of 147 stars in the catalogues of Besseland Pond, and not included
among those considered by Argelander, found the co-ordinates of
the point to be ^=252° 24'-4, Dec. + 14° 26'-l ; and Otto Struve,
still more recently, from the comparison of about 400 of Bradley's
stars with the positions determined at the Dorpat Observatory, ob-
tained the result ^=261° 23'-l, Dec. + S7° 35'-7. The mean of
those results taken with respect to their probable errors, was found
by O. Struve to be ^=259° 9'-4, Dec. + 34° 36'-5.

All the stars included in the calculations of Argelander, Lundahl,
and O. Struve being situated to the north of the tropic of Capri-
corn, it appeared to be a point of some interest to determine whe-
ther the southern stars agree with the northern in their indication
of the direction of the solar motion, or afford any confirmation of

76 Royal Society,

the hypothesis of the sun's translation. Unfortunately, we have
no observations made in the southern hemisphere in the last century
equal in precision to those of Bradley, but the catalogue given by
Lacaille in his * Astronomise Fundamenta,' furnishes a means of
comparison of considerable value in reference to the present in-
quiry. In Mr. Johnson's 'Catalogue of 606 Stars in the Southern
Hemisphere' (London, 1835), there are sixty-one which, on com-
paring their places in 1830 with those of Lacaille reduced to the
same epoch, appear to have sliifted their positions not less than 8"
in space in the interval of eighty years between the epochs of the
catalogues, or to have an annual proper motion of not less than
one-tenth of a second in space. Prof. Henderson's catalogue (Mem.
11. Astron. Society, vols. x. and xv.) furnishes thirty-six stars, which,
on a like comparison, appear to have an annual proper motion ex-
ceeding the same limit. Of these, however, thirty-two are contained
in Mr. Johnson's catalogue, but Henderson gives the proper motions
of sixteen other stars (in the southern hemisphere), from the com-
parison of his own places with those of Bradlej'-. On the whole,
therefore, the two catalogues furnish eighty-one different stars whose
proper motions are given both in right ascension and declination.
The method of investigation is the same as that of Argelander. From
the differences of M. and Dec. given by comparison of the cata-
logues, the direction of the apparent motion of each star is com-
puted. It is then assumed that the sun is moving towards a point
whose right ascension A=259° 46'*2 and declination D=-f-32°
29'*6 ; and the direction in which each star would appear to move,
if it were itself at rest, is computed on this hypothesis. The differ-
ence of these two directions is treated as an error of observation,
and its numerical value substituted for the differential of the angle
which determines the direction of the parallactic motion ; this diffe-
rential being expressed by a formula containing the differentials of
A and D multiplied by known coefficients. An equation is thus
obtained of the form

0 = ac? A -fMD+w,
in which a, b, and n are known quantities. Each star furnishes a
similar equation ; and the equations, being first multiplied respec-
tively by the sine of the star's distance from the point assumed as
the apex of the sun's motion, in order to give them all the same
weight, are solved by the method of least squares, and the result-
ing values of c? A and dT) applied as corrections to the assumed
values of A and D. The results are as follows: — the whole of the
eighty-one equations give (for 1790) as co-ordinates of the point
towards which the sun's motion is directed,

^=263° 38'-0 + 5° 14'-5 ; Dec.=:-f 37° 15'-0+6° 17'-6.

But two of the stars compared with Lacaille move in a direction so
nearly opposite to that of their motion on the assumed hypothesis,
that (in one case especially) a slight error of observation would
change the sign of n in the equations of condition. It therefore
appears necessary to reject those two stars ; and a further reason

Intelligence and Miscellaneous Articles. 77

for rejecting them is, that they are both situated within 8° of the
pole, in which position Lacaille's determination of the right ascen-
sion is probably not to be depended upon. Setting aside, there-
fore, the two stars in question, the remaining seventy-nine equa-
tions give

M^'ISQ" 51'-5 + 4° 45'-l ; Dec.= + 34° 14.'-3+5° 36'-2.
The author further observes, that one of the stars compared with
Bradley's catalogue is also remarkable as appearing to move in a
direction nearly opposite to the mean direction of the whole, and
that if this star be rejected also on account of the great probability
there is that the parallactic motion is in this case concealed by the
larger proper motion of the star itself in an opposite direction, the
co-ordinates of the solar apex become

^=259° 47H+4° 31'-9 ; Dec.= +34° 19''5 + 5° 17'-7,
a result differing less than a degree either in right ascension or de-^
clination from the mean, as above stated, of the three previous de-
terminations.

XIII. Intelligence and Miscellaneous Articles.

ACTION OF CHLORINE ON ALCOHOL. — FORMATION OF ACETAL.

MSTAS states that he has observed that the causes which give
• rise to acetal are not always oxidating causes. When
chlorine is made to act upon alcohol, acetal is the principal product,
as long as it does not act by substitution, and it is at once a dehy-
drogenating and an oxidizing body. This discovery, the author is
of opinion, throws great light on the hitherto obscure action of
chlorine upon alcohol.

In order to obtain acetal by the action of chlorine upon alcohol,
it is sufficient to pass a current of chlorine into alcohol of 80 per
cent., cooled to 50° or 60° F. The action is to be discontinued
when chlorinated bodies commence formation by substitution : this is
readily ascertained, for the alcohol then becomes turbid on the addi-
tion of water ; the liquid, which has become very acid, is to be di-
stilled, and one-fourth of the quantity is to be preserved. This is to
be neutralized by means of chalk, and by a fresh distili3,tion one-
fourth of the product is again to be obtained ; in this fused chloride
of calcium is to be dissolved, which immediately separates a large
quantity of a very volatile fluid, containing, like common rough
acetal, aldehyd, acetic aether and alcohol ; by the addition of more
chloride of calcium, the utmost quantity of alcohol and acetic aether
are separated ; the purification of the acetal is to be completed.

The analysis of the acetal thus obtained was similar to that pro-
cured in the usual way ; and thus the chlorine acts, as already
stated, both as a dehydrogenating and oxidizing body: C''^H'*0^-f-
2Ch-2HO=C'2H'<0* + 2CH + 2HO.— Jw». de Chim. et de Fhys.,
Feb. 1847.

78 Intelligence and Miscellaneous Articles.

BISILICATE OF IRON OR FERRUGINOUS PYROXENE.

This new mineral is described in a memoir presented by M. Du-
fr^noy to the Academy in the name of M. Gruner, mining engineer,
and Professor in the School of Mines at St. Etienne. It corresponds
in composition to a pyroxene with a base of iron.

M. Gruner states that this mineral resembles certain varieties of
asbestos, or more nearly fibrous amphibole. Its specific gravity is
3* 7 13, which exceeds that of the densest epidotes, amphiboles or
pyroxenes. By analysis M. Gruner obtained —

Silica 43-9

Protoxide of iron 52*2

Lime '5

Magnesia , . . . 1 ' 1

Alumina 1*9

99-6
Admitting that the greater portion of the foreign bases is derived
from a small quantity of the gangue, it will be seen that this mineral
is bisilicate of iron, or ferruginous pyroxene with one base only. —
Comptes Rendus, Mai 5, 1847.

CHLOROSULPHURET OF SILICIUM.

M. Isidore Pierre states that M'hen hydrosulplmric acid and chlo-
ride of silicium in vapour are passed through a porcelain tube heated
to redness, they react upon each other : much hydrochloric acid is
produced, which is disengaged with excess of hydrosulphuric acid
gas and a little chloride of silicium, which escapes the reaction.

If the products of this reaction be passed into a U-shaped tube
immersed in cold water, a fuming liquor condenses, which has a
sharp foetid odour, resembling that of hydrosulphuric acid and chloride
of sulphur. The liquor thus obtained was slightly opake by sul-
phur suspended in it: this was deposited by being left forty-eight
hours in a well- stoppered bottle. There were also deposited on the
sides of the bottle, clear lemon-yellow crystals, which were sulphur
in the form of oblique rhombic prisms, without any modification.

The condensed liquor has consequently the power of dissolving
sulphur, and of depositing it in crystals belonging to the same system
as those which are obtained in the dry way. The smallness of these
crystals prevented the author from determining their angles ; but he
reckons upon being able soon to do so. No sensible traces of sul-
phuret of silicium were found in the minute deposit produced in the
porcelain tube.

The liquid condensed in this operation was distilled in an oil-bath
from a retort furnished with a thermometer : the more volatile por-
tions, which usually distil from 140° to 176° F,, were rejected.
They consist principally of chloride of silicium mixed with a small
quantity of chlorosulphuret. Afterwards there is obtained a limpid
colourless liquid which fumes in the air, and has an odour resembling
that of chloride of silicium and hydrosulphuric acid.

Meteorological Observations. 79

Its specific gravity at 60° F. is about 1 "45 ; that is, a little less
than that of chloride of silicium. When it is thrown into water, it
occasions an abundant disengagement of sulphuretted hydrogen gas
and a slight deposit of sulphur. It boils at above 212" F. ; but the
small quantity obtained did not allow of ascertaining its exact boil-
ing-point.

By analysis, it yielded such proportions of its constituents as to
indicate for its formula CP S Si, which would give —

Chlorine 65-47

Sulphur 14-83

Silicium '. 19-70

10000
M. Pierre proposes the name of chlorosulphuret of silicium for this
compound. — Ibid, Mai 5, 1847.

METEOROLOGICAL OBSERVATIONS FOR MAY 1847.

Chiswick. — May 1, Very fine. 2. Cloudy. 3. Rain, 4. Cloudy. 5. Cloudy
and fine. 6. Slight fog : fine. 7. Overcast: showery. 8. Rain. 9. Fine:
cloudy: densely overcast : rain. 10. Very fine : slight showers, 11. Cloudy.
12. Very fine. 13. Cloudy and fine t showers. 14. Showery. 15. Fine: rain
at night. 16. Rain: cloudy: rain at night. 17. Cloudy. 18. Fine: rain.
19,20. Cloudy and fine. 21,22. Very fine. 23. Very hot and sultry. 24.
Cloudy and fine. 25 — 27. Very fine. 28. Slight haze : sultry. 29. Cloudy :
thunder and heavy rain. 30. Clear and fine. 31. Cloudless: exceedingly fine.

JNlean temperature of the naonth 56°*83

Mean temperature of May 1846 56-16

Mean temperature of May for the last twenty years ... 55 '01
Average amount of rain in May 1 "84 inch.

Boston. — May]. Fine. 2. Cloudy: rain early a.m. : rain p.m. 3. Cloudy:
rain A.M. and P.M. 4. Cloudy. 5. Fine: rain p.m. 6. Cloudy. 7. Fine: rain
P.M. 8, Cloudy : rain P.M. 9. Cloudy. 10. Cloudy : rain early a.m. II. Rain.
12. Fine: rain, with thunder i'.m. 13. Fine: rain p.m. 14, 15. Fine: rain
early a.m. 16. Rain: rain, with thunder p.m. 17. Cloudy. 18. Cloudy;
rain p.m. 19, 20, Cloudy, 21—24. Fine. 25. Windy. 26, 27. Fine. 28.
Fine : 1 o'clock p.m. thermometer 82°. 29. Rain : 4 o'clock a.m. thunder, hail
and rain : rain all night. 30. Fine : rain early a.m. 31. Fine.

Sandmck Manse, Orkney. — May 1. Bright: clear. 2. Bright: drops. 3.
Bright : clear. 4. I3right : damp. 5. Fine. 6, 7. Cloudy : damp. 8, 9. Drizzle:
fog. 10. Clear : fine. 11. Cloudy : rain. 12. Rain ; cloudy. 13. Cloudy.
14. Rain: fog. 15. Damp: rain : fog. 16. Bright : cloudy. 17, 18. Cloudy :
clear. 19. Showers: drizzle. 20. Fog : cloudy. 21. Bright: rain. 22. Showers.
23. Clear. 24. Fine. 25. Bright : cloudy. 26. Bright : showers. 27. Fine :
clear. 28. Fine : cloudy : fine. 29. Rain: thunder: cloudy: fine. 30. Clear:
fine. 31. Cloudy: fine.

Applegarlh Manse, Dumfries-shire. — May 1. Fine summer day. 2. Mild:
showers. 3. Cloudy : keen. 4. Spring, but keen. 5. Cold : wet p.m. 6. Grow-
ing: wet P.M. 7. Dull : showers. 8. Dull : wet p.m. 9. Mild: dull: wet p.m.
10. Fine growing day. 11 — 14. Dull: showers. 15. Fine summer day. 16.
Stormy : wet all day. 17. Wet and cold. 18. Wet and stormy. 19. Dull:
wet. 20. Sunshine : fine. 21. Dry : cloudy. 22. Cloudy : showers. 23. Warm :
thunder: rain. 24. Fine : clear : wet p.m. 25. High wind : clear. 26. Fine:
clear : light : cloudy. 27. Fine : clear: thunder. 28. Fine : wet p.m. 29. Fine :
heavy rain P.M. 30. Fine : warm. 31. Remarkably fine.

Mean temperature of the month 51°*1

Mean temperature of May 1846 52 '6

Mean temperature of May for twenty-five years 5i "1

Mean rain in May for twenty years 1*69 inch.

'9JIIJS

-saujiund

•jjoiAvsiqo

« : «s o o

TfiC'^'* •'-<<^^ooo^o^ • CJ
— — oo :ooo — o :o

: in

'Xau^JO

I i S S' « i i « « « « c; cs « « « S s 1 1 ^* s:' s' ^ " «• i « ^ S 2

•ajiqs
-sauj'iuna

^i^^^'^il^i^^'i'^iii^iilii^^i ^ ^ ^ i i

s s

i^^^ii .' ^i-^ii^iiill i i i ^ i ^ ^ii^ ^'-g-g

•5lO!aiS!q3

Sf A aJ to a3 u5 is fe' ^* ^ ^ SJ 5f fe ^ w ^ S a 6t B: fe > ^ a; m w ^ g"

O »!

?6

Sic
Q

•XBJ\[

•W8 ¥8

•uo^soa

•U!H

-eig

00 ONfooo o <M lOO in(M iou:>oo mr^oo o^o "o ^^o^^^o^-- OMn-^-^ i>-t~o

-«>g

•UOJSOa

JO sXbq

2S

O^O^O^O^C> O^O^O^C^^O^O^C?^O^O^O^O^o^C C^O^O^O^O^a^O^O^O o^o^O O

O^St^C•^OOIO^o«^C^Olomt:^coO<^lTtl>.o^^OO^OOO^C^^O•^-HlO
■^d <p ip -rf n rp o (N cool «N cocT^Ol ro»pcOCO>p'piprO'P'7<'0 "^r^oo 0^

r^\ rT\ (T\ fTN 0^ ON (?T\ f^T\ fT\ 0\ f?s f^s CTS CTS ON 0\ '5N 0\ ON ON 0\ ON 0\ ON ON ON (TN rTN CTN C^N /TN

C^O^O^O^O^O^o^o^O^O^CNO^O^O^O^^^O^O^O^O^C^C^O^O^O^c^^o^(^O^ONC^
C<<M0»<N0l<N(M0l(Me<<N(NC«0ltNC»C»<N0l<N0t0l<MC<O?«<N(MCNCN(M

0^ t^oo 00 >

lO t^OO lO <

■<* CO 00 00

-^ •^ cj us

QO t^OO 00

On -^ CO

in Ti>oo

CO lO (N ON

> CO r^r^ip

\ ON Cn On On
<N ?N ei «N

^ii^cTco "^

t-H CO — • 00
lO 00 00 o

— in r^-^

t^OO (N -^

00 t^ 00 00

-^ ON (N 00
O 00 o -^

O OlON O

■^ '*0 —

-< O 00 lO

-- o r^oo

■^00 o lo
CO "^r^oo

— O O 00

0> CO 00

— o vs
T* T 9^
6 6 ON

COCOOI

CO -^00

(N <N O

o 6 6

CO CO CO

o lo

00 t^

com

ON-H

— ON

in CO
CO -^
6 6
coco

C< CO •^"0^0 t-»oo o^O^<NcoTt•ln^o t^oo o^o — iNco'^m'o t>QO o^ o —

^!pH^>-<rtr-^FH^rt<MCI(MC«C«<NC»OIC«C^coC0

*=* • <^ o

THE
LONDON, EDINBURGH and DUBLIN

PHILOSOPHICAL MAGAZINE

AND

JOURNAL OF SCIENCE.

[THIRD SERIES.]

AUGUST 184.7.

XIV. On a tiew Voltaic Battery, cheap in its construction and
use J arid more powetful than any Battery yet made; and on
a cheap substitute for the nitric acid o/' Grove's Platina
Battel y. By the Rev. N. J. Callan, Professor of Natural
Philosophy in the Royal College, Maynooth*.

QOME time ago, whilst I was reflecting on the principle of
^ action of Grove's and Bunsen's batteries, it occurred to me
that lead might be substituted for the platina of the former and
the carbon of the latter. I put into the porous cell of a Grove's
battery a piece of lead about y^thof an inch thick, two inches
broad and six inches long. I found that the voltaic current
produced by the lead excited by a mixture of concentrated
nitric and sulphuric acid was very powerful. 1 afterwards
compared the power of this leaden battery with that of a pla-
tina one of the same size, by sending through the helix of a
galvanometer, at the same time, but in opposite directions, the
currents produced by the two batteries. Both batteries were
charged with the same acids ; the lead and platina were ex-
cited by concentrated nitric and sulphuric acid, and the zinc
by dilute sulphuric acid. The current from the platina bat-
tery destroyed the deflection produced by the leaden one, and
caused an opposite deflection, which indicated that the former
current was about twice as strong as the latter. The two
batteries were left working for about three hours and a half.
At the end of that time the current from the lead was about
twice and a half as powerful as the current from the platina.
The quantity of lead dissolved during these three hours and
a half was very small.

It struck me that by diminishing the action of the acids on
the lead, I might increase the power of the battery. I there-
fore covered a leaden plate with gold leaf, and coated another

* Comniunicated by the Author.
Phil. Mag. S. 3. Vol. 3 1 . No. 206. Jug. 1 84-7. G

82 The Rev. N. J. Callan on a new Voltaic Battery.

of the same size with chloride of gold, in the same way in
which sheet silver is platinized for Smee's battery. These
plates and a platina one of the same size were put successively
into the porous cell of a Grove's battery, and the voltaic cur-
rent sent through the helix of our large electro-magnet, in
which the iron bar is about thirteen feet long and two and a
half inches thick; the copper wire is about 500 feet long and
one-sixth of an inch diameter. The magnetic power given to
the electro-magnet by the leaden plate coated with chloride of
gold, appeared to be equal to that which was produced by the
platina plate. The magnetic effect of the current from the
leaden plate covered with gold leaf was not so great. A coat-
ing of chloride of platina was afterwards found to answer as
well as one of chloride of gold.

Some days after a leaden and platina battery of the same
size were left working for four hours and a half. At the end
of that time the lead plate acted fully as well as the platina.
"When the nitric acid was so much exhausted that the lead
was barely capable of magnetizing the large electro-magnet so
as to make it sustain a certain weight, the leaden plate was
taken out of the porous cell, and a platina plate of the same
size put in its stead. The platina plate was not able to make
the electro-magnet sustain the weight which the lead had
caused it to sustain.

The magnetizing power of the platinized or gilded lead and
platina batteries was compared several times in working an
electro-magnetic machine. On these occasions the power of
the leaden battery was evidently superior to that of the pla-
tina one. Sometimes the platina plate was taken out of the
porous cell, and a platinized or gilded lead plate of the same
size put in its place : the velocity of the machine was instantly
and considerably increased. The same effect was produced
when the platina plate was taken out of the cell and a plati-
nized platina one put in its stead. Hence it appears that a
leaden plate coated with chloride of platina or gold, or a pla-
tinized platina plate, produces a more powerful voltaic cur-
rent than a platina plate does. On the 24th of last May, a
small platinized lead battery and a Grove's battery of the same
size, were exhibited before the Royal Irish Academy. The
power of the former was obviously superior to that of the
latter. By using double leads and single zincs instead of
double zincs and single leads, the power of the battery appears
to be increased. When the lead plates have been used for a
long time, they require to be newly gilded or platinized. After
being used they should be rinsed in water, and dipped into a
weak solution of chloride of gold or platina.

The Rev. N. J. Callan on a neuo Voltaic Battery. 83

Seeing that the concentrated acids, by dissolving the lead,
removed the gold or platina powder, and that the nitric acid
was very expensive, 1 endeavoured to find in its stead a cheap
substitute which would not act on the lead. The first that
occurred to me was common nitre. I dissolved about the
eighth of an ounce of it in sulphuric acid, which I diluted with
nearly an equal bulk of water. 1 poured the mixture into
the porous cell of a Grove's battery, and put into it a pla-
tinized leaden plate. I then sent the voltaic current through
the helix of our large electro-magnet: the magnetic power
given to the magnet appeared to be greater than that which
was given to it by a Grove's battery of the same size, in which
the platina was excited by concentrated nitric and sulphuric
acid. I afterwards compared the heating power of the two
batteries, and found the power of the platinized lead battery
to be evidently superior to that of the other. I charged a
platinized leaden battery with a mixture consisting of about
five parts of sulphuric acid, five of solution of nitre, and one
of nitric acid, and a Grove's battery with equal parts of nitric
and sulphuric acid. The former fused a piece of steel wire
which the latter only raised to a white heat. When a platina
plate is excited by a mixture of sulphuric acid and a solution
of nitre, the voltaic current appears to be as powerful as that
which is produced by the plate when excited by concentrated
nitric and sulphuric acid. The cost of the nitre necessary
for charging a battery is about the twentieth part of that of
the nitric acid. The power of the former declines sooner than
that of the latter: but from the results of several experiments,
I have come to the conclusion that the expense of doing a
given amount of work by a platina battery excited by con-
centrated nitric and sulphuric acid, would be three or four
times as great as if the work were done by a platinized lead
battery excited by a mixture of sulphuric acid and a solution
of saltpetre. I have tried nitrate of soda, or cubic nitre, and
nitrate of ammonia, as substitutes for nitric acid ; but although
they give great power, they do not answer as well as the com-
mon nitre. A solution of common nitre and cubic nitre along
with sulphuric acid, forms a mixture scarcely inferior to the
solution of common nitre and sulphuric acid. The most
powerful mixture for the platina or platinized lead battery
consists of about four parts of sulphuric acid, two of nitric
acid, and two of a saturated solution of nitre. When no nitric
acid is used, at least one half of the mixture should consist of
sulphuric acid, and the remainder of nitre and water : the
solution need not be saturated with nitre. Four parts of sul-
phuric acid, two of a solution of chromate of potash, and two

G2

84 The Rev. N. J. Callan on a new Voltaic Battery.

of the solution of nitre, make a most powerful exciting mix-
ture for platina, but give comparatively little power to plati-
nized lead. I endeavoured to find among the sulphates a
substitute for sulphuric acid, but did not succeed. The vol-
taic current from a platinized lead battery, excited by two
parts of sulphuric acid, three of sulphate of soda, and three
of nitrate of potash, is very powerful, but considerably inferior
to that which is produced by the concentrated acids.

On finding that platinized or gilded lead and platinized
platina were superior to platina, I saw that the cause of the
superiority was that, in the platinized or gilded lead, and in
the platinized platina batteries, the acting metals were, not
lead or platina and zinc, but gold or platina powder, and
zinc; and that the gold or platina powder was more negative
compared with zinc than platina is. Hence I inferred, first,
that a leaden plate coated with any of those substances which
are more negative and cheaper than platina or gold, would
act as powerfully as platinized or gilded lead ; and secondly,
that any other metal to which the platina or gold powder
would adhere might answer as well as lead. I therefore
coated, by the galvanic process, leaden plates with antimony,
arsenic, chromium, molybdenum and borax. The plates
coated with arsenic and molybdenum were much inferior to
platina : those that were coated with antimony and borax
appeared fully equal to platinized lead, but they soon lost their
power. The first plate which I chromed acted as well, and
retained its power nearly as long as platinized or gilded lead.
I afterwards coated a great number of plates with chromium ;
but all of them were far inferior to the first. The power of a
leaden plate is greatly increased by being coated with mer-
cury, or even with clay boiled in aqua regia, or with any other
substance which I tried ; but I have not found any substance
to answer as well as the chloride of gold or platina.

I have compared with platinized lead, the other cheap me-
tals coated with gold or platina, or chromium ; and with the
exception of cast iron, they were all inferior to it. Platinized
or chromed cast iron answers as well as platinized lead ; and
without being chromed or platinized, cast iron appears to act
as powerfully as platina. The power of a cast iron battery in
magnetizing our large electro-magnet, and in driving an elec-
tro-magnetic machine, was compared with that of a Grove's
battery of the same size. In the two batteries the exciting-
mixture was the same. The power of the former appeared to
be fully equal to that of the latter.

From the results of several experiments which I have made
on the relative power of platinized silver and platinized lead,

The Rev. N. J. Callan on a new Voltaic Battery. 85

I feel confident that the latter may, without any diminution
of power, be substituted for the former in Smee's battery.
Cast iron does not take the coating of platina powder (at least
until the hard surface is worn away) so well as lead or silver,
and on that account it does not act as powerfully as either.
But I have found zinc and cast iron excited by dilute sul-
phuric acid as constant in their action as zinc and platinized
lead. A platinized lead, or cast iron plate six inches square,
may be had for the twelfth part of the cost of a platinized
sheet of silver of the same size.

From the experiments which have been described, I infer,
first, that a battery superior in power to Professor Grove's
nitric acid battery may be made by substituting platinized
platina or lead for platina, and nitrosulphuric acid and nitrate
of potash for nitric and sulphuric acid; and secondly, that a
battery equal in power to the nitric acid battery may be con-
structed by the substitution of cast iron for platina.

The advantage of what I may call the nitre platina battery
over the nitric acid one is, that the expense of working the
former is, as has been already stated, considerably less than
that of working the latter.

The advantage of the cast iron or platinized leaden batteries
over Professor Grove's is, that they are far less expensive in
their construction. A plate of cast iron or platinized lead
may be had for a shilling, whilst a platina plate of the same
size will cost nearly three pounds. Besides, a cast iron or
platinized lead battery may be worked by a mixture of nitre
and sulphuric acid for one hour for about the tenth part of
the expense of working a Grove's battery for the same time.

The cheapness of cast iron and platinized lead will enable
every one to procure a powerful voltaic battery. A platinized
lead battery is about fifteen times as powerful as a common
Wollaston battery of the same size. A cast iron battery is a
little less powerful than the platinized lead one; but I prefer
the former, because the cast iron does not require to be chromed
or platinized. I am now preparing two large cast iron bat-
teries for the College : one will contain about thirty-three
square feet of zinc and sixty-six of cast iron, the other will
contain eighty square feet of zinc and a hundred and sixty of
cast iron. These batteries will be more powerful than any
battery ever constructed. The expense will be very mode-
rate; for the zinc plates and Wedgwood troughs of our for-
mer batteries will answer for the new ones.

Maynooth College, July 3, 1847.

[ 86 ]

XV. On the Perturhatioyis of Planets moving in Eccentric
and Inclined Orbits. By Sir J. Lubbock, Bart., F.R.S.

[Continued from page 6.]

¥ N the last Number of the Philosophical Magazine I de-
■*- scribed tables by which the development of the disturbing
function R is greatly facilitated. I shall now describe other
tables which have been calculated for me by Mr. Farley,
and which also facilitate the numerical solution. The ad-
vantages which the employment of tables presents wherever
they can be applied are well known. Not only the march of
the figures affords security against error, but the computer
acquires facility in such calculations systematically undertaken,
while the operations are more easy than they would be if the
quantities required were not connected by a common origin,
or so troublesome as they would be if undertaken by different
individuals, or by the same individual at different times. The
use of tables is out of the question in a literal or algebraic
development; but, on the contrary, it is an important pro-
perty of the numerical development that it can thus be mate-
rially facilitated.

All developments whatever may be resolved into three
classes, which I call literal, quasi-literal, and arithmetic. Li-
teral or algebraical are those which result when the numerical
values of the constants are inserted last, and after the deve-
lopment is complete. Quasi-literal are those which result
when either a part only of the constants are expressed by means
of general symbols, or when the development is made up of
several distinct processes, and when the numerical values are
inserted after a portion of these, but not all have been accom-
plished. Finally, arithmetic or numerical developments are
those which result when the numerical valuesof the constants
are inserted in place of the general symbols before any step
of the development is attempted.

A literal development is generally preferable, for this rea-
son, that if it can be performed, the development which re-
sults serves for every possible value which can be assigned to
the constants. Such, for instance, is the development of the
disturbing functiondue to M. Binet; and if such a development
in terms of the requisite variables could be accomplished and
carried out to a sufficient extent, and if, being accomplished,
numerical values of the constants could be easily introduced,
it would be preferable to any other. M. Hansen's develojj-
ment, in his Memoir on the Perturbations of Encke's Comet
by Saturn, is a quasi-literal development, because a portion
only of the processes is general. The conversion of the quan-

Sir J, Lubbock on the Perturbations of Planets. 87

tities P;^, ; into explicit functions of sines and cosines of mul-
tiples of / is an arithmetical process ; while the calculation of
the quantities A„,,- in p. 29 of M. Hansen's paper, is literal
or algebraical,

I regard as very difficult any development of the disturbing
function, either literal or quasi-literal^ when the eccentricity
of the disturbed body is considerable and the perturbations
are large, as in the case of Encke's comet disturbed by Jupiter;
and if such were possible, the replacement of the numerous
symbols by numbers at the conclusion, would be an operation
of almost insurmountable difficulty. On the other hand, in
performing an arithmetical development according to the rules
which I have invented, not only no quantity can be introduced
which has a numerical value beneath any given limit (say
beneath unity in a given decimal place), but it is equally im-
possible, except by a numerical mistake, that any quantity
which is above that limit can be omitted. The developments
may also be effected by mechanical quadratures, as explained
by M. dePontecoulant(T//eor.y47m/.,vol.iii.), or by the method
given by M. Le Verrier in the first number of the Developpe-
ments sur plusieurs points de la Theorie des Perturbations des
Planetes.

If ar, J/, z are rectangular co-ordinates of a comet or planet
7W, and/' the true anomaly,

x-r\v cos/+a- sin/},

^=;'{^cos/+(7sin/},

2= rig- cos/+^ sin/ }.

*Fj 8^5 O'i (7' Sr> i ^''s constants which depend only on the
elliptic constants of the planet tw, and such that

^ = cos TT + 2 sin^ — sin (tt — v) sin v,

3"= — sin 7r+2 sin^— cos (tt — v) cos V,

^ = sin TT — 2 sin^ — - sin (w — v) cos V,

(J = cos TT— 2 sin^ — cos (tt — v) cos v,
3"= sin(7r— v) sin i,

^= cos (tt— v)sin I.

* I have had occasion to use so many alphabets in the course of the
work from which this is extracted, that I have had recourse to this artifice
of reversing the letters in order not to use the same symbol in two different
significations.

88 Sir J. Lubbock on the Perturbations of Planets

Mr. Farley has calculated for me a table of the values of
these quantities for all the planets, and also for the comet of
Encke, the comet of Bieia, and the comet of Halley.

ciR m' a' f gtt' 7- J, ,, , ui-^l

— = r^ 0-2--, -cosff — »){! +>•} * >•.

[X, }L r \^ r a ^ 'J

l+P=l-^^{{^^Cos/-l3^sin/}f;cos/'

+ {C^cos/+^rsin/}isin/'}

+

cH ci' V 7' I r

I call the quantities -^ cos/', -jSin/', - cos^ — sm/, -5,

a'2

--J25 &c. the elementary quantities, because they are the elements

which, by means of various Combinations, form the disturbing

function IL —z—, &c., and if the numerical values of the con-
ar

slants are introduced before the development is begun are
alone required. Mr. Farley has calculated the coeflicients of
these quantities when they are developed in terms of the mean
motions for the planets, and also other tables for eccentricity,
•1, -2, ... . •?, which show the convergence to be so slight,
that such mode of development can only be employed when
the eccentricity is small. These tables have all been con-
structed by means of mechanical quadratures. These tables
are not wanted for the comets, because their co-ordinates
cannot be developed in terms of their mean anomalies in suf-
ficiently converging series.

When the method of mechanical quadratures is applied to
the determination of the perturbations of comets, a correction
is required ; but when that method is used for the determina-
tion of coeflicients of this nature, the limits of the integral are
0 and 360°, and the correction vanishes ; so that by means of
several particular values, rigorous values of the coeflicients
are easily obtained. Nor does the width of the interval matter,
provided it is not made too large. It is difficult to give pre-
• cise rules to regulate the width that should be employed ; but
in the formation of these tables it was easy to employ various
modes of verification. As this inquiry is in its infancy, I
considered it sufficient to retain only those terms which are

moving in Eccentric and Inclined Orbits. 89

due to the elliptic motion ; but hereafter it may be desirable
to reconstruct the tables of the elementary quantities for each
of the planets, retaining some of the principal inequalities due
to the disturbing force.

Pingre, in his work On comets rtiariy years since, gave a list
of comets, with their elements. At that time, however, the
method of finding the orbit, or even the distance of a comet,
was understood by so few persons, that, from that and other
causes, the numbers contained in that table may not be accu-
rate: many other comets have been discovered since, and such
a table brought Up to the present tittle appears tO be an impor-
tant desideratum in astronomy.

Mr. Hind has kindly favoured me with the following list of
comets which have been made out to be periodic: —

Elements of Halley's comet, by Westphalen, for 1835. Ast.
Nach.^ No. 588.

^=•96739 7r = 304° 31' 32"-19 v = 55° 9' 59"'S^
i=\T ^b' 5"'\S a=17'9879i Retrograde.

Elements of Encke's comet by Encke, for 1829. Ast. Nac/i.,
No. 489.

f =-84462 ff=157° 17' 53"-35 v = BW 29' 3i"*62

i = 1 3° 20' 34"'49 a = 2-22394 Direct.
Elements ofBiela's comet, 1846j by Prof. Plantamour. Ast.
Nach., No. 584.

^ = •75700 7r=109° 2' 20"-10 v = 245° 54' 38"-8

»=12° 34' 53"-47 a = 3-52452 Direct.
Elements of the comet of Faye, by M. Le Verrier, for 1844,
omitting the terms multiplied by jw,". Ast. Nach., No. 541.
€='55596 7r = 49° 34' 19"-39 v = 209° 29' 19"'26

1 = 11° 22° 31"-40 a = 3-81179 Direct.
Elements of DeVico's first comet,byDr.Brunnow, for 1844.
Ast. Nack., No. 563.

^=•61765 7r = 342° SO' 49"-64 v = 63° 49' 0"-l 1

« = 2° 54' 50"-33 « = 3- 10295 Direct.
Elements of Brorsen's first comet, byDr.Brunnow,forl846.
Ast. Nach., No. 557.

^ = •79362 7r=116° 28' 34" i/= 102° 39' 36"'3

» = 30° 55' 6"-6 a = 3-1502l Direct.
The following are the elements of the comet of Encke for
1829 used by M. Hansen : Additions a la Conn, des Temps,
1847, p. 54.

^ = •844676 ' *f=157° 18' 24"'6 v^334° 29' 28"-8
»=i3° 20' 40"-2 a = 2-21997.
Ast. Nac/i.y No. 5^1.

[ 90 ]

XVI. On the Heat of Vapours.
By Sir J. Lubbock, Bart.^ F.R.S.*

T ET V be the quantity of absolute heat, considered as a
-■-' function of the sensible heat or temperature 9,

p being the density, p the pressure, k and u constants,
dp _ «p dp _ up

dQ~~ l+uQ dS "" l+«fl*

If c is the specific heat of a gas, the pressure being con-
stant, and Cf its specific heat when the volume is constant, so
that

_dFdp _dfd^ _£

"""dpdfl ""''dpdQ '^^ c,

dV dV ^

^d7+y^d^=^-

Laplace evidently considered 7 constant, and he integrated
this equation upon that hypothesis, " En supposant cette quan-
tite rigoreusement constante, &c.," Mec. Cel. vol. v. p. 127.
Again, Poisson, in repeating the same theory, Traite de M^c,
vol. ii. p. 6^Q, "En regardant y comme une quantite constante,
&c." If y is constant,

2.
V=A + B^^=A + B-(-+^p^-\

p a \u /

(see vol. xviii. p. 507) which is identical with the equation
given in the Comptes Rendus, Seance de 31 Mai 184'7, p. 920,

q=m-\-n{a-\-t)p-'^,

m — A, n= — , a= —, t=9. 2;=! , A: = y;

a. a. y

but if, as Professor Holtzmann maintains (see Taylor's Sci-
entific Memoirs, vol. iv. part 14), z is variable, the integral
of Laplace does not necessarily obtain, nor does the equation
{Comptes Rendus, p. 920)

obtain ; because if s is a function of t,

-^^^7ip-^-n{a-\-t)p-<'\ogp-^,
* Cuiumunicated by the Author.

Mr. Grove on the Decomposition of Water by Heat. 91

and

q—q^=zn{a-\-t)p-'' —n{a-^t^)p^-\.
It has not, I believe, been remarked, that the integral

?
will however still satisfy the differential equation

If

or

pAV , AV ^

f> dp <yp dp

XVII. On ceiiain Phcenomena of Voltaic Ignition and the
Decomposition of Water into its constituent Gases hy Heat,
By^. R. GuovE, Esq., M.A., F.B.S.
[Continued from p. 35.]

I WAS now anxious to produce a continuous development of
mixed gas from water subjected to heat alone, in other
words, to succeed in an experiment which should bear the same
relation to experiment fig. 9 as fig. 5 did to fig. 7 ; for this pur-
pose the apparatus shown at fig. 10 was constructed: a and

Fig. 10.

b are two silver tubes 4 inches long by 0*3 inch diameter ;
they are joined by two platinum caps to a platinum tube c,
formed of a wire one-eighth of an inch diameter drilled
through its entire length, with a drill of the size of a large

92 Mr. Grove on the Decomposition of Water hy Heat.

pin ; a is closed at the extremity, and to the extremity of h is
fitted, by means of o coiled strip of bladder, the bent glass
tube d. The whole is filled with prepared water, and having
expelled the air from a by heat, the extremity of the glass
tube is placed in a capsule of simmering water. Heat is now
applied by a spirit-lamp, first to h and then to a, until the
whole boils; as soon as ebullition takes place, the flame of an
oxyhydrogen blowpipe is made to play upon the middle part
of the platinum tube c, and when this has reached a high
point of ignition, which should be as nearly the fusing-point
of platinum as is practicable, gas is given off, which, mixed
with steam, very soon fills the whole apparatus and bubbles up
from the open extremity, either into the open air or into a gas
collector. Although by the time I had devised this apparatus
I was from my previous experiments tolerably well assured of
its success, yet 1 experienced a feeling of great gratification
when on applying a match to one of the bubbles which were
ascending, it gave a sharp detonation ; 1 collected and ana-
lysed some of it; it was 0*7 oxyhydrogen gas, the residue
nitrogen, with a trace of oxygen.

Those who have endeavoured to deprive water of air, will
have no difficulty in accounting for the residual nitrogen, or
nitrogen mixed with a small portion of oxygen, which has
occurred in all my experiments. De Luc pointed out the
impossibility of practically depriving water of air, and Priest-
ley, from observing the obstinacy with which water retained
air, was led to believe that water was convertible into nitrogen
(phlogisticated air). I have repeated several of Priestley's
experiments under much more stringent circumstances, and
have never been able to free water from air, or so to boil
water that for every ebullition of vapour a minute bubble of
permanent gas was not left, which appeared to have been an
indispensable nucleus to the vapour.

The difficulty of boiling water increases, as M. Donny has
proved, in proportion to its freedom from air, and at last the
bursts of vapour become so enormous that the vessels em-
ployed are generally broken. There appears to me a point
beyond which this resistance does not extend ; but even at
this point a minute bubble of air is left for each burst of va-
pour, though they are so few and distant that the aggregate
amount of gas is very trifling. I have produced from water
which had been previously carefully deprived of air by the
ordinary methods, three-fourths of its own volume of perma-
nent gas, which proved to be nitrogen ; but as the water in
this experiment was boiled under a long column of oil, it is
probable that if any oxygen were present, it might have been

Mr. Grove on the Decomposition of Water by Heat, 93

absorbed by the oil ; I have, however, always found the pro--
portion of oxygen to decrease as the boiling was continued.
It may be vvorth noticing, as having bad some influence on
my mind, that many months ago, when considering the expe-
riments of Henry and Donny on the cohesion of water, I
mentioned to Mr. Gassiot, and also to Mr. Bingham my
assistant (to whose assiduity I am much indebted), that I was
inclined to think if water could be absolutely deprived of air,
it would be decomposed by heat, a result which I have now
attained by a totally different series of inductions. It is a cir-
cumstance worthy of remark, that I find the greater part of
the air to be expelled at a comparatively low temperature, and
when the water has come in contact with the platinum, while
the decomposition all takes place when the platinum is sur-p
rounded by an atmosphere of steam, if steam it may be called,
for the state of this atmosphere at the first immersion of the
platinum is at present very mysterious.

I think I may now safely regard it as proved, that platinum
intensely ignited will decompose water, and several considera-
tions press on the mind in reflecting on this novel phaeno-
menon.

First of all, to those who are attached to the cui bono argu-
ment, and estimate physical science in proportion only to its
practical applications, I would say that these experiments
afford some promise of our being, at no distant pei'iod, able
to produce mixed gases for purposes of illumination, &c. by
simply boiling water and passing it through highly ignited
platinum tubes, or by other methods which may be devised;
we in fact by this means, as it were, boil water into gas, and
there appears theoretically no more simple way of producing
chemical decomposition.

To pass however to more important considerations: the
spheroidal state, which has lately attracted the attention of
philosophers, appears to be closely connected with these re-
sults, and is rendered more deeply interesting. The last
experiment but two which I have mentioned, shows that the
spheroidal state is intermediate between ordinary ebullition
and the decomposing ebullition ; it is probably therefore a
state of polar tension, coordinate in some respects with that
which takes place in the cell of a voltaic combination before
decomposition, or when the power employed not being of suf"
ficient intensity to produce actual decomposition, the state
commonly called polarization of the electrodes, obtains. The
phajnomenon brings out also a new relation between heat,
electricity, and chemical affinity; nitherto many electrical
phpenomena could be produced by heat and chemigal action,

94- Mr. Grove on the Decomposition of Water by Heat.

the difference being that in the effects produced by the last
two forces there was no polar chain, but every minute portion
of the matter acted on gave rise to the phaenomena which in
the electrical effects are only observable at the polar extremi-
ties ; thus in decomposing water by iron and sulphuric acid,
or by passing steam over heated tubes of iron, parallel results
are obtained to the electrolysis of water with an iron anode ;
but in the former cases every portion of the iron oxidated
gives off its equivalent of hydrogen, in the latter the equiva-
lent is evolved from the cathode at a point distant from that
where the oxidation takes place. Hitherto electricity has been
the only force by which many compounds, and particularly
water, could be resolved into their constituents without either
of these being absorbed by another affinity. The decompo-
sition by ignited platinum removes this exception, and pre-
sents the parallel effect produced by heat alone.

Although there is no substance except platinum and some
of the more rare metals, such as iridium, which promise much
success in a laboratory experiment made for the purpose of
producing the effect I have described, as the greater number
of substances which will bear a sufficient heat, are fragile,
oxidable, or affected by water, yet general considerations from
the nearest analogies in chemistry would lead us to expect a
similar effect from all matter in a state of intense ignition ;
even assuming the presence of solid matter to be necessary,
the catalytic effects of platinum are shared in different degrees
by other substances : it therefore appears probable that at a
certain degree of heat water does not exist as water or steam,
but is resolved into its constituent elements. If, therefore,
there be planets whose physical condition is consistent with
an intense heat, the probability is, that their atmosphere and
the substances which compose them ai*e in a totally different
chemical state from ours, and resolved into what we call ele-
ments, but which by intense heat may be again resolved into
more subtle elements. The same may be the case in the
interior of our planet, subject however to the counter agency
of pressure.

The experiments strongly tend to support the views of
Berthollet, that chemical and physical attraction are affinal,
or produced by the same mode offeree. All calorific expan-
sions appear to consist in a mechanical severance of the mole-
cules of matter; and if heat produce effects of decomposition
merely by increase of intensity, there seems no reason why
we should assign to it in this case a different mode of action
from its normal one. On this view physical division carried
on indefinitely must ultimately produce decomposition, and

Mr, Grove on the Decomposition of Water hy Heat. 95

chemical affinity is only another mode of molecular attraction.
Thus a high degree of rarefaction, as at the bounds of the
atmosphere, or in the interplanetary spaces, may entirely
change the chemical condition of matter.

In a paper published in the Philosophical Transactions for
1843, p. Ill, I have shown that we may oppose a chemical
action by a physical one (electrolysis by a vacuum), that an-
tagonizing chemical by physical tension, they mutually oppose
each other. 1 believe the converse of this experiment has
been made by M. Babinet, who by physical compression has
prevented the development of chemical action.

I have also described in the Philosophical Magazine for
November 1845, certain phaenomena which appear to me to
be irreconcileable with received chemical views ; and though
I then believed that the theory of Grotthus would be obliged
to give way, I now incline to think that some of our chemical
doctrines must ere long undergo a revision.

It is rather surprising that the valuable applications of which
the phaenomena of voltaic ignition are capable, and the fertile
field which (as I believe) it presents for discoveries, both phy-
sical and chemical, should have been so completely neglected.
It is true that until a recent period the imperfection of the
voltaic battery rendered accurate and continued experiment
on this subject difficult of performance, but still much might
have been done. Davy made several experiments on the
voltaic disruptive discharge, which in many points may be
regarded simply as very intense ignition \ but I am only aware
of two experiments of his on voltaic ignition; one, in which
he employed it in an exhausted receiver to examine to what
extent the i*adiation of heat was carried on in vnaio; and
another, already alluded to, in which, by immersing a portion
of an ignited wire in water, he observed that it conducted in
some inverse ratio to its heat.

I have made a vast number of experiments on the voltaic
arc or disruptive discharge, in various media*; when this is
taken in a medium incapable of acting chemically on the elec-
trodes, the phaenomena are those of intense ignition of the
terminals, which are dissipated in vapour and condensed upon
the interior of the vessel in which the discharge is taken. 1
have examined some of these deposits, and they appear to
consist of the metal of the terminals in a finely-divided state ;
this is strikingly shown with zinc. If the arc be taken between
zinc points in an exhausted receiver, a fine dark powder,
nearly black, is deposited on the interior, which, when col-
lected, proves to be pure zinc, and on the application of a

* Phil. Mag., June 1840 ; Literary Gazette and AthenEeum.Feb. 7, 1845.

96 Mr. Grove on the Decomposition of Water by Heat.

gentle heat, takes fire in the open air and burns into the white
oxide: to casual observation the zinc would appear to be
burned twice. The experiment appears to me to present an
argument in favour of the dynamic theory of heat.

With charcoal, on the other hand, there is little or no de-
posit, but the charcoal continually yields carbonic oxide and
hydrogen, and this for hours after the presence of water would
be deemed impossible. I have taken the arc between pieces
of well-burned charcoal for eight or nine successive hours,
and there was still gas generated ; indeed it appeared to be
given off as long as there was any charcoal remaining, and a
conversion of the carbon into inflammable gas might have
been supposed, Much still remains to be done with this
powerful agent, the voltaic arc: where, however, the object
is simply to expose gases to an intense heat, the ignition of a
conjunctive wire of platinum is more simple in its application,
more uniform in its action, and instead of requiring a power-
ful battery, the effect can be satisfactorily produced by five or
sIk cells, in many cases by two.

The heat is not so intense as that of the arc, but as it can
be brought to within a few degrees of the fusing-point oC pla-
tinum, it is far more intense than, any heat usually employed
in laboratories, certainly than any which can be applied to
minute, I may say microscopic portions of gas or vapour.

In conclusion, I must express my sincere thanks to the
managers of the London Institution, for having permitted me,
as an honorary member, to carry on these experiments in the
laboratory of the Institution.
London Institution, Aug. 21, 1846.

XVIII. Supplementary Paper on pertain PhcBnomena of Voltaic
Ignition, and the Decompositio?i of Water into its constituent
Gases by Heat. BijW. R. Grove, Esq."^

IN selecting the above title, I endeavoured to give as clear
an enunciation of the phEpnomena to be described in the
paper as was consistent with the brevity usual in a title.

An exception has, however, been taken to it, that as the
effects of decomposition are produced by ignited platinum,
the phsenomena may result from that obscure mode of action
called catalysis. That I did not intend to exclude from con-
sideration any possible action of the substance employed, will
be evident from the paper itself, in which I have called attention
to the general production of catalytic effects by solid bodies.

* From the Philosophical Transactions for 1847, part i.; having heen
received by the Royal Society November 26, and read November 26, 1846.

Mr. Grove on the Decomposition of Water by Heat. 97

Whatever value or novelty there may be in the facts I
have communicated, is the same whether they be regarded
as resulting from catalytic or from thermic actions. If the
action be catalytic, it is one absolutely the reverse of that
usually produced by platinum, and therefore just as much at
variance with received experience as decomposition of water
by heat would be ; the effect of platinum, like that of heat, on
the elements of water having been hitherto known only as
combining them. With regard to any theoretic views I may
have advanced, I by no means attach the same importance to
them as 1 do to the facts themselves, though I consider it
necessary for the collation of facts, and desirable for the pro-
gress of science, that an author pretending to communicate
new results should give with them the impressions which led
to their discovery, and the inferences which he regards as im-
mediately deducible from them. No expression can be given
to facts which does not involve some theory, and admitting
the difficulty (perhaps insuperable) of correctly enunciating
new phaenomena, and the probability of future discoveries
entirely changing our views regarding them, I cannot at pre-
sent see that the title of ray paper could be altered without
being open to greater objections. I am of this opinion, not
so much because other bodies than platinum will produce the
effect, as I shall presently show, nor from the fact that the
electrical spark will decompose aqueous vapour, though these
are arguments in its favour ; but from the following conside-
rations. The catalytic action of platinum will induce or en-
able combination to take place where there is already a strong
affinity or tendency to combine, as with mixed oxygen and
hydrogen gases; it will also induce decomposition where the
affinities are extremely weak, or in a state of unstable equili-
brium, as in Thenard's peroxide of hydrogen ; again, where
there are nicely-balanced compound affinities, it may change
the chemical arrangement of the constituents of a compound,
but 1 do not know of any case in which a powerful chemical
affinity can be overcome by catalytic action ; to effect this we
require some natural force of greater intensity than that to be
overcome. We might as well say that the platinum electrodes
of a voltaic battery decompose water, as to say that platinum
decomposes it in the case in question ; there, the force of
electricity acts only by means of matter, and matter of a pecu-
liar description ; its action also is only perceptible at the sur-
face of this matter. I seek to use the expression in my title
with reference to heat in a similar sense to that in which we
use similar terms with reference to electricity, i. e. to regard
heat as the immediate dynamic force which overcomes the
PJiil. Mag. S. 3. Vol. 3 1 . No. 206. Aug. 184.?. H

98 Mr. Grove Ofi the Decomposition of Water by Heat.

affinity ; thus, as we say when employing the voltaic battery,
that we decompose water by electricity, so here we should say
that we decompose it by heat.

If it be said that heat so weakens or antagonizes the affinity
of the elements of water as to enable catalytic action to sepa-
rate them, this amounts to the same theory, as heat is then
regarded as the antagonizing force, and in this case the action,
both thermic and catalytic, is the reverse of the normal action.
I have thought it desirable shortly to discuss this question as
likely to lead to further investigation, though I liave been
somewhat embarrassed by the want of definite meaning in the
term catalysis; I must plead guilty to have frequently used
the term, but notwithstanding, or perhaps on account of, its
convenience, it has I fear had an injurious effect on scientific
perspicuity.

The following experiments were made to ascertain whether
platinum was the only substance by which the effect could be
produced. A knob or button of the native alloy of iridium
and osmium of the size of a small pea was formed by the vol-
taic battery; to this was attached by fusion another smaller
knob of the same metal one-fourth the size of the former, and
to this smaller one was attached a stout platinum wire ; the
object of the second knob was both to prevent the fusion of the
platinum wire and also to avoid the possibility of any surface
of platinum being exposed to the recipient tube or alloyed with
the metal to be heated. The preparation of this simple in-
strument was very troublesome, but when made it answered
the purpose well ; the larger button could be fully ignited to
an intense glow, while on account of the narrow neck which
united them, the smaller was barely red-hot, and the platinum
wire not perceptibly ignited. An experiment having been
made with this metallic button and prepared water, similar to
that previously made with platinum, gas was given off which
averaged 0*3 of mixed gas ; the residue was nitrogen mixed
with varying small quantities of oxygen. The effect, upon the
whole, was decidedly inferior to that of the platinum. Indeed
as platinum is the most dense and unalterable of all known
substances, it would be likely, upon any received theory of
heat, to produce the greatest effects.

I tried palladium in the same manner ; the gas yielded was
hydrogen with small quantities of oxygen, and the water was
stained with the oxide of the metal.

I now tried silica and other oxides, but the results were
not very satisfactory. A spheroid of silica was formed by
fusing pulverized silica on to a platinum wire, so as to cover
it for the length of 0*4; of an inch ; when this was plunged into

Mr. Grove on the Decomposition of Water by Heat. 99

the hot water and again fused in the oxyhydrogen blowpipe,
it constantly became frothed with small bubbles of vapour,
and after a few experiments generally separated in fissures ;
in the experiment which was continued for the longest time
without disintegration, the gas given off contained 0*15 of
oxyhydrogen gas ; from the whole result I believe there is an
action of the water on the silica (probably forming a hydrate
decomposable by heat) which is a bar to satisfactory results.
With other oxides, at least such as would bear an intense
heat, the difficulties were still more insuperable. Priestley
has shown that water will corrode glass, and if I mistake
not, others have shown the same effect produced on silica.

Although, as applied to the facts detailed, I attached no
further meaning to the title of my paper than that which I
have above stated, yet in one or two theoretical inferences I
have certainly gone further ; for instance, when I suppose the
possibility or probability of mechanical rarefaction producing
the same effects as heat, here (although I do not, indeed I can-
not conceive the existence of heat without matter) I certainly
abstract from the proposition any consideration of solid matter.
In order to ascertain how far this view might be founded on
truth, I had thought of making a few experiments on the
effect of mechanical rarefaction on the tendency of gases to
combine, but (in addition to the interference of necessary
occupations) I find that M. de Grotthus has already experi-
mented on the point; his experiments, as far as they go, cor-
roborate the views I have put forth.

He finds* that mixed gases, such as chlorine and hydrogen,
or oxygen and hydrogen, when rarefied either by slow incre-
ments of heat or by the air-pump, do not take fire (" ne s'en-
flamment pas") by the electric spark. From the context, he
evidently means that the gases will not detonate or unite in
volumes, as he states that a partial combination ensues. Grott-
hus appears to have considered the combination of gases by
the electric spark as an effect of sudden compression or mole-
cular approximation, certain particles being brought within
the range of their affinities by the sudden dilatation of others.
Although he did not pursue the subject far enough to ascertain
whether a degree of rarefaction could be reached which would
be an actual bar to combination, still his experiments strengthen
those views which assimilate mechanical and thermic molecular
repulsion, and regard chemical affinity as being antagonized
by physical repulsion.

Pursuing the series of analogies from the decomposition of
euchlorine at a low temperature, that of ammonia at a higher,
• Annales de Chimie, vol. Ixxxii.

H2

J 00 Mr. Grove on the Decomposition of Water by Heat,

that of metallic oxides at a higher, and so on to oxide of hy-
drogen, there appears to be an extensive series of facts which
afford strong hope of a generalized antagonism between ther-
mic repulsion and chemical affinity, and a consequent esta-
blishment of the law of continuity in reference to physical and
chemical attraction.

The deposit from chlorine, to which I have alluded in my
paper, I have since examined, and though it differs in colour
from that described in books, I find it is a protochloride of
platinum, formed at the expense of the platinum wire. The
larger portion of the chlorine in the tube combines with the
hydrogen of the aqueous vapour, and the muriatic acid is
absorbed by the water; when the experiment terminates the
gaseous volume is reduced to nearly one-half, and this residue
is oxygen.

This effect induced me to try an ignited wire on other ana-
logues of chlorine, and I tried bromine and chloride of iodine
in the apparatus (fig. 5). The tube was filled with the liquid,
and its extremity was in the first experiments immersed in
another narrow tube of the same liquid as that which filled it.
When the platinum wire was ignited, permanent gas was
given off both from the bromine and from the chloride of
iodine, which gas on examination proved, to my surprise, to
be oxygen. In one experiment I collected half a cubic inch
of gas from an equal volume of chloride of iodine. As the
experiment in this form required too large a quantity of the
liquid to enable me to observe any change which might take
place in its character, I repeated it with a tube five feet long,
bent in two angular curves. A small quantity of the liquid
was placed in the extremity of the tube containing the wire,
which was so arranged as to be the lowest point; the angles
were placed in cold water and the experiment proceeded with ;
my object was to enable the dense vapour of the liquids to
shelter them from the atmosphere, there being no satisfactory
method of shutting them in and yet allowing room for the
elimination of the liberated gas, or of absorbing the latter by
combination without also absorbing the vapours.

1 had hoped by the above means to proceed with the ex-
periments until all the oxygen was liberated that could be
driven off, and then to have examined the residua ; but I found
that after experimenting for a short time, both the platinum
wire and the glass in proximity to it were attacked by the
liquids ; this difficulty, similar to those which have hitherto
prevented the isolation of fluorine, I have not yet been able
to conquer, though I hope to resume the experiments.

As chloride of iodine is decomposed by water, it cannot

Sir David Brewster on the Structure of Topaz, 101

contain any notable quantity of the latter, but, until the expe-
riments ai'e carried further, it must remain a question whether
the oxygen results from a small quantity of water contained
in the liquid, the hydrogen combining with the liquid itself, or
from a decomposition similar to that of the peroxides. The
experiments certainly add a new and striking analogy to those
already known to exist between the peroxides and the halo-
gens, l3ut they do not, as far as I have hitherto carried them,
necessarily prove analogy of composition.

In conclusion, I would call attention to a point which I
omitted to notice in my original paper, viz. the explanation
afforded by the results contained in it of the hitherto myste-
rious phaenomena of the non-polar decomposition of water by
electrical discharges, as in the experiments of Pearson and
Wollaston. This class of decompositions may now be car-
ried much further. With the exception of fused metals, I
know of no liquid, which, when exposed to intense heat such
as that given by the electric spark, the voltaic arc, or incan-
descent platinum, does not give off permanent gas; phos-
phorus, sulphur, acids, hydrocarbons, water, salts, bromine
and chloride of iodine, all yield gaseous matter.

Viewing these effects simply as facts, and without entering
on any theoretical explanations or speculations, I cannot but
think that there is a remarkable generality pertaining to them
worthy of the most careful attention.

The apparatus I have described, particularly that repre-
sented by fig. 5, and the numerous applications of voltaic
ignition which will occur to those who duly consider the sub-
ject, promise, I venture to believe, new methods and powers
of investigating the molecular constitution of matter, and will,
1 trust, lead to many novel and important results.

Nov. 10, 1846.

XIX. On the Modification of the Doubly Refracting and Phy-
sical Stnicture of Topaz, by Elastic Forces emanating from
Minute Cavities. By Sir David Bkewster, K.H.i D.C.L.,
F.B.S., and F.P.R.S. Edin.'^

[With a Plate.]

WHILE examining, in polarized light, the form and
structure of the numerous crystals which I had dis-
covered in the fluid cavities of topaz, my attention was par-
ticularly called to certain optical phaenomena exhibited in
other parts of the specimen. These phaenomena, when first

* Read before the Royal Society of Edinburgh on the 20th of January
1845, and published in their Transactions, vol. xvi. part l.p. 7.

102 Sir David Brewster on a Modification of the Doubly

presented to me, were very indefinite in their character, and
very imperfectly developed ; but after a diligent examination
of nearly 900 specimens of topaz, 1 succeeded in obtaining the
most satisfactory exhibition of them under various forms, and
in various degrees of intensity.

When an elastic force is propagated from a centre, in a soft
and compressible medium, an increase of density is commu-
nicated to the surrounding mass,— of a temporary nature if
the medium is a hard solid, like glass, but of a permanent
nature if the medium is soft, and becomes indurated during
the continuance of the compressing force. Both these effects
may be exhibited experimentally ; the first by a pressure upon
glass, and the second by the action of an expanded bubble of
air upon gum in a state advancing to induration.

The physical change thus produced in the transparent me-
dium, whether it be temporary or permanent, may be exhibited
to the eye in two ways ; either by the property of the com-
pressed parts in depolarizing light, or in the unequal refraction
of common light produced by a varying density, and conse-
quently a varying refractive power. In ihejirst of these cases,
the depolarizing action is displayed in the production of four
quadrants of light, separated by the radii of a black rectan-
gular cross, similar to the central portion, or the tints of the
first order, in the uniaxal system of polarized rings ; and, in
the second case, the inequality of refractive density is shown
by the mirage of a luminous point, in the form of concentric
circles surrounding the centre of force, each circle marking
successive actions of the central force.

When the four luminous quadrants of depolarized light,
shown at A, B, C, D in Plate I. fig. 1, first presented them-
selves to me, 1 had some difficulty in perceiving the seat of
the force, by which I believed that they were produced. The
centres, or intersections of the black cross, were either too
deep beneath the surface of the topaz, or too much covered
by fluid cavities, to be seen ; but by removing the part of the
crystal which contained these cavities, I succeeded in finding
that in every case there was a minute cavity in the centre of
the luminous quadrants, or at the intersections of the arms of
the black cross, from which the compressing force had ema-
nated. One of these cavities is shown at E, fig. 2. It is of
a quadrangular form, like the section of a rhbmboidal prism,
sometimes elongated, and sometimes of a slightly irregular
shape. When perfectly regular, these cavities are between
the SOOOdth and the 4000dth of an inch in diameter. They
are always dark, as if the elastic substance which they con-
tained had collapsed into an opake powder ; and I have met

Refracting and Physical Structure of' Topaz. 103

with only one case in which there seemed to be a speck of
light in the centre. The degree of compression to which the
topaz has been subjected is measured by the polarized tint
developed in the luminous quadrants. It varies from the
faintest pale blue to the white of the first order. In one case
I found the luminous quadrant of one cavity coinciding with
a luminous quadrant of another cavity, and thus producing
the sura of their separate tints. This effect is shown in fig. 3.

In the phaenomenon now described, the elastic force has
spent itself in the compression of the topaz. The cavity itself
has remained entire, without any fissure by which a gas or a
fluid could escape. I have discovered, however, other cavities,
and these generally of a larger size, in which the sides have
been rent by the elastic force ; and fissures, from one to six in
number, propagated to a small distance around them. These
fissures have modified the doubly refracting structure pro-
duced by compression ; but, what is very interesting, no solid
matter has been left on the faces of fracture, such as that which
is invariably deposited, when an ordinary cavity, containing
one or both of the two new fluids, is exploded by heat. The
form of some of the cavities which have suffered this disrup-
tion is shown in fig. 4.

The influence of the compressing forces in altering the
density, and consequently the refractive power of the topaz, is
so distinctly seen in common light as to indicate the phseno-
mena that are seen under polarized light. When the cavity
is most distinctly perceived, it is surrounded with luminous
and shaded circles, as shown in fig. 5; and traces of these are
distinctly seen, as shown in fig. 6, when the specimen is ex-
amined in polarized light.

The cavities now described have obviously no resemblance
whatever to those which I have described in previous papers
as containing two new fluids. When any of the latter are
either burst by heat, or exposed under high temperatures to
the compressing forces of the fluids which they contain, they
exhibit none of the phaenomena peculiar to the former. The
doubly refracting structure suffers no change ; and when the
cohesive forces of the crystal are overpowered, the faces of
most eminent cleavage separate, and are covered with trans-
lucent crystalline particles, which the evaporated or discharged
fluids leave behind.

The peculiar character of the pressure cavities, as we may
call them, is still further evinced by the nature of the speci-
mens in which they occur. I have never found them accom-
panying the ordinary cavities with two fluids. The specimens
which contain them have imbedded in them numerous crystals,

104' Sir David Brewster ott the Structure of Topaz.

differing little in their refractive power from topaz, and ex-
hibiting in polarized light the most beautiful colours, varying
with the thickness of the crystal, and diminishing in intensity
as their axes approach to the plane of primitive polarization.

It is impossible to review the preceding facts without arri-
ving at the conclusion, that the topaz must have been in a
soft and plastic state when it yielded to the compressing force
which emanated from the cavities ; and that a mineral body
thus acted upon could not have been formed, according to
the received theory, by the aggregation of molecules having
the primitive form of the crystal.

In a letter to Sir Joseph Banks, printed in the Philosophical
Transactions for 1805, I deduced, from my experiments on
depolarization, the existence of a new " species of crystalliza-
tion, which is the effect of time alone, and which is produced
by the slow action of corpuscular forces;" and I have re-
marked that "this kind of crystallization will probably be
found to have had an extensive influence in those vast arrange-
ments which must have attended the formation of our globe."
These views have been confirmed by various new facts, wholly
independent of each other ; — by the existence of crystals im-
bedded in topaz, and having their axes in all possible direc-
tions, but especially by the nature and form of the strata of
fluid cavities in that mineral. These strata cut at all inclina-
tions the primary and secondary planes of the crystal. They
are bent in the most capricious manner, forming planes of
double curvature; and, what is also true of individual cavities
stretching in every possible direction, they could never have
been formed but when the topaz was in a soft and plastic
state.

An objection to these views may be drawn from the fissures
which proceed from the pressure cavities. The topaz must,
doubtless, have been indurated when these fissures took place ;
but it is equally obvious that the depolarization produced by
compression must have previously existed, and it is probable
that the fissures were produced after the crystal had been
removed from its matrix, and when, from cleavage or other-
wise, its cohesive forces had been diminished.

St. Leonard's College, St. Andrew's,
January 16, 1845.

C 105 ]

XX. Researches on the Compositioji and Characters of certain
Soils and Waters belonging to the Flax districts of Belgium,
and on the Chemical Constitution of the Ashes of the Flax
Plant. By Sir Robert Kane, M.D., M.B.I.A.

[Continued from p. 45.]

3. Results of the Examination of the Ashes of Flax grown
upon the Soils previously analysed,

A. This was coarse flax ; and the flax of this district is
usually of rather poor quality. It is however in most cases
sown late, about the 15th of May.

On incineration, this flax was found to give of pure ash, in
average, 4'237 per cent.

The stem, dried at 212°, and analysed, was found to con-
tain 0*982 per cent, of nitrogen.

The ash contained, per cent., after deducting the sand and
charcoal, which can be considered but as accidentally present :

Potash 7-697

Soda 19-186

Lime ........ 15-379

Magnesia 3 •446

Oxide of iron 4*501

Alumina 0-444

Oxide of manganese ... a trace

Sulphuric acid 6-280

Phosphoric acid .... 11*206

Carbonic acid 20-599

Chloride of sodium . . . 8-213
Silica 3-056

100000

B. This flax was of the very best description, and was
grown from first-class seed.

The stem, dried at 212°, and analysed, was found to con-
tain per cent 0*756 of nitrogen.

On incineration, the plant, dried at 212°, yielded in average
5-434 per cent, of pure ash.

After deducting the sand and charcoal accidentally present,
the ash was found to contain per cent., —

106 Sir Robert Kane on the Chemical Constitution of the

Potash 22-897

Soda none

Lime . 16-4.83

Magnesia 3*332

Peroxide of iron .... 1*523

Alumina 0-438

Oxide of manganese ... a trace

Sulphuric acid 6* 174

Phosphoric acid .... 11*802

Carbonic acid 25*235

Chloride of sodium . . . 8*701
Silica 3*409

99*994

C. This flax was very fine, and was said to be as good as
any grown in that season.

The stem, dried at 212°, and analysed, was found to con-
tain, per cent., 0*876 of nitrogen.

On incineration, the plant, dried at 212°, yielded in average
3-670 per cent, of pure ash.

After deducting, as usual, the sand and charcoal, the ash
was found to contain per cent., —

Potash . 22*303

Soda 14*116

Lime 18*525

Magnesia 3*933

Peroxide of iron . . . , 1*100

Alumina 0*725

Oxide of manganese ... a trace

Sulphuric acid 6*833

Phosphoric acid .... 8*811

Carbonic acid 16*383

Chloride of sodium . . . 4*585
Silica 2*678

99-992

D. This flax, of a rather coarse quality, had been sown
May 2nd, and pulled July 29.

The plant, dried at 212° and analysed, yielded 0*901 per
cent, of nitrogen.

On incineration after desiccation, it gave 4*543 per cent, of
ashes.

The composition of the ash per cent, was —

Ashes of the Flax Plant. 107

Potash 25-790

Soda -4.29

Lime 19-098

Magnesia 3-648

Peroxide of iron , . ♦ . 2*281

Alumina none

Oxide of manganese . . . none

Sulphuric acid 12-091

Phosphoric acid 10-983

Carbonic acid 9-895

Chloride of sodium . . . 12*751

Silica 3-030

99-996
Loss -004

100-000

H. The flax grown upon the Dutch soil yielded, on ana-
lysis, roOO per cent, of nitrogen, when dried at 2 12° Fahren-
heit.

It also gave, by incineration, 5*151 per cent, of ashes, of
which the composition per cent, was found to be as follows :—

Potash ........ 18-410

Soda 10-912

Lime 18-374

Magnesia ....... 3-023

Peroxide of iron .... 2-360

Alumina . 1-439

Oxide of manganese . . . none
Sulphuric acid . . . . . 9-676
Phosphoric acid .... 11 -058

Carbonic acid 13-750

Chloride of sodium . . . 5*655
Silica 5*327

99*984
Loss *016

100*000

If we examine somewhat in detail the results of the ash
analyses above given, there will be found several points worthy
of attention, in reference to the probable laws of replacement
of acids and bases, as mineral constituents of plants ; and also
with regard to the necessary presence of certain materials.

It will be seen that in all cases a large proportion of the
bases of the ash had been combined with organic acids, and
were hence found in the ash as carbonates. This quantity is,

108 Sir Robert Kane on the Chemical Constitution of the

however, variable ; and it will be seen that a variation takes
place in the quantity of sulphuric acid exactly of an opposite
character ; so that in the plant, the proportions of organic
salts and of sulphates would appear to have been such, that an
increase in one replaced any deficiency of the other. Thus
when the quantity of carbonic acid in the ash was 25*235, the
sulphuric acid was 6"174-; but when the sulphuric acid was
12'091, the carbonic acid fell to 9'895. I do not however
mean absolutely to assert that the sulphuric and the organic
acids of the plant are, in all cases, or exactly, mutually re-
placing.

The small quantity, as well as the narrow limits of fluctua-
tion of the silica, is worthy of notice ; particularly when com-
pared with that which I shall have to notice as regards the
composition of Irish flax. It does not appear connected with
any of the bases in particular, nor to follow any special varia-
tion among them.

There is nothing more peculiarly characteristic in the com-
position of the ashes of the flax plant, than the quantity of
phosphoric acid which is found therein. In order to bring
this into full evidence, I shall extract from the works of other
chemists the determination of the quantity of phosphoric acid
in the ash yielded by the stems of other plants.

Tobacco stalk and leaves . . 2*73

Wheat stems 3*10

Oat stems 3-00

Clover plants 6*30

The stems of flax are, then, more than double as rich in
phosphoric acid as the stems of even the cereal grasses or
leguminous plants ; and if we even look to the constitution of
the ash of many substances used as food by man, we shall find
that, in 100 parts, there are from the ash of —

Oats . . . .14*9 phosphoric acid
Potatoes . . . 11'3
Turnips ... 6*1

whilst the average of the analyses of Belgian and Dutch flax
ashes show that there are present no less than 10*77 per cent.
It was this enormous quantity of the most valuable ingredient
of manure that first impressed me with the importance of its
oeconomy, and induced me to endeavour to fix the attention of
agriculturists upon the fact ; for if we calculate, from the pro-
duce per acre, the quantity of phosphoric acid taken from a
statute acre of ground by an ordinary crop of any of the usual
kinds, we shall find that it amounts in the case of flax to very
nearly as much as with any of the ordinary grain or root crops ;

Ashes of the Flax Plant. 109

and that whilst the mineral elements of these are what the
value really consists in, the value of the flax is altogether in-
dependent of those constituents, which are thus so much real
loss to the farmer.

Hence, under the ordinary plan of cultivation, farmers were
certainly in the right to consider it one of the most exhausting
crops ; and that its place in rotation should be equivalent to
that of a grain crop, which it ought by no means to follow, or
be followed by ; whereas, under a system of management which
should allow of the proper oeconomy of its mineral constituents,
that are separated in the processes of watering and dress-
ing, the phosphoric acid and other materials might be restored
to the manure heap or to the field, and the crop of flax be
thus deprived of those permanently exhausting qualities which
it now possesses.

It will be interesting further to notice the constitution of
these ashes, under a point of view which has been put forward
by some chemists, as possessing the character of a general rule
or law; to wit, that although the individual bases present in
an ash may vary very much, and even some (as in one of the
ashes analysed, B soda) may be totally absent, yet the sum of
the oxygen present in the bases will be found to be constant.
If we apply that rule to the ashes above analysed, we shall
find —

Title of ash. Quantity of oxygen in bases.

A 13-73

B 10-95

C 14-65

D 13-45

H 13-60

Average . 13*28

There is certainly a close agreement among these numbers;
and if we excluded one analysis (B), which is also exceptional
in containing no soda, it should decidedly appear that the
quantity of oxygen present in the bases of 100 parts of ash
was represented by a constant number (13*86). It will be
found that the analyses of Irish flax lend support to this view;
but I think that we shall require very many more analyses
before we can fix upon it as a positive law.

In order to afford comparison with the results above given,
I have extended my analyses of Irish flax ; and as there appear
one or two remarkable points of difference between them, I
shall notice also my prior results.

The flax I originally experimented on was grown at my
own residence, a short distance from Dublin. It yielded,

110 Sir Robert Kane on the Chemical Constitution of the .

when dried at 212°, 0*56 of nitrogen per cent., and 5 percent,
of ashes, consisting of, in 100 parts, —

Potash 9-78

Soda 9-82

Lime 12*33

Magnesia 7*79

Alumina 6'08

Phosphoric acid . . . 10*84

Sulphuric acid .... 2*65

Carbonic acid .... 16*95

Chlorine 2*41

Silica 21*35

100-00

I selected for another analysis a specimen of flax given to
nie by William Blacker, Esq., which had received a prize at
the Market-hill show by the tenants of the Earl of Gosford.
When dried at 212°, this flax yielded 0*672 per cent, of nitro-
gen, and 5*572 per cent, of ashes, which contained per cent. —

Potash 6*332

Soda 6*350

Lime ........ 22*699

Magnesia ...;... 4.*058

Peroxide of iron .... 13*520

Oxide of manganese . . . 1*092

Alumina none

Sulphuric acid .... . 8*929

Phosphoric acid .... 7*002

Carbonic acid 4*107

Chloride of sodium . . . 0*901

Silica 24-978

99*968
There is first to be remarked the very curious circumstances
of both Irish specimens containing a large quantity of silica,
from 21 to 25 per cent., whilst the Belgian and Dutch flax
contained only from 3 to 5 per cent. In the Dublin flax there
is no particular replacement to which this could be attributed;
but in the Armagh flax, the small quantity of carbonic acid,
only 4 per cent., shows that the organic acids had been but
little generated in the plant, and probabh' a quantity of silica
was substituted for them. The question of whether this large
quantity of silica, which, however, is mostly removed from the
fibre along with the other materials during its dressing, could
produce in it any degree of hardness or brittleness, is very
well worthy of the attention of the philosophical agriculturist.

Ashes of the Flax Plant. Ill

It is remarkable also, that in both Irish flaxes the potash
and soda are present in equal quantities, though not in the
same quantity in each ash. This, however, may be only a
coincidence, though still a remarkable one.

A more interesting peculiarity is the presence, in the Armagh
flax, of the very large quantity of peroxide of iron, 13'5 per
cent. In the Dublin flax I have not formerly counted iron
as an ingredient, although I did find in the analyses a small
quantity, because I had burned the plants on a sheet of iron
wire-gauze, and I feared that a minute quantity of iron might
be derived from that ; and also that, in that analysis, my only
object was to show the presence of large quantities of valuable
ingredients, which the farmer ought to oeconomise. I there-
fore did not separately determine that minute trace of iron,
which, however, could in no way aflect the numerical results.
The occurrence of the large quantity of iron in the Armagh
flax is, therefore, the more curious ; and it will be interesting
to examine, by other analyses of the flax sown in the sand-
stone districts of the north of Ireland, whether the same pro-
portion of oxide of iron will be found.

Notwithstanding the great difference in the quantity of silica
in the Irish flaxes from the Belgian, the proportion of oxygen
per cent, in the bases comes out nearly the same. Thus the
bases contain of oxygen, —

Flax from Dublin .... 13-41
Flax from Armagh . . . . 13*66
closely coinciding with the number already found for the
Belgian and Dutch flax.

It is not unimportant to correct a statement recently made,
that prepared fibre of flax is not so destitute of mineral con-
stituents as I have assumed in the preceding investigations.
In order to arrive fully at the truth, I have instituted some
additional experiments, with the following results : —

A. Very imperfectly dressed flax from the county Clare
gave, by incineration, with proper precautions, 0*97 per cent,
of ashes, containing principally oxide of iron and lime.

B. A specimen of perfectly dressed flax from Belfast gave,
on incineration, 0*62 per cent, of ashes.

C. A specimen of fine dressed linen gave, on incineration,
0*24 per cent, of ashes, principally lime, with some oxide of
iron. Hence it is evident that my former results on this point
were precisely confirmed by these new trials,

4. Results of the Examiyiation of the Waters selected for steep-
ing Flax in Belgium.
No. 1. This water is from a large pond near the bank of

112 Sir Robert Kane on the Chemical Constitution of the

the Scheldt, which has been most likely* formed by digging
out peat for fuel, as the soil near it is peat, and as in neigh-
bouring ponds peat is now scraped up from the bottom, and
prepared for fuel by drying in the sun. This water is renewed
by the overflowing of the Scheldt, and is apparently not at all
peaty.

This water was pretty clear, but contained some suspended
matter. When 100,000 grains were evaporated to dryness
there was obtained 51*70 grains of residue, consisting of, in
100 parts, —

Protoxide of iron . . . . '514

Lime 6*940

Magnesia '856

Soda 28-620

Potash 8-740

Sulphuric acid ..... 8*054

Muriatic acid 25-765

Phosphoric acid .... no trace
Carbonic acid, with organic! gO-Sll
matter and loss . . J

100-000

No. 2. Water from one of the best Bloe retting pits, near
Hamme Log, in Belgium. This water is also supplied from
the Scheldt annually, before the retting season commences,
and left to stand in the pit for six or eight weeks. The top
becomes covered with green weeds which are cleared oif im-
mediately the flax is put in. This causes the water to be
muddy, as there is a considerable thickness of mud at the
bottom which is disturbed, the workmen standing in the pit
when cleaning the top of the water. The flax is then laid in ;
and after laying two or three layers, they shovel up some of
the mud in the bottom to put on the flax to sink it ; and when
the pit is full, the flax is covered by about an inch thickness
of mud. This sample was taken from a pit which had just
been disturbed and mudded by cleansing the top of weeds,
preparatory to putting the flax in.

This water was found very muddy, but the suspended mat-
ter was principally organic.

100,000 grains left by evaporation 139-69 grains of solid
matter} of ochrey appearance, and consisting, per cent., of —

Ashes of the Flax Plant. 113

Protoxide of iron .... 6*633

Lime 8-435

Magnesia ]'369

Soda 11-607

Potash . 4-181

Sulphuric acid ..... 8-435

Muriatic acid . . . . . 8-682

Phosphoric acid .... no trace

Carbonic acid, with organic 1 -/% /- m
,. , , ^ > 50*658

matter and loss . . J

100-000

No. 3. This water is from a large pond similar to that from

which No, 1 is taken, but from a different part of the country,

and a much larger body of water.

It was clear, containing but very little suspended matter.

100,000 grains left on evaporation 50*68 grains of solid

residue, which consisted of, per cent., —

Protoxide of iron .... 2*584

Lime 17*829

Magnesia 1-530

Soda 30*232

Potash ....... 15-762

Sulphuric acid 11-627

Muriatic acid 2-580

Phosphoric acid .... no trace

Carbonic acid, with organic^! , >, ^ -^
11 ° ^ 17*856

matter and loss • • J

100-000

No. 4. This water is from the river Lys, so celebrated for

its steeping qualities. It was taken from the river in France

before it had reached the highest retting place. The specimen

was clean, but there was some suspended matter, principally

organic.

100,000 grains, evaporated to dryness, left a residue of 45-1 1

grains, consisting of, in 100 parts, —

Protoxide of iron .... 6-200

Lime 5-484

Magnesia 1-192

Soda 28-298

Potash 5-405

Sulphuric acid 9*300

Muriatic acid 7*754

Phosphoric acid .... -079

Carbonic acid, with organic"!

matter and loss . . J___^

100-000

Phil. Mag. S. 3. Vol. 31 . No. 206. Aug. 1 847. I

114 Mr. J. P. Joule on the Theoretical Velocity of Sound.

No. 5. This watei'' was from a retting pit in Holland.

100,000 grains, evaporated to dryness, gave a residue of

42*4< grains, which consisted, per cent., of —

Protoxide of iron . . . . 1*183

Lime ... . . . . . 3-613

Magnesia 7-601

Soda 19-277

Potash 8-205

Sulphuric acid 5*607

Muriatic acid ..... 9*439

Carbonic acid, with organic! a ^ r^^r-
a\ f 45-075

matter and loss. . . J

100-000
With regard to the constitution of these several specimens
of water, it can only now be remarked, that in all there was
present a large quantity of mineral impurities ; and that in
Nos. 2 and 4, the very samples which are of the most remark-
able and celebrated steeping waters in Belgium, a large quan-
tity of iron is present, so that they might be in a degree termed
chalybeate waters. How this regards their excellence for
preparing flax 1 do not pretend to say, and indeed it will
require much more extended investigation before a satisfactory
solution of it can be given.

All these waters are further remarkable for containing a
larger quantity of potash than ordinary waters are found usu-
ally to have. I shall not, however, enter minutely into the
discussion of their constitution, as I shall have to resume the
subject at another time; and I wish only to place on record
for the present the analytical results which the samples of
waters forwarded to me from Belgium by Mr. Marshall, had
afforded.

XXI. On the Theoretical Velocity of Sound. By J. P. Joule *.

THE celebrated French mathematician De Laplace has,
it is well known, pointed out that the heat evolved by
the compression of air is the cause of the velocity of sound,
according to the theory of Newton, being so much less than
that actually observed. He has also given a formula by which
the velocity may be determined when the ratio of the specific
heat of air at constant pressure to that at constant volume is
known. The determination of the elevation of temperature in
air by compression has however been hitherto attended with
difficulty, and hence the theorem of De Laplace has never yet
been fairly compared with experiment. I was therefore anxious
* Communicated by the Author.

Mr. Nicholson on the Composition of Caffein. 115

to ascertain how far the mechanical equivalent of heat, as de-
termined by my recent experiments on the friction of fluids,
might be able to contribute to clear up this question.

The capacity of air at constant pressure, according to the
experiments of De la Roche and Berard, is 0*2669. Conse-
quently a quantity of heat capable of increasing the tempera-
ture of a lb. of water by 1°, will give 1° also to 3'747 lbs. of air,
while the air will be expanded i^^j', an expansion in which a
force equal to 200*7 lbs. through a foot is expended in raising
the atmosphere of the earth. The equivalent of a degree of
heat per lb. of water, determined by the careful experiments
brought before the British Association at Oxford, is 775 lbs.
through a foot. Hence 200*7 lbs. through a foot is equal to
0°-259.

We see, therefore, that for every degree of heat employed
by Qe la Roche and Berard in expanding and heating air,
0°'259 was occupied in producing the mechanical effect, leaving
0°*741 as that actually employed in raising the temperature of
the air. Hence the actual specific heat (commonly called
capacity at constant volume) is 0*2669 x 0*741 = 0*1977. Ta-
king this as the specific heat of air and the equivalent 775, it
follows that if avolume of air of 17r6 cubic inches be com-
pressed to 170*6 cubic inches, it will be heated 1°, a quantity
of heat which will occasion an increased pressure of ^|y. So
that the celerity of sound will be increased by this means in the
subduplicate ratio of 491 to 661*6, or in the simple ratio of
2216 to 2572, which will bring it up from Newton's estimate
of 943 to 1095 feet per ", which is as near 1130, the actual
velocity at 32°, as could be expected from the nature of the
experiments on the specific heat of air, and fully confirms the
theory of Laplace.
Oak Field, near Manchester,
July 17, 1847.

XXII. On the Composition of Caffein, and of some of its
Compounds. By Edward Chambers Nicholson, Esq."^

/^AFFEIN was first analysed by Professors Liebig and
^^ Pfafft in 1832. The result of this investigation was
confirmed by a subsequent analysis of Prof. Wohler %.

In 1838 Professor Liebig induced M. Jobst § to analyse
thein, who proved this body to be identical with caffein.
His analyses gave the same results as his predecessors. The
same remark apphes to the experiments of Mulder I| on thein,

* Communicated by the Chemical Society; having been read Feb. 15,

t Liebig's Jnnalen, i. 17. % Ibid. § Ibid. xxv. 63.

II Bulletin des Sciences Phys, et Nat. de Neerlande. 1838, p. 32.

12

116 Mr. Nicholson on the Composition of Caff em,

and also to an analysis which M. Martins * made of guaranin,
a substance, the identity of which with cafFein and thein had
previously been pointed out by Berthemot and Dechastelusf.
Lately Dr. Stenhouse :j:, when examining Paraguay tea, has
also made some analyses of thein.

The following table, in which I have recalculated these ana-
lyses according to the atomic weights, carbon 6 and hy-
drogen 1, allows a comparison to be made of the results ob-
tained by these chemists.

Mean of the Analyses.
CafFein. Thein. Guaranin.

->

Liebig & Pfaff. Wohler. Mulder. Jobst. Stenhouse. Martius,
Carbon » 49-30 49-25 49-18 49-47 48-95 49-23
Hydrogen 5*22 5-43 5-49 5-20 5*15 5-08
Nitrogen. 28-86 28-53 28-90 28-83

The most simple expression which can be deduced from
these numbers is

Cs Hg Ns O^.
Stenhouse's analysis however of the platinum compound
proves that this formula must be doubled, and that the atom
of cafFein or thein is

CjsH.oN^O^.
The theoretical numbers of this formula are the following: —
16 eqs. Carbon ... 96 49-48

10 ... Hydrogen . . 10 5-15

4 ... Nitrogen . . 56 28-86

4 ... Oxygen . . . _32 16-51

] 94 100-00

From these numerous experiments the composition of caf-
fein might have been considered as perfectly established. In
a recent investigation of coffee, however, M. Payen§ states
that he has obtained results which differ very sensibly from
those obtained by his predecessors, and which he has trans-
lated into the formula

which contains 1 equiv. of oxygen less than the formula up
to the present time admitted.

The theoretical numbers of Payen's formula are —
16 eqs. Carbon . . . 96 51*43

10 ... Hydrogen . . 10 5*35

4 ... Nitrogen . . 56 30-34

3 ... Oxygen . . . ^ 12-88

186 100-00

* Liebig's Annalen, xxxvi. 93. t Ibid, xxxvi. 90.

% Mem. Chem. Soc, vol. i. pp. 215, 237. [Phil. Mag., xxiii. p. 426.]
§ Compies Rendus de I'Academie, tome xxiii. 8.

and of some of its Compounds, 117

We observe here a difference of 2 per cent, of carbon, which
M. Pay en has obtained over the results of the above-men-
tioned chemists.

In order to elucidate this discfTepancyy Dr. Hofmann in-
duced me to make some experiments under his direction,
partly m ith a quantity of beautiful caffein \vhich he gave me*
and partly with a specimen which I have prepared myself.

Caffein,

To ensure perfect purity of the substance it was crystal-
lized three times from dilute alcohol, washed and dried. Thus
purified, it formed very beautiful long white prisms, perfectly
transparent when dried in the air, but which became opake
if exposed to a higher temperature. The crystals dried in the
water-bath lost no weight when kept in an air-bath for four
hours at a temperature of 130° C.

The specimen which I had prepared myself was obtained
from Costa Rico coffee, by boiling the bruised fruit in water,
precipitating the decoctions by basic acetate of lead and treat-
ing the filtrate with hydrosulphuric acid ; after the whole of
the lead had been removed, I evaporated the liquid to dry-
ness in a water-bath, in order to get rid of acetic acid, and
dissolved the residue in a small quantity of boiling water :
upon cooling, the caffein crystallized out of a dark colour,
and very impure. To purify it, it was washed and recrystal-
lized three times from water, and finally from alcohol. It was
then perfectly white, and had exactly the same appearance as
the specimen which I obtained from Dr. Hofmann.

Analysis.

I. 0*3827 grm. of substance, dried at 100° C. and burnt
with chromate of lead, gave 0*6948 grm. of carbonic acid, and
0'1800 grm. of water.

II. 0*417 grm. of substance, burnt with chromate of lead
and chlorate of potash, gave 0*7552 grm. of carbonic acid,
and 0*1965 grm. of water.

III. 0*3934 grm. of substance of my own preparation gave
0*7123 grm. of carbonic acid and 0*1878 grm. of water, which
calculated in 100 parts gives —

I. II. III.

Carbon . . . 49*51 49*39 49*37

Hydrogen . . 5*22 5*23 5*30

* I owe this specimen, of great beauty, to the well-known kindness of
Mr. E. Merck of Darmstadt.— A. W. H.

118 Mr. Nicholson on the Composition of Caffein,

which agrees with Professor Liebig's formula, as is seen by
the following : —

Mean of
experiments.
49-42
5-28

Theory.

16 eqs. Carbon . .

. 96

49-48

10 ... Hydrogen .

10

5-15

4 ... Nitrogen .

. 56

28-86

4 ... Oxygen . .

. 32

16-51

194

100-00

Caffein and Bichloride of Platinum,

On precipitating a solution of cafFein in hydrochloric acid
with bichloride of platinum, as Dr. Stenhouse has shown, a
precipitate of an orange-yellow colour is obtained. If the two
solutions are mixed hot, the fluid on cooling deposits the com-
pound in beautiful granular crystalline tufts, which, when
thrown on a filter and washed with alcohol, are perfectly pure.
This double salt is only sparingly soluble in alcohol, aether,
and water. It does not alter when exposed to light, nor does
it lose in weight when kept at 100° C. for a considerable time.

The analyses of salts, all prepared at different periods and
dried at 100° C, gave the following results: —

I. 0'5382 grm. of substance, burnt with chromate of lead,
gave 0-4765 grm. of carbonic acid, and 0-1387 grm. of water.

II. 0*4881 grm. of substance gave 0*1196 grm. of platinum.

III. 0-4779 grm. of substance gave 0-1172 grm. of pla-
tinum.

IV. 0*6022 grm. of substance gave 0-1482 grm. of pla-
tinum.

V. 0*5781 grm. of substance gave 0-1425 grm. of pla-
tinum.

VI. 0-5246 grm. of substance gave 0*1293 grm. of pla-
tinum.

VII. 0-3847 grm. of substance made of caffein of my own
preparation, gave 0-0945 grm. of platinum.

Which give the following per-centages : —

I. II. III. IV. V. VI. VII.

Carbon . 23-80
Hydrogen 2-86
Platinum ... 24-51 24-52 24-60 24*64 24-64 24-56

leading exactly to the formula given by Dr. Stenhouse, viz.

Ci6H,oN4 04HCl,PtCl2,

as is seen when placed in comparison with the calculated
numbers.

and qf some of its Compounds. 119

Theory.

Mean of my
experiments.

Dr. Stenhouse's
mean.

16 eqs. Carbon .

96-0

23-97

23-80

24-22

11 ... Hydrogen .

11-0

2-74

2-86

2-89

4 ... Nitrogen

56-0

13-98

4 ... Oxygen . .

32-0

8-02

3 ... Chlorine . .

106-5

26-59

*

1 ... Platinum

98-9

24-70

24-58

24-49

1 ... CafFein and'

bichloride

-400-4

100-00

of platinum

The analysis of cafFein, as well as that of the platinum
compounds, agree so perfectly with the numbers of Professor
Liebig's formula, that there can be no doubt about its accu-
racy.

Assuming 1 equiv. of oxygen less in the equivalent of caf-
fein, as is proposed by M. Pay en, the platinum compound
should contain not less than 24-46 per cent, of carbon and
25-12 of platinum. Now three determinations by Dr. Sten-
house, and six which I have made, never gave more than 24-64
per cent., that is, 0-6 per cent, less of platinum.

Not satisfied, however, with these proofs, I have tried to
find some other compounds by which the atomic weight of
cafFein could be determined with equal accuracy.

In what follows a description of several new double salts
of cafFein will be given, the analyses of which correspond
equally well with the original formula of this substance.

Caffein and Nitrate of Silver.

This compound is obtained when a solution of nitrate of
silver is added in excess to an aqueous or alcoholic solution
of cafFein. If the solutions are concentrated it falls down in
white hemispherical nodules, which adhere firmly to the side
of the vessel.

When washed with water and crystallized from alcohol it is
absolutely pure. This compound is indistinctly crystalline,
of a perfectly white colour, and if dry undergoes no change
when exposed to light, but if moist acquires a purplish hue.
It is very soluble in hot M'ater and alcohol, sparingly soluble
in cold, and may be boiled in either solvent without vmder-
going decomposition. It loses no weight in the water-bath,
but at a higher temperature it is decomposed, cafFein sublimes,
and metallic silver is left.

Analysis. — When burnt with chromate of lead —

I. 0-4514 grm. of substance gave 0-4345 grm. of carbonic
acid, and 0'1162 grm. of water.

120 Mr. Nicholson on the Composition of Caffein,

II. 0*2500 grm. of substance gave '0744 grm. of silver.

III. 0*2716 grm. of substance gave '0810 grm. of silver,
which give the following per-centages : —

I. II. II.

Carbon . . . 26*45

Hydrogen . . 2*86

Silver 29*76 29*82,

and the formula — Cjs Hjo N4 O4 + AgO, NO5,
as may be seen by the following calculation : —

Theory.

Found.

16 eqs. Carbon . .

. 96

26*37

26*45

10 ... Hydrogen .

. 10

2*74

2*86

5 ... Nitrogen .

. 70

19*23

10 ... Oxygen

. 80

22*00

1 ... Silver . .

. 108

29*66

29*79

364

100*00

The only analogues to this singular compound which I
know are those of urea and nitrate of silver, analysed by Wer-
ther* : the formulae of which are— Co H4 Ng Og + Ag O, NO5,
andC,H4N,0, + 2(AgO,N05).

These compounds, however, in consequence of the peculiar
nature of xarea, are not very stable, being decomposed when
boiled with water into nitrate of ammonia and cyanate of
silver. There likewise exists a compound of nitrate of silver
and glycocoll, lately described by Horsfordfj having the for-
mula

C4H4N03+AgO,N05;
and, according to H. Rose, a compound of nitrate of silver
with ammonia, 3 equivs. of this gas being absorbed by 1
equiv. of the former salt.

Chloride of Mercury and Caffein.

This beautiful compound is obtained w^hen an aqueous or
alcoholic solution of caffein is added to a solution of chloride
of mercury ; the latter being kept in excess, the fluid remains
perfectly clear, but after the lapse of a few seconds solidifies
into a mass of very small crystals, W'hich when recrystallized
from water or alcohol and washed on a filter, are quite pure.

"When pure and crystallized from water it is very similar in
appearance to caffein, the crystals not being however quite so
large. It is very soluble in alcohol and water, hydrochloric,
nitric, and oxalic acids, and seems to form with the latter a
crystalline compound. It is nearly insoluble in aether. In
reference to its constitution, it is distinguished from the dou-

* Liebig's Annalen, Ivi. 262, t Ibid. Ix. 36.

and of some of its Compounds. 121

ble salt of platinurrij for in this instance the cafFein is in direct
combination with the chloride of mercury, and is exactly
analogous to the corresponding compounds of leucoline and
aniline investigated by Dr. Hofmann*. The mercurial com-
pounds of this kind are generally easily decomposed, but the
compound of chloride of mercury and cafFein is so stable, that
it may be boiled in water for a considerable time without un-
dergoing the slightest change in its properties. It may be
dried at 100° C. and loses no M'eight at that temperature.

I endeavoured to combine the determination of the carbon,
hydrogen and mercury of this substance in one combustion,
and have perfectly succeeded. The operation was conducted
as follows : — The substance was mixed with chromate of lead
and introduced into a combustion-tube of at least 26 inches in
length. About 6 inches of copper turnings are placed above
the mixture, leaving a space of 8 inches from the copper to
the anterior end of the tube. A receptacle for the mercury
is formed out of the tube itself by contracting it about an
inch from the copper turnings, and again so as to leave an
elongated bulb of an inch in length. At the close of the ope-
ration the tube is cut with a file at the posterior contraction.
In order to separate the water from the mercury, the chloride
of calcium tube (which has not been detached) is connected
with an aspirator and air admitted through chloride of cal-
cium, the bulb being kept at a temperature of 100'^ C.

I obtained in my analysis the following numbers : — 0'7833
grm. of substance gave 0*5832 grm. of carbonic acid, 0*1639
grm. of water, and 0*3365 grm. of mercury, corresponding to
the following per-centage, which I place in comparison with
the theoretical numbers : —

Theoiy. Expt.

16 eqs. Carbon 96 20*68 20*30

10 ... Hydrogen 10 2*15 2*32

4 ... Nitrogen 56 12*11

4 ... Oxygen 32 6*89

2 ... Chlorine 70 15*08

2 ... Mercury 200 43*09 42*91

464 100*00

Caffein and Terchloride of Gold.

This compound is formed when a solution of terchloride of
gold is added in excess to cafFein dissolved in dilute hydro-
chloric acid. If concentrated solutions are employed, the
whole immediately solidifies into a mass of a most splendid
lemon-yellow colour; this is to be washed with cold water

♦ Liebig's Annalen, xlvii. 37.

132 Mr. Nicholson on the Composition of Caffein.

and crystallized from alcohol, and finally dried in the water-
bath.

The crystals from on alcoholic solution are in the form of
long needles, of an orange-yellow colour and a very bitter
metallic taste; they are soluble in alcohol and water. When
boiled in water for a short time, the 'salt is decomposed, a
yellow flocculent matter precipitating, which is insoluble in
alcohol, ether and water, but soluble in hydrochloric acid.
If an aqueous solution is kept on the sand-bath for some
hours at a temperature of about 68° C. it is also decomposed,
and metallic gold separates in shining scales.

It is not altered when exposed to light, and when dry may
be heated to 100° C. without undergoing decomposition.

Analysis. — When burnt with chromate of lead —

I. 0*8530 grm. of substance gave 0*5525 grm. of carbonic
acid and 0*1622 grm. of water.

II. 0*3224 grm. of substance gave 0*1197 grm. of metallic
gold.

III. 0*301 9 grm. of substance gave 0*1115 grm. of metallic
gold, which give the following per-centages : —

I. II. III.

Carbon . . 17*72
Hydrogen .2*11
Gold . . . 37*12 36*93

corresponding to the formula Cjg Hjo N4O4 HCl Au CI3, as
may be seen by the following table, where the calculated
and quantities found are placed in comparison : —

Theory. Found.

16 eqs. Carbon . 96*00 17*98 17-72

11 ... Hydrogen 11*00 2*06 2*11

4 ... Nitrogen . 56*00 10*50

4 ... Oxygen . 32*00 6*01

4 ... Chlorine . 142*00 26*60

1 ... Gold . . 196*66 36*85 37*02

533*66 100*00

The caffein compounds which I have analysed are there-
fore—

Caffein Cjg Hjo N4 O4.

Platinum compound Cjg Hiq N4 O4, HCl Pt Clg.
Silver compound . Cjg Hjq N4 O4, AgO, NO5.
Mercury compound Cjg Hjo N4 O4, 2(Hg CI).
Gold compound . . Cjg Hjo N4 O4 HCl, Au CI3.

There exist several other double compounds of caffein,
which I have however not subjected to analysis.

Prof. Young oti the Extension o/'Euler's Theorem. 123

On mixing a hot alcoholic solution of caffein with an alco-
holic solution of cyanide of mercury, beautiful needles of a
double salt are deposited upon cooling, which correspond
most likely to the mercury salt I have just described. A so-
lution of caffein in hydrochloric acid gives a beautiful brown
precipitate with chloride of palladium ; and the filtered solu-
tion deposits another compound in the form of yellow scales,
very similar in appearance to iodide of lead.

CaiFein gives no precipitate with solutions of sulphate of
copper, chloride of tin, acetate of lead, and nitrate of suboxide
of mercury. When boiled with sesquichloride of iron, a red-
dish-brown precipitate subsides upon cooling, which is per-
fectly soluble in water, and is most likely a double compound
of caffein and sesquichloride of iron.

XXIII. Note in reference to the exteiision o/Euler's Theorem,
By J. R. Young, Professor of Mathematics in Belfast College.

To Richard Taylor^ Esq.
Dear Sir,

IN the Philosophical Magazine for June last a communica-
tion of mine was published respecting an extension of a
certain theorem of Euler concerning the products of the sums
of squares. At the time that notice was written, I was under
the impression that the theorem admitted of an extent of ge-
neralization which a further investigation of the matter proves
to me has not place. I am now prepared to show that the
proposition does not hold beyond the case for eight squares,
the formulae for which I have already printed in the Proceed-
ings of the Royal Irish Academy ; in the Transactions of
which body it is probable that the entire investigation of the
theorem for eight squares, and the proof that it does not apply
beyond that number, will hereafter appear.

It may perhaps be interesting to algebraists to find the real
limits to this theorem demonstrably established ; and thus to
know — in any attempts that may hereafter be made to extend
Sir W. R. Hamilton's remarkable and very fertile theory of
quaternions — beyond what boundaries such attempts must
prove fruitless.

I remain, dear Sir,

Very faithfully yours,

Belfast, July 16, 1847. J. R. YouNG.

[ 124 ]

XXIV. On the Precipitate produced in Spring and River Waters
by Acetate of Lead, By A. Connell, Esq., Professor of
Chemistry in the University of St. Andrews*.

^V^YLYj white precipitate which it is well known is usually
JL produced in spring and river waters by acetate of lead,
has been commonly attributed to the presence of sulphates,
chlorides and carbonates. The comparatively trifling action
of silver salts, however, shows that it is very rarely, unless in
the case of what are called mineral waters, due to chlorides ;
and the ready solubility of the precipitate in acetic acid in
whole or in great part, proves that it is not due to sulphates
or phosphates, except in so far as it may be insoluble in acetic
acid. Carbonates therefore remain as the probable cause ;
and this is established by the circumstance, that although
effervescence cannot be noticed on the immediate addition of
acetic acid, effervescence will be observed if the precipitate is
allowed to subside, and the greater part of the solution de-
canted, and an acid then added. 1 have found on investiga-
tion that carbonate of lime is the usual source of the reaction.
The remarkable fact however on this view is, that the reaction
is scarcely diminished by boiling and filtering the water; and
indeed in some instances does not take place unless these steps
are had recourse to, and acetic acid still dissolves the whole
or great part. If the waters referred to are boiled and filtered
and then largely concentrated by evaporation, they usually
deposit carbonate of lime, and do not indicate any such alka-
line reaction as shows an alkaline carbonate. The carbonate
of lime causing the reaction is therefore evidently held dis-
solved in the water independently of the presence of free
carbonic acid ; and I do not think that chemists, generally
speaking, are aware that common water may still retain enough
of carbonate of lime to give, with acetate of lead, a consider-
able precipitate of carbonate of lead, although they may have
been boiled and filtered. If in any such case the precipitate
should be found to dissolve in acetic acid truly mthout effer-
vescence, the probable cause would be the presence of a suffi-
cient quantity of some organic matter, such as crenic or apo-
crenic acid, which precipitates lead salts ; for it is not the least
likely that fluorine, which has been found in some spring
waters, should ever be present in sufficient quantity to affect
lead salts, and fluoride of lead would very likely not be so-
luble in acetic acid.

The question then arises, whence proceeds this carbonate

* Communicated by the Author.

On the Precipitate produced in Water by Acetate of Lead. 125

of lime. To know whether it arises from the water redis-
solving carbonate of lime, which had been held dissolved by
carbonic acid and then precipitated by boiling, I transmitted
a current of carbonic acid through lime water tilLit completely
redissolved the precipitate which had at first formed. I then
boiled the solution for a short time, as in experimenting
with the spring waters, and filtered the liquid; but although
it was slightly precipitated by acetate of lead, the effect was
very much less than that on common water ; showing that we
cannot account for the effect on common water by supposing
that all the carbonic acid had not been driven off by the ebul-
lition. Again, when distilled water was left in contact with
marble in impalpable powder for several days, both acetate of
lead and oxalate of ammonia showed less lime than in the
common waters, although rather more than in the lime-water
experiment. I incline therefore to think that the carbonate
of lime owes its origin to double decomposition between an
alkaline carbonate and a lime salt, such as a chloride. If to
a few ounces of distilled water a drop or two of muriate of
lime and a drop or two of carbonate of soda be added, the
liquid remains quite transparent ; and the reaction of common
water with acetate of lead and acetic acid may be exactly imi-
tated with this liquid. And in all the common waters yielding
the reaction, I could detect alkalies in union with acids.

The common water of the town of St. Andrews, I found,
after being boiled and filtered, to yield by evaporation -^j^-^-^
of carbonate of lime; and other well and river waters may
contain still more. Fresenius has stated that water is capable
of holding in solution yy^oT ^^ carbonate of lime, after being
saturated with that salt by long-continued boiling, and left
in contact for four weeks with the deposit formed on cooling.
Nature of course does not take such pains to charge spring
waters with lime ; and I think the method I have suggested
affords a much more simple and probable means of effecting
this end.

The St. Andrews' water also contains a trace of carbonate
of magnesia after being boiled and filtered ; and it is probable
that this substance may sometimes be in part the cause of the
reaction referred to, but to a much less extent*.

* I have given fuller details on this subject in a paper inserted in the
Transactions of the Royal Society of Edinburgh for the present year.

[ 126 ]

XXV. On the Action of a mixt^ire of Red Prussiate of Potash
and Caustic Alkali upon Colouring Matters. By John
Mercer, Esq.^

A BOUT ten years since I discovered and used extensively
-^^ in calico-printing the oxidizing properties of a mixture
of red prussiate of potash and caustic alkali. For many
years I have been in the habit of communicating to my friends
several applications of this interesting reaction, among whom
I may mention Mr. Crura of Glasgow and Dr. Lyon Playfair.
Since then Boudault t has directed attention to the oxidizing
power of the same mixture, as far as relates to metallic oxides,
but has not shown any important practical application of the
knowledge thus acquired.

There are but few processes known in the arts for bleach-
ing indigo, the principal of these being that in which chromic
acid liberated from the bichromate of potash by means of an
acid is used. In certain cases this process is attended with
various disadvantages, and the cloth requires to be subjected to
a clearing process to remove the oxide of chromium. The
topical application of a mixture of red prussiate of potash and
an alkali at once effects the same pm*pose, and in a most com-
plete manner, leaving a brilliant white on the spot where the
colour is discharged without rendering any injury to the
fabric. The manner of applying this discharge may be ar-
ranged to suit the conditions of the calico-printer. As a class
experiment for a lecture-table it is convenient to impregnate
the indigo-blue calico with a solution of prussiate of potash,
and then dip it into a weak solution of alkali.

This action is a beautiful illustration of those double affi-
nities which we frequently find at play in combinations or
decompositions. Thus, though neither chlorine nor charcoal
can decompose alumina per se, the same gas passed over a
mixture of alumina and charcoal combines with the metallic
radical ; the charcoal in this case having aided the combina-
tion by withdrawing the oxygen. It is the same kind of
action in the case under consideration. Red prussiate of
potash, Feg Cyg 3K, differs from the yellow prussiate, Fcg
Cyg 4K, by containing one atom less potassium. When pot-
ash is presented to the former, this deficient atom of potas-
sium is supplied, but the affinity is not strong enough to
liberate the oxygen. When however a second body having
an attraction for oxygen, such as litharge or indigo, is pre-
sented to the potash and red prussiate, this second affinity

* Communicated by the Chemical Society; having been read Feb. J,
1847.

t Journal de Pharmacie. [Phil. Mag., vol. xxvii. p.307.]

Dr. W. Gregory on the Preparation ofHippuric Acid. 127

acting in a different direction withdraws the oxygen and
allows the potassium to unite with the compound radical fer-
rocyanogen ; thus Fe^ Cyg 3K + KO + PbO = Fcg Cyg 4K
+ Pb02, the decomposition being of the same kind when an
organic matter is substituted for the oxide capable of further
oxidation. Soda and ammonia may be substituted for potash
in the above decomposition, producing the oxidation or dis-
charging the indigo. This is curious in the case of ammonia,
for it cannot be explained by any other than by the ammo-
nium theory, and shows the complete analogy between the
oxide of ammonium and the oxide of the simple metallic ra-
dicals, potassium and sodium. It is interesting also to ob-
serve that the last member in the formula Fe^ Cyg 4R, may
be substituted by any alkaline base. Thus, that it may either
be Fe^ Cyg 3K K, or Fe^. Cyg 3K Na,or Fe^ Cyg 3K NH4. This
circumstance points to important theoretic^ considerations
in the atomic constitution of the prussiates, which would be
foreign to the present paper, the principal object of which is
to furnish a means of discharging indigo, and thus supply
a process much wanted in the art of calico-printing, and which
I have followed for many years with success.

XXVI. On the Preparation of Hippuric Acid.
By William Gregory, M.D."^

SINCE the discovery of hippuric acid by Liebig, that body
has at all times attracted much attention. Its composition
and the products of its decomposition, among which were ben-
zoic acid and benzamide, rendered it interesting, and various
ingenious views were entertained of its constitution. Its
detection in human urine by Liebig gave it additional im-
portance.

The beautiful discovery of Dessaignes, that hippuric acid,
when heated with strong acids, is resolved into benzoic acid
and glycocoU, has greatly increased the interest already at-
tached to hippuric acid, which now affords the best means of
obtaining glycocoll, and has enabled Horsford, in his elabo-
rate researches on that substance, to fix its formula in a very
satisfactory manner.

If to hydrated hippuric acid . Cjg N Hg Og,
we add 1 equiv. water ... HO,

and from the sum Cjg N H,q O7,

subtract 1 equiv. glycocoll . . C4 N H4 O3,

there remain C14 Hg O4,

which is hydrated benzoic acid.

* Communicated by the Chemical Society; having been read March 15,
1847.

128 Dr. W. Gregory on the Preparation of Hippuric Acid.

There cannot, I think, be any longer a doubt that C4 N H4 O3
is the true formula of glycocoll, and Hor^ford has, in esta-
bUshing this point, at the same time confirmed and explained
in the most satisfactory manner the observation of Dessaignes.

The researches of Horsford, however, have also demon-
strated that glycocoll is in itself one of the most interesting
compounds known to chemists, and it is evident that the fur-
ther study of this most singular body will lead to very va-
luable results.

I have already stated that glycocoll is best obtained from
hippuric acid, but as soon as I began to prepare for this pur-
pose a considerable quantity of hippuric acid, I found, as all
who have done so must have found, that the operation as pre-
scribed in books is not only tedious and troublesome, but un-
certain.

The usual process consists in evaporating the urine of the
horse or cow at a moderate temperature to about one-eighth
of its bulk, and adding hydrochloric acid, when on standing
a few hours, crystals of impure hippuric acid are deposited.
But it is well-known that if the temperature should rise too
high, although still to a point short of boiling, the hippuric
acid will partially or totally disappear, and benzoic acid will
be found in its place. Now when we bear in mind that the
urine contains but little hippuric acid, it is evident that to
obtain this acid in quantity we must operate with a very large
bulk of urine, and those who have done so well know how
tedious the evaporation is, since if we attempt to hasten it
by raising the temperature, we run the risk of losing the
whole ; and this indeed frequently happens.

The impure, highly-coloured acid first obtained has been
purified by different chemists in a great variety of ways.
Some have used chloride of lime ; but this method is not
easily managed, and often converts the whole into benzoic acid.

The last and by far the best method of purification is that
of Schwarz, who boils the impure acid with an excess of milk
of lime, and strains the alkaline liquid from the undissolved
lime. It passes rapidly and clear through calico, and the lime
retains the colouring matter, so that the addition of acid to
the filtered liquid causes the deposition of crystals of hippuric
acid nearly white. Schwarz recommends the addition of
chloride of calcium to the filtered or unfiltered liquid, and the
precipitation of the lime as carbonate by carbonate of potash
or soda, when the precipitated carbonate of lime carries with
it the last traces of colouring matter. I have not found this
necessary, as a repetition of the process with the milk of lime
never fails to yield colourless crystals.

Dr. W. Gregory on the Preparation of Hippuric Acid. 129

As it was clear that the hippuric acid was not in the slight-
est degree decomposed by boiling with excess of lime, al-
though so easily metamorphosed by acids, I thought that by
applying the same principle to the urine directly, I might be
enabled to boil it down, and thus shorten the process, and at
the same time prevent the decomposition of the hippuric acid,
since it would appear that hippurate of lime is not affected
by boiling, nor by excess of lime.

Accordingly, I took some urine of the horse, mixed it with
excess of milk of lime and boiled for a few minutes. I then
strained the solution, which was very materially decolorized,
and boiled the clear liquid as rapidly as possible down to the
requisite bulk. On adding hydrochloric acid I obtained a
copious deposit of crystals, which when pressed had a slight
red colour. I then treated them by Schwarz's method and
obtained an abundant crop of almost colourless crystals,
which consisted entirely of the needles of hippuric acid, with-
out a visible trace of benzoic acid, the crystallization of which
is easily recognized. A second treatment with milk of lime,
which was hardly needed, and probably would have been
quite unnecessary had a greater excess of lime been used in
the previous one, yielded snow-white crystals of the utmost
beauty and purity.

The improvement which I have thus introduced in the
preparation of hippuric acid may seem trifling, and is indeed
only the application of Schwarz's method to the urine, in-
stead of to the crude acid ; but any one who tries to prepare
some ounces, not to say pounds, of hippuric acid, will soon
find that the difference is practically important. By my me-
thod it is possible to extract in one day the hippuric acid
from as much urine as would require a week to operate upon
on the usual plan, so that the quantity of hippuric acid which
we can thus obtain is only limited, as it were, by the quan-
tity of urine to be procured. The tedious evaporation at low
temperatures is got rid of, and we are sure of obtaining the
whole hippuric acid originally present ; whereas, on the for-
mer plan, however carefully the evaporation is conducted,
and it requires constant superintendence, it almost ahvays
happens that some of the hippuric acid is decomposed ; while
a very slight accidental rise of temperature may destroy the
whole of it, as I have often seen.

On the whole, I am satisfied that all who wish to study
hippuric acid and glycocoll will find on trial that what was
formerly a disagreeable and troublesome operation is now a
very easy and short one ; and that they may now easily ob-
tain these remarkable compounds in any desired quantity.
Phil. Mas. S, 3. Vol. 3 1 . No. 206. Aus. 1847. K

[ 130 ]
XXVII. Proceedings of Learned Societies.

CAMBRIDGE PHILOSOPHICAL SOCIETY.

[Continued from vol. xxx. p. 367.]

Nov. 9.#^N the Structure of the Syllogism, and on the application
1846. ^-^ of the Theory of Probabilities to Questions of Argument
and Authority*. By Professor De Morgan.

The object of this paj^er is twofold : first, to establish two distinct
theories of the syllogism, both differing materially from that of Ari-
stotle, and each furnishing a general canon for the detection of all its
legitimate forms of inference ; secondly, to investigate the mode in
which the distinctive character of the two great sources of convic-
tion, argument and authority, affects the application of the notion of
probability to questions not admitting of absolute demonstration.

The two theories of the syllogism arise out of simple notions con-
nected with the /onns of propositions and their quantities. The dif-
ference between a positive and negative assertion is not essential,
but depends on the manner in which objects of thought are described
by language. If Y and y be names so connected that each contains
everything which is not in the other, and the two have nothing in
common (a relation which is described by calling them contrary

* Upon this paper a controversy has arisen, which, up to the present
time, may be summed up as follows : — ^piilSO. Mr. De Morgan published
a statement in answer to an assertion ot SirW. Hamilton of Edinburgh, to
the effect that the second, or quantitative, system of syllogism, was a wilful
plagiarism from certain letters which Sir W. Hamilton had written to Mr.
De Morgan. Maj/ 22. Sir W. Hamilton replied at length in another
pamphlet, retracting the assertion o^ wilful plagiarism, but maintaining that
the system was taken, unconsciously, from those letters. This was followed
by a letter from Mr. De Morgan in the Athenaeum of May 29, and another
from Sir W. Hamilton in the same publication for June 5. The point at
issue now seems to be as follows : — Mr. De Morgan challenges Sir W. Ha-
milton to show anything in his second system which was not substantially
contained in a digressive section of the description of his first system, ad-
mitted to have been sent to Cambridge before any communication had
taken place. SirW. Hamilton, in reply, contends that the digression above-
mentioned contains nothing to the purpose. Mr. De Morgan defers further
reply until he publishes a work which he states himself to be preparing on
logic.

In the Athenaeum of June 19, appeared a letter from Mr. James Broun,
asserting certain mistakes on the part both of Sir W. Hamilton and Mr.
De Morgan, and giving certain extensions to the quantitative forms of the
latter. Again, June 26, appeared in the same publication a letter from
Mr. De Morgan, dated June 19, stating that he also had arrived at Mr.
Broun's forms, giving reasons for their rejection in favour of certain simpler
forms, giving the heads of an extended system of quantitative syllogism,
and asserting that he had materially extended both his systems. So the
matter stands. The subject of the structure of the syllogism seems to be
likely to excite some attention ; and, without pronouncing any opinion on
the personal claims or conflicts of the several parties, we recommend the
attention of our readers to this rather neglected branch of pure science. —
Ed. Phil. Mag.

Cambridge Philosophical Society. 131

names), the propositions ' Every X is Y ' and ' no X is y ' are sim-
ply identical. In the same manner, the particular and universal
proposition are only accidentally distinct. If in ' some Xs are Ys '
the Xs there specified had had a name belonging to them only, say
Z, then the preceding proposition would have been identical in mean-
ing with * every Z is Y,'

From the above it is made to follow, that every legitimate syllo-
gism can be reduced to one of universal affirmative premises, either
by introduction of contrary terms, or invention of subgeneric names.

In considering the nature of the simple proposition, Mr. De Mor-
gan uses a notation proposed by himself. Thus —
Every X is Y is denoted by X)Y A
NoXisY . . . . X.Y E

Some Xs are Ys . . . XY I

Some Xs are not Ys . . X:Y O
and names which are contraries are denoted by large and small let-
ters. Aristotle having excluded the contrary of a name from formal
logic, and having thereby reduced the forms of proposition to four,
these forms (universal affirmative, universal negative, particular affir-
mative, particular negative) the writers on logic in the middle ages
represented by the letters A, E, I, O. Thus X)Y and Y)X are
equally represented by A. When contraries are expressly intro-
duced, all the forms of assertion or denial which can obtain between
two-terras and their contraries, are eight in number ; and the most
convenient mode of representing them is as follows : — Let the letters
A, E, I, O have the above meaning, but only when .the order of sub-
ject and predicate is XY. Then let a, e, i, o stand for the same
propositions, after x and y, the contraries, are written for X and Y.
The complete system then is —

A=X)Y a=x)y~Y)X

0=X:Y o=a^:y=Y:X

E:=X.Y e-=-x.y

I=XY i-=-xy

and every form in which subject and predicate are in any manner
chosen out of the four X, Y, x, y, so that one shall be either X or x,
and the other either Y or y, is reducible to one or other of the pre-
ceding.

The propositions e and i, which are thus newly introduced, are
only expressible as follows, with reference to X and Y.

(t.) There are things which are neither X nor Y.

(e.) There is nothing but is either X. or Y or both.

The connexion of these eight forms is fully considered, and the
various syllogisms to which they lead. Rejecting every form of syl-
logism in which as strong a conclusion can be deduced from a weaker
premise ; rejecting, for instance,

Y)X+Y)Z=XZ

because XZ equally follows from Y)X4-YZ, in which YZ is weaker
than Y)Z — all the forms of inference are reduced to three sets.
1 . A set of two, called single because the interchange of the terms

K2

132 Cambridge Philosophical Society.

of the conclusion does not alter the syllogism. Neither of these
forms are in the Aristotelian list. One of them is

X)Y+Z)Y-xz;

or if every X be aY, and also every Z, then there are things which are
neither X nor Z ; namely, all which are not Ys.

2. A set of six, in which the interchange produces really different
syllogisms of the same form, and in which both premises and con-
clusion can be expressed in terms of three names, without the con-
trary of either. This set includes the whole Aristotelian list, except
those in which a weaker premise will give as strong a conclusion, or
the one in which the same premises will give a stronger conclusion.

3. A set of six resembling the last in everything but this, that no
one of them is expressible without the new forms e and i ; that is,
requiring three names and the contraries of one or more of them.

Those of the third set are not reducible to Aristotelian syllogisms,
as long as the eight standard forms of assertion are adhered to.

The second theory of the syllogism has its principles laid down in
the memoir before us ; but those principles are only applied to the
evolution of the cases which are not admitted into the Aristotelian
system. The formal statement of the manner in which the ordinary
cases of syllogism are connected with those peculiar to this second
system is contained in an Addition.

In providing that premises shall certainly furnish a conclusion,
the common system requires that one at least of the premises shall
speak universally of the middle term ; that is, shall make its asser-
tion or denial of every object of thought which is named by the middle
term. Mr. De Morgan points out that this is not necessary : m
being the fraction of all the cases of the middle term mentioned in one
premise, and n in the other, all that is necessary is that m + n should
be greater than unity. In such case, the real middle term, being
the collection of all the cases by comparison of which with other
things inference arises, is the fraction m + w — 1 of all the possible
cases of the middle term. Thus, from the premises ' most Ys are
Xs ' and ' most Ys are Zs,' it can be inferred that some Xs are Zs,
since m and n are both greater than one-half. The assignment of
definite quantity to the middle term in both premises, gives a canon
of inference, of which the Aristotelian rule is only a particular case.

In the addition above alluded to, this same canon, namely ' that
more Ys in number than there exist separate Ys shall be spoken of
in both premises together,' is made to take the following form : — If
in an affii'raation or negation, in ' As are Bs ' and ' As are not Bs,'
definite numerical quantity be given to both subject and predicate, if
it be stated how many As are spoken of and how many Bs — the
number of effective cases of the middle term is seen to be the nnici-
her oi subjects in an affirmative proposition, whether the middle term
be subject or predicate. Hence, defining the effective number of a
premise to be the number of subjects if the proposition be affirmative,
and the number of cases of the middle term if it be negative, all
that is necessary for inference (over and above the usual condition

Cambridge Philosophical Society. 133

that both premises must not be negative) is that the sum of the
effective numbers of the tvro premises shall exceed tlie number of
existing cases of the middle term ; and the excess (being the fraction
denoted by m + « — 1 in the Memoir) gives the number of cases in
which inference can be made.

To attempt to combine these two systep^ oiform and of quantity
is rendered useless by language not possessing the forms of mixed
assertion and denial, which the syllogisms deduced from the combi-
nation would require. As far as the combination can, in Mr. De
Morgan's opinion, be made, nothing is required but a distinct con-
ception of, and nomenclature for, the usual modes of expressing a
logical form, and implying one or the other of the alternations which
the mere expression leaves unsettled. Mr. De Morgan proposes the
following language.

Two names are identical when each contains all that the other
contains : but when all the first (and more) is contained in the second,
then the first is called a subidentical of the second, and the second
a superidentical of the first. Two names are contrary when every-
thing (or everything intended to be spoken of) is in one or the other
and nothing in both. But when the two names have nothing in
common, and do not between them contain everything, they are
called subcontraries of one another. And again, if everything be in
one or the other, and some things in both, they are called supercon-
trari.es of one another. Lastly, if the two names have each some-
thing in common and something not in common, and moreover do
not between them contain everything, each is called a complete par-
ticular of the other. A table is then given, which contains every
form of complex syllogism.

If X and Z be the terms of the conclusion, and both be described
in terms of Y, the middle term : it can be seen from this table what
can be affirmed and what denied, of X with respect to Z. For in-
stance, if X be supercontrary of Y, and Z subcontrary, then X must
be a superidentical of Z : but if X and Z be both subidenticals of Y,
nothing can be affirmed ; only it may be denied that X is either
contrary or suj^ercoiitrary of Z.

The remaining part of this paper relates to the application of the
theory of probabilities above-mentioned. Mr. De Morgan asserts
that no conclusion of a definite amount of probability can be formed
from argument alone ; but that all the results of argument must be
modified by the testimony to the conclusion which exists in the mind,
whether derived from the authority of others, or from the previous
state of the mind itself. The foundation of this assertion is the
circumstance that the insufficiency of the argument is no index of
the falsehood of the conclusion. Various cases are examined; but
it must here be sufficient to cite one or two results.

If n be the probability which the mind attaches to a certain con-
clusion, a the probability that a certain argument is valid, and b the
probability that a certain argument for the contradiction is valid :
then the probability of the truth of the conclusion is

(1-%
(l-%-f-(l-a)(l-j«)-

134 Cambridge Philosophical Society.

If J=0, or if there be no argument against, and if the mind be

unbiassed, or if u= — , this becomes
t* 2'

or « + ^^ —.

2— fl 2— a

For this writers on lo^c generally substitute a, confounding the
absolute truth of the conclusion with the validity of the argument,
and neglecting the possible case of the argument being invalid, and
yet the conclusion true.

Nov. 23. — On a New Notation for expressing various Conditions
and Equations in Geometry, Mechanics and Astronomy. By the
Rev. M. O'Brien.

If A, P, P' be any three points in space, whether in the same
straight line or not, and if the lines AP and AP' be represented in
magnitude and direction by the symbols u and u' , then,, according to
principles now well-known and universally admitted, the line PP' is
represented in magnitude and direction by the symbol u' —u. Now
if AP and AP' be equal in magnitude, and make an indefinitely small
angle with each other, PP' is an indefinitely small line at right angles
to AP, and u' —u becomes du. Hence it follows, that, if u be the
symbol of a line of invariable magnitude, du is the symbol of an in-
definitely small line at right angles to it ; and therefore, if X be any
arbitrary coefficient, \du is the general expression for a right line
perpendicular to u.

The sign \d therefore indicates perpendicularity, when put before
the symbol of a line of invariable length. The object of the author
is to develope this idea, and to show that it not only leads to a
simple method of expressing perpendicularity, but also furnishes a
notation of considerable use in expressing various conditions and
equations in geometry, mechanics, astronomy, and other sciences
involving the consideration of direction and magnitude.

The author first reduces the sign \d to a more convenient form,
which not only secures the condition that u is invariable in length,
but also defines the magnitude and direction of the perpendicular
which Kdu denotes. This he does in the following manner. He
assumes

u=^xa,-\-y^ + zy,

(where a- ^ y reipresent three lines, each a unit in length, drawn at
right angles to each other, and if y z are any arbitrary numerical
coefficients,) and supposes that the differentiation denoted by d affects
a j3 y, but not x y z. This secures the condition that u is invariable
in length, and leads to the following expression for Kdu, viz,

Xdu={zy' —z'y)(x, + (oaz' —x'z)^ + (yx' —y'a;)y,
x' y' z' being arbitrary coefficients.

Assuming u'=x'a,+y'fi + z'y, it appears from this expression for
Xdu, that du=0 when u=u', and therefore that c? denotes a differen-
tial taken on the supposition that u' is constant.

On this account the author substitutes the symbol Du' in place of
Ac? ; he then shows that the operation D^/ is distributive with respect
to u' (i. e. that Dm'+m"s=D„/4-Dm")> and to indicate this he elevates

Cambridge Philosophical Society, 186

the subacript index m', and writes Dm'.m instead of Dm/m, Thus he
obtains the expression

Du' .u=i{zy'~z'y)a,-\-ixz' —x'z)^-\-iyx' —y'x)y.

From this it follows that Dm'.m is a line perpendicular both to u'
and u, and that the numerical magnitude of Dm'.m is rr' sin ^, where
r and r' are the numerical magnitudes of u and m', and fl the angle
made by u and u' .

Having investigated the principal properties of the operation Dm'.,
the author, by a similar method, obtains another notation, Au'.u,
which represents the expression xx' -{-yi/'+zz', or rr' cos Q. He then
gives some instances of the application of these two notations to
mechanics, which may be briefly stated as follows : —

1st. If U, U', U", &c. be the symbols* of any forces acting upon
a rigid body, and m, m', m", &c. the symbolsf of their respective points
of application, then the six equations of equilibrium are included in
the two equations

2U=0 and I:Dm.U=0.

2nd. That these two equations are the necessary and sufficient
conditions of equilibrium, may be proved very simply from first prin-
ciples by the use of the notation Dm.

3rd. The theory of couples is included in the equation 2Dm.U=0.
In fact the symbol Dm.U expresses, in magnitude and direction, the
axis of the couple by which the force U is transferred from its point
of application U to the origin.

4th. Supposing that the forces U, U', TJ", &c. do not balance each
other, and putting 2U=V, 2;Dm.U=W, we may show immediately,
by the use of the notation Am, that the condition of there being a
single resultant is

AV.W = 0;

and when there is not a single resultant, the axis of the couple of
minimum moment is

AV.W y

AV.V. ■ *

5th. The three equations of motion of a rigid body about its
centre of gravity are included in the equation

-('2Du. — Sm\='LDu.VSm; (1.)

dt\ dt J

u being the symbol of the position of any particle hm of the body,
and U the symbol of the accelerating force acting on Jm.

6th. If u) be assumed to represent the expression cWia + cwojS + Wjy,
where Wy, w.^, Wj are the angular velocities of the planes of yz, zx,
xy about the axes of x, y, z respectively, then the symbol of the

* By the symbol of a force is meant the expression X«+Y/3+Zy, where
X Y Z are the three components of the force.

t By the symbol of a point is meant the expression jf«-t-^/3+2y> where
X y z are the coordinates of the point.

1 36 Cambridge Philosophical Society.

velocity of ^m is Dcu.w ; from which follow immediately the three
well-known equations,

dx dy dz _

The symbol w represents in direction the axis of instantaneous
rotation, and in magnitude the angular velocity about that axis.
7th. The equation (1.) maybe reduced to the form

-^ { A w la + BwzjS + C Way } = SDm .U Jot,

which includes Euler's three equations of motion about a fixed point.
8th. If the forces U, U', U", &c. arise from the attraction of a
distant body, the symbol of whose position is u' , this equation may
be further reduced to the form

^ / Aw.a+BwajS-i- Cwgy") = ^' DM'.(Aa?'a+By/3 + CzV)-

9th. In the case of the earth attracted by the sun or moon, this
equation becomes

y being the polar axis, and A= .

10th. The mean daily motion of y is given by the equation

5 = ^X(A«'.7)(D«'.y);
dt nr^

which equation gives immediately all the well-known expressions
for solar and lunar precession and nutation, for -5- is the symbol of

the velocity of the north pole, representing that velocity both in
magnitude and direction.

Supplement to a Memoir on some cases of Fluid Motion. By G.
G. Stokes, M.A., Fellow of Pembroke College, Cambridge.

In a former paper the author had given the mathematical calcula-
tion of an instance of fluid motion, which seemed to offer an accurate
means of comparing theory and observation in a class of motions, in
which, so far as the author is aware, they had not been hitherto com-
pared. The instance referred to is that in which a vessel or box of
the form of a rectangular parallelepiped is filled with fluid, closed,
and made to perform small oscillations. It appears from theory that
the effect of the inertia of the fluid is the same as that of a solid
having the same mass, centre of gravity and principal axes, as the
solidified fluid, but different principal moments of inertia. In this
supplement the author gave a series for the calculation of the prin-
cipal moments, which is more rapidly convergent than one which he
had previously given. It is remarkable that these series, though
numerically equal, appear under very diflferent forms, the wthterm of

Cambridge Philosophical Society. 137

the latter containing exponentials of the forms e"** and e " , while
the «th term of the former contains exponentials of the second form
only. In conclusion, the author referred to some experiments which
he had performed with a box, such as that described, filled with
water, employing the method of bifilar oscillations. The moment of
inertia of the fluid about an axis passing through its centre of gra-
vity (?'. e. the moment of inertia of the imaginary solid which may
be substituted for the fluid), was a little greater as determined by
experiment than as determined by theory, as might have been ex-
pected, since the friction of the fluid was not considered in the cal-
culation. The diff"erence between theory and experiment varied in
difi^erent cases from the Jgth to the -J^st part of the whole quantity.

Dec. 7. — On the Principle of Continuity in reference to certain
results of Analysis. By Professor Young of Belfast College.

The object of this paper is to inquire into the influence of the law
of continuity, as it affects the extreme or ultimate values of variable
functions, more especially those involving infinite series and definite
integrals.

The author considers that this influence has hitherto been impro-
perly overlooked ; and that to this circumstance is to be attributed
the errors and perplexities with which the different theories of those
functions are found to be embarrassed. He shows that every parti-
cular case of a general analytical form — even the ultimate or limiting
case — must come under the control of the law implied in that form ;
this law being equally efficient throughout the entire range of indi-
vidual values. Except in the limiting cases, the law in question is
palpably impressed on the several particular forms ; but at the limits
it has been suffered to escape recognition, because indications of its
presence have not been actually preserved in the notation.

It is in this way that the series 1 — 1 + 1 — 1-|- &c. has been con-
founded with the limits of the series \—x-\-x°—x^+ &c. ; these
limits being arrived at by the continuous variation of x from some
inferior value up to x=\, and from some superior value down to
x=si\. It is shown however that the series 1 — 1 + &c. has no equi-
valent among the individual cases of 1— a? + a?^— &c., with which
latter, indeed, it has no connexion whatever.

By properly distinguishing between the real limits, and what is
generally confounded with them, the author arrives at several con-
clusions respecting the limiting values of infinite series directly op-
posed to those of Cauchy, Poisson, and others. And to prevent a
recurrence of errors arising from a neglect of the distinction here
noticed, he proposes to call such an isolated series as 1 — I + 1 — &c.
independent or neutral; and the extreme cases of l—x+x''^ — &c.,
dependent series : the difference between a dependent and a neutral
series becomes sufficiently marked, as respects notation, by introdu-
cing into the former what the author calls the symbol of continuity,
which indeed is no other than the factor, whose ascending powers
Poisson introduces — and, as here shown, unwarrantably — into the
successive terms of strictly neutral series ; thus bringing such series

138 Cambridge Philosophical Society,

under the control of a law to which in reality they owe no obedi-
ence.

An error somewhat analogous to this is shown to be committed
in the treatment of certain definite integrals, which are here submit-
ted to examination and correction, and some disputed and hitherto
unsettled points in their theory fully considered. The author is thus
led to what he considers an interesting fact in analysis ; viz. that the
differentials of certain forms require indeterminate corrections, in a
manner similar to that by which determinate corrections are intro-
duced into integrals ; and he attributes to the neglect of these the
many erroneous summations assigned to certain trigonometrical
series. This is illustrated by a reference to the processes of Poisson.

The paper concludes with some observations on what has been
called discontinuity ; a term which the author thinks is sometimes
injudiciously employed in analysis, and prefers to treat discontinuous
functions as implying distinct continuities ; and by considering these
in accordance with the principles established in the former part of
the paper, he arrives at results for definite integrals of the form

/+»
x~P dx totally diflferent from those obtained by Poiseon. Two
■m

notes are appended to the paper ; one explaining what the author
denominates insensible convergency and insensible divergency, and the
other discussing some conclusions of Abel in reference to certain
trigonometrical developments.

March 1, 1847. — On the' Theory of Oscillatory Waves. By G.
G. Stokes, M.A., Fellow of Pembroke College.

The waves which form the subject of this paper are characterized
by the property of being propagated with a constant velocity, and
without degradation, or change of form of any kind. The principal
object of the jiaper is to investigate the form of these waves, and
their velocity of propagation, to a second approximation ; the height
of the waves being supposed small, but finite. It is shown that the
elevated and depressed portions of the fluid are not similar, as is the
case to a first approximation ; but the hollows are broad and shallow,
the elevations comparatively narrow and high. The velocity of pro-
pagation is the same as to a first ajiproximation, and is therefore
independent of the height of the waves. It is remarkable that the for-
ward motion of the particles near the surface is not exactly compen-
sated by their backward motion, as is the case to a first approxima-
tion ; so that the fluid near the surface, in addition to its motion of
oscillation, is flowing with a small velocity in the direction in which
the waves are propagated ; and this velocity admits of expression in
terms of the length and height of the waves. The knowledge of
this circumstance may be of some use in leading to a more correct
estimate of the allowance to be made for leeway in the case of a ship
at sea. The author has proceeded to a third approximation in the
case in which the depth of the fluid is very great, and finds that the
velocity of propagation is increased by a small quantity, which bears
to the whole a ratio depending on the square of the ratio of the
height of the waves to their length.

Cambridge Philosophical Society, 139

In the concluding part of the paper is given the velocity of pro-
pagation of a series of waves propagated along the common surface
of two fluids, of which the xipper is bounded by a horizontal rigid
plane. There is also given the velocity of propagation of the above
series, as well as that of the series propagated along the ujjper sur-
face of the upper fluid, in the case in which the upper surface is free.
In these investigations the squares of small quantities are omitted.

March 15. — Contributions towards a System of Symbolical Geo-
metry and Mechanics. By the Rev. M. O'Brien.

The distinction which has been made by an eminent authority in
mathematics between arithmetical and symbolical algebra, may be
extended to most of the sciences which call in the aid of algebra.
Thus we may distinguish between symbolical geometry and arithmC'
tical geometry, symbolical mechanics and arithmetical mechanics. This
distinction does not imply that in one division numbers only are
used, and in the other symbols, for symbols are equally used in both ;
but it relates to the degree of generality of the symbolization. In
the arithmetical science, the symbols have a purely numerical signi-
fication ; but in the symbolical they represent, not only abstract
quantity, but also all the circumstances which, as it is expressed,
aj^ect quantity. The arithmetical science is in fact the first step of
generalization, the symbolical is the complete generalization.

In this view of the case, the author has entitled his paper Contri-
butions towards a System of Symbolical Geometry and Mechanics.
The proposed geometrical system consists, first, in representing
curves and surfaces, not by equations, as in the Cartesian method,
but by single symbols ; and secondly, in using the differential notation
proposed in a former paper* to denote perpendicularity, and to ex-
press various equations and conditions. The proposed mechanical
system is analogous in many respects. Examples of it have already
been given in the paper just quoted.

The author uses the term direction unit to denote a line of a unity
of length drawn in any particular direction ; and he employs the
symbols a j3 y to denote any three direction units at right angles to
each other.

He defines the position of any point P in space by the symbol re-
presenting the line OP (O being the origin) in magnitude and direc-
tion. U X y zhe the numerical values of the coordinates of P, and
a ^ 7 the direction units of the coordinate axes, the expression

xoc+y^ + zy

represents the line OP in magnitude and direction, and therefore
defines the position of P. This expression he calls the symbol of the
point P.

If r be the numerical magnitude, and e the direction unit of OP,
we have

rs—xa.+yfi+zy:

re is therefore another form for the symbol of the point P.

* Read Nov. 23, 1846.

140 Cambridge Philosophical Society,

The following is the method by which the author represents curves
and surfaces.

If the symbol of a point involves an arbitrary quantity, or, as it is
called, a variable parameter, the position of the point becomes inde-
terminate, but so far restricted that it will be always found on some
line or curve. Hence the symbol of a point becomes the symbol of
a line or curve when it involves a variable parameter.

In like manner, when the symbol of a point involves two variable
parameters, it becomes the symbol of a surface.

The parameters here spoken of are supposed to be numerical
quantities. An arbitrary direction unit is clearly equivalent to two
such parameters ; and therefore, when the symbol of a point involves
an arbitrary direction unit, it becomes the symbol of a surface.

The following are examples of this method : —

1 . If M be the symbol of any particular point of a right line whose
direction unit is e, then the symbol of that right line is

M + rg,
r being arbitrary.

2. If M be the symbol of the centre of a sphere, and r its radius,
the symbol of the surface of a sphere is

e being an arbitrary direction unit.

3. If M be the symbol of any particular point of a plane, e and s'
the direction units of any two lines in the plane, the symbol of the
plane is

M + rg+r'g',
r and r' heing arbitrary.

4. If £ be the direction unit and r the numerical magnitude of the
perpendicular from the origin on a plane, the symbol of the plane is

re + Dv-e.
V being an arbitrary line symbol, i. e. denoting in magnitude and
direction any arbitrary line.

5. If M and u' be the symbols of two points, the symbol of the
right line drawn through them is

u + m(u' — m),
m being arbitrary.

6. If M be the symbol of any curve in space, the symbol of the
tangent at the point u is

u -\- mdu,
m being arbitrary.

7. The symbol of the osculating plane at the point u is

u+mdu + m'd"u,

m and m' being arbitrary.

8. If s denotes the length of the arc of the curve, and e the direc-
tion unit of the tangent, then

du

Cambridge Philosophical Society, 141

9. _ or — d{ — \ represents a line equal to the reciprocal of the
ds ds \ dsj

radius of curvature drawn from the point u towards the centre of
curvature, i. e. it represents what may be called the index of curva-
ture in magnitude and direction. "

Hence, since « = xa + y/3 + zy, the numerical magnitude o^ y ^ ( ;r- )

which is the general expression for the reciprocal of the radius of
curvature.

10. The symbol of the normal which lies in the osculating plane is

u+md

/du\
[Ts)'

m being arbitrary.

1 1 . The symbol of any normal at the point u, i. e. the symbol of
the normal plane, is

M + Dy.rfw,

V being an arbitrary line symbol.

12. The symbol of the normal perpendicular to the osculating
plane is

u-\-rnDd"~u.du,
m being arbitrary.

13. If M be the symbol of a surface, involving therefore two vari-
able parameters, A and /* suppose, then the symbol of the normal at
the point u is

TV du du
u + mD -— . — -,
d\ dji

m being arbitrary.

14. The symbol of the tangent plane at the point u is

J , du , du

u+mdu, or u-\-m 1- n — ,

d\ d^h

m and n being arbitrary.

15. The symbol of the plane which contains the three points
u tt' u" is

u + m{u'—u) + n{u"—u).

16. If M be the symbol of a right line, the symbol of the plane
containing it and the point u' is

u+m(u'—u).

The following are examples of the proposed mechanical system in
addition to those given in the paper already quoted.

1 . If r be the radius vector of a planet, and a /3 y be chosen so
that a is the direction unit of the radius vector, and y perpendicular
to the plane of the orbit, it may be shown immediately by the sym-

142 Cambridge Philosophical Society.

bolical method, that the symbol of the force acting on the planet is

\dt° / r dt

where la is the angular "velocity of the planet, and to' that of the
plane of the orbit about the radius vector. The expressions for the
three component forces along r, perpendicular to r, and perpendicular
to the plane of the orbit, are the coefficients of a /3 7 in this expres-
sion.

2. The equation of motion of the planet, when the force is the
attraction of a fixed centre varying as the inverse square of the di-
stance, is

d'^u _ jua

It is curious that this equation is immediately integrable, the in-
tegral being the two equations

r h

The latter equation is the symbolical equation of a conic section,
the origin being focus, h c and s being the arbitrary constants intro-
duced by integration.

3. The application of this method to the case of a planet acted on
by a disturbing force is worthy of particular notice, as it expresses
the variations of the elements of the orbit with great facility, in the
following manner : —

If U be the symbol of the disturbing force, we have

^M:=VU,\J (1.)

dt

^)=if!:/3Aj8.U + U. ..... (2.)

dt h^^ ^

These two equations determine with great facility all the elements
of the orbit. For y is a direction unit perpendicular to the plane
of the orbit {i. e. it is the symbol of the pole of the orbit), and there-
fore it defines completely the position 01 the plane of the orbit. Also
8 is a direction unit in the plane of the orbit at right angles to the
axis major, and therefore it determines the position of the axis major ;
in fact the direction unit of the axis major is Dy.g. The letters h
and e have their usual signification.

To find h and y separately from (1.), suppose that we obtain by
integration of (1.)

then A2=:AW.W ; and h being thus found, we have y= — . The
same observation applies to (2.).

Royal Astronomical Society. H3

4. The expression for the parallax of the planet is

These instances suffice to show the nature of the proposed sym-
bolical method.

ROYAL ASTRONOMICAL SOCIETY.
[Continued from vol. xxx. p. 211.]

April 9, 1847. — On an important error in Bouvard's Tables of
Saturn. By Mr. Adams.

Having lately entered upon a comparison of the theory of Saturn
witli the Greenvirjch observations, I was immediately struck with the
magnitude of the tabular errors in heliocentric latitude, and the more
so, since the whole perturbation in latitude is so small, that it could
not be imagined that these errors arose from any imperfection in the
theory. In order to examine the nature of the errors, I treated them
by the method of curves, taking the times of observation as abscissae,
and the corresponding tabular errors as ordinates. After elimina-
ting, by a graphical process, the effects of a change in the node and
inclination, a well-defined inequality became apparent, the period of
which M^as nearly twice that of Saturn. One of the principal terms
of the perturbation in latitude (viz. that depending on the mean lon-
gitude of Jupiter minus twice that of Saturn) having nearly the same
period, I was next led to examine whether this term had been cor-
rectly tabulated by Bouvard. The formula in the introduction ap-
peared to be accurate; but on inspecting the Table XLIL, which
professes to be constructed by means of this formula, I was surprised
to find that there was not the smallest correspondence between the
numbers given by the formula and those contained in the table, the
latter following the simple progression of sines, while the formula
contained two terms. The origin of this mistake is rather curious.
Bouvard's formula for the terms in question is

9"-67sin{(p-2^'-60°-29}-f.28"-19sin{2^-4^'-h66°-12};

but in tabulating the last term he appears to have taken the simple
argument ^ — 2^' instead of 2f — 4ip', so that the two parts may be
united into a single term,

25"-85 8in{^-2^' + 43°-88}.

which I find very closely to represent Bouvard's Table XLII.

After correcting the above error, and making a proper alteration
in the inclinations and place of the node, the remaining errors of
latitude are in general very small. I subjoin a correct table to be
used instead of Bouvard's. The constant added being 36"'0 instead
of 26"'0, it will be necessary to subtract 10''"0 from the final result.

144

Royal Astronomical Society.

Table XLII. — Argument III. de la Longitude.

Argiunent.

Equation.

Aliment.

Equation.

Argument.

Equation.

Argument.

Equation.

0

5^-4

2500

17-4

5000

68-1

7500

u

100

54-4

2600

16-2

5100

69-4

7600

40

200

56-0

2700

15-5

5200

70-2

7700

2-3

300

57-2

2800

15-2

5300

70-5

7800

11

400

580

2900

15-2

5400

70-4

7900

0-4

500

58-3

3000

15-7

5500

69-8

8000

01

600

58-3

3100

16-6

5600

68-7

8100

0-4

700

57-8

3200

17-9

5700

67-2

8200

1-0

800

56-9

3300

19-6

5800

65-3

8300

2-2

900

55-7

3400

21-7

5900

62-9

8400

3-7

1000

541

3500

241

6000

601

8500

5-7

1100

52-2

3600

26-7

6100

57-1

8600

8-0

1200

500

3700

29-7

6200

53-7

8790

10-7

1300

47-5

3800

32-8

6300

50-0

8800

13-7

1400

44-9

3900

36-2

6400

46-2

8900

16-8

1500

421

4000

39-6

6500

421

9000

202

1600

39-2

4100

431

6600

38-0

9100

23-7

1700

36-2

4200

46-5

6700

33-9

9200

27-3

1800

33-3

4300

500

6800

29-8

9309

310

1900

30-4

4400

53-3

6900

25-7

9400

34-5

2000

•21-7

4500

56-5

7000

21-8

9500

380

2100

25-1

4600

59-4

7100

181

9600

41-4

2200

22-8

1 4700

621

7200

14-6

9700

44-6

2300

20-6

4800

64-5

7300

11-4

9800

47-5

2400

18-8

1 4900

66-5

7400

8-5

9900

50-1

2500

17-4

5000

681

7500

61

1000

52-4

Constante ajoutee 36"'0.

On the Development of the Disturbing Function R. By Sir John
Lubbock.

The greatest practical difficulty which is encountered in the pla-
netary theory consists in the development of the expression for the
reciprocal of the linear distance between the disturbed and disturbing
planets. The algebraical expression of this development may be
obtained either by means of the binomial theorem or by Taylor's
theorem applied to several variables ; by the latter method M. Binet
has carried the development as far as terms of the 7th order. But
when high powers of the eccentricities and inclinations are retained,
the expressions become excessively complicated, so that further pro-
gress in this direction appears utterly hopeless.

The numerical coefficients of the series may also be obtained by
quadratures ; but to determine all the coefficients in this way would
involve very great labour.

In considering the problem of the perturbations of bodies whose
eccentricities and inclinations are considerable, the author has been
led to another mode of development, which he conceives to possess
great advantages over those just mentioned, and the use of which
may be greatly facilitated in all cases by special tables, which may
be prepared beforehand.

Royal Astronomical Society. 145

The principle of this method consists in expressing the square of
the ratio of the distance between the two planets to the radius vector
of the more distant planet, under the form of P— Q, in which P is the
product of any convenient number of factors of the form 1 + A cos a,

and all the terms in Q have small coefficients. Then (P— Q)""''
may be developed by the binomial theorem in a series of ascending
powers of Q, which consequently converges rapidly, and the values of

the quantities P"*^, P""*, &c., which enter into the successive terms
of this series, may be found by multiplying together the developments

1 3

of the several factors (1 + A cos a) ^, (1 + A cos a) ^, &c. If then
tables were prepared, giving for different values of A the coefficients
of the development of these quantities in cosines of multiples of a,
all the operations requisite for the development of the disturbing
function might be performed with great facility.

The author remarks, in conclusion, that instead of developing, as

is usually done, in powers of the ratio of the mean distances ( — )'

it would be preferable to develope according to powers oi ,-^

which is much less than the former when a and a' do not
widely from each other.

Observations of Hind's Second Comet in full Sunshine*. By Mr.
Hind.

I take the liberty to send you two positions of the comet discovered
here Feb. 6, obtained yesterday in full daylight, and about five hours
before the perihelion passage. The visibility of a comet in the day-
time, and within 2° distance from the sun, is a phBenomenon of so
rare occurrence, that it may in some measure interest you if I give
very briefly the particulars of our observations.

I had determined, by theory, that the intensity of light on March
30 ought to be 100 times stronger than that of a star of fourth mag-
nitude, and was induced to make preparations for a daylight obser-
vation. I first saw the comet about 11 a.m. When the sky was
perfectly cloudless about the sun, it had a whitish appearance, which
rendered it a matter of no little difficulty to see the comet ; but
during the passage of some cumuli clouds over the sun, and between
the breaks, I obtained some excellent views of the comet, and several
observations, which will no doubt be of great assistance in the accu-
rate determination of the elements. The nucleus was nearly, if not
perfectly, round, beautifully defined and planetary, the diameter 8"
or 1 0". Two faint branches of light formed a divided tail, extend-
ing about 40" from the head, like two longish erect ears or horns
rising from each side of the disc. At times I felt certain that the
nucleus twinkled. The tail resembled a thin smoke.

With respect to the observations for position, I can only add that
they were as good as could possibly be made, under the circum-

* The comet was seen at noon near the sun by two other observers, at
Truro and in the Isle of Anglesey.

Phil. Mag, S. 3. Vol. 31. No. 206. Aug. 1847. L

146 Intelligence and Miscellaneous Articles.

stances, by instrumental comparisons. The index errors are very-
constant, and were accurately determined last evening.

Greenwich M.T. Comet's R.A. Comet's Dec. Weight,

h m s o I II o I II

March 30 1 23 40 7 32 27 -1-1 48 43 1

1 55 8 7 33 56 +1 45 21 2

In the observations for the first position the centre of the field was
estimated, and nine single results are tolerably accordant. The
second place depends on one good observation with cross wires,
clouds preventing any further comparisons.

Had the sky been free from the whiteness which is so fatal to
vision by daylight, I should have obtained much better places.

I communicated an ephemeris to Mr. Dawes, who has observed
the comet with extreme care, but I do not know at present whether
he saw it yesterday in daylight.

XXVIII. Intelligence and Miscellaneous Articles.

ON A NEVr TEST FOU PRUSSIC ACID, AND ON A SIMPLE METHOD
OF PREPARING THE SULPHOCYANIDE OF AMMONIUM. BY
PROF. LIEBIG.

WHEN some sulphuret of ammonium and caustic ammonia are
added to a concentrated aqueous solution of prussic acid, and
the mixture heated with the addition of pure flowers of sulphur, the
prussic acid is converted in a iew minutes into sulphocyanide of am-
monium. This metamorphosis depends on the circumstance, that
the higher sulphurets of ammonium are instantly deprived by the
cyanide of ammonium of the excess of sulphur they contain above
the monosulphuret ; for instance, if a mixture of prussic acid and
ammonia be added to the pentasulphuret of ammonium, the solution
of which is of a deep yellow colour, and the whole gently heated,
the sulphuret of ammonium is soon decolorized ; and when the clear
colourless liquid is evaporated, and the admixture of sulphuret of
ammonium expelled, a white saline mass is obtained, which dissolves
entirely in alcohol. The solution yields, on cooling or evaporation,
colourless crystals of pure sulphocyanide of ammonium. Only a
small quantity of sulphuret of ammonium is requisite to convert, in
the presence of an excess of sulphur, unlimited quantities of cyanide
of ammonium into sulphocyanide ; because the sulphuret of ammo-
nium, when reduced to the state of monosulphuret, constantly reac-
quires its power of dissolving sulphur and transferring it to the
cyanide of ammonium. The following proportions will be found to
be advantageous: — 2 oz. of solution of caustic ammonia of 0'9^
spec. grav. are saturated with sulphuretted hydrogen gas ; the
hydrosulphate of ammonia thus obtained is mixed with 6 oz.
of the same solution of ammonia, and to this mixture 2 oz. of
flowers of sulphur are added ; and then the product resulting from
the distillation of 6 oz. prussiate of potash, 3 oz. of the hydrate of
sulphuric acid, and 18 oz. water. This mixture is digested in the

Intelligence and Miscellaneous Articles. 147

water-bath until the sulphur is seen to be no longer altered and the
liquid has assumed a yellow colour; it is then heated to boiling, and
kept at this temperature until the sulphuret of ammonium has been
expelled and the liquid has again become colourless. The deposited,
or excess of, sulphur is now removed by filtration, and the liquid
evaporated to crystallization. In this way from 3^ to 3^ oz. of daz-
zling white dry sulphocyanide of ammonium are obtained, which may
be employed as a reagent, and for the same purposes as the sulpho-
cyanide of potassium. Of the 2 oz. of sulphur added, ^ an oz. is
left undissolved.

The behaviour of the higher sulphurets of ammonium towards
prussic acid furnishes an admirable test for this acid. A couple of
drops of a prussic acid, which has been diluted with so much water
that it no longer gives any certain reaction with salts of iron by the
formation of prussian blue, when mixed with a drop of sulphuret of
ammonium and heated upon a watch-glass until the mixture is be-
come colourless, yields a liquid containing sulphocyanide of ammo-
nium, which produces with persalts of iron a very deep blood-red
colour, and with persalts of copper, in the presence of sulphurous
acid, a perceptible white precipitate of the sulphocyanide of copper.
— Liebig's Annalen, Jan. 1847.

ON THE FUSION OF IRIDIUM AND RHODIUM. BY R. HARE.

This communication respects mainly my success in fusing both
iridium and rhodium, neither of which, in a state of purity, had been
previously fused. It may be supposed that the globule of iridium,
obtained by Children's colossal battery, forms an exception ; but
the low specific gravity and porosity of that globule may justify a
belief that it was not pure, and at any rate the means employed were
of a nature not to be at command for the repetition of the process,
so that iridium might as well be infusible, as to be fusible only by
such a battery.

The first specimen of the last-mentioned metal on which I ope-
rated was one given me by Mr. Booth, a former pupil of Wohler,
whom he had assisted in obtaining it by the excellent process de-
vised by that distinguished chemist. This specimen was fused in
the presence of Mr. Booth. Subsequently I procured specimens,
warranted pure, severally from the house of Pelletier at Paris, and
from Messrs. Johnson and Cock, London. Another specimen was
given to me by a friend, who had received it as pure, from a source on
which reliance may be placed ; and lastly, I obtained myself, by
Wohler's process, a specimen of about sixty grains, from the inso-
luble residue of platinum ore. All the specimens thus procured
were found to be fusible under my hydro-oxygen blowpipe. The
specimen obtained from Messrs, Johnson and Cock, after repeated
fusions, by which it was much consolidated, weighed sixty-seven
grains. During fusion there appeared to be an escape of volatile
matter, supposed to be osmic acid, arising from the presence of a
minute portion of osmium, between which and iridium an affinity of

L2

148 Intelligence and Miscellaneous Articles,

a peculiar degree of energy exists. At a certain point of the process
a reaction took place sufficiently explosive to throw a portion of the
metal, in globules, off from the support. One of these, about twice
as large as the head of a common brass pin, proved to be hollow.
By prolonged and repeated fusion the metal became more compact
and more fusible.

Fused iridium has nearly the grain of soft cast steel, with the pale
whiteness of antimony, and appears to be susceptible of a fine polish.
Although as hard as untempered steel, it is somewhat sectile, since,
when split by means of a cold chisel, the edge penetrated about the
eighth of an inch before a division was effected. By light ham-
mering a corner was flattened without fracture, although under
heavier blows the mass cracked. I infer that although nearly un-
malleable and very hard, iridium may be wrought in the lathe.

I have already mentioned that I fused into a globule a specimen
of iridium obtained by me from the insoluble residuum of platinum
ore by Wohler's process. From this globule, while congealing, a
portion ran out from the inside, leaving a cavity and covering one
of its sides externally with an incrustation, among which crystalline
spangles, or facets, were discernible. The specific gravity of the
globule of iridium, from the specimen furnished by Messrs. John-
son and Cock, was taken by Mr. T. R. Eckfelt of the United States
mint at Philadelphia, and by Dr. Boye, both having balances of the
greatest accuracy, and being very skilful in the employment of
them. In the first instance there was a perfect coincidence in the
results obtained, 21*83 being the numbers found by both of these
gentlemen. Agreeably to another trial made by Dr. Boye, using
river-water instead of distilled, the number was 21*78, water being
in either case about sixty-eight, with allowance for the diflference of
the water, and the temperature being above the standard of 60°.
The specific gravity of the specimen may then be estimated at 21*80.

The specific gravity of fused platinum, purified according to the
instructions of Berzelius, before subjection to the hammer, proved
in one specimen to be not more than 19*70, although by hammering
it became equal to 21*23. It is with fused platinum that fused
iridium should be compared. Of course the specific gravity of the
last-mentioned metal, when both are obtained by fusion, may be
assumed to be one-tenth greater than that of the former. Moreover,
as this metal is the only impurity existing in the standard platinum
of London, of Paris, or of St. Petersburg, it follows that a high
specific gravity is not to be viewed as a proof of purity. Accord-
ingly a specimen of platinum, purified from iridium by the Ber-
zelian process, and which had proved eminently susceptible of being
beaten into leaf, was found only to be of the gravity of 21*16, while
that of a specimen of standard Russian platinum, very briUiantly
white but inferior in malleability, presented to me by his Excellency
Count Cancrine, as a specimen of the purest platinum of the Russian
mint, was 21*31.

Of rhodium I have fused two specimens, one of five pennyweights,
purchased of Messrs. Johnson and Cock, the other received through

IntelUsence and Miscellatieous Articles. 149

"b

the same channel as the specimen of iridium above-mentioned*.
Rhodium is at least as fusible as iridium, both of the specimens
alluded to having been converted into fluid globules. That pro-
cured from Messrs. Johnson and Cock gave a globule weighing
ninety grains. On a second fusion it formed a perfect globule as
fluid as mercury ; and yet in congealing lost its brilliancy by be-
coming studded with crystalline facets all over its surface, excepting
the portion in contact with the support. The facets had the ap-
pearance of incipient spangles. The rapidity with which they were
formed seemed anomalous. The mass being split by a cold
chisel and viewed by a microscope, it appeared porous immediately
beneath the facets. When the mass was first fused, I found by the
gravimeter the specific gravity to be 11 '0, which coincides with the
observations of WoUaston. Yet by a careful trial made at the United
States mint by Mr. Eckfelt, after the second fusion and the forma-
tion of the facet, the specific gravity proved to be only 10'8. This
is sufficiently explained by the porosity above mentioned. In fact
the porosity to which rhodium and iridium are liable may render it
difficult to find specimens of precisely the same specific gravity.

In sectility, malleability and hardness, rhodium did not appear to
differ much from iridium, but it is not of so pale a white as iridium.
The one has the pale white of antimony, the other the ruddy hue of
bismuth.

Osmiuret of iridium, as existing in the native spangles associated
with platina ore, or as otherwise obtained, is far more difficult of
fusion than pure iridium. The propensity to assume the crystalline
form, and to adhere to it, is even greater in this alloy than in the
last-mentioned metal. On first exposure to the most intense heat
of the hydro-oxygen blowpipe some slight appearances of fusion
may be seen, and the spangles or grains may be made to cohere.
Nevertheless it yields very slowly, aijd requires an expenditure of
gas too great to be incurred unless it were for the purpose of once
well determining the question of its ultimate fusibility. This ob-
ject was obtained completely as respects a globule of 45 grains in-
weight. The specific gravity of this globule appeared to be 20*4,
but this result was evidently less than tliat whicii would have been
obtained had there not been some minute cavities, which, after
splitting the globule, were detected by a magnifier.

The specific gravity of some large spangles of osmiuret of iridium
from South American ore was, by Dr. Boye, found to be 19*835.
That of some grains heavier but not so flat, presented to me by
Count Cancrine, was found to be 20'938.

That the alloy of iridium with osmium should be more difficult to
fuse than pure iridium, leads to the inference that osmium must be
the most infusible of the metals, although, like carbon, very sus-
ceptible of combustion, and capable, like tliat infusible non-metallic
radical, of forming a volatile peroxide. Of course its liability to
oxidizement would render it impossible to fuse it by the hydro-

* One other larger specimen from the same source has been fused since the
above was written.

150 Intelligence and Miscellaneous Articles.

oxygen blowpipe, of which the efficacy requires the simultaneous
presence of oxygen and the most intense heat. It might be fused
by exposure in vacuo to the discharge of a powerful voltaic series,
by means of the apparatus of which a description with engravings
has been given in a recent volume of the Transactions of the Ame-
rican Philosophical Society, and republished in ' Silliman's Journal'
for 1841, vol. xl. p. 303.

I have obtained osmium by heating the osmiate of ammonia in a
glass tube with sal-ammoniac, agreeably to the instructions given by
Berzelius. In this way a result was obtained which the information
given by that distinguished chemist had not led me to anticipate.
The tube became coated with a ring of osmium, which it would be
impossible by inspection merely to distinguish from the arsenical
ring on the peculiar features of which reliance has been placed for
the detection of arsenic.

It follows from my experiments and observations, that of all
metallic bodies, osmiuret of iridium is the most difficult to fuse ;
that rhodium and iridium are both fusible by the hydro-oxygen blow-
pipe, properly employed ; that the former has the rosy whiteness of
bismuth, the latter the pale white of antimony ; and that both of
them are slightly sectile, though extremely hard and nearly un-
malleable ; that iridium merely fused is heavier than platinum con-
densed by the hammer. Thus it follows from my experiments, and
from the recent observations of Breithaupt, on some specimens of
native iridium, that the metal, whether in this state or pure as ob-
tained by chemical skill and consolidated by fusion, must be allowed
that pre-eminence in density, which, until of late, was given to
platinum.

It may be proper to add, that subsequently to the writing of the
preceding narrative, receiving some large quantities of iridium and
rhodium from Messrs. Johnson and Cock, my experiments were
successfully repeated on a larger scale, but without any result be-
sides that of confirming the facts above stated. — Silliman's Journal
for Nov. 1846, p. 365.

NOTE ON THE MEANS OF TESTING THE COMPARATIVE VALUE
OF ASTRINGENT SUBSTANCES FOR THE PURPOSES OF TAN-
NING. BY ROBERT WARINGTON, ESQ.

Having been frequently called upon to examine the value of
astringent substances imported into this country for the purposes of
tanning, such as valonia, divi-divi, sumac, cutch, &c.,-I am induced
to believe that the detail of the manipulation adopted may not be
without interest to some of the members of the Society. As the
manufacture of leather was the object of the purchaser of these
materials, gelatin was selected as the basis for the estimation of
their comparative value ; and after several trials with various kinds of
natural and manufactured gelatin, such as varieties of isinglass, glue,
patent gelatin, &c., the finest long staple isinglass was found to be
the most constant in its quality and least liable to undergo change.

Intelligence and Miscellaneous Articles. X&\

With this therefore the test solution was prepared, of such a
strength, that each division, by measure in the ordinary alkalimeter
tube, should be equivalent to the one-tenth or one-fourth of a grain
of pure tannin, and thus the number of divisions used would indicate
the proportion of available tannin or substance precipitable by ge-
latin contained in any specimen. A given weight of the sample
under trial was then infused in water, or if necessary the astringent
matter extracted by boiling, and the clear liquid precipitated by
the test solution until no further deposit occurred.

It was necessary in the course of this operation to test at intervals
a portion of the solution under examination, to ascertain the pro-
gress of the trial ; and this, from the nature of the precipitate, was
attended at first with some little difficulty : paper filters were inad-
missible from the quantity of the solution they would absorb, and
thus introduce a source of extensive error ; subsidence rendered the
operation very tedious. The plan I have adopted is as follows : — a
piece of glass tubing, about twelve inches in length and about half
an inch internal diameter, is selected, and this has a small piece of
wet sponge loosely introduced into its lower extremity, and when
it is wished to abstract a part of the fluid under investigation for a
separate testing, this is immersed a few seconds in the partially pre-
cipitated solution ; the clear liquid then filters by ascent through the
sponge into the tube, and is to be decanted from its other extremity
into a test glass ; if on adding a drop of the gelatin solution to
this a fresh precipitate is caused, the whole is returned to the ori-
ginal bulk, and the process proceeded in, and so on until the opera-
tion is perfected ; this method of operating is facilitated by conduct-
ing the examination in a deep glass. After a few trials the mani-
pulation will be found extremely easy, and in this way considerable
accuracy may be arrived at. — From the Proceedings of the Chemical
Society. ________

ON THE TWO VARIETIES OF ARSENIOUS ACID. By M. BUSSY.

The author first gives a new process for determining the quantity
of arsenious acid. This process is based on the employment of stand-
ard reagents. The reagent which he uses is permanganate of pot-
ash, which M. Marguerite has already successfully employed for the
quantitative determination of iron.

When a solution of permanganate of potash is poured into a solu-
tion of arsenious acid, it becomes arsenic acid, and the red colour of
the reagent disappears. The liquor begins to become coloured only
when the transformation of arsenious acid is complete. When, then,
a standard solution of permanganate of potash is prepared, the quan-
tity of arsenious acid contained in any solution may be determined
by that of the permanganate required to convert it into arsenic acid.

M. Bussy states that the two varieties of arsenious acid, the vi-
treous and opake, absorb the same quantit}'' of permanganate, and
consequently that the diiFerences observed in their solubility is not
derived from any difference of oxidizement.

With respect to the solubility of the two varieties of arsenious

152 Intelligence and Miscellaneous Articles.

acid, M. Bussy has arrived at the following conclusions : — 1st. The
vitreous, so far from being less soluble in water than the opake acid,
as stated by chemists, is, on the contrary, much more soluble. This
difference is nearly in the proportion of 3 to 1, at about 53° to 55°
of F. ; the same quantity of water which dissolves 36 to 38 parts of
the vitreous acid, will take up only 12 to 14 of the opake. 2nd.
The vitreous acid dissolves much more rapidly than the opake acid.
3rd. Neither of the varieties possesses a degree of solubility which
is to be regarded as strictly peculiar to it. 4th. The opake acid
is converted into vitreous acid by long boiling in water ; that is to
say, it then acquires the same degree of solubility as the vitreous
arsenious acid, which is such that 11 parts are dissolved by 100 of
water. 5th. Under the influence of water and a low temperature,
the vitreous acid is converted into opake acid ; that is to say, a solu-
tion of vitreous acid becomes reduced after a certain time to the
point of saturation which belongs to the opake acid. 6th. The mix-
ture of the two varieties of acid in the same solution explains the
anomalies observed in the solubility of arsenious acid, which in fact
offers nothing opposed to the principles admitted by chemists. 7th.
Division, which facilitates the solution of the opake acid, without
however increasing its solubility, considerably diminishes that of the
vitreous acid ; and to such an extent, that this acid, reduced to fine
powder and levigated, is not sensibly more soluble in water than
the opake acid ; this resulting unquestionably from a transformation
which it undergoes, either at the moment of pulverization, or of its
contact with water. 8th. Acid which has been rendered opake by
the action of ammonia, and acid crystallized in water, act similarly
with water, and appear to belong to the samevariety. 9th. The opake
acid dissolves more slowly than the vitreous in dilute hydrochloric
acid. This circumstance, which thus modifies the nature of the
products formed during solution, explains why the luminous phseno-
mena observed by M. Rose in the crystallization of the vitreous acid,
are not in general observable with so great intensity in the solution
of the opake variety. ] 0th. The difference which has been observed
in the action of the two arsenious acids on tincture of litmus is
merely apparent. If the opake acid does not redden the tincture, it
is on account of its slight solubility, and especially because it dis-
solves slowly ; whilst the vitreous acid, which dissolves quickly,
immediately reddens the tincture. But if comparative experiments
be made, and the tincture be exposed to the action of the powder,
it becomes gradually red, and no difference is perceptible at the ex-
piration of three or four days. — Comptes Rendus, Mai 1847.

ON THE PREPARATION OF GUN-COTTON.

Mr. Coathupe recently forwarded to the Chemical Society two
specimens of gun-cotton, with a view to illustrate the greatly in-
creased explosive effects that are to be derived from a subsequent
immersion of the gun-cotton, when properly prepared in the ordi-
nary way, in a saturated solution of chlorate of potash.

Intelligence and Miscellaneous Articles. ] 53

" Having experimented with solutions of nitrate of ammonia, ni-
trate of potash, nitrate of soda, bichromate of potash, &c. &c., for
the purpose of increasing the explosive properties of this interesting
substance, I can affirm that none of the results will bear the slight-
est comparison with those obtained from the solution of chlorate of
potash, either in rapidity of ignition or in intensity of flame. The
process adopted for preparing the inclosed specimens was as follows :
viz. into a mixture of equal measures of strong nitrous acid and of
oil of vitriol, spec. grav. 1*845, the cotton was immersed and stirred
with a glass rod during about three minutes : it was then well-
washed in many waters and dried ; a portion of it was then soaked
for a few minutes in a saturated solution of chlorate of potash, well-
squeezed and dried."

ON BALSAM OF TOLU, AND SOME PRODUCTS DERIVED FROM IT.

M. E. Kopp states that the experiments which he has made on
this substance confirm the greater number of the results previously
obtained. He remarks that the balsam is composed of a very small
quantity of tolene C'oH"', C=75H=6"25; of free cinnamic acid,
CIS H'6 O* ; of a resin very soluble in alcohol, C^e H?8 O^ ; of a resin
slightly soluble in alcohol, C'« H^o O*, or C^e H^o Qio.

Tolen. — This carburetted hydrogen was prepared by exactly fol-
lowing the plan proposed by M. Deville. It is colourless, very fluid,
of a penetrating taste somewhat like pepper, and its smell resembles
that of elemi. Its density at 60° F. is 0'858 ; its boiling-point is
between 310° and 320° F. Exposed in an imperfectly closed tube,
it gradually becomes resinous and very slightly coloured. M. De-
ville gives as its formula C'^H'^. M. Kopp states that his analysis,
which diff'ers but little from that of M. Deville, indicates 0'° H'^,

Cinnamic Acid. — The free acid of balsam of Tolu, as observed by
M. Fremy, is merely cinnamic acid. This fact was proved by ana-
lysis, and by its conversion into nitrocinnamic acid, very slightly
soluble in cold alcohol ; whereas benzoic and nitrobenzoic acids are
very soluble in it. The results obtained by M. Deville are probably
derived from his having examined the acids procured by the distilla-
tion of the balsam, or extracted by concentrated alkaline solutions.

M. Kopp has shown that, under these two circumstances, the resins
of balsam of Tolu are so changed as to give rise to a large proportion
of benzoic acid. The resins, cautiously distilled with caustic soda, yield
pure benzoen, and a coaly residue which contains much benzoate of
soda. Cinnamic acid, mixed with cold concentrated caustic soda,
and submitted to a current of chlorine, is converted into chlorocin-
namic acid C'** (H^* Cl^) O*. If however the temperature be raised
and the action is very strong, the chlorinated oil described by Mr.
Stenhouse is disengaged, and chlorobenzoic acid, C* (H'o Cl^) O*, is
formed.

These two acids strongly resemble each other ; but the latter is
more soluble in water and in alcohol, and its salts crystallize more
readily. Cinnamic acid, treated with concentrated nitric acid, is at

1 54 Intelligence and Miscellaneous Articles.

first converted into nitrocinnamic acid, then into benzoic acid, and
finally into nitrobenzoic acid.

Cinnamic and benzoic sethers are both, though with great diffi-
culty, converted into nitrocinnamic acid and nitrobenzoic aether.
There is almost always a great part of the sether decomposed, and
the acids are set free.

Nitrobenzoic sether is solid, colourless, and of an aromatic odour
and taste. It crystallizes in fine rhombic lamina;. Its melting-point
is 11 6°, and its boiling-point 664°. It is easily obtained by exposing
an alcoholic solution of nitrobenzoic acid to a current of hydrochloric
acid gas. Its formula is C* (H« N^ O*) 03 + C* H'oOrrC's H'^ N^ O^.

Nitrocinnamic acid dissolved in an alcoholic solution of sulphuret
of ammonia is reduced with the assistance of a gentle heat. Sulphur
is deposited, and two distinct substances are formed, one of which is
of a yellowish colour and belongs to the class of resins, and the other
to that of alkaloids. The latter is solid, colourless, crystallizes in
small indistinct masses, insoluble in water, soluble in alcohol and
in aether, and forms difficultly crystallizable salts.

Resin a, C^^ H^s O^. This substance is brown, translucent, brittle
when cold ; its powder agglomerates at 59° F. and fuses perfectly at
140° F. Concentrated sulphuric acid imparts a purple colour to it.
When dissolved in potash and exposed to the air, it is readily oxi-
dized, and is converted into resin /3. By dry distillation it yields
benzoen and benzoic acid. It dissolves readily in alcohol and in
aether.

Resin jd, C's H^° O^ Colour dull brownish-yellow, without taste or
smell, slightly fusible (above 212° F.), but little soluble in alcohol
or ajther. It is less alterable than the preceding resin. Sulphuric
acid renders it of a violet colour ; potash dissolves it with a brown
colour.

The mixture of the two resins treated with nitric acid yields, as
gaseous products, carbonic acid, nitrous vapours and nitric oxide ;
as volatile products, hydruret of benzule, hydrocyanic acid, and a
little benzoic acid ; as residue, a flocculent yellowish substance,
which is benzoic acid intimately combined with a yellow colouring
matter of a resinous nature, which destroys its crystallizing power,
and accompanies it in all its combinations, even in that of aether.
By the action of heat, especially by distillation, the resinous matter
is destroyed, and perfectly pure benzoic acidis obtained. The resin
yields nearly one-third of its weight of benzoic acid.

As to the constitution of balsam of Tolu, it seems very simple.
Primarily it is formed of the soft resinous matter C^'^ H^s O'', or of
that which gives rise to it. This resin, under the influence of the
air, is converted into cinnamic acid and resin /3 : C^^ H38 + 0^=C'8
H'" O^ + C'sH^oO^-fH'^O. In fact it is observed that in time
balsam of Tolu becomes hard, and contains a larger quantity of cin-
namic acid. The resin C'^ W^ O* may itself easily furnish benzoic
acid for C'sH^o 0^=0'^ Hi^O^ + H^ O + C* H^ The carburetted
hydrogen perhaps gives rise to tolene ; but it is more probable that
it is converted by the action of oxidizing bodies into resinous colour-

Intelligence and Miscellaneous Articles. 156

ing matter, or perhaps into water and carbonic acid. — Ann. de Ch.
et de Phys., Juillet 1847.

ON THE EQUIVALENT OF TITANIUM. BY M. ISIDORE PIERRE.

The author remarks that chemists generally agree that it would
be difficult to add to the precision of the numbers which represent
the equivalents of hydrogen, carbon, chlorine, bromine, iodine, phos-
phorus, arsenic and silicon, as determined by the researches of Du-
mas, Marignac and Pelouze.

M, Pierre thinks however that this is not the case with titanium ;
and that if the labours of different periods respecting this substance
be examined, it will be evident that its equivalent requires renewed
examination.

M. H. Rose originally obtained, by various methods, numbers
which varied between 380 and 450 ; but he afterwards found that
the sulphuret of titanium which he employed in his experiments,
was procured free from titanic acid with great difficulty.

In his last experiments, M. Rose made use of chloride of titanium,
which he decomposed by water. He precipitated with ammonia the
titanic acid derived from this decomposition, and afterwards treated
the filtered liquor with nitrate of silver, in order to separate the
chlorine in the state of chloride of silver : this method gave him
303*686 as the equivalent of titanium.

The chloride of titanium used by M. Pierre was not prepared from
rutil, but from calcined artificial oxide of titanium : it was free from
oxide of iron and from chloride of silicium, and its boiling-point was
perfectly stationary. The chloride employed had been kept in a small
tube from the time of its preparation hermetically sealed : it was
broken by agitation in a stopped bottle, one quarter filled with distilled
water. By frequent agitation, Avithout unstopping the bottle, the
whitish cloud at first produced above the liquid disappears. Without
this precaution there would be a probable loss of hydrochloric acid
in opening the bottle too soon, or by introducing the solution of
silver, which would expel a small quantity of this vapour.
The following results were obtained : — gr.

I. Chloride of titanium employed. . 0"8215

Silver 1'84523

indicating Chlorine 0'60623

Titanium by difference 0*21727

These results gave 314*76 as the equivalent of titanium.
II. Chloride of titanium employed. . 0*774

Silver 1*73909

indicating Chlorine 0*57 136

Titanium by difference 0*20264

These numbers give for the equivalent of titanium 314*37.
III. Chloride of titanium employed. . 0*7775

Silver 1*74613

indicating Chlorine 0*57367

Titanium by difference 0*20383

The equivalent of titanium deduced from this experiment is 31 4*94 .

156 Intelligence and Miscellaneous Articles,

IV. Chloride of titanium employed. . 0*716

Silver 1-61219

indicating Chlorine 0'52966

Titanium by difference 0*1 8634

Equivalent of titanium 31 1*84 .

V. Chloride of titanium employed. . 0"8085

Silver 1-82344

indicating Chlorine 0*59907

Titanium by difference 0-20943

Equivalent of titanium 309*38.

The three first numbers agree perfectly, but the two latter are
notably less, especially the last, since it differs from the three first
by five whole numbers, or more than 1^ per cent. It was difficult
to attribute this difference entirely to deficient precision in the me-
thod used. It occurred to the author that it might be owing to the
partial decomposition of the chloride of titanium, by the moisture of
the air during manipulation, and this was soon found to be the case
by direct experiment.

M. Pierre proposes to adopt, as the nearest approximation to truth,
314-69, the mean of the three first experiments, as the equivalent
number for titanium.

This number is very different from 355 deduced from 6-536, the
density of the vapour of the chloride of titanium observed by M,
Dumas. Its density, calculated from 314-69, would be 6-614. —
Ann. de Ch. et de Phys., Juillet 1847.

ON A MODIFICATION OF THE APPARATUS OF VARRENTKAPP AND
WILL FOR THE ESTIMATION OF NITROGEN. BY WARREN DE
LA RUE.

My attention having been called to a communication by Mr. Alex.
Kemp in the number of the ' Chemical Gazette' for the 1st of April
1847, in which he describes a modification of Messrs. Varrentrapp
and Will's tube for nitrogen determinations, of a very similar con-
struction to one I employed as far back as November 1845 in the
laboratory of the Royal College of Chemistry, and which I have re-
peatedly shown to my friends, I am induced to lay before the So-
ciety a description of my form of apparatus, which differs somewhat
from that described by Mr. Kemp.

By the drawing, it will be seen that the tube B E, instead of
opening immediately into the bottom of the flattened bulb C, is pro-
longed and rises for some distance into the bulb curving over to-
wards its side ; in this respect Mr. Kemp's apparatus does not differ
materially from mine. I found it necessary however to have a third
bulb (D) blown (which is best of a spheroidal form), in order to ef-
fectually prevent the acid from being drawn into the tube G when-
ever a sudden absorption took place ; this third bulb communicates
with C by a narrow neck. If the apparatus be constructed without

Intelligence and Miscellaneous Articles. 1 57

the third bulb D, a portion of fluid generally passes into the tube G
from the rotary motion induced in the fluid in C.

The dotted lines indicate the height of fluid in the bulbs, and this
quantity is quite sufficient for the condensation of all the ammonia
likely to be formed. I would remark, that if during the progress of
the combustion a cessation of the production of gas should occur,
the construction of the apparatus is such as to prevent the whole of
the acid ever being carried over into the bulb C, so that on the evo-
lution again commencing no fear need be entertained for the com-
plete condensation of the ammonia.

It only remains for me to add, that though this new form of ap-
paratus is not so readily rinsed out as the original one of Messrs.
Varrentrapp and Will, no great inconvenience is experienced from
that cause, as the acid can, at the close of the operation, be easily
caused to flow into the bulb C and out at the tube G, by properly
inclining the bulbs, &c., and when this is done water or alcohol may
be introduced by a pipette through the limb H.

From the Proceedings of the Chemical Society.

ON THE DETECTION OF COTTON IN LINEN. BY G. C. KINDT.

This subject has frequently engaged the attention of commercial and
scientific men ; many experiments have been made in order to detect
cotton thread in linen ; many processes have been recommended, but
none have hitherto proved satisfactory. I was therefore much sur-
prised when a stranger, a few weeks ago, showed me a sample of
linen from the one-half of which all the cotton filaments had been
eaten away. He had obtained it in Hamburg, and asked me whether
I could give him a process for eflfecting this purpose. Now since,
as far as I am aware, nothing has been published on this subject,
and it is of very general interest, I consider it a duty to communi-
cate the results of my experiments. I had already observed, in ex-
perimenting with explosive cotton, flax, &c., that these two sub-
stances behave somewhat diflferently towards concentrated acids ;
and although it has long been known that strong sulphuric acid con-

158 Intelligence and Miscellaneous Articles^

verts all vegetable fibre into gum, and when the action is continued
for a longer period, into sugar, I found that cotton was metamor-
phosed much more rapidly by the sulphuric acid than flax. It is
therefore by means of concentrated sulphuric acid that cotton may
be removed from linen when mixed with it ; and this object may be
obtained by the following process : —

The sample to be examined must be freed as perfectly as possible
from all dressing by repeated washing with hot rain- or river-water,
boiling for some length of time, and subsequent rinsing in the same
water; and I may expressly observe, that its entire removal is
requisite for the experiment to succeed. When it has been well-
dried, the sample is dipped for about half its length into common
oil of vitriol, and kept there for about half a minute to two minutes,
according to the strength of the tissue. The immersed portion is
seen to become transparent. It is now placed in water, which dis-
solves out the gummy mass produced from the cotton ; this solution
may be expedited by a gentle rubbing with the fingers ; but since
it is not easy to remove the whole of the acid by repeated washing
in fresh water, it is advisable to immerse the sample for a few in-
stants in spirits of hartshorn (purified potash or soda have just the
same eifect), and then to wash it again with water. After it has
been freed from the greater portion of the moisture by gentle press-
ure between blotting-paper, it is dried. If it contained cotton, the
cotton threads are found to be wanting in that portion which had
been immersed in the acid; and by counting the threads of the two
portions of the sample, its quantity may be very readily estimated.

If the sample has been allowed to remain too long in sulphuric
acid, the linen threads likewise become brittle, or even eaten away ;
if it were not left a sufficient time in it, only a portion of the cotton
threads have been removed ; to make this sample useful, it must be
washed, dried, and the immersion in the acid repeated. When the
tissue under examination consists of pure linen, the portion im-
mersed in the acid likewise becomes transparent, but more slowly
and in a uniform manner, whereas in the mixed textures the cotton
threads are already perfectly transparent, while the linen threads
still continue white and opake. The sulphuric acid acts upon the
flax threads of pure linen, and the sample is even somewhat trans-
parent after drying as far as the acid acted upon it, but all the
threads in the sample can be seen in their whole course.

Cotton stuffs containing no linen dissolve quickly and entirely in
the acid; or if left but one instant in it, become so brittle and
gummy that no one will fail to recognise it as cotton when treated
in the above manner. — Liebig's Annalen, Feb. 1847.

THE PLANET HEBE*.

On July 1, M, Henke of Driessen in Prussia, discovered another
planet, which appears to belong to the singular group lying between
the orbits of Mars and Jupiter. It was first observed accurately at

• Communicated by J. R. Hin(|, E8q.,F.R.A.S.

Meteorological Observations.

159

Berlin by Prof. Encke on July 5, and since that date observations
have been made very generally at the different European observato-
ries. The following are the elements according to different calcu-
lators :— -

Galle and d' Arrest.

Neumann.

H*nd.

Epoch.

July 10-0 Berlin.

July 8-41864 Berlin.

July 00 Greenwich.

Mean anomaly

Long perihelion

^L^scending node

Inclination

268 5.5 50-6
19 4 14-9

139 5 31
14 88 58-5
10 41 16-7
0-3772450

283 9 4^-6
9 3 9-6

138 12 16-2
14 49 536
13 5 48-2
0-3955266

283 56 54-0
8 17 24-1

137 25 351
15 2 561
13 49 200
0-4016899

Sin-i e

Log. semi-axis major

The longitudes in first and second set are counted from M. Equi-
nox of 1847-0 ; in the third set from M. Equinox of July 0.

METEOROLOGICAL OBSERVATIONS FOR JUNE 1847-

Chiswick. — June I — 3. Clear and very fine. 4. Light clouds and fine. 5.
Cloudy. 6. Light clouds : clear, 7. Clear : cloudy. 8. Rain : thunder-showers.
9. Clear and fine. 10. Rain: cloudy: clear. 11, 12. Clear and very fine.
18. Rain : cloudy. 14. Densely clouded : showery. \5. Rain: thunder and
heavy shovt^ers. 16. Cloudy : rain. 17,18. Rain. 19. Cloudy and fine. 20.
Cloudy: slight showers. 21. Cloudy: fine. 22. Very fine. 23. Very fine :
heavy showers, with thunder. 24. Cloudy and fine. 25. Rain : cloudy and
fine. 26. Very fine. 27. Drizzly : cloudy and fine, 28. Fine. 29. Very fine,
30. Light clouds : very fine : overcast.

Mean temperature of the month 58°*46

Mean temperature of June 1846 66 '63

Mean temperature of June for the last twenty years 66*90

Average amount of rain in June l-88inch,

^os<o«.— June 1 — 4. Fine. 5, 6. Cloudy. 7. Fine, 8. Fine : rain early a. m*
9. Fine. 10. Cloudy: rain early a.m. : showery all day, 11,12. Fine. 13.
Cloudy: rain early a.m. 14. Cloudy : rain early a.m. : rain p.m. 15. Fine;
rain P.M. 16. Fine: rain a.m. and p.m. 17. Fine. 18. Cloudy: rain early
A M. : heavy rain p.m. 19. Cloudy : rain early a.m. 20. Cloudy : rain a.m. and
p.M, 21. Cloudy: rain p.m. 22,23. Fine: rain p.m. 24. Rain: rain p.m.
25. Fine : rain p.m. 26. Fine. 27, Cloudy. 28, Fine. 29, 30. Cloudy. —
This month has been the coldest since 1843, and the wettest since June 1841.

Sandwick Manse, Orkney. — June I, 2. Clear : fine. 3. Cloudy: fog. 4. Bright :
cloudy. 5. Showers : cloudy. 6. Bright : cloudy. 7. Showers. 8. Bright :
drops. 9. Cloudy: rain. 10. Showers: sleet-showers, 11. Bright: cloudy.
12. Cloudy. 13. Cloudy : rain. 14. Rain : damp. 15. Cloudy : rain : cloudy.
16. Cloudy: fine. 17,18. Bright : fine. 19. Clear : fine, 20, Bright : rain.
21. Showers: clear. 22. Bright : showers : fine. 23. Bright: showers. 24.
Bright : thunder : drops. 25. Bright : thunder. 26. Clear: fine, 27. Damp.
28. Cloudy. 29. Fog : cloudy. .30. Damp : fog.

Applegarth Manse, Dumfries -thire. — June 1^—3. Very fine. 4. Warm, but
overcast.' 5. Fair a.m. : showers p.m. 6. Fair a.m. 7. Threatening : rain p.m.
8. Slight'shower, 9. Fair : thunder : rain. 10. Fair : clear. 11. Fair, but cool.
12. Cloudy: rain P.M. 13. Rain. 14. Fine: thunder: rain. 15. Drizzly:
thunder. 16. Bright a.m. : rain. 17. Drizzly. 18. Fair and fine, 19. Fine:
a few drops. 20. Rain p.m. 21. Wet a.m. : cleared, 22. Showery. 23. Fine,
very : slight shower, 24, Showery : thunder. 25. Showers a.m. : thunder. 26.
Slight shower P.M. 27. Shower a.m. : fair. 28—30. Very fine.

Mean temperature of the month 55°'2

Mean temperature of June 1846 63*2

Mean temperature of June for 25 years 56 '10

Mean rain in June for 20 years , 2*32 inches.

^-3

2 3

•« ea
« 3

<i o

s §

•3lDlil8!ll3

•JIoiJipuBg

•aiiqs
-saujoinci

•5joiMsiq3

co»c : -^ CO — 00 ■<* — rt Tt — •^ o vo Ti< ^o

7>7fi :>;-07<0— i-Jt^OOOO-^iT'Oi^

OOOOOC«OOI'7<-7iOOOO

: to CI lo >-< — '
; coo o o o

^ 01 B S S C

0) & fe « g ^

" a '^ S2 S S

^ ^ : aJ

^ ^'

s' ^

i i ^ i ^ ^ " ^ ^

I ^ ^ ^

C cn c/3 {/I

fe fc ^'

V C

« « ^' S S c c -3

^-

. . . s s . .

a fe a
2 ^^'

• a •

q S o

2 oJ aJ 2 aj J aJ ^' (i ^ > ^* g:* g: 6- B: ^' j-' ^ ^ B= fe" ^ ^ fe 6t ^* aJ <u aJ

•ta'd
?8

•uiKT

S2

\-xeifi

•uojsoa

CO'iO C« O t^-QO t>.00 00^Oa\-<--^<yi'^C0dlCOOdiO<NlOO>O'OtC'^

«llM HS «I(M «!« mIo Hd h|s ^IS «]«

0^■^C^cooo^»0^a^OO^(M^OOO<^^coco^o'o — -"t^O-^-^-^io-HOOovo

"I« «I« rtldi-tel

<M t^C^fS VO -H 0\(0 CI O O 00 O O OMO t^O TfCO COOO (N

-^ CO -^ -^ ^^ lO -^ ■^ •^ lO IC ■<* lO IT) ■^ rt Tt lO Tf in ITS lO IfS

,,,,,,, '^l« i-tlW "^IN m1« f^«

a\ o 00 lo 00 lo ON (>i o lo «N vo — < o^ r^ >-4 o^ — CO r^ CO •^ CO

IC O lO IC lO LfS »0 to <0 lO ^ <0 lO lO m lO lO VO lO lO !> !>. «>.

ip ip ip ip »p

OM3NONC0— a>t--'^inTtibd o — o co'o6>qo ctn— ' co-^'-icsincs lO'-" (m

lOiO»0^'0»OiOlOlOir>»0'£>*OVOlO>0>OlO>OlO>O^OVO^OlO>OlOO*OlO

•u'W

•XBH

Hwg

-Wg

O0000OM3^OO'-<l>-(TiQ0r-CClt>.O00MO<OOON'-'C<l«N(yi0000OO\

^fcooi <ji o oop ir~"p r^op ip coco-^'^ipr^op ip>pco'^co« oi^ (n cooi
O O O O O O O^O^C^O^O^O^O^O^O^O^C^O^C^O^O^C^O^O^O^O^O O O O
COrorocOCOCO<NOIOld<M01CM<M«M01CI<NO)OI<N(MOIO»<N(NCOCOCOCO

«MO-^O'»:l<0^C>(MO00'^OOOt0OmiCQ0i0O-i«MOI:^«M(N0MCCNO

■^^cococ-i o o^ioop ipopcp >pco(N kp-^ip ir-i>--<a<-j)'Tj<ro-«i'c» o o» oi co
O O O O O O 0^0^©^O^o^O^O^O^CT^o^0^0^0^6^0^o^o^o^0^6^0 O o O
COCOCOCOCOCOOI<N(NOIdCI<NCNOICNOICNO<<M<N<NtN<N<N(NCOCOcOO

•WB ?8
"UOJSOa

vomdr-ooc^io-^co-^o-'a'ONtNoodaicooO'^iMooooTfooNCMO^oo
ONON(»r^r--ip>pco'^?<ip>pc<>^o07<c^co'^cO'^';H— ic<-«a<r^i^

o^o^0^o^o^c^o^0^o^o^o^o^o^o^o^o^O^O^O^C^o^c^O^O^O^c^^c^

<M<N010J<M<N<NOICNtNCM<NOIOI<N<M<NClCJ«NCNCNOI<N<NOI(NO)(M<M

^i

SOO (N
NCOCN
XN CI

. 6 6

)COCO

00 >o
lO-'S'

— o

00 •^vo t^

O lO

o^oo

COOM3>IO
CPiO O «
ip ip ip ip

CTi o^ cr\ ON

CI d d CI

lOO^
t--in

ON-<*

CO CI

6 6

coco

O CO
d o

00 lO IC -^
O co*0 lO
^ 00 00 00

o 0\
6 d>

cOd

00 00 -< 00

t^io d in

C--0 t^t^

•«* o 00 d -rf

-H Tf -^lO d

O d d d d

d O IC t^d

ON*© *0 lO I>.

-< d d d d

6 6 6 6 6

CO CO CO CO CO

JO SiiBQ

F^C(C0TfiO(Ot^00ONO'-<dC0'*imot^00O^O--dC0'<!j'lrtlOI-^00O\O

,H,-li-<rtr-<F-^.-lr-lpHdddddClddddCO

te* O (^ o

THE
LONDON, EDINBURGH and DUBLIN

PHILOSOPHICAL MAGAZINE

AND

JOURNAL OF SCIENCE.

[THIRD SERIES.]

SEPTEMBER 184-7.

XXIX. On certain Products of Decomposition of the Fixed
Oils in contact xvith Sulphur. By Thomas Anderson, Esq.^
31. D.) F.R.S.E., Lecturer on Chemistry, Edinhirgh^.

]VrUMEROUS researches have established as a general rule
-'- that the products of the decomposition of organic sub-
stances vary with the circumstances of the experiment, and the
nature of the agents under the influence of which it is performed.
If, for instance, we examine the action of heat alone, we find it
causing a set of decompositions specially characterized by the
evolution of carbonic acid, formed by the union of partof tlie
carbon of the substance with the whole or part of its oxygen;
and this action is rendered more definite, and the number of
the products circumscribed by all circumstances facilitating
the formation of carbonic acid, such as the presence of abase,
which will even cause its evolution when heat alone is inca-
pable of producing decomposition. Acids, on the other hand,
have a precisely opposite effect ; they, in some instances, alto-
gether prevent the formation of carbonic acid, and cause the
oxygen to exert its action on ihe hydrogen of the compound,
and to eliminate one or more atoms of water which do not
generally exist ready formed in it.

In these particular instances, decomposition takes place at
the expense of the constituent atoms of, the compounds them-
selves, the extraneous substances serving merely as disponents
to the oxidation; in the one case of part of their carbon, in
the other of their hydrogen. But there is another class of
agents, which, besides eliminating one or more substances, are
capable at the same time of entering into union with the resi-
dual atoms, and forming a new derivative of the original com-
pound. The best investigated of this class of agents are chlo-

* Read before the Royal Society of Edinburgh on the 19th of April,
1847, and published in their Transactions, vol. xvi. part 3, p. 363.
Phil. Mag, S. 3. Vol. 31. No. 207. Sept, 1 847. M

162 Dr. T. Anderson on certain Products of Decomi^o&ition

rine, bromine, nitric acid and ammonia; the three former of
which exert their action on the hydrogen, the latter on the
oxygen of the substance, and form compounds, the complete
investigation of which is important, not merely in a purely
chemical point of view, but also from the light which they
seem likely to throw on the general question of the atomistic
constitution of matter. In fact, the great object of the re-
searches of organic chemistry at the present moment is that
of developing the relations which the individual atoms bear
to the molecules of their compound, by a knowledge of which
we hope eventually to arrive at some definite conclusions with
regard to the mode in which the elementary atoms are grouped
together in a complex molecule. Almost, all the scanty in-
formation which we possess on this subject has been derived
from investigating the products of the action of different agents
upon organic substances ; and it is sufficiently obvious, that
the more varied the circumstances, and numerous the points
of view under which these reactions can be examined, so much
the more likely are we to arrive at definite results.

It was the consideration of these points which led me to
undertake an investigation into the nature of the action of
sulphur in the free state upon organic compounds, a subject
hitherto totally uninvestigated, unless we except the curious
researches of Zeise* on the simultaneous action of ammonia and
sulphur upon acetone, which yields a variety of remarkable
products, the properties of which he has described, without
however determining their constitution. The results at which
I have already arrived in these researches are contained in
the following pages. They are, however, to be considered
only as the commencement of the investigation ; and I am
desirous of submitting them to the Society even in their present
very imperfect state, as it is impossible to fix a period within
which a series of researches, surrounded by so many difficul-
ties, can be completed. No one who has not been specially
occupied with such experiments can have any conception of
tlie numerous sources, of annoyance which they present, and
of the expenditure of time and labour which is necessary for
their performance. Indeed, I have more than once felt in-
clined altogether to abandon a subject occupying so much
time in proportion to the results obtained, and the completion
of which is further protracted by the nauseous odour of the
compounds, which is so disgusting that it is impossible to
pursue the investigation for any length of time continuously.

At the commencement of these researches I endeavoured to

examine the action of sulphur upon some of the simpler

* Forhandlingar vid de Skandinaviska Naturforskarnes trcdje miite^ p. 303.

of ike Fixed Oils in. contact with Sulphur. 163

organic compounds, in the hope of arriving at results of cor-
responding simplicity. My expectations, however, vfere dis-
appointed, and I was obliged to have recourse to the fixed
oils, on which sulphur has been long known to exert an action;
the product obtained by heating together olive oil and sulphur
until a uniform balsam-like substance was formed, having
been employed in medicine by the older physicians under the
name of the balsam of sulphur.

The phaenomena which manifest themselves during the
mutual action of sulphur and a fixed oil are these : — At the
first application of heat, the sulphur melts and forms a stratum
at the bottom of the oil ; but as the temperature rises it slowly
dissolves, with the formation of a thick viscid fluid of a dark
red colour. As the heat approaches that at which the oil
undergoes decomposition when heated alone, a violent action
takes place attended by the evolution of sulphuretted hydrogen
in such abundance, that the viscid mass swells up and occupies
a space many times its original bulk. If at this point the
mixture be allowed to cool, it concretes into a tough sticky
tenacious mass, adhering strongly to the fingers, and having
a disagreeable sulphureous odour; if however the heat be
sustained, the frothing and evolution of sulphuretted hydrogen
continue, and at the same time an oil of a peculiar disgusting
odour, resembling that of garlic, but more disagreeable, passes
into the receiver.

In the investigation of the products of this action, the first
and most essential step was to determine the particular con-
stituents of the oil from which they are derived. In order to
do this, it was necessary to examine separately the action of
sulphur upon each of its components. I commenced therefore
by making use of stearic acid, which can be readily obtained
in a pure state : experiment however showed that none of the
peculiar products were derived from it; for when mixed with
half its weight of sulphur and distilled, mere traces of sulphu-
retted hydrogen were evolved, and the products were identical
with those obtained from the unmixed acid. The nauseous
smelling oils being then obviously derived either from the
oleic acid, or the glycerine of the oil, I prepared a quantity
of pure oleic acid, by the decomposition of the aethereal solu-
tion of the oleate of lead. This, when mixed with half its
weight of sulphur, and distilled in a capacious retort, under-
went decomposition precisely as the crude fixed oil did ; sul-
phuretted hydrogen was developed in great abundance, and
the product of the distillation could not be distinguished from
that which I had previously obtained. I was unable to obtain
glycerine in sufficient qunntij.y to make a separate investiga-

M 2

164; Dr. T. Andersoh on certain Products of Decomposition

tion of the products of its decomposition ; but these must also
be peculiar, as I could not distinguish the presence of acroleine
during any period of the distillation of an oil with sulphur.

The product of the distillation of oleic acid was in the form
of a reddish-brown oil, having an extremely nauseous odonr,
in which that of sulphuretted hydrogen was apparent. When
rectified, this sulphuretted hydrogen was driven off, and the
first portions which distilled were perfectly transparent and
colourless. As the process continued, however, the products
became gradually darker in colour, and the last portions which
distilled became semisolid on standing, from the deposition of
a quantity of wiiite crystalline plates. These were separated
by filtration through cloth, expressed strongly, and purified
by successive crystallizations from alcohol, until they were
entirely free from smell and colour. The product was then
in the form of white pearly scales, which possessed acid pro-
perties, and were totally insoluble in water; they were not
therefore sebacic acid, no trace of which could be discovered
among the products, but, on the contrary, possessed all the
properties of margaric acid. These crystals were obtained
from quantities of oleic acid, prepared at different times, and
with the greatest possible care, and must have been formed
during the decomposition. In order however to set this point
at rest, some of the same oleic acid was distilled alone, when
abundance of sebacic acid was obtained, and the latter portions
of the rectified product did not deposit any crystals on cooling,
but remained perfectly fluid. As this solid acid is produced
only in comparatively small quantity, and I was unable to
obtain enough of oleic acid, I made use, in preparing it on the
large scale, of pure almond oil, which, according to Schiibler
and Gasserow, is entire!}' free of margarine. The oil which
I employed was expressed specially for these experiments, at
a temperature slightly above 32°; and in order to satisfy
myself of the absence of margaric acid in the products of its
ordinary decomposition, a quantity was distilled alone, and
the product rectified. The latter portions being collected
apart did not deposit margaric acid ; and this 1 have also
found to be the case with the ordinary almond oil of com-
merce, in the expression of which a moderate degree of heat
is employed.

In distilling the oil and sulphur on the large scale, it be-
came impossible to perform the process by the simple admix-
ture of the substances, the frothing being so great as inevitably
to expel the materials from the retort. After a trial of various
methods, I found it most convenient to employ the apparatus,
of which this is a sketch. The oil was introduced into a large

of the Fixed Oils in contact with Sulphur.

165

glass balloon, to the mouth of which two tubes were adapted,
one descending to near the middle, and furnished with a cork

at the upper end, the other which constituted the neck of the
distilling apparatus passed into a tubulated receiver, kept cold
by immersion in water or ice. To the tubulature, a doubly
bent tube was affixed, which descended into a vessel of alcohol,
for the purpose of retaining any of the more volatile portions
which might be carried over by the rapid current of sulphu-
retted hydrogen. The heat must be applied by means of an
open charcoal fire ; and the furnace should be so constructed
that the fire may be rapidly withdrawn in the event of the
action becoming too violent. It is very desirable too that the
balloon should go down into the furnace, so that it may be
entirely surrounded by hot air. The oil is introduced into
the balloon, of which it must not occupy more than a fifth, or
a fourth at most, along with a few small pieces of sulphur, and
heat is gradually applied. So soon as effervescence com-
mences, the cork of the small tube is withdrawn, and a small
piece of sulphur is introduced ; and this is continued gradually,
so as to keep up a uniform action. A dark reddish-brown
oil passes into the receiver, and at the same time sulphuretted
hydrogen passes in torrents through the alcohol ; it there
deposits a certain quantity of oil, and on escaping, may be
kept burning during the whole operation, with a flame eight
or nine inches high. The principal difficulty of this process
consists in regulating the heat, so as to keep up a steady action.
If the heat be allowed to fall, the contents of the balloon be-
come so viscid as inevitably to boil over ; and at the same time
too high a temperature causes the whole action to go on with

166 Dr. T. Anderson on certain Products of Decomposition

excessive violence. I have generally operated on quantities
of three pounds, each of which requires a complete day for
its distillation, during vv^hich time the operator must never
leave it, but constantly attend to the regulation of the heat,
and the gradual addition of sulphur in small quantities. When
a quantity equal to about half the oil employed has distilled
over, the remaining mass becomes excessively viscid ; and just
at this point the balloon frequently cracks, the contents escape,
and the whole catches fire, and blazes off with a bright yel-
low flame and smell of sulphurous acid.

The product of this distillation, which exactly resembled
that of the pure oleic acid, was rectified, and the crystals which
deposited from the latter portions were expressed and purified ,
by successive crystallizations in alcohol. They then presented
all the characters of margaric acid, and gave the following
results of analysis : —

{5*275 grains of the acid gave
14-*558 ... carbonic acid, and
5-919 ... water.

r 6*358 grains of the acid gave
II. -< 17*578 ... carbonic acid, and
L 7*212 ... water.

Which gives the following results per cent.: —

Experiment.

Calculation.

A

(

A

( ,

^

>

I.

11.

Carbon

75-27

75*40

75-55

^34

2500-0

Hydrogen

12*51

12-66

12*59

H34

425-0

Oxygen .

12-22

11*94-

11*86

O4

400-0

100-00 100-00 100-00 3325-0

These results agree completely with the formula for margaric
acid, and were further confirmed by the analysis of its silver
salt and aether.

4*643 grains of the silver salt gave 1*325 of silver =28*53
per cent.

7*926 grains of the silver salt gave 2*284 of silver =28*70
per cent.

The calculated result for margarate of silver gives 28*65
per cent.

The aether was prej:>ared in the usual manner, by dissolving
the acid in absolute alcohol, and passing dry hydrochloric
acid gas through the solution. The product, which possessed
all the properties of margaric (ether, gave the following re-
isults of analysis: —

of the Fixed Oils in contact with Sulphur. 167

{5-596 grains of the asther gave
15*662 ... carbonic acid,
6-399 ... water.

Experiment. Calculation.

Carbon
Hydrogen
Oxygen .

. 76-33
. 12-70
. 10-97

76-51
12-74
10-79

C38

O4

2850-0
475-0
400-0

100-00 100-00 3725-0

Tiiese analyses establish, in a satisfactory manner, that the
acid produced was margaric acid. It is scarcely possible how-
ever, in the present state of the investigation, to give anything
like a rational explanation of the mode in which it is here
formed. Its production from oleic acid has been already
observed by Laurent as the first product of oxidation by nitric
acid; but the action of sulphur is certainly of a, very different
character, and cannot be considered as bearing any analogy
to that of an oxidizing agent. The quantity of margaric acid
produced does not appear to be constant, but varies with the
rapidity of the distillation, and is always most abundant when
it is slowly performed.

The oil which distils previous to and along with the mar-
garic acid, and constitutes by far the most abundant product
of the action of sulphur upon oleic acid and oil of almonds, is
a very complex substance, and contains some of its constituents
in very small proportion. On this account I found it neces-
sary to prepare it in very large quantity; and in doing so I
abandoned the use of almond oil and employed linseed oil
instead, which is a much cheaper substance, and yields the
same fluid products. When the product of the action of sul-
phur is carefully rectified, the first portions which pass over
are perfectly transparent and colourless, highly limpid and
mobile, and boil at the temperature of 160 Fahr. Only a
small quantity however passes at this temperature, and the
immersed thermometer gradually rises without indicating any
fixed boiling-point for the fluid. My first attempts to purify
this oil and separate it into its various constituents, did not
afford any satisfactory conclusions. Numerous analyses of
the more volatile portions were made without obtaining com-
parable results, although all indicated the presence of carbon
and hydrogen nearly in the proportion of equal atoms. The
following are the details of three of these analyses : —

f 4-657 grains of the most volatile oil gave
I. ■< 12-688 ... carbonic acid, and

I 5-127 ... water.

168 Dr. T. Anderson on certain Products of Decomposition

{5*50 J grs. of an oil less volatile than the preceding gave
15*762 ... carbonic acid, and
6*292 ... water.

r 4*191 grains of another portion of oil gave
III. < 12*185 ... carbonic acid, and
L 4*720 ... water.

Which correspond to the following results per cent. : —
I. II. III.

Carbon . . 75*03 78*79 79*95

Hydrogen . 12*20 12*72 12*75

All these oils, when treated with fuming nitric acid, yielded
an abundant precipitate of the sulphate of barytes; but as the
results of the combustion were not constant, no quantitative
determination was made.

The action of precipitants however upon this oil afforded
a more satisfactory method of obtaining some of its consti-
tuents. It gives with corrosive sublimate a bulky white pre-
cipitate, and with bichloride of platinum a yellow compound,
the characters of which vary slightly, according as it is pre-
pared from the more or less volatile portion of the oil. Ni-
trate of silver and acetate of lead, mixed with the alcoholic
solution of the oil, produce only a slight cloudiness, but on
boiling the solutions, the sulphurets of silver and lead are de-
posited.

The Mercuri) Compound, — In order to obtain this substance
in the pure state, the oil was dissolved in alcohol, and an
alcoholic solution of corrosive sublimate added. The preci-
pitate which fell was collected on a filter, and washed with
lEther until the oil was thoroughly extracted, for which pur-
pose a considerable quantity of aether is required. It is then
boiled with a large quantity of alcohol, which dissolves a part
of it ; and the solution being filtered hot, allows the compound
to deposit, on cooling, in the pure state. It is then in the
form of a white crystalline powder, having a very fine pearly
lustre, and exhibiting under the microscope crystals of a very
peculiar form. They are six-sided tables, two opposite angles
of which are rounded off, so as to give them a very close re-
semblance to the section of a barrel. It possesses, even after
long-continued washing with aether, a peculiar slight sickening
smell, which becomes more powerful on heating, and its pow-
der irritates the nose. It is insoluble in water, which moistens
it with difficulty. It requires several hundred times its weight
of boiling alcohol for solution, and is almost entirely deposited,
on cooling, in microscopic crystals. In aether it is almost in-
soluble. When heated, it is decomposed with the evolution

of the Fixed Oils in contact with Sulphur,

169

of a peculiar nauseous smelling oil. The sparing solubility
of this compound in alcohol renders its preparation in sufficient
quantity for analysis an extremely tedious process, and I have
sought in vain for a more abundant solvent. The only sub-
stance which I have found capable of taking it up in larger
quantity is coal-tar naphtha; but its employment is inadmissible,
as the best which can be procured is an extremely impure
substance, and the crystals of the compound deposited from
it always acquire a rose or violet tint from some of its impu-
rities. Oil of turpentine likewise dissolves it, but not more
abundantly than alcohol.

By many successive solutions in alcohol, I obtained enough
of this substance for an analysis, of which the following are
the results : —

r 12*302 grains, dried in vacuo, gave
< 6'592 ... of carbonic acid, and
L 3-018 ... ofwater.

8*061 grains deflagrated with a mixture of nitre and car-
bonate of soda, gave 7*297 grains of sulphate of baryta =
1-0067 = 12*48 per cent, of sulphur.

The mercury and chlorine were determined together by
mixing the substance with quicklime, and introducing the
mixture into a combustion-tube. The end was then drawn
out into an elongated bulb, into which the mercury sublimed,
and which was afterwards cut off, dried in the water-bath,
and weighed, both with and without the mercury; the chlo-
rine was determined in the usual way from the residue in the
tube.

9*958 grains gave 5'976 mercury =60*01 per cent., and
4'*310 grains chloride of silver =10-67 per cent, of chlorine.

5*797 grains gave 2*409 of chloride of silver = 10*25 per
cent, of chlorine.

These results correspond closely with the formula C^q Hjg
S,^ Hg4 CI2, as is shown by the following comparisons : —

Experiment.

Calculation.

-T-^

11.^

A

(

1

Carbon .

14*61

...

14*46

C16 1200*0

Hydrogen

2*72

...

2*42

His 200-0

Mercury .

60*01

...

60*32

Hg4 5003*6

Chlorine .

10*67

10*25

10*67

CI2 885*3

Sulphur .

12*48

...

12*13

S5 1005*8

100*49

100*00

8294*7

It is sufficiently obvious that the formula O^q Hig S5 Hg4 Clg

170 Dr. T. Anderson on certain Products of Decomposition

cannot be supposed to represent the rational formula of
this substance. On the contrary, the remarkable analogy
between its properties and those of the mercury compound of
sulphuret of allyle appear clearly to indicate a similarity in their
chemical constitution, — a similarity which, as we shall after-
wards see, is borne out by the properties of the platinum com-
pound. I consider this substance to contain an organic sul-
phuret analogous to sulphuret of allyle, the constitution of
which must be represented by the formula Cg Hg Sg, to which
I give the provisional name of sulphuret of odmyl (from oS/aij,
odow)^ and that the rational formula of the mercury com-
pound is —

(Cg Hg S,+ Hg, CI,) + (Cg Hg S, + Hg, S).

On contrasting this with the formula of the allyle compound,
which is —

(Cfi H5 CI + Hg, CI,) + (Ce H5 S + Hg, S,), .

two important points of difference are apparent, namely, that
in the new compound we have the sulphuret, and not the
chloride, of the base in union with corrosive sublimate, and
the presence of subsulphuret in place of sulphui-et of mercury
in the second member of the compound. It is even possible
to approximate more closely the formulae of the allyle and
odmyle compounds, by assuming the sulphuret of odmyle to be
represented by C4 H4 S; in which case the oiercury compound
becomes —

{3(C4 H4 S) + Hg, S,} + (C4 H4 CH- Hg, CI).

This formula is however incompatible with its reactions, as
it involves the presence of calomel in the compound. Treat-
ment with caustic potash however shows that this is not the
case, as it immediately becomes yellow, from the separation
of oxide of mercury, while the black suboxide would have
been formed had calomel been present.

When a current of sulphuretted hydrogen is passed through
the mercury compound suspended in water, it becomes rapidly
black, a peculiar smell is observed, along with that of sulphu-'
retted hydrogen ; and by distillation an oil passes over, which
is obtained floating on the surface of the water. It is per-
fectly transparent and colourless. Its smell is peculiar, and
resembles the nauseous odour developed by crushing some
umbelliferous plants. When dissolved in alcohol, it gives
with corrosive sublimate a white precipitate, soluble in hot
alcohol, from which it is deposited in crystals precisely similar
to those from which it had been originally separated, and with
bichloride of platinum a yellow precipitate, slightly soluble in

of the 'Fixed Oils in contact mth Sulphur, 1*71

hot alcohol and aether. This oil is in all probability the
sulphuret of odmyle Cg Hg Sg j but the small quantity in which
I have been able to obtain it, has prevented my performing
any analysis of it.

The Flaiinum Compound. — When a solution of bichloride
of platinum is added to the alcoholic solution of the crude oil,
a yellow precipitate makes its appearance, which does not fall
immediately, but goes on gradually increasing for some time,
precisely as is the case with the allyle compound. The pro-
perties of this precipitate are not however perfectly constant,
but vary according to the portion of the oil employed to yield
it. That obtained from the more volatile portion has a fine
sulphur-yellow colour, but the less volatile oil gives an orange
precipitate. It is insoluble in water, sparingly soluble in
alcohol and aether. When heated it becomes black, an oil is
evolved smelling exactly like that obtained from the mercury
compound, and sulphuret of platinum is left behind, which
requires a high temperature to drive off all its sulphur, and
leaves metallic platinum as a silver-white mass. When treated
with hydrosulphuret of ammonia, it is converted into a brown
powder, exactly like that obtained under similar circumstances
from allyle.

The analysis of the yellow compound has not hitherto given
results of a satisfactory character. I have found the amount
of platinum to oscillate between 43'06 and 49*66 per cent.
The former of thes'e was obtained from the most volatile oil,
the latter from that which boiled between 300° and 400° F.,
and intermediate results were obtained at intermediate tem-
peratures. T!ie results obtained from the oil which boiled at
a high temperature were remarkably constant; thus I have
found, in different experiments, 49*00, 49*51, and 49*66 per
cent, of platinum, which appear to indicate the presence of
some compound of rather sparing volatility. The precipitate
obtained from the most volatile oil appears to be that corre-
sponding to the mercury compound which has just been de-
scribed. Of it I have been able only to perform a very incom-
plete analysis, which is insufficient to establish its constitution,
especially as it is impossible to ascertain whether it is a homo-
geneous substnnce. As the results, however, approximate to
a formula analogous to that of the mercury compound, I give
the details, such as they are.

["9*155 grains of the platinum compound gave
■< 7*474 ... carbonic acid, and
1^3*294 ... water.

5*701 grains gave 2*455 grains of platinum = 43*06 per cent.

t72 On certain Products of Decomposition of the Fixed Oils.

These results approximate to a formula similar to that of
the mercury compound ; viz. —

(C8H8S2+PtCg + (C8H8S, + PtS).

Experiment.
. 22-26

Ci

20-83

alculation.

Carbon

Cjg 1200-0

Hydrogen

3-99

3-47

Hie 200-0

Platinum .

. 43-06

42-84

Pt2 2466-6

Chlorine .

. •••

15-38

CIg 885-3

Sulphur .

.

17-48

Ss 1005-8

100-00 5757*7

The analogy which those substances bear to allyle is exceed-
ingly interesting, as showing the possibility of forming, by
artificial processes, substances similar in constitution to so
remarkable a compound, which is not a product of decompo-
sition, but exists ready-formed in a variety of different vege-
tables, where it must obviously be produced under circum-
stances very different from the artificial substance ; for allyle
cannot exist at all at a high temperature, and is entirely de-
composed at, or even below, its point of ebullition. Unfor-
tunately, however, the examination of this substance is much
complicated by the necessity of examining its compounds in
place of itself. Had it been possible to separate it directly
from the crude oil, the determination of its constitution and
that of its compounds would have presented comparatively
little difficulty, and been arrived at with much less labour than
that expended upon the imperfect details I have been able to
accumulate. Another point worthy of observation, is the total
alteration of the products of decomposition of oleic acid pro-
duced by the presence of sulphur ; no sebacic acid, and, in
fact, none of its ordinary products being evolved, although all
the substances produced contain carbon and hydrogen in the
proportion of equal atoms, just as they exist among the ordi-
nary products, — a circumstance which, taking into considera-
tion the abundant evolution of sulphuretted hydrogen, we
certainly should not have anticipated.

The oil which remains after the separation of the mercury
compound, likewise contains sulphur as one of its constituents ;
but I have not yet had time to commence the investigation of
this part of the subject. The discussion of it, as well as va-
rious other points connected with the compounds already de-
scribed, I hope to make the subject of a future communica-
tion.

[ 173 ]

XXX. On the Mechanical Equivalent of Heat^ as determined
by the Heat evolved by the Friction of Fluids. By J. P.
Joule, Secretary to the Literary and Philosophical Society
of Manchester *,

IN the Philosophical Magazine for September 1845 I gave
a concise account of some experiments brought before the
Cambridge Meeting of the British Association, by which I
had proved that heat was generated by the friction of water
produced by the motion of a horizontal paddle-wheel. These
experiments, though abundantly sufficient to establish the
equivalency of heat to mechanical power, were not adapted to
determine the equivalent with very great numerical accuracy,
owing to the apparatus having been situated in the open air,
and having been in consequence liable to great cooling or
heating effects from the atmosphere. I have now repeated
the experiments under more favourable circumstances, and
with a more exact apparatus, and have moreover employed
sperm oil as well as water with equal success.

The brass paddle-wheel employed had, as described in my
former paper, a brass framework attached, which presented
sufficient resistance to the liquid to prevent the latter being
whirled round. In this way the resistance presented by the
liquid to the paddle was rendered very considerable, although
no splashing was occasioned. The can employed was of cop-
per, surrounded by a very thin casing of tin. It was covered
with a tin lid, having a capacious hole in its centre for the axle
of the paddle, and another for the insertion of a delicate ther-
mometer. Motion was communicated to the paddle by means
of a drum fitting to the axle, upon which a quantity of twine
had been wound, so as by the intervention of delicate pulleys
to raise two weights, each of 29 lbs., to the height of about 5^
feet. When the weights in moving the paddle had descended
through that space, the drum was removed, the weights wound
up again, and the operation repeated. After this had been
done twenty times, the increase of the temperature of liquid
was ascertained. In the second column of the following
table the whole distance through which the weights descended
during the several experiments is given in inches. I may
observe also that both the experiments on the friction of water,
and the interpolations made in order to ascertain the effect of
the surrounding atmosphere, were conducted under similar
circumstances, each occupying forty minutes.

* Read before the Mathematical and Physical Section of the British
Association at Oxford, and communicated by the Author.

174- Mr. J. P. Joule on the Mecha?iical Equivalent of Heat.
Table I.— Friction of Distilled Water.

Nature of
experiment.

Total descent of

each weight of
29 lbs. in inches.

Mean tem-
perature of
the room.

Difference.

Tempcrature of the
water.

Gain or
loss of
heat.

Before ex-
periment.

After ex-
periment.

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

1268-5
0

1266-1
0

1265-8
0

1265-4
0

1265-1
0

1265-3
0

1265-2

e

1262-4
0

1262-3
0

60°839
61-282

61-007
61-170

57-921
58-119

58-152
58-210

57-860
58-162

57-163
57-602

57-703
58-091

56-256

56-888

57041
67-612

0-040-
0120-

0-408+
0-570+

0-809-
0-628-

0-293-
0-003+

0-215+

0'25G+

0-220+
0-121 +

0-359+
0-304+

0-015-
0-285 -

0-078-
0-285-

60'-452
61-145

61-083
61-752

56-752
57-472

57-51 1

58-207

57-735
58-416

57-050
57-716

57-731
58-393

55-901
56-590

56-617
57-310

6f-145
61180

61-748
61-729

57-472
57-511

58-207
58-219

58-416

58-420

57-716
57-731

58-393
58-397

56-582
56-617

57-310
57-344

0-693 gain,
0-035 gain.

0-665 gain.
0-023 loss.

0-720 gain.
0-039 gain.

0-696 gain,
0012 gain.

0 681 gain.
0-004 gain.

0-666 gain,
001 5 gain,

0-662 gain.
0-004 gain,

0-681 gain.
0027 gain.

0-693 gain.
0-034 gain.

Mean friction
experiments

j 1265-13

0-0037-

0-6841 gain.

Mean of the
interpolations

} »

0-0071-

00163 gain.

Corrected re-

1 1265-13

0-6680 gain.

We see then that the weights of 29 lbs., in descending
through the altitude of 1265'13 inches, generated 0°-668 in the
apparatus. But in order to reduce these quantities, it became
necessary in the first place to ascertain the friction of the pul-
leys and that of the twine in unwinding from the drum. This
was effected by causing the twine to go once round a roller of
the same diameter as the drum, working upon very fine pivots,
the two extremities of the twine being thrown over the pul-
leys. Tlien it was found that, by adding a weight of 3150
grain.s to either of the two weights, the friction was just over-
come. The actual force employed in the experiments would
therefore be 40G000 grs. —3150 grs. = 402850 grs. through
1265*13 inches, or 6067*3 lbs. through a foot.

The weight of water being 77617 grs., that of the brass
paddle-wheel 24800 grs., the copper of the can 11237 grs.,

Mr. J. P. Joule on the Mechanical Equivalent of Heat. 175

and the tin casing and cover 19396 grs., the whole capacity
of the vessel and its contents was estimated at 77617 + 2319
+ 1056-4-363 = 81355 grs. of water. Therefore the quantity
of heat evolved in the experiments, referred to a pound of
water, was 7°-7636.

The equivalent of a degree of heat in a pound of water was
therefore found to be 781-5 lbs. raised to the height of one
foot.

I now made a series of experiments in which sperm oil was
substituted for the water in the can. This liquid, being that
employed by engineers as the best for diminishing the friction
of their machinery, appeared to me well-calculated to afford
another and even more decisive proof of the principles con-
tended for.

Table II. — Friction of Sperm Oil.

Nature of
experiment.

Total descent of
each weight of
29 lbs. in inches.

Mean tem-
perature of
the room.

Difference.

Temperature of the
oil.

Gain or
loss of
heat.

Before ex-
periment.

After ex-
periment.

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

Friction

Interpolation

1263-8
0

1269-0
0

1268-7
0

1268-5
0

1268-1
0

1268-3
0

1268-7
0

1267-6
0

12680
0

56-677
57-316

56198
56-661

57-958
57-051

58-543
57-153

59097
57-768

56-987
57-156

57-574
57-336

58-537
59-641

59-131
60-164

0-453 +
0-595 +

1-024+
1-221 +

0-588+
0-773+

1-685-
1-504-

0-534-
1-927-

0-186-
0-413+

0-734+
0-237+

0-829-
0-364+

0-148+
0-138-

56-354
57-906

56-516
57-929

57-813
57-836

55-951
55-568

57-766
55-731

56-029
57-573

57-581
57-565

56-884
60-026

58-532
59-984

57-906
57-917

57-929
57-836

59-280
57-813

57-766
55-731

59-361
55.951

57-573

57-565

59036
57-581

58-532
59-984

60026
60-069

1-552 gain.
0-011 gain.

1-413 gain.
0-093 loss.

1-467 gain.
0023 loss.

1-815 gain.
0-163 gain.

1-595 gain.
0-220 gain.

1-544 gain.
0 008 loss.

1-455 gain.
0-016 gain.

1-648 gain.
0-042 loss.

1-494 gain.
0-085 gain.

Mean friction
experiments

1 1267-85

0-034+

1-5537 gain.

Mean of the
interpolations

} "

0-004+

0-0366 gain.

Corrected re-
sult

j 1267-85

1-5138 gain.

176 Prof. Schoenbein on a tiew Test for Ozone.

In this instance, the force employed, corrected as before for
the friction of the pulleys, was equal to raise eOSO*^ lbs. to the
height of one foot.

In estimating the capacity for heat of the apparatus, it was
necessary in this instance to obtain the specific heat of the
sperm oil employed. For this purpose I employed the method
of mixtures. 43750 grs. of water were heated in a copper
vessel weighing 10403 grs. to 82°-697. I added to this 28597
grs. of oil at 55°" 593, and after stirring the two liquids
together, found the temperature of the mixture to be 76°'583.
Having applied to these data the requisite corrections for the
cooling of the liquids during the experiment, and for the capa-
city of the copper vessel, the specific heat of the sperm oil came
out 0*45561. Another experiment of the same kind, but in
which the water was poured into the heated oil, gave the spe-
cific heat 0'46116. The mean specific heat was therefore
0-45838.

The weight of oil employed was 70273 grains, and the
paddle, can, &c. were the same as employed in the first series
of experiments ; consequently the entire capacity in this in-
stance will be equivalent to that of 35951 grs. of water. The
heat evolved was therefore 7°'7747 when reduced to the ca-
pacity of a pound of water.

Hence the equivalent deduced from the friction of sperm
oil was 782" 1, a result almost identical with that obtained from
the friction of water. The mean of the two results is 781*8*,
which is the equivalent I shall adopt until further and still
more accurate experiments shall have been made.

XXXI. Letter from Prof. Schcenbein to Prof. Faraday,
F.E.S., on a neiso Test for Ozo7ief.

My dear Faraday,

HAVING a good opportunity for sending you a few lines,
I will make use of it to tell you something about my
little doings. You are no doubt struck with the peculiarity
of the ink in which this letter is written, and I am afraid you
will think it a very bad production ; but in spite of its queer
colour, you will like it when I tell you what it is, and when I

* This number is slightly different from 775, the equivalent stated at
Oxford, and used by me as one of the data for calculations on the velocity
of sound. The reason of the difference was that by an oversight 1 had
taken the friction oiboth pulleys as the correction of each weight instead
of both weights. The whole of the experiments are exactly the same as
those presented to the Oxford meeting. The slight alteration in the equi-
valent will make only a very trifling alteration in the theoretical velocity
of sound given in the last Number of this Magazine.

t Communicated by Professor Faraday.

Dr. Wilson on the Decomposition of Water by Platinum. 177

assure you that as long as the art of writing has been practised
no letter has ever been written with such an ink. Dealing now
again in my ozone business, I found out the other day that all
manganese salts, be they dissolved or solid, are decomposed
by ozone, hydrate of peroxide of manganese being produced
and the acid set at liberty. Now to come round again to my
ink, 1 must tell you that these lines are written with a solution
of sulphate of manganese. The writing being dry, the paper
Is sus})encled within a large bottle, the air of which is strongly
ozonized by means of phosphorus. After a few minutes the
writing becomes visible, and the longer you leave it exposed
to the action of ozone the darker it will become. Sulphurous
acid gas uniting readily with the peroxide of manganese to
form a colourless sulphate, the writing will instantly disappear
when placed within air containing some of that acid ; and it
is a matter of course that the writing will come out again
when again exposed to ozonized air. Now all this is certainly
mere playing; but the matter is interesting in a scientific
point of view, inasmuch as dry strips of white filtering paper
drenched with a weak solution of sulphate of manganese fur-
nish us with rather a delicate and specific test for ozone, by
means of which we may easily prove the identity of chemical,
voltaic and electrical ozone, and establish with facility and
certainty the continual presence of ozone in the open air. I
have turned brown my test-paper within the electrical brush,
the ozonized oxygen obtained from electrolysed water and
the atmospheric air ozonized by phosphorus. The quantity
of ozone produced by the electrical brush being so very small,
it requires of course some time to turn the test-paper brown.
As it is rather inconvenient to write with an invisible ink,
I will stop here; not however before having asked your kind
indulgence for the many blunders and faults which my ozone
bottle will no doubt bring to light before long.

Yours most truly,

Bale, July 1, 1847. C. F. SCH(ENBEIN.

XXXII. On the Decomposition of Water by Platinum and the
Black Oxide of Iron at a white heat, with some observations
on the theory of Mr. Grovels Experiments. By George
Wilson, ilf.D.*

THE remarkable discovery recently made public by Mr.
Grove, that water in certain circumstances, when raised
to a white heat, is resolved into its constituent gases, has na-

♦ Communicated by the Chemical Society; having been read March 15,
1847.
Phil. Mag. S. 3. Vol. 31. No. 207. Sept. 1847. N

178 Dr. Wilson on the Decomposition of Water by Platinum

turally excited much attention. It furnished the unexpected
confirmation of the truth of an opinion expressed by James
Watt so far back as 1783, that if steam could be made red
hot [white hot] so that all its latent heat should be converted
into sensible heat, either the steam would be converted into
permanent air, or some other change would take place in its
constitution *.

In the greater number of Mr. Grove's experiments, water
was raised in temperature through the medium of platinum ;
and it became a question accordingly, as Sir John Herschel
and my friend Dr. Lyon Playfair suggested, how far the de-
composition of water observed was owing to the mere heat of
the metal, how far to the peculiar surface-influence, or so-
called catalytic force, which has been so long recognized as
possessed by platinum and the other noble metals. jDr. Play-
fair also referred to the fact, " that many bodies at high tem-
peratures exhibited a great affinity for oxygen, which they did
not possess at lower temperatures ; as, for instance, silver,
gold, and even platinum itself, which metals absorb oxygen
when intensely heated, and give it out again on cooling. If
the experiments had been tried in tubes of quartz or silica,
they would not have been open to the objection which the
use of so peculiar a metal as platinum appeared to involve -f."

There was indeed one form of Mr. Grove's experiment not
liable to the exception urged against those where platinum was
used. He found it quite possible to decompose steam by
sending Leyden-jar discharges through it, and refers the de-
composition solely to the heat evolved by the electric spark.
The same view has been suggested as not improbable by
Faraday, in relation to the decomposition of water in the
liquid form by electric discharges J. With great diffidence,
however, I would remark, that the spark decomposition of
water cannot be regarded as an experimentum crucis. Al-
though the electric spark cannot decompose steam electroly-
tically, we may not at once infer that it cannot decompose it
in another way. I have no wish to assert that it can, but it
is possible that it may, and a crucial experiment should be
unexceptionable. Again : the spark discharge of a Leyden
jar exerts a great disruptive force, and acts topically with
much violence. There is reason moreover to believe that
mechanical agitation or disturbance of a chemical compound
can in many cases cause the separation of its elements. It
may seem an extravagant idea to suppose that oxygen may
be torn or detached from hydrogen by the action of a dis-

» Phil. Trans. 1783, p. 41G.

t Athenagum for September 19th, 1840, p. 966.

J Researches in Electricity, 3rd series, paragraph 337.

and the Black Oaeide of Iron at a white heat. 179

ruptive force on the molecules of water, as if chemical affinity
were but a kind of mechanical cohesion, which may be over-
come by division. On the other hand, however, it must not
be forgotten, that v/e are how acquainted with a large num-
ber of fulminating compounds, which can be decomposed by
friction, by a touch, or a stroke. These compounds are all
fragile, and water is a very stable combination ; but fragility
and stability are but terms of degree, in relation to stability
of union : and if it shall appear that a feeble mechanical force
can overcome a small intensity of affinity, it will be acknow-
ledged as quite possible that a powerful mechanical agency
may overcome a great one. We have no means perhaps of
making an unexceptionable experiment as to the decomposing
power of mechanical force ; for we cannot bring it into play
without calling into action other agencies. If we touch, or
rub, or strike a fulminate, for example, we cause the evolution
of heat, and add its decomposing power to that of the mecha-
nical impulse. It would be a mere petitio principii, however,
to assume that the heat produced alone effects the decompo-
sition observed. It seems to me, therefore, that the decom-
position of steam by the electric spark furnishes a more
complex problem for solution than the action of M'hite-hot
platinum on the same compound does ; and that the experi-
ments made with the metal are more likely to throw light on
those tried with the spark, than to be explained by them.

Whilst thinking over these difficulties, and the objections
to Mr. Grove's conclusions suggested by Herschel and Play-
fair, I had occasion to perform the familiar class-experiment
of burning iron wire in oxygen. I observed with an interest
I had not felt previously, although I had carelessly noticed
the phenomenon before, that bubbles of apparently perma-
nent gas rose from the globules of white-hot oxide of iron as
they fell into the water. It seemed to me possible that this
gas might be a mixture of oxygen and hydrogen separated by
the influence of the metallic oxide, acting as platinum did in
Mr. Grove's experiments. It was certain, moreover, that if
this should prove to be the case, it would supply a powerful
argument in favour of that gentleman's conclusion, which
seems, in spite of all the objections noticed, in the highest
degree probable, namely, that heat, apart altogether from the
medium through which it is applied, can resolve water into
its elements.

As the following experiments were made solely in the hope
of substantiating Mr. Grove's view, which unfortunately, how-
ever, they leave exactly as they found it, I trust that gentle-
man will not consider their publication an interference with
his researches. I was led to try them incidentally, and

N2 *

180 Dr. Wilson on the Decomposition of Water by Platinum

abandoned them as soon as I found I could render Mr. Grove
no assistance by means of them.

It would be difficult to conceive a more rapid and effectual
way of raising a body to a white heat than that afforded
by the combustion of iron in oxygen. I took for granted
also (as it afterwards appeared, too hastily) that the metal could
not but be saturated with oxygen and converted into a de-
finite oxide, which would be chemically ijidifferent to each of
the elements of water, and if it decomposed it at all, would
reject both its constituents. The convenient way, moreover,
in which the globules of oxide detach themselves and fall
into the water, and the rapidity with which the whole process
goes on, make it a very easy matter to collect in considerable
quantity whatever gases are evolved. A stoppered bottomless
jar of the ordinary construction for the iron-wire experiment,
and of 291 cubic inches' capacity, was made use of in the fol-
lowing trials. Eighteen experiments were made with it, and
from 100 to 110 grains of fused globules were obtained from
each combustion. A test-tube, with a funnel fixed into it
by a perforated cork, and filled with water, was arranged so as
to receive the gas. In some experiments it was placed within
the oxygen jar, so that the coil of wire when introduced hung
close to it, a piece of tin plate being arranged so as to guide
the globules within the edge of the inverted funnel. In the
greater number of trials however the tube and funnel were
placed outside of the vessel containing the oxygen, and an
inclined plane of tin plate was so placed as to carry the
globules past tlie edge of the jar, and within the mouth of
the funnel. No difference of result was observed in experi-
ments made in both ways, but the latter arrangement was
preferred as more convenient, and as enabling more oxygen
to be employed at each trial.

In all the experiments, permanent gas was evolved when
the fused globules fell into the water. This statement is to
be considered as applying to each combustion considered as
a whole ; for individual globules were frequently observed to
give off no gas at all, or to evolve so very little, that it might
be air separating from the water, in which it had previously
existed in solution. The quantity of gas obtained at each
combustion varied greatly. Sometimes as much as a cubic
inch was procured, more frequently only half that quantity,
and occasionally less. The globules from thick coils of wire
gave off a larger volume of gas than those from thin ones.

Portions of the gas were transferred to a Grove's eudio-
meter over water, and exposed to a white-hot platinum wire.
They did not kindle or detonate, nor were they sensibly
diminished in volume. Other portions were subjected to

and the Black Oxide of Iron at a white heat. 181

electric sparks and discharges in a syphon eudiometer over
water, with the same negative results ; but when air or oxy-
gen was mingled with the gas, it exploded sharply with
heated platinum or the electric spark. When a match was ap-
plied to the open end of a tube containing the unmingled gas,
it burned rapidly with a pale blue flame, but did not explode.
The gas given off during the action of the fused globules on
water was not then a mixture of oxygen and hydrogen.

Its freedom from all but a trace of oxygen Avas ascertained
in other ways. To. one portion of the gas standing over water
nitric oxide was added, but no ruddy fume or yellow colora-
tion showed itself. When phosphorus was introduced into
the gas, in one instance it did not smoke, but in the greater
number of cases it fumed for a brief period, and occasioned
an amount of contraction barely perceptible. The gas ap-
peared to be nearly pure hydrogen. To ascertain if it cer-
tainly were so, a portion of it was carefully dried, by chloride
of calcium, and transferred to a eudiometer over warm mer-
cury. Dry oxygen was then added, and the mixtiu'e exploded.
When the whole had cooled, the walls of the eudiometer ap-
peared dimmed by a very thin layer of moisture, but the
quantity of gas operated on was too small to admit of visible
drops being produced. Another portion of the gas was mixed
with half its volume of oxygen and fired by the electric spark.
The contraction which followed explosion varied in different ex-
periments, but was frequently such as to leave not more than
one-twentieth part of the mixed gases unconsumed. Phos-
phorus smoked in this residue for a short time, showing that
excess of oxygen had been made use of, and left a minute
volume of gas which was not diminished by caustic potash,
and must have been nitrogen.

It seemed possible that the trace of carbon present even
in malleable iron might affect the quality of the gas resulting
from the action of the globules of oxide on water, and that
carburetted hydrogen, carbonic oxide or carbonic acid might
be produced. It seemed desirable to know whether the latter
were present or not, as the oxygen might have gone to form
them. It was impossible to be certain that carbonic acid
was absent, for the gas from the globules being necessarily
collected over water, the temperature of which was low, car-
bonic acid would be retained in solution by that liquid. All
that I can say on this point is, that lime-Avater was not ren-
dered muddy or in the slightest degree opalescent by the
gas. It was several times detonated with oxygen over lime-
water, but the latter remained quite transparent, so that nei-
ther carbonic oxide nor carburetted hydrogen can have been
present. In short, the gas evolved from water by the white-

182 Dr. Wilson on the Decomposition of Water by Platinum

hot globules of oxide of iron, was hydrogen mingled with a
small quantity of air, previously no doubt in solution in water.

As only the hydrogen, then, of the water decomposed was
obtained, it became necessary to account for the absence of
the oxygen. I was tempted for a moment to think it pos-
sible that the black oxide of iron might have changed into the
red oxide of the same metal, by combining with the oxygen
not obtained in the elastic form : ex. gr. thus 2 Fcg 04 + 0 =
3Fe2 O3.

But the proto-peroxide of iron is known to be a very stable
compound, little if at all prone to become the peroxide ; and
it seemed more likely that unoxidized iron might be present
in the fused globules, which occasioned the evolution of hy-
drogen when it came in contact with water. To ascertain this
point, portions of the globules were dissolved in dilute muriatic
and sulphuric acids, and were found in most cases to evolve hy-
drogen. Some specimens of the globules gave off not a trace
of gas when they dissolved, and must have consisted of the
definite oxide ; a point of interest in connection with the
fact already mentioned, that globules were frequently ob-
served to drop into water without any bubbles of gas rising
from them.

The volume of hydrogen however given off in some of the
trials, when the product of combustion was placed in acid,
was very considerable. A graduated gas jar was filled with
dilute sulphuric acid, and inverted over a small capsule con-
taining 100 grains of the crushed globules, which was placed
in a basin also containing dilute acid. By this arrangement
the gas was collected and measured at the same time, Mithout
risk of mixing with air, or necessity for watching the process,
which is a slow one. 100 grains treated in this way gave off
16 cubic inches of hydrogen, corresponding to 9 grains of
iron. The experiment was accidentally stopped at this point
whilst the gas Mas still rising in undiminished quantity.

Metallic iron, then, was certainly present in many of the
globules, and of this I had direct ocular demonstration. On
crushing some of them in a mortar, they were found to sepa-
rate into a shell of pulverizable oxide, and a core of iron
which formed a nearly spherical pellet. In one case 50 grains
of the globules were crushed, the pellets separated, and the
residue placed in diluted sulphuric acid. It did not evolve a
trace of hydrogen in the course of twenty-four hours. The
pellets were then added to the same acid, and gave off 12
cubic inches of gas = 13*6 per cent, of iron in the globules*.
The shell of oxide is frequently imperfect or perforated, so

• In none of the experiments was the thermometer or barometer spe-
cially observed, as minute accuracy was not aimed at.

and the Black Oxide of Iron at a white heat, 183

that water or any other liquid penetrates to the iron core, and
is subject to its influence. When this becomes known, it
need not surprise us that most of the globules should rapidly
decompose water. After observing this fact, I tried the
effect of thin and thick coils of wire, and found that the latter
invariably gave off the greater volume of gas. When the coil
is so thin that the metal all oxidizes, no gas is evolved at all.
A thick coil indeed furnishes a striking mode of illustrating
to a class the principle of Lavoisier's mode of decomposing
water, and forms a beautiful addition to the iron-wire expe-
riment.

From these observations then, it would seem that white-
hot oxide of iron cannot decompose water in the way white-
hot platinum does. But before any conclusion can be drawn
from this fact inimical to Mr. Grove's views, or favourable to
the opinion that a specific property of the platinum has more
to do with the decomposition of water than its mere tempera-
ture has, we should require to know how far the two white-
hot bodies are to be considered as at the same temperature.
In Mr. Grove's experiments, platinum is raised to as high a
heat as it can bear without fusing. It must then be elevated
to a temperature much above that necessary to make iron
white-hot, or to fuse its oxide, for our forges can melt iron
and its oxides, but do not fuse platinum. It may also be re-
marked, that bright as the light emitted by burning iron is,
it falls short in intensity of that given off by platinum on the
verge of fusion. It seems accordingly probable, that during
the combustion of iron in oxygen the temperature never rises
high enough to confer upon the resulting oxide the power of
decomposing water. The question admits of direct decision,
by ascertaining whether oxide of iron, heated by the oxy-
hydrogen blowpipe to as high a temperature as fusing pla-
tinum, acquires the power of decomposing water without ap-
propriating to itself either of its elements. But it would have
been an interference with Mr. Grove's own researches to have
made experiments of this kind, and I have accordingly left the
question undecided.

Meanwhile the experiments I have recorded are of some
little interest, as at least showing that not only a white heat,
but a high white heat, is essential to the successful perform-
ance of Mr. Grove's experiments. Unfortunately, we have
not at present any method of measuring high temperatures
which admits of ready application or secures great accuracy.
" W^hite heat " is in fact a vague expression for a range of
temperature, of the extremes in either direction or extent of
which we have no very precise knowledge. The power of
the eye to measure the relative intensities of the light evolved

184) Dr. Wilson ow the Decomposition of Water by Platinum.

by white-hot bodies is very limited, and varies greatly in
different individuals. But the experiments 1 have recorded
seem to supply the means of so far at least defining the white
heat requisite for the separation of the elements of water,
inasmuch as they show that it must at least exceed the tem-
perature necessary for the fusion of malleable iron or its black
oxide. If, moreover, the decomposing powers of the electric
spark be solely referable to its temperature, we seem entitled
to conclude, from the experiments I have detailed, that the
heat of the smallest spark that can decompose water is at
least equivalent to that of fusing platinum. They appear also
to Avarrant another conclusion. It was suggested by Dr.
Leeson and by Mr. Hunt, that the bursting of steam-boilers
might occasionally be owing to the metal they consist of be-
coming white-hot and decomposing water like platinum, with
the rejection of both its elements*. This ingenious sugges-
tion seemed to myself, before making experiments with iron,
likely to prove just; but as fusing white-hot iron appears
•unable to decompose water, otherwise than by combining
with its oxygen, it is impossible that the walls of a boiler can
ever be raised to a temperature sufficiently high to enable
them to separate the elements of water in the way platinum
does.

I may now be permitted to make some comments on the
rationale of the results obtained by Mr. Grove. That gentle-
man, if I understand him aright, considers the decomposition
of water by white-hot platinum not only, as assuredly it is, a
remarkable and unexpected result, but as evidencing on the
part of heat a power to produce opposite or dissimilar chemi-
cal effects in the same circumstances. He is reported in the
Atheneeum (Sept. 19th, 1846, p. 966) to have "announced
his discovery that all the processes by which water may be
formed are capable of decomposing water" (p. 966). If by
this statement be simply meant, that heat combines oxygen
and hydrogen into water, and decomposes water into these
gases, it will be admitted to be a just conclusion ; but it may
be questioned, I think, whether Mr. Grove's experiments
add anything to our knowledge of the power of heat to effect
chemical changes, except in so far as they supply an addi-
tional very remarkable example of its twofold analytical and
synthetical agency, which has been so long recognised. Hy-
drogen, which as a gas is probably the vapour of a very vola-
tile metal, may be compared with mercury, also a volatile
substance. If mercury and oxygen be heated together to the
temperature of 662° F., they combine and form the red oxide of
the metal. If this resulting oxide be raised to a low red heat,
* Athenaeum, Sept. 19th, p. 9Q6.

and the Black Oxide of Iron at a white heat. 185

it is decomposed into mercury and oxygen. In like manner,
if hydrogen and oxygen be raised together to the tempera-
ture of 660° F.*, they unite and form water. If the resulting
Mater be raised to a white heat, it is resolved into hydrogen
and oxygen. Both metals (?) present the same phaenomena.
At one temperature (nearly the same in both cases) combina-
tion with oxygen occurs ; at a higher temperature, decompo-
sition of the oxide happens. Many other examples might be
given in illustration of the same fact. Such cases, however,
do not seem to warrant a conclusion as to heat exhibiting
anything like a polarity of force, by which I understand the
manifestation in opposite directions of opposite powers of
equal intensity. At all events, if the opposite effects of dif-
ferent intensities of the same agent be considered equivalent
to a polarity of action, it is difficult to see what force may not
be called a polar one. The decomposing and combining
power of heat of different intensities, seems exactly compara-
ble to the opposite effects of different intensities of mechanical
impulse.

If two pieces of smooth glass are laid together and struck
gently or compressed slightly, they unite or cohere. If the
united pieces are thereafter exposed to a sharp blow or to
great compression, the union is dissolved, or they are shat-
tered to fragments. Here the same force effects mechanical
synthesis and mechanical analysis. But in these contrasted
actions, as seems to be the case also in Mr. Grove's experi-
ments, the results are occasioned by a difference in degree of
intensity of the same power, not as in the opposite effects of
a polarizing force like electricity, by a difference in the kind
of power which appears, whatever be its intensity. There is
one form, indeed, of Mr. Grove's experiment which at first
sight does not appear to admit of the explanation proposed
in reference to the other trials — I allude to the decomposition
of steam by the electric spark, which is well known to have
the poM^er of combining hydrogen and oxygen into water.
A similar experiment w^as made in perhaps a still more in-
structive form in the latter part of last century by Beccariaf,
Pearson and Van Troostwyk, and more recently by WoUas-
ton J, in his well-known decompositions of water with guarded
poles. In certain of these trials it was found that Leyden
jar discharges sent through water, decomposed it till the ac-
cumulation of permanent gas left the wires bare ; after which
the first spark that passed recombined the gases into water,
which again covered the wire, when decomposition could

* Graham's Elements, 1st edit. p. 259.

t Lettere dell' Electrecismo, quoted in Lardner's Electricity, vol. i.p. 78.

j Faraday's Electrical Researches, series 3, paragraph 328.

186 Dr. Wilson on the Decomposition of Water by Platinum

anew be obtained. Here, to appearance, the same agent act-
ing with the same intensity, alternately decomposed and re-
composed water. For argument's sake, let it be acknow-
ledged that the heat alone of the spark was the cause of che-
mical change. Nevertheless it may be questioned, whether
it acted with equal intensity in both cases. The electric spark
must be conceived, according to the results already given, to
be at first at a high white heat, and whilst retaining this tem-
perature we may believe it to possess a power of disuniting
the elements of water, and of preventing their union. But as
soon as the spark falls to the temperature of 660° F., it loses
its power of decomposing water, and, on the other hand, ac-
quires a power of uniting hydrogen and oxygen. Although
therefore the spark is always furnished of the same intensity,
its action may change, and even be reversed, as its intensity
diminishes. Moreover, even when the spark is white-hot, it
is only the amount of matter directly in its track that will be
raised to a white heat. Contiguous portions will have their
temperature much lower, so that in the case of hydrogen and
oxygen, at some little distance from the route of the spark,
the temperature will be 660° F., and there combination will
begin, and ultimately extend through the whole mass of gas.
In like manner, when a platinum wire is made white-hot in a
mixture of hydrogen and oxygen, it causes their combination.
Here we may suppose that union occurs as soon as the tem-
perature of the metal I'ises to 660" F., and before it acquires
a white heat. Or if we were to arrange matters so that the
wire should be made white-hot in a vacuum and hydrogen
and oxygen afterwards admitted to it, still union of the gases
should happen ; for although the wire might prevent com-
bination immediately around itself, at no great distance where
the temperature was below 700° F. it would compel union.
In all such experiments the combining effect of heat will be
much more manifest than its decomposing power ; not that
perhaps the former is in reality greater than the latter, but
because flame is propagated through a mixture of hydrogen
and oxygen by a series of combustions. The hot wire or the
electric spark kindles only the portions of gas immediately
adjacent to it, but the combustion of those sets fire to the
molecules contiguous to them, and these in their turn to their
neighbours, till all are made to burn. Thus the flame travels
after the original cause of combustion has ceased to operate
directly, and the momentary action of a small spark, or the
transient heat of a red-hot capillary wire may suffice to fire
an infinitely large mass of hydrogen and oxygen. There is
no provision for a similar propagation of decomposition
through water or steam when either is made white-hot j the

and the Black Oxide of Iron at a white heat. 187

absolute amount accordingly of disunion of the elements of
M'ater occasioned is very small.

If allowance, however, be made for the apparent difference
in extent of effect which heat shows in uniting and in dis-
uniting the elements of water, the phaenomena otherwise
seem referable solely to the intensity of the temperature to
which hydrogen and oxygen are exposed. The opposite pro-
cesses might go on simultaneously, union or disunion being
determined simply by the different temperatures to which
different portions of the gases were raised. At least it seems
not improbable that if a mixture of steam and of hydrogen
and oxygen were exposed to electric discharge, decomposition
of the steam and combination of the hydrogen and oxygen
might be effected by the same spark, provided the volume of
steam were not large. In the track of the spark decompo-
sition would occur, so long as a white heat prevailed. When
the temperature fell, combination would happen where the
spark had passed, if it had not already commenced in the
neighbourhood of its direct route. Similar remarks apply
mutatis mutandis to the action of a hot platinum wire on a mix-
ture of steam with oxygen and hydrogen.

It may be objected to this view, that Mr. Grove decora-
poses steam in his eudiometer, and obtains a permanent bub-
ble of gas, consisting of hydrogen and oxygen. The bubble
however obtained in this way is very small, and could not
probably be greatly increased. Mr. Grove has not mentioned
how large a volume of hydrogen and oxygen he could obtain
in the same eudiometer, by alternately boiling the water till
the steam produced caused the liquid to fall below the wire,
and allowing the steam to condense till the water rose above
the metal. But I venture to say that no large volume of per-
manent gas could be procured by this process if the same
eudiometer were employed many times successively. The
combining action of the wire might not take effect on the
hydrogen and oxygen when their quantity was small, and
they were diluted through a large volume of steam, for in
virtue of the law of diffusion, the molecules of hydrogen and
oxygen would be separated from each other by molecules of
water-vapour; but when the latter diminished in bulk, it
seems impossible to doubt that kindUng of the gases would
occur.

Mr. Grovels experiments then do not appear to prove that
heat of the same intensity is able in the same circumstances
to form water and to decompose it. When therefore it is
stated that water can be produced by the processes that dis-
unite its elements, the word ^process ^ can only be understood
to signify that the general arrangement in both cases is the

188 Dr. Wilson on the Decomposition of Water by Platinum.

same, not that the intensity of the agent called into play, or
its mode of action is identical. If this could be affirmed, we
should be able to announce as a general proposition, that
manifestations of the same force absolutely identical as to
quality, quantity and intensity, could produce totally oppo-
site results, which would be tantamount to affirming that un-
like effects may flow from the same cause, without any altera-
tion in the qualities or conditions of the latter.

The last observation I would make refers to the curious
fact noticed by Mr. Grove, namely, that when a platinum wire
is heated white-hot in steam, " in a few seconds a small bub-
ble of gas is formed; but if the action be continued for a
week, it does not increase in quantity *.^^

Are we to suppose that the wire is at the same time decom-
posing water around itself, and producing water at a little
distance, undoing in one place what it effects in another, so
that no permanent accumulation of gas is allowed to take
place ? This is possible, but I think not likely. The ob-
servation made by Mr. Grove seems sufficiently explicable,
on the supposition that as soon as the wire is completely en-
veloped in steam, the thermo-circulatory currents which the
high temperature occasions in the vapour prevent it from
remaining long enough in contact with the wire to become
heated white-hot. The steam probably circulates endlessly
around the wire without a trace of decomposition occurring
in it. It seems not unlikely indeed that in Mr. Grove's ex-
periments with his eudiometer it was not steam that yielded
the hydrogen and oxygen obtained, but the last film of water
below the wire, which could not escape from the metal, but
tended rather, in consequence of its expansion, to rise towards
it, and was thus compelled to acquire a white heat, and to
break up into its elements. If this view be correct, an ar-
rangement where a white-hot wire or sheet of platinum foil
was kept grazing the surface of water, might be found to
effect a continuous decomposition of the liquid in question.

It is no objection to this view that an electric spark decom-
poses steam readily, for the duration of the spark is so short,
that there is no time for the production of thermo-currents,
nor any possibility of the steam escaping from the powerful
topical action of the discharge. The spark may be compared
to fulminating silver, whose action is instantaneous and vio-
lent, but quite local, — the heated platinum to gunpowder, the
effect of which is cumulative and more general.

* Athenaeum, Sept. 19tli, 1846, p. 966.

[ 189 ]

XXXIII. An account of a Discovery in the Theory of Numbers
relative to the Equation A.x^ + Bj/^ + Cs^ = Tixyz. By J. J.
Sylvester, Esq-^ M.A., F.R.S.^

First General Theorem of Transformation.
TF in the equation Kx^-\-^y^->tC^ = 'Dxyz . . (1.)

A and B are equal, or in the ratio of two cube numbers to
one another, and if 27 ABC — D^ (which I shall call the De-
terminant) is free from all single or square prime positive
factors of the form 6n+ 1, but without exclusion of C7ibic factors
of such form, and if A and B are each odd, and C the double
or quadruple of an odd number, or if A and B are each even
and C odd, then, I say, the given equation may be made to
depend upon another of the form

A'm3 + Wv" + C'«)3 = B'xyz ;
where A'B'C' = ABC

D'=D
u.v .iv =some factor of z.
The following are some of the consequences which I deduce
from the above theorem. In stating them it will be convenient
to use the term Pure Factorial to designate any number into
the composition of which no single or square prime positive
factor of the form 6 n+1 enters.

The equations x^ + 2^ + 22^=Dxys:
a^ + y^ + 4!Z^ = T)xyz
2a^ + ly^ + z^ = Yixyz
are insoluble in integer numbers, provided that the Determi-
nant in each case is a Pure Factorial.

The equation a^-\-y^-\- K^'rs.^Vtxyz
is insoluble in integer numbers, provided that the Determinant,
for which in this case we may substitute A — 27B^, is a pure
factorial whenever A is of the form 9w+l, and equal to
2p3!±i Qp 4.p3i±i^ jp being any prime number whatever.

I wish however to limit my assertion as to the insolubility
of the equations above given. The theorem from which this
conclusion is deduced does not preclude the possibility of two
of the three quantities a?, ?/, z being taken positive or negative
units^ either in the given equation itself or in one or the other
of those into which it may admit of being transformed. Should
such values of two of the variables afford a particular solution,
then instead of affirming that the equations are insoluble, I
should affirm that the general solution can be obtained by
equations in finite differences f.

* Communicated by the Author.

t Take for instance the equation ifi-\-y^-\-^z^'=9xyz. The Determinant

1 90 Ofi a discovery in the Tlieory of Numbers.

Second General Theorem of Transformation,
The equation f^oi^-\-^y^-{-h^z^=^^xyz . . . (2.)
may always be made to depend upon an equation of the form

Aw^ -f Btfi + Cwr^ = Tiuim,
where ABC=:R3-S3

D = 3R;
and u.v.vo =. some factor o^fx-^gy + hz.
R representing K + 6fgh
S ... K-3/^A.

I have not leisure to show the consequences of this theorem
ef transformation in connexion with the one first given, but
shall content myself with a single numerical example of its
applications: x^-{-'y^-{-z^=—Qxyz

may be made to depend on the equation

and is therefore insoluble.

It is moreover apparent that the Determinant of equation
(2.) transformed is in general — 27R% and is therefore always
a Pure Factorial, and consequently the equation

f^a^ +g^y^ + h^i^ = ^xyz
will be itself insoluble, being convertible into an insoluble form,
provided that K + Gfgh is divisible by 9, and provided further
that {K + Qfghf — {K — Sfghf belongs to the form m^.Q,
where Q is of the form 9«+ 1, and also of one or the other of
the two forms ^Ip^'^^^ 4;>^'*>, p being any prime number what-
ever.

Pressing avocations prevent me from entering into further
developments or simplifications at this present time.

It remains for me to state my reasons for putting forward
these discoveries in so imperfect a shape. They occurred to
me in the course of a rapid tour on the continent, and the
results were communicated by me to my illustrious friend M.
Sturm in Paris, who kindly undertook to make them known
on my part to the Institute.

Unfortunately, in the heat of invention I got confused about

27"25 is a Pure Factorial : consequently if the solution be possible, since
in this case the transformed must be identical with the given equation, this
latter must be capable of being satisfied by making x and y positive or ne-
gative units. Upon trial we find that x=.\ i/=.\ z=2 will satisfy the equa-
tion. I believe, but have not fully gone through the work of verification,
that these are the only possible values (prime to one another) which will
satisfy the equation. Should they not be so, my method will infallibly
enable me to discover and to give the law for the formation of all the others.
Here, then, under any circumstances, is an example, the first on record,
of the complete resolution of a numerical equation of the third degree be-
tween three variables.

On a new Kite- Apparatus for Meteorological Observations. 191

the law of oddness and evenness, to which the coefficients of
the given equation are in the first theorem generally (in order
for the successful application of my method as far as it is yet
developed) required to be subject. I stated this law erro-
neously, and consequently drew erroneous conclusions from
my Theorems of Transformation, which I am very anxious
to seize the earliest opportunity of correcting. I venture to
flatter myself that as opening out a new field in connexion
with Fermat's renowned Last ' Theorem, and as breaking
ground in the solution of equations of the thii'd degree, these
results will be generally allowed to constitute an important
and substantial accession to our knowledge of the Theory of
Numbers.
26 Lincoln's Inn Fields,
August 24, 1847.

XXXIV. Experiment made at the Kew Observatory on a new
Kite- Apparatus for Meteorological Observations^ or other
purposes^,

MR. W. R. BIRT (on the Uth of this month) took some
kites, &c. to the Kew Observatory, for the purpose of
endeavouring to ascertain how far it might be practicable to
measure the force of wind at various elevations by their means,
and (in the mere manipulation of his experiments) was assisted
by Mr. Ronalds. After several trials, &c. they agreed that
the sudden variations, horizontal and vertical, in the position
of the kite, the great difficulty of making a kite which should
present and preserve a tolerable approximation to a plane,
that of measuring, with sufficient accuracy, at any required
moment, its inclination, and lastly, the influence of the tail,
would always tend to render the observation somewhat unsa-
tisfactory. Mr. Ronalds then proposed to try the following
method of retaining a kite in a quasi invariable given position.
Three cords were attached to an excellent hexagonal kite of
Mr. Dirt's construction : one in the usual manner, and one
on each side (or wing). The kite was then raised as usual ;
the two lateral cords were hauled downward by persons stand-
ing at the apices of a large equilateral triangle (described upon
the ground) until the ascending tendency became considerable
(even when the force of the wind was at its minimum), and the
three cords were made fast to stakes or held in the hand.
He had entertained no expectation of the favourable result of
this simple and obvious contrivance. 1 he place of the kite
did not seem to vary so much as one foot in any direction, and
it really appears to him probable that a very large kite or
kites might be employed in this kind of manner often and very
* Comraunicated by Mr. Ronalds.

1 92 Dr. Playfair on Transformations

cheaply as a substitute for a captive balloon in meteorological
inquiries, or even (on a very extensive scale) for other require-
ments in military science, &c. An anemometer, a thermo-
meter, an hygrometer, &c. of some registering kinds, &c.,
might be hauled up and lowered at pleasure (like a flag) by a
person standing in the centre of the triangle (above referred
to), and by means of a line passing through a little block
attached to the kite. The cords and kite should of course be
of pure silk, for»the sake of lightness, combined with extreme
strength, and the size and thickness in some measure adapted
to the breeze or lighter air. The silk might be advantageously
covered with a very light coat of elastic varnish.

XXXV. On Transformations produced by Catalytic Bodies.
By Lyon Playfair, Esq,^

T> ERZELIUS rendered a most useful work to science, when
■^-^ he collected into one class those varied phaenoraena of
chemical action resulting from causes certainly very different
from the ordinary manifestations of those affinities, which
produce combinations or promote decompositions. This phi-
losopher believes the power f? which causes decomposition
without the acting body participating in its result, to be a
distinct electro-chemical agency different from other recog-
nised powers, and he named it the " Catalytic force." Ac-
cording to this view, catalytic bodies do not act by chemical
affinity, but they excite inherent affinities in other substances,
in consequence of which new combinations or decomposi-
tions ensue.

Mitscherlich J, adopting this view, considered a number of
catalytic decompositions in detail, and showed the important
influence exerted by the state of surface of bodies in favouring
this peculiar action, which he denominates decomposition by
contact. The examples, adduced in this interesting memoir,
of the favourable action of an extended surface upon combi-
nation, fully prove that the physical condition of bodies ex-
ercises an important influence upon the action of this force ;
but they do not remove the necessity for studying the force
itself, as it may either be a vis occulta, entirely distinct from
powers already recognised, as Berzelius supposes, or may be
modified forms of those in continual operation.

Liebig§ views the catalytic power as a dynamical action

* Communicated by the Chemical Society ; having been read April 5,
1847.

"t" Jahresbericht, xv. 237.

X Taylor's Scientific Memoirs, Part xiii. ; or Fogg. Ann, xxxi. 281.

§ Liebi^'s Chem. of Agriculture, 4th edit., p. 284.

produced by Catalytic Bodies. 193

on the atoms of a complex molecule, conceiving that the ac-
tivity of the atoms of a body in a state of motion may be
communicated to those of another body in a state of rest.
The atoms of a compound, according to this view, if in a state
of exact statical equilibrium, arrange themselves according to
new affinities, when the vis inertice is overcome by motion.
In proof of this view, Liebig carefully examines a large num-
ber of decompositions, and accounts for some of the most
difficult transformations in organic chemistry.

But there are many instances, to which I shall have to
draw attention in the present memoir, where catalytic de-
compositions ensue when there is no intestine motion in the
atoms of the exciting body ; and hence we cannot do more
than consider motion as favourable to the development of
dormant affinities, in a manner similar to the surface action
described by Mitscherlich. The power of peroxide of hy-
drogen and of pyruvic acid to reduce oxide of silver is cer-
tainly a singular phaenomenon, and appears favourable to
Liebig's views ; but the cause of the original decomposition
of the peroxide of hydrogen cannot be ascribed to motion, as
the atoms of the oxide of silver are not in that state, and
those of the peroxide of hydrogen either not at all or only
slightly so. Neither will it suffice to suppose that the escape
of gas during such decompositions is due to the presentation
of angular points from which the gas may escape *, because
solutions of alkalies equally effect the decomposition, accord-
ing to Thenard -f. The cause, therefore, which enables cer-
tain substances to hasten the decomposition of such bodies
as peroxide of hydrogen or persulphuret of hydrogen, al-
though favoured by the state of surface and by motion, is in-
dependent of mere physical condition.

In further proof of the importance of motion in causing com-
bination or decomposition, Liebig cites the favourable effects
of agitation on the precipitation of potash by tartaric acid. It
may be questioned, however, whether this is not either a me-
chanical breaking up of a combination or the simple effect of
cohesion. Thus when water is saturated with a gas, a brisk
agitation with a rod causes the separation of bubbles of gas
previously dissolved. The mechanical force may here be sup-
posed to have broken up the compound molecule of water and
gas by detaching the former, and thus enabling the gas to
escape by its elasticity. In the precipitation of potash by
tartaric acid, cohesion may effect the same result that elasti-
city does in the case of gas, the agitation knocking off the

• Ann. der Pharm., ii. 22.

f Ann. de Chim. et de Phys., xlviii. 79.

Phil. Mag. S. 3 . Vol. 3 1 . No. 207. Sept. 1 847. O

194 Dr. Play fair on Transformations

atoms of water which are feebly attached. In fact we know
that the addition of alcohol equally aids the precipitation,
the action here being a chemical separation of the water,
as in the other it is mechanical. The diminished solubility
of the salt, after it has been influenced by cohesion and sepa-
rated from water, has its counterpart in many similar in-
stances ; for example, in the small solubility of anhydrous
sulphate of iron. The effect of agitation on a solution of
sulphate of soda, saturated while hot and allowed to cool, I
ascribe to the same cause. The supposed effect of cohesion
or elasticity in these cases is nothing more than that con-
stantly observed in ordinary phagnomena, when the gravity
of a substance is different from that of the medium in which
it exists. The vesicles of water in the atmosphere may be so
small that they float in it and produce fogs ; but when ag-
gregated together by the motion of the air, they form drops,
which precipitate to the ground with a rapidity proportionate
to their size : the converse of this is also true. Thus, the
particles or aggregated atoms of carbonic acid in water may
be so very small, that, with the slight affinity of the latter
added, they may be enabled, when in a state of rest, to re-
main without resuming their elastic form; but agitation
causes a larger system of aggregated atoms, and the gas now
escapes in small bubbles.

The first instance of cohesion applies in the precipitation
of tartar. At the moment of formation the particles may be
so widely apart, that, aided by their slight affinity for water,
they remain without aggregating to any considerable extent.
Brisk agitation, and the presentation of an extended surface,
effect their aggregation and cause a speedy precipitation. It
may be that these are really instances of combination favoured
by motion ; but presuming that they are, the general argu-
ment is not aflfected, that other decompositions perfectly ana-
logous are produced where the exciting body is in a state of
rest.

The third theory of these decompositions is, that catalytic
bodies act by exerting a feeble chemical affinity on one of the
constituents of the body decomposed. This view was intro-
duced by Mercer *, and supported by several very ingenious
experiments communicated to the British Association at its
meeting in Manchester. One of these was, that protoxide of
manganese had the singular power of hastening the oxidation
of starch in nitric acid f. The metallic protoxide, from its

* Reports of British Association, vol. xi. 2d Part, p. 32.
f The experiment is easily made by dissolving 1 ounce oxalic acid in
J- a pint of water at 180° F., and adding to this 1 oz. colourless nitric acid of

produced by Catalytic Bodies. 195

disposition to pass into the state of peroxide, aids the oxalic
acid to decompose the nitric acid, the united affinities of
both being able to accomplish what neither by itself could do.
The protoxide remains unaffected at the end of the experi-
ment, because, under the circumstances (the presence of acid),
it cannot gratify its desire to become peroxide, and, therefore,
it passes over its oxygen to the carbon, which escapes as
carbonic acid. Mercer cited, as further examples, the action
of protoxide of copper in eliminating oxygen from a solution
of hypochlorite of lime, and of peroxide or binoxide of nitro-
gen in commencing the oxidation of a mixture of protochlo-
ride of tin and nitric acid. Mercer implied by these instances,
that catalysis is an affinity of the catalytic agent for an ele-
ment in the body acted upon, that affinity being feeble and
incapable of gratification under the circumstances.

It would be advantageous to science if we could arrange
under a known power the cases of decomposition which ap-
peared so mysterious as to induce the great Berzelius to
ascribe them to the action of a new force. It may not be
possible in the present state of our knowledge to comprehend
the whole of the instances observed, but, if most are included
in one category, we have a right to suppose that the others
may be embraced as our knowledge progresses. I shall there-
fore endeavour to show that many catalytic decompositions are
merely cases of chemical affinity exerted under peculiar con-
ditions.

In no instance of chemical union does there seem to be
such a complete gratification of affinity as to suppress the at-
tractions of the elements. The inherent affinities still remain
more or less powerful, for, if it were not so, the compound
would be permanent under all circumstances and not liable
to further change by the action of external agents. When
manganese unites with 1 atom of oxygen, the affinity of the
metal for oxygen is not wholly merged, but is still strong
enough to attach to itself 1, 2 or 3 atoms more oxygen.
When the oxide is one of the lowest of the series, this affinity
exhibits itself in a basic power by attaching itself to any com-
plex highly oxygenized molecule, such as the oxygen acids,
or of radicals playing the part of oxygen. When, on the
other hand, the manganese or other radical becomes highly
oxygenized, we find it possessing acid properties, that is, the

1-30 sp. gr. No action ensues on this mixture, but it immediately com-
mences on the addition of a protosalt of manganese, which for simplicity
may be the oxalate or nitrate. The action is also strikingly shown by heat-
ing a mixture of oxalic acid until tlie action commences, then diluting it
till all action ceases. A little protosalt of manganese now added to the
solution causes an immediate renewal of the oxidation.

02

196 Dr. Play fair on Transformations

additional atoms of oxygen, being less firmly attached, are
capable of gratifying the disposition of a less oxygenized atom
(the base) to attach itself to a higher oxide, or, to use the
convenient phraseology of Graham, the base becomes zincous
to the acid, which is now chlorous.

On heating the nitrates, nitric acid is not given off, but
NO4 + O. The decomposition readily results from the dis-
position of the base to appropriate more oxygen and pass
into the higher oxides. If the base be oxide of nickel, the
oxygen becomes attached to the oxide and remains ; if, how-
ever, an oxide which has but a feeble affinity for oxygen
at an elevated temperature, the elasticity of that element
is able to overcome the affinity, which succeeded in break-
ing up the nitric acid. The final action is so obviously de-
pendent upon the oxygenous part of the acid, as to make
Schonbein believe that salts contain peroxides ready-formed ;
thus that NO5, IIO=N04 + H02, or PbO, N05 = N04-f
PbOg. This however is an unnecessary supposition, the pre-
vious view accounting sufficiently for the decomposition of a
nitrate, so as to produce NO4 and O. Admitting this view
to be correct in the expression that the preponderating quan-
tity of a chlorous element in an acid renders the latter chlo-
rous to a base, the mechanical attachment being to the chlo-
rous element, we can understand why the number of atoms
of oxygen in a base should regulate the number of atoms of
acid attached to it. Thus RO presents only one chlorous
element of attachment to the acid, and therefore the latter
adheres to it in one proportion ; whereas Rg O3, which pos-
• sesses three atoms of a chlorous element equally distributed
round a zincous nucleus, presents three points of attach-
ment, and therefore produces a salt Rg O3, 3A. This view
in result gives all the simplicity of the acid radical theory,
both views entertaining the idea that the oxygenous atoms of
the base and acid are attached to each other. We have cer-
tain instances, as for example KO, CIO5 ; PbO, NO5, where
the elastic atoms of oxygen combine as closely together as
non-elastic atoms, such as lead or silver.

Although to aid conception we may suppose the atoms of
oxygen of the base and of the acid to be in mechanical con-
nexion, the true arrangement is probably not so, seeing that
in a base there is always a part more zincous than the oxy-
genous atom, although the base as unity is zincous to the
acid. We see many instances in chemistry of union of atoms
in pairs, or what may be called dual affinity. This Graham*

* Trans. Royal Soc. Edin. vol. xiii. ; Phil. Trans. 1837, p. 47 etseq,-,
Phil. Mag. Third Series, vol. xxiv. p. 401 et seq.

produced by Cdtalytic Bodies. 197

has proved to be the case with regard to atoms of water, and
we know of numberless instances in the case of oxides. Thus
RO uniting with oxygen forms ROg. In this case RO + O
corresponds to RO + A, the acid here representing the chlo-
rous element from its oxygenous character. It is not neces-
sary to suppose that A and O are associated in one continu-
ous line, the probability being that the molecule may really
be represented by ARO. Thus also in Rg O3, where the O3
are probably grouped equally round Rg, there is room for
three more of a chlorous element to gratify the dual affinity,
and the general formula Rg Og, 3A is the result, the 3A here
representing three of a simple chlorous element. The result,
as regards affinity;, will still however be the same, the whole
'. depending upon the attraction of the central nucleus R. It
ig therefore only for simplicity of expression in studying the
phaenomena of catalysis, that I view the atoms of oxygen of
an acid as associated in mechanical continuation with the
atoms of oxygen of the base, the ejf'ect being represented by
this expression : the whole views of molecular or atomic con-
stitution of bodies are in my opinion only convenient fictions
to enable us to study the forces themselves, and the concep-
tion of a mechanical arrangement I only adopt as expressive
of the manifestations of powers residing in matter.

To show that the tendency of bases to N O5, even without
being combined, is to attach themselves to the oxygenous
part of the acid, a curious phaenomenon observed by Mercer
may be cited.

A portion of alumina may be taken and placed at the bot-
tom of a vessel containing warm NO5 ; no action ensues, ex-
cept partial solution ; a slip of calico coloured in indigo-blue
may now be introduced into the mixture, and remains unaf-
fected in the clear acid, but is immediately discharged when
pressed with a glass rod into the alumina. Here the alumina
acts by placing the oxygen of the nitric acid in a state of ten-
sion without however succeeding in decomposing it, but the
moment an assistant affinity comes into play, that state is
shown by the decomposition of the nitric acid and oxidation
of the indigo. The alumina in the presence of the acid could
not oxidize (in fact, we know of no higher oxide), and there-
fore the indigo appropriates the oxygen. I find that various
other oxides, such as calcined Crg O3 and SnOg, have the same
power, the latter showing this disposition more strongly than
any of the other oxides. The best mode of trying these ex-
periments is to heat a certain quantity of nitric acid, and then
dilute it till indigo cloth ceases to be bleached. The oxide of
tin is now added and allowed to fall to the bottom. On in-

198 Dr. Playfair on Transformations

troducing a slip of indigo-blue calico, the portion in the clear
acid will be found to remain unaffected, while that in contact
with the insoluble oxide will be bleached in a few seconds.
That this decoloration of the indigo is due to the assistant
affinity of another body acting in the same direction, i. e. also
having a disposition to unite with oxygen, may perhaps best
be shown by the following experiment : — Warm nitric acid is
diluted to such extent that it just ceases to discharge indigo-
blue calico ; it is then divided into two portions, with slips of
coloured calico in each, and through one of these binoxide
of nitrogen is passed. In the latter the indigo becomes
quickly bleached, while it remains unaffected in the former,
the action obviously being due to the accessory affinity of the
nitric oxide for more oxygen. In the same w'ay indigo-blue
is discharged during the decomposition of a nitrate by heat,
other kinds of organic matter being oxidized under like cir-
cumstances ; in these instances the decomposition of the ni-
tric acid is much facilitated, — 1. by the affinity of the base for
oxygen ; 2. the affinity of the organic matter for oxygen,
which unites with it at the elevated temperature. There are
many similar instances of this kind, where the behaviour of
NO2 or NO4 as an assistant is too clearly contrasted with the
action of other bodies to permit mistake. Thus urine when
kept is unfit for the preparation of urea, that substance ha-
ving been converted into carbonate of ammonia during the
action of the air upon the mucus or colouring matter con-
tained in the fluid. Colourless nitric acid unites wdth urea
and may be heated with it without decomposition ; but nitric
acid containing any of the lower oxides of nitrogen, such as
NOg or NO4, immediately decomposes urea into carbonic acid
and ammonia*. We cannot conceive that a lower oxide can
more readily oxidize urea than a higher oxide, and hence we
can only view the NO4 as aiding the urea to oxidize itself, as
the mucus does in urine. In the same way, the action of
pure nitric acid on colourless uric acid is to form alloxan, if
the operation has been conducted so as to prevent the forma-
tion of nitrous acid (NO4) during the oxidation. But if NO4
has been evolved, or if the colouring matter of the urine be
still contained in the uric acid, the products are only carbo-
nate and oxalate of ammonia. The colouring matter of the
urine and NO4 are thus seen to possess a similar action,
which is exactly the same as that of protonitrate of manga-
nese on a mixture of starch and nitric acid, no oxalic acid
being formed in the presence of this salt, the only product of

* A solution of urea in nitric acid is immediately decomposed with lively
effervescence when a little NOj is passed through it.

produced by Catalytic Bodies. 199

oxidation being carbonic acid. The NO4 or NOg acts in these
cases clearly by aiding the compound ready to oxygenate, but
which, under the conditions, has not sufficient power to de-
compose the nitric acid without additional aid. The same ex-
planation probably applies to the singular discovery of Pro-
fessor Graham*, that the addition of NO4 to non-accendible
phosphuretted hydrogen renders it inflammable. In this case
the two combined affinities produce the union of oxygen
with one of the bodies. The presence of the small quantity
of another compound of phosphuretted hydrogen in the spon-
taneously accendible gas, as described by Leverrierf and by
ThenardJ, may probably act in the same manner.

The action of this compound (PHc2) corresponding to ami-
dogen (NH2) may be conceived so to disturb the attraction of
the phosphorus to the hydrogen in the gas PH3 as to produce
the inflammability. Both the elements of this gas are highly
combustible, uniting with oxygen at a low temperature.
Their mutual attractions are sufficiently strong to prevent the
oxygen breaking up this union ; but when the second body
is present, the desire of PHg for another atom of hydrogen
may be supposed so far to draw the third atom of hydrogen
from the PHg, that oxygen has now the power to unite with
the two inflammable elements. In disturbing the existing
equilibrium, it is presumed to act just as a spark would do by
elevating the already strong affinities of the two elements for
oxygen. When a solution of hypochlorite of lime is poured
into a solution of muriate of ammonia in excess, a very pun-
gent volatile compound results, which has no bleaching pro-
perties, and therefore does not contain hypochlorous acid.
The decomposition is expressed by the equation NH4 CI +
CaO, CIO = NH2 CI + 2HO + Ca CI. The volatile com-
pound NHg CI has an affinity for hydrogen in order to pass
into NH4 CI. This body was well-fitted to test the view of the
cause of the inflammability of phosphuretted hydrogen (even
supposing PH2 is not spontaneously inflammable, as it is
stated to be by Thenard), On placing gas (which had en-
tirely lost its inflammability by standing several days over
water), in contact with the above mixture, in about an hour it
acquired the property of smoking strongly in the air, although
it did not inflame spontaneously. This showed that the
affinity of PH3 for oxygen was much elevated, although the
attraction was not sufficient for inflammation.

* Trans, Royal Soc. Edin. vol. xiii. p. 5.
t Ann. de Ch. et de Phys. Ix, 174.

X Comptes Rendus des Seances de VAcademie des Sciences, t. xviii.
pp. 252, Ol-l ; t. xix. p. 313.

200 Dr. Playfair on Transformations

There cannot be any doubt that the atoms of a body may
be placed in a greater or less degree of tension by varying
conditions. The experiments of Mr. Joule* and myself on
AUotropism have fully proved that the space occupied by the
same body alters under different circumstances. It is there-
fore not an unreasonable assumption that the affinity of one
body for a particular element may be sufficiently great to
produce a tense state of the atoms without effecting decom-
position! : hence the added affinity of a second body acting
in the same direction may cause that change which each alone
could not effect. Anything that disturbs the state of statical
equilibrium in such a body will often effect its decomposition.

This accessory affinity is recognised when both bodies
enter into union. Charcoal and chlorine decompose alumina
at a red heat, though neither can do so separately. In the
same way BoudaultJ has shown that a mixture of potash
or soda and red prussiate of potash oxidizes various me-
tallic oxides, while Mercer has for many years made use
of this mixture to discharge indigo-blue on calico §. Red
prussiate of potash (FcgCygSK) has a great disposition to
attach to itself another atom of potassium to become yellow
prussiate of potash (Fe2Cyg4K). It cannot gratify this de-
sire without aid ; but when assisted by a substance having
an affinity for the oxygen of the potash, and capable of
appropriating it, decomposition follows. There are often
cases in which the body exercising the accessory affinity may
be unable to effect the union, either by the influence of un-
favourable chemical conditions or of cohesion or elasticity.
Thus, in the case with which we first started, the affinity of
protoxide of manganese for oxygen aids in the decomposition
of nitrate of protoxide of manganese, and sesquioxide of man-
ganese remains. If the temperature during the decomposition

* Memoirs of Chemical Society, vol. iii. p. 93.

t The alteration in volume is best seen in those oxides which contract
and increase in specific gravity by tlie application of heat, for example,
when the brown oxide becomes the green oxide of chromium. The two
oxides must have a different molecular constitution, and this may be sup-
posed to result from the elastic powers of one of its elements and the cohe-
sive force of the other. The first eflTect of heat on oxide of chromium
must be to expand the atoms of oxygen, and removing them further from
the two atoms of chromium, permit the cohesive attraction of the latter to
be gratified. Hence the compound acquires properties dependent upon
cohesion, such as indiflference to union and diminished solubility.

X Journal de Pharmacie, tome vii., 437. [Phil. Mag. Third Series, vol.
xxvii. p. 307.]

§ Phil. Mag. Third Series, vol. xxxi, p. 126. In justice to Mercer, al-
though this does not remove Boudault's claim of prioi'ity of publication, I
cannot refrain from stating that the former chemist pointed out to me tlie
oxidizing powers of the prussiates four or five years since.

produced by Catalytic Bodies. 201

be elevated, the oxygen resumes its elastic state and refuses to
form this higher oxide, as in fact we know is the case in Mer-
cer's experiment Avith oxalic acid and nitric acid, where the
presence of hot NO5 is an unfavourable chemical condition
to the existence of Mug Og, and therefore it is not formed,
but in its stead the oxygen is passed over to the organic mat-
ter, which is able to unite with it under the circumstances.
A similar instance of the effect of such conditions is seen
when the peroxides of copper, manganese or lead, are thrown
into a solution of bleaching powder. The affinity of these
oxides for an additional quantity of oxygen enables them to
decompose the hypochlorite of lime, converting it into chlo-
ride of calcium. When the protoxides are used, this liberated
oxygen unites and converts them to peroxides. The latter
themselves have sufficiently strong affinity for oxygen to cause
the decomposition to proceed ; but not uniting with it, pure
oxygen is given off in the gaseous state. Here elasticity has
come into play, and being more powerful than the feeble che-
mical affinity, causes the oxygen to escape as a gas. When
the solution is cool the gas goes off in a succession of small
bubbles ; but when hot, the escape is tumultuous, the heat
aiding the oxygen to enter into the elastic state*. A solution
of chloride of lime evolves oxygen slowly at the boiling-point ;
but the decomposition is much accelerated by the accessory
agents referred to.

The action of certain oxides upon peroxide of hydrogen is
exactly similar to that on a solution of hypochlorite of lime.
Thus peroxide of manganese, the protoxides of cobalt and
lead, minium, peroxide of iron, and the protoxides of nickel,
copper and bismuth, all exert this action on peroxide of hy-
drogen with a force indicated by their orderf. In none of
these cases does the oxide unite with a further proportion of
oxygen. The violence of the action is however in proportion
to their power of uniting with more oxygen. The first five
oxides in the list have higher oxides of definite composition
and of a certain degree of stability, with the exception of ferric
acid; while the protoxides of copper and bismuth, although
possessing the power of uniting with more oxygen, do not
present superior oxides of a marked character. We should have

* The best mode of instituting the experiment is to make a mixture of
chloride of soda and caustic soda, lieat this to a temperature near ebuUition,
and add sulphate of copper. The oxide of copper precipitated in the fine
state of division causes such a copious evolution of oxygen gas that the con-
tents are apt to be thrown out of the vessel : a mixture of chloride of lime
and lime, or the ordinary unfiltered bleaching-powder of commerce, are
also well-fitted to show the action.

f Thenard's Traite de Chetnie, 6th edit. vol. i. p. 216.

202 Dr. Playfair on Transformations

expected the oxides of nickel and cobalt to have exerted the
same power, but from Thenard's description of the former
being in the state of a black powder, it may have been the
oxide of increased specific gravity, to which attention has
already been drawn*. In all these cases the affinity is sup-
posed to be sufficiently strong to break up the atoms of a
body yielding to the slightest disturbance of its state of stati-
cal equilibrium. Two affinities are at play in these decompo-
sitions, viz. the attraction of the metaUic oxide for oxygen
and that of the water for the same body ; both these affini-
ties resist the union, and therefore, elasticity coming into
operation, robs both oxides of the gas. The affinity causing
the decomposition is so slightly preponderating in its in-
fluence, that a second cause coming into operation is quite
sufficient to alter the conditions under which it was originally
exerted, and to draw one of the elements of the body acted
upon beyond the sphere of its affinity.

The balance of affinities in all such cases is so near that we
not unfrequently find apparently contradictory effects result-
ing from their gratification. Thus the addition of oxide of
silver to peroxide of hydrogen expels oxygen from the latter,
but at the same time it is robbed of its own oxygen and re-
duced to the metallic state. In this case we have two feeble
compounds instead of one, with affinities very nearly balanced,
and with atoms so tense as to yield readily to the first dis-
turbing cause. We can scai'cely adopt as sufficient the ex-
planation of Thenard and Mitscherlichtj that the reduction
is due to the elevation of temperature accompanying the de-
composition, because even when that is lowered by the ad-
dition of much water to the peroxide of hydrogen, the silver
still becomes metallic.

It is a point yet undetermined, whether a lower oxide is to
be considered as unity to a higher oxide, or whether all the
atoms of oxygen are held by equal attractions. We know
that tartaric acid is able to separate potash from nitric acid
in forming a bitartrate, and yet acetic acid is sufficient to re-
move the second atom of potash from the neutral tartrate.
But in a bibasic acid, like tartaric acid, it may be either atom
of potash that is abstracted, and the superior affinity for the
remaining one may be owing to attractions resulting after
the expulsion of the first. Thus MnOg may have its atoms
of oxygen distributed round the central nucleus Mn, and held
by equal attractions, and the stability of the red oxide pro-
duced by its calcination does not show that it pre-existed

* Memoirs of the Chemical Society, vol. ii. p. 381, and vol. iii. p. 81.
f Poggendoi-fF's ^nwa/e«, Iv. 321.

produced by Catalytic Bodies. 203

in the black oxide, but merely that the attraction became
stronger when one of the elements which divided it was
removed. If it be admitted that the attraction of a radical
for oxygen is equally divided between all the atoms of that
element associated with it, the action to ^^ hich we have al-
luded becomes comprehensible. In an oxide mc have the
attraction of affinity opposed by the elasticity of its oxygen
and by the cohesive force of the metal. If a be the attraction
of the central nucleus or radical, c the cohesive force of the
metal, and e the elasticity of the oxygen, then the molecular

formula of a protoxide will be , of a sesquioxide — j

and of a binoxide r— . Now if, as in oxide of silver, the a

c + 2e

and e are nearly equal, or the a only slightly preponderating,
and the c or cohesive force very powerful, we can readily con-
ceive that the added force of a second e may overcome the
small amount of preponderating force in favour of a. Thus,
when oxide of silver is placed in contact with peroxide of
hydrogen, its affinity for more oxygen is sufficient to draw
the second atom of oxygen beyond the sphere of attraction of
H, and deliver it over to its own elasticity. But in doing
this the attraction of silver for oxygen has been divided be-
tween its own oxygen and that of the peroxide of hydrogen.
Scarcely at any time capable of retaining its own oxygen,
this division of its attractive force has been fatal to the exist-
ence of its oxide, and the water in statu nascens at the same
time exerting an affinity for the oxygen just ready to escape ;
all these causes combined result in the reduction of the
silver*.

When pyruvic acid is in contact with oxide of silver, it
unites and forms a salt; but when acting on carbonate of
silver, a certain quantity of oxygen also leaves the oxide
during the escape of carbonic acid, and metallic silver re-
mains-}-. As Liebig J suggests, motion may aid this result;
but were this the only explanation, we should expect that

* During the passage of this paper tlirough tlie press, Mr. Brodie, in a
lecture at the Royal Institution, showed that peroxide of potassium reduces
chloride of silver, the two atoms of oxygen passing off in the gaseous state,
while chloride of i-utassium and metallic silver remain behind, a singular
decomposition, when the behaviour of potash is remembered. But the
action is strictly the same as that here described ; the atoms of oxygen, -
being liberated at the same time, are presented to the silver, which, dividing
its attractive force between them, is not able to overcome the influence of
elasticity of the oxygen and its own cohesion, and therefore remains in a
metallic state.

f Berzelius, Lehrhuch der Chemie, fifth edit. vol. iv. p. 231.

I Chemistry of Agriculture, 4th edit., page 283.

204 Dr. Playfair on Transformations

silver would constantly be reduced during the action of other
feeble acids on carbonate of silver. If, however, we suppose
that the pyruvic acid, Cg Hg O5, from its affinity for more
oxygen, exerts an attraction for that element at the moment
of the liberation of the carbonic acid, the decomposition
would be similar to those we have already considered, espe-
cially if the previous view of the molecular constitution of
salts be admitted. In that case the oxygen of the oxide being
attached to that of the carbonic acid, will be made highly
tense during the escape of the latter, and may therefore be
detached by a very feeble force, its elasticity finally over-
coming the weak affinity. An extension of the explanation
however strikes me as more probable, but it would be prema-
ture to insist upon it without being supported by experiments
which I have not yet been able to conclude.

The action of metals and of charcoal on peroxide of hydro-
gen may be explained by the same feeble affinity. Alkalies
also, from their attraction for oxygen, as indicated both by
their capability of uniting with more oxygen and by their
basic power or disposition to attach themselves to a com-
pound behaving as an oxygenous or chlorous element, favour
the decomposition of HOg, while acids, on the other hand,
render it more stable, perhaps, as Thenard himself suspected*,
from there being an inferior oxide (Hg O3?). In this instance
the elasticity of the oxygen tends to conceal the play of affi-
nities by preventing combination.

When the acting body is present in large quantity, or ex-
hibits an increased surface, the action goes on with propor-
tionate rapidity. Thus, when nitric acid is in contact with
starch, the action is moderate until a certain quantity of per-
oxide of nitrogen has been evolved by decomposition, after
which it proceeds with a violence difficult to control. The per-
oxide of nitrogen surrounding every particle of starch aids it
in the decomposition of the nitric acid. That this is the real
cause of the phaenomenon may be proved by the following
simple experiment. Nitric acid is heated with starch to a tem-
perature at which the action has a tendency to commence but
has not yet begun. A stream of N04or NOg is then passed
through the liquid, when action immediately begins with an
activity proportionate to the quantity of gas added. The ele-
vation of temperature due to the progressive action influences
the decomposition, by causing the atoms of nitric acid to be-
come more tense. Exactly the same accessory affinity is
used by the manufacturer of oxymuriate of tin, when he
adds a fragment of tin to the mixture of chloride of tin and
* Traite de Chemie, p. 211.

produced by Catalytic Bodies. 205

nitric acid. The tin eliminating some nitric oxide quickens
the action, which commences with difficulty with pure nitric
acid ; nitric oxide gas passed through the solution answers
the same purpose.

This accessory affinity also enables oxide of copper or per-
oxide of manganese to evolve copious streams of oxygen from
chlorate of potash in a state of fusion. The heat of fusion
decomposes the compound slowly, but on adding a body ha-
ving an affinity for the element acted upon by the heat (oxy-
gen), the decomposition proceeds with greatly increased ra-
pidity. We cannot ascribe this action to the presentation of
points from which the gas may escape, as in the lowering of
the temperature of ebullition by particles of sand, because
silica has no influence in accelerating this decomposition*.

In the examples previously given we have the decomposi-
tions aided by the tendency of one of the bodies to assume
the elastic form. But when the body acted upon has two
elements, one of which is influenced by elasticity, the other
by cohesion, we find it peculiarly liable to be acted upon by
external agents. Persulphuret of hydrogen is a compound of
this class, and has been closely studied in its decompositions
by Thenardf. The same bodies which decompose peroxide
of hydrogen act catalytically upon this sulphuret. The de-
composition cannot be due to points for the escape of gas, as
suggested by LiebigJ, to explain the decomposition of per-
oxide of hydrogen, because solutions of the alkalies act with
equal power. The sulphurets, especially those of the alkaline
metals, decompose it very readily. As in the case of per-
oxide of hydrogen, the acids afford stability to its sulphur
analogue. In the view of acids given, they are supposed to
have become chlorous or electro-negative, representing and
behaving as oxygen, and therefore exerting no affinity, we
should anticipate that they would not show any disposition
to break up an oxygenous compound or its analogue of sul-
phur. Another instance of accessory affinity is seen in the
nitrosulphates § ; the formula (RO, SO^ + NOg) given by Pe-
louze to these compounds does not allow us to understand
their decompositions, which however becomes intelligible if
we view nitrosulphuric acid as nitric acid, in which the fifth
atom of oxygen has been replaced by one of sulphur (RO,
NO4 S). In this acid we have two elements — the nitrogen
and the sulphur — sharing the oxygen, their mutual affinities
being nearly balanced when the acid is united with an al-

* Taylor's Scientific Memoirs, vol. iv. p. 9.

t Ann. de Ch. et de Ph. xlviii. 79. + Ann. der Pharm, ii. 22.

4 Ann, de Ch. et de Ph. Ix. 151.

206 Dr. Playfair on Transformations

kali, although in a free state, the sulphur exhibits a superior
affinity, as shown by the decomposition which then results,
NO4S = NO + SO3. Now any substance which acts as an
accessory to the sulphur by aiding the withdrawal of oxygen
from the nitrogen decomposes it. This instability is especially
exhibited in NH4O, NO4S ; the 3 atoms of hydrogen of the
ammonia in their attraction for oxygen introducing another
affinity, which accelerates decomposition. And, in fact, we
do find that the same agents which so readily decompose the
oxygenous compounds, chloride of lime and peroxide of hy-
drogen, do equally cause the disruption of nitrosulphate of
ammonia into protoxide of nitrogen and sulphate of ammonia.
Alkalies are an exception to this rule, as they render the ni-
trosulphates more stable, while they make the peroxide of hy-
drogen prone to decomposition ; but the cases are different,
the latter substance having none of the properties of an acid.
The basic character of alkalies, defined as their power of
uniting with more oxygen, or with an acid playing the part
of an oxygenous element, is illustrated by several curious
decompositions. Thus, though grape-sugar reduces sulphate
of copper with ease, cane-sugar alone does not readily do so,
but M'hen mixed with potash and boiled with the salt, sub-
oxide of copper is produced, as in the mode of preparation
of that oxide suggested by Boettger*, or the reduction of
chloride of silver as proposed by Levolf. Here the disposi-
tion of the organic matter to unite with oxygen is able to
gratify itself when aided by the accessory affinity of the pot-
ash for oxygen. That the potash in this state acts by aiding
the oxidation, is seen by heating CugO with a solution of
caustic potash, exposed to the air, when it oxidizes much
more rapidly than when boiled with water alone J. When
suboxide of copper is dissolved in ammonia it oxidizes with
surprising rapidity. In this instance the hydrogen of the
ammonia adds to its disposition as an alkali to absorb oxy-
gen. The quick oxidation is not merely due to the fact of

♦ Ann. der Pharm. und Chemie, xxxix. 176.

f Berzelius, Jahreshericht, vol. xxv.

X This experiment may be simply made as follows : — Three shallow eva-
porating basins of the same size and form, each containing the same quan-
tity of suboxide of copper, are taken, and to one is added a solution of
potash or soda; to the second, a solution of chloride of manganese; to the
third, common water, taking care that the same volume of each fluid is
added. The whole are now placed on a sand-bath, so as to be exposed to
equal temperatures, and stirred occasionally. The suboxide of copper in
the basin containing chloride of manganese oxidizes very rapidly ; that in
contact with the potash more slowly; and that with simple water is scarcely
effected when both the others have lost their red colour. These actions are
strictly in accordance with theory.

produced by Catalytic Bodies. 207

the suboxide being in a state of solution, because the soluble
salts of the suboxide do not oxidize with such extraordinary
ease, nor is it to be expected that they should, if we admit that
the acid itself plays the part of oxygen. The accessory affi-
nity of alkalies for oxygen is exhibited in many other cases of
chemical action. Thus, colouring matters, such as deoxidized
logwood, Brazil-wood, peach-wood, japan, fustic and catechu
are oxidized more rapidly in contact with alkalies than in
water alone ; and various dyeing principles, such as orcine and
erythrine, absorb oxygen with great avidity in the presence of
ammonia. Sugar may be boiled with potash without de-
composition, but when air is admitted, formic, melassic, and
glucic acids are produced. Hydruret of benzyle when ex-
posed to air gradually absorbs oxygen and passes into benzoic
acid, but in contact with potash this absorption is very much
accelerated. The rapid decomposition of the gallates and of
hematine in the presence of free alkali and air is aphaenome-
non of the same kind. In fact, numberless instances of this
catalytic action of the alkalies are known to chemists.

We find the influence of an accessory oxidation in many
cases of chemical union. Thus Campbell has shown * that
the transformation of cyanide of potassium into cyanate of
potash is much accelerated by the presence of the iron in yellow
prussiate of potash, the iron being converted into oxide during
the transformation. Here the iron plays the part of the
protoxide of manganese in the cases of oxidation already re-
ferred to, or it perhaps bears a more direct relation to the
action of lead in communicating a tendency to the base metals
to seize oxygen during the process of cupellation. The in-
fluence exerted by peroxide of manganese in first converting
cyanide of potassium into cyanate of potash and afterwards
into the carbonate of that base, is another instance of accessory
affinity ; for only a portion of the oxygen is derived from the
oxide employed. The solution of an alloy of silver and pla-
tinum in nitric acid may be supposed to be a similar affinity.
It is not necessary to believe that this is a case proving the
communication of intestine motion to the atoms of platinum,
by which it acquires the power of decomposing nitric acid f ;
for an equally simple explanation is given by assuming that
the united affinities of platinum and silver are able to decom-
pose nitric acid, both these affinities acting in one direction
at the same time, and enabling the platinum to dissolve. We
have only to suppose that the atoms of nitric acid are placed
by the silver in a state of such tension that the platinum can

* Phil. Mag. Third Series, vol. xix. p. 513.

t Liebig's Elements of Agriculture, 4th edit., p. 280.

208 Dr. Playfair on Transformations

now seize oxygen, which it could not do from the nitric acid
when in a less tense state. The quartation of gold is ob-
viously a phaenomenon of the same kind. In these instances
the interposing silver much reduces the cohesive or aggregative
force of the platinum or gold, which opposes so strongly the
action of nitric acid upon them. But when we have every
atom of platinum or of gold separated by one of silver, great
facility is given to the nitric acid to act upon these metals,
especially when the silver at the same time aids them by its
assistant affinity.

We have seen, in the consideration of the previous in-
stances of catalysis, that the play of affinities was occasionally
so nearly balanced, that a second disturbing cause determined
the direction of the action. In the case of non-accendi-
ble phosphuretted hydrogen, the addition of another oxi-
dizable body, NO4, decided the union of oxygen with the gas.
In accendible phosphuretted hydrogen the compound PHg
played the same part. When the accessory agent is present
in small quantity, the preponderating affinity of the body
acted upon shows itself in the result. But, as the ac-
tion is due to two affinities nearly equal in amount, it is easy
to conceive that the increased quantity of the accessory agent
may exactly balance affinities, and that the catalytic phaeno-
menon will be prevented. Thus one-twentieth of the volume
of binoxide of nitrogen, according to Graham*, added to ac-
cendible phosphuretted hydrogen, does not deprive it of in-
flammability, the bubbles of gas escaping into the air with a
kind of explosion, although one-tenth volume of the same
gas altogether prevents the accendibility. This nitric oxide,
when pure, does not, like protoxide of nitrogen, render phos-
phuretted hydrogen spontaneously inflammable, the reason
obviously being that its own affinity for oxygen is more
powerful than that of the phosphuretted hydrogen. When
added however in such small proportion to the accendible
gas that the foreign constituent in it preponderates, then it
becomes an accessory to the oxidation, though an increase of
the quantity renders it more powerful, and prevents accendi-
bility by itself seizing oxygen. Thus also larger volumes of
gas, having an affinity for oxygen, but incapable like NO4 of
gratifying that desire under ordinary circumstances, may
exactly balance the feeble affinity of the foreign accessory
body and prevent oxidation. Five volumes of hydrogen,
2 volumes of carbonic acid, 1 volume of olefiant gas, and
1 volume sulphuretted hydrogen, deprive 1 volume of phos-

* Phil. Mag., Tliird Series, vol. v. p. 405.

produced by Catalytic Bodies. 209

phuretted hydrogen of its spontaneous inflammability*. The
very conception of a catalytic agent, on the view adopted,
implies the exertion of an affinity, which is passed over or
added to that of the body acted upon. If, therefore, a third
body claim this added affinity, the increase of power being
divided, may be insufficient to exert the force which it did
when wholly applied to aid the affinity of one body. It may
be this balancing of affinities which prevents the action of
platinum on a mixture of oxygen and hydrogen. The plati-
num by its surface affinity condenses oxygen, and presenting
it to hydrogen in a condensed form produces union. But in
the presence of small quantities of certain oxidizable gases,
such as sulphuretted hydrogen, carbonic oxide, and defiant
gases f , it ceases to exert this action, the assumption in this
case being that the affinity of the added gases for oxygen
balances that of hydrogen for the same gas.

This balancing of affinities may account for several phaeno-
mena otherwise inexplicable. On the decay of vegetable
mould we find the hydrogen constantly diminishing in quan-
tity until a certain period of decomposition, when the affinity
of the carbon of the humus for its hydrogen balances the
affinity of the surrounding oxygen. It seems to be the same
balancing of affinities which renders corrosive sublimate so
antiseptic in its properties ; but, in this case, the balance re-
sults from the affinity of the second atom of chlorine in the
bichloride of mercury for the hydrogen of the organic sub-
stance, thus preventing its union with oxygen. It is probable
that the same affinity of chlorine for hydrogen causes turpen-
tine and the volatile oils to act cataly tically in exploding chlo-
ride of nitrogen. The chlorine attracted by the hydrogen of
these substances is drawn without the sphere of its attraction
for nitrogen, and a disruption of the elements consequently
ensues, compounds such as this resting on the very verge of
separation between physical and chemical attraction. The
antiseptic action of corrosive sublimate is very different from
that exerted by sulphurous acid and sulphate of iron, these
bodies acting by their superior affinity for oxygen, and neu-
tralizing the power of the ferments or accessory oxidizers
present in the organic body.

There is no difficulty in applying these notions of catalysis
to organic compounds, which from the complexity of their

• The influence which the vapours of turpentine exert in jjreventing
the oxidation of phosphorus in the air is probably another instance of this
balancing of affinities.

t Faraday, Phil. Mag., Third Series, vol. v. p. 405 ; Turner, Jameson's
Journal, xi. 99 and 311.

Phil. Mag. S. 3. Vol. 31. No. 207. Sept, 184.7. P

210 Dr. Play fair on Transformations

molecules are peculiarly liable to change. If it once be ad-
mitted that an assisting affinity may exist in the sense defined
in the present paper, then we see the same cause operating
upon organic as well as inorganic molecules. When nitric
acid acts on oxalic acid or starch, an inorganic body (a pro-
tosalt of manganese) lowers the temperature necessary for
the oxidation, and exerts its influence until all the starch
is converted into carbonic acid, being equally efficacious
on the addition of more nitric acid and starch. Here the
body acting as an assistant remains unchanged, and there-
fore continues its action ad infinitum, rendering it impossible
to prepare oxalic acid from nitric acid and starch or sugar,
carbonic acid being the only product*. Had the assistant
oxidizer passed from solution during the progress of the oxida-
tion, it could not of course continue its favourable effect, and
a new portion of it must have been added. Here the inor-
ganic salt enables the sugar to oxidize itself from the sur-
rounding medium just as yeast does, the only difference being
that the yeast itself suffers change, and therefore can only
continue its action for a limited period. It is exactly in the
same condition as a mixture of nitric acid and binoxide of
nitrogen made to act on protochloride of tin. A. small portion
of the latter added to such a mixture is oxidized, but when
the solution is heated until all the NOg is expelled, oxidation
does not ensue on the addition of a new portion at the same
low temperature as before. Now Saussure and Colin have
shown that j^east only induces fermentation when it is in a
position to absorb oxygen. It acts therefore strictly as bin-
oxide of nitrogen, or a protosalt of manganese, in the previous
instances, by adding its affinity for oxygen to that of the
sugar, the added affinities of both completing the union. The
only difference between these two decompositions is, that in
one case the oxidizing agent is nitric acid, in the other it is
water. The composition of sugar shows it to contain the
elements of alcohol and carbonic acid minus an atom of water.
In such a compound we have the affinity of carbon for hy-
drogen and of carbon for oxygen. The yeast by its nitrogen
also exerts an affinity for hydrogen, and by its carbon for
oxygen. The united affinities of the sugar and of the yeast
acting upon water decompose it, its elements on their libera-
tion being shared by the carbon of the sugar, for which it
may be supposed to have the strongest affinities, Cjg IIn On

• In this it resembles the action of oxalic acid in converting an unlimited

quantity of oxamide into oxalate of .ammonia, with this difference, that the
oxalic acid, which causes the change, may not be the same, but a regene-
rated portion, while the salt of manganese always remains unchanged.

produced by Catalytic Bodies. 211

+ HO=4C02 + 2 (C4 Hg O2). To show the exact similarity
of the two processes of oxidation when the assisting body is
either organic or inorganic, I may cite the curious manufac-
turing process for oxidizing oils in the method of dyeing
Turkey-red used in this country, and included in Mercer's
patent for that colour. It consists in oxidizing oils by blow-
ing hot air through them, the oils being in contact with a
solution of a salt of copper or of bran ; the contact of either
of these solutions is found very materially to accelerate the
oxidation. The catalytic action of oxide of copper in evolving
oxygen from hypochlorite of lime was adduced as showing its
affinity for more oxygen, and this feeble affinity is well known
and used empirically by all calico-printers, who are in the con-
stant habit of mixing a salt of copper with their colours for
the purpose of ageing them more speedily ; in other words, of
causing them to unite with oxygen. This also is the assisting
cause in Mercer's process for oxidizing oils ; bran in solution
answers the same purpose from its affinity for oxygen. The
addition of common salt or muriate of ammonia favours the
oxidation in all the cases referred to, the oxidation proceeding
much more quickly in their presence. No sub-chloride is
ever formed, the action being purely catalytic, and probably
depending on the conversion of the salt of copper into a
chloride, the chlorine of which may be supposed to exert
a slight affinity for the hydrogen of the compound, thus
withdrawing it somewhat from the sphere of its own special
attractions in the body; the copper now aiding the chlorine,
delivers the hydrogen more easily into the power of the oxy-
gen of the atmosphere. It is therefore immaterial whether
the body exercising the assistant affinity be organic or inor-
ganic, if the conditions be favourable to the exercise of this
influence. The action of a body in acetous fermentation on
the transformation of bi'andy into vinegar must be recognised
as a phenomenon of a like kind. We know that brandy may
trickle m ithout change over a large surface of wood shavings,
through which air circulates at the heat of the human body,
but that it is quickly converted into vinegar if brandy in the
act of oxidation be mixed with it. Here the added ferment
exerts its assisting affinity in precisely the same way as the
salt of copper, when it aids the oxidation of oils or colours,
or as protonitrate of manganese or peroxide of nitrogen
during the oxidation of starch. The conversion of hydrogen
and oxygen into Mater by the action of fermenting silk, cot-
ton, or woody fibre, as observed by Saussure, is obviously a
phaenomenon of the same kind, and can only be exerted
slowly and in the immediate vicinity of the assisting oxidi-

P2

212 Dr. Playfair oh Transformations

zers, just as a ball of spongy platinum silently effects the
union of these two gases.

In these cases we must admit that the action is indepen-
dent of a state of intestine motion of the atoms of one com-
pound molecule imparted to those of another, or, if we do not
allow this, we must create two new powers and separate de-
compositions caused by inorganic bodies from those produced
by organic compounds, although all the phaenomena of the
decomposition show them to belong to one category.

In a body in a state of such incessant change as the blood
of living animals, it would naturally be expected that an added
agency, such as that described, would render it prone to
abnormal actions and oxidations, and in fact we do recognise
by all the recent progress in the study of public hygeine
that the addition of any oxidizing miasm or putrid matter
to the blood does produce those changes which are known
by their results in the different forms of disease. These and
other catalytic agents no doubt exercise most important in-
fluence on the processes of animal life and on the action of
medicaments on the system, but it would be foreign to the
object of this paper to examine them in detail.

The limits of a paper such as this compel me to avoid
including many other instances of catalytic decompositions
which come under this explanation, or of drawing special
attention to those which cannot be included in the present
state of our knowledge. Thus diastase, acting on starch,
converts it into sugar, but we have so little knowledge of
the composition or properties of the first body, that it would
be unwarrantable to embrace a case such as this. But
in analogous changes produced by bodies which are under-
stood, the same power is recognised. Sulphuric acid in
converting starch into grape-sugar offers an example of
combination Avhich may fairly be examined by the same
method employed in investigating other decompositions.
Graham has shown* that heat is evolved even on the addi-
tion of the 48th atomic proportion of water to sulphuric acid,
or, in other words, that the affinity of that acid for water is
not gratified as long as our instruments of research can follow
the change. This is merely another proof of the doctrine
with which I started, that there is no evidence of such a
complete gratification of affinity as ever to merge entirely the
attractions of the elements of any body. In the case referred
to, the development of heat on each successive addition
proves that the water is condensed on entering into union
with the acid. When the heat of the sulphuric acid is arti-
* Phil. Mag. Third Series, vol. xxii. j). 334.

produced by Catalytic Bodies. 213

ficially increased, this compound is broken up, for distil-
lation drives off the water and concentrates the acid. Now
when starch is in the presence of this weak combination
of sulphuric acid and water, at a temperature at which the
latter is just able to exert its affinity and again have it de-
stroyed by heat, it is not at all extravagant to suppose that
the starch may seize the water in its nascent state at the mo-
ment of expulsion, or even that it may be able to unite with
the last atoms of the series of acid and water when presented
in that condensed state, although it cannot do so when the
water is free and not nascent. Any such union would explain
the transformation of starch into grape-sugar, the change
merely being in the acquisition of water, Cjg Hiq Oio + 4HO
= C,2 Hi4 O14. The action here is not the same, but the very
reverse of that which ensues in the preparation of aether. In
the one case the sulphuric acid abstracts water, in the other
it is the means of adding it, and the difference of the action
depends on the relative strength of the acids employed.
Without at all giving an opinion in favour of the necessity
for the formation of sulphovinic acid, as supposed by Liebig*,
or as to its not being an essential condition, as argued by
Mitscherlichf, the final result is simply of the order now
under consideration. In this decomposition the sulphuric or
phosphoric acid is so strong that it combines with the water
instead of yielding it, and the elevation of temperature essen-
tial to the change may either be due to the formation and
after decomposition of sulphovinic acid, or it may be simply
owing to the necessity of rendering the molecule of alcohol
tense by heat, the elasticity of the aether and water both
tending to break up the hydrate, the decomposition of which
is determined by the presence of the strong acid now also
aiding and abstracting the water. The final result is certainly
purely catalytic in whatever light it is considered, although
there may be more than one step in the process.

In conclusion, facts have been brought forward to show
that there is at least as m\ich probability in the view that the
catalytic force is merely a modified form of chemical affinity
exerted under peculiar conditions, as there is in ascribing it
to an unknown power, or to the communication of an intes-
tine motion to the atoms of a complex molecule. Numerous
cases have been cited in which the action results when the
assisting or catalytic body is not in a state of change, and
attempts have been made to prove by new experiments that
the catalytic body exercises its peculiar power by acting in

* Geiger's Pharmacie, vol. ii. p. 711 ei seq.
t Lehrbuch der Chemie, vol. i. p. 247 et seq.

^14 Sir W. Rowan Hamilton on Qiiaternions,

the same direction as the body decomposing or entering into
union, but under conditions in which its own affinity cannot
always be gratified. The catalytic body is therefore a sub-
stance which acts by adding its own affinity to that of an-
other body, or by exerting an attraction sufficient to effect
decomposition under certain circumstances, without being
powerful enough to overcome new conditions, such as elasti-
city and cohesion, which occasionally intervene and alter the
expected result.

At the same time the theory is far from being fully proved ;
but if I have succeeded in rendering probable that the ca-
talytic force is only chemical affinity recognised under an
aspect which chemists have not been accustomed to view it,
and exerted under conditions which can only be developed
by close attention to details, it will not have been useless to
direct increased study to this interesting class of phenomena.

XXXVI. 0« QiiaterJiions / or on a Nexv Si/stem of Imaginaries
in Algebra. By Professor Sir William Rowan Hamilton,
LL.D., V. P.R.I. A.^ F. R.A.S., Correspofidhig Member of the
Institute of France^ and of other Scientifc Societies in British
and Foreign Countries^ Andrews^ Professor of Astronomy in
the University of Dublin ^ and Royal Astronomer of Ireland.
[Continued from vol. xxx. p. 461.]

33. T^OR the sake of those mathematical readers who are
•*- familiar with the method of co-ordinates, and not with
the method of quaternions, the writer will here offer an inves-
tigation, by the former method, of that general property of
the ellipsoid to which he was conducted by the latter method,
and of which an account was given in a recent Number of
this Magazine (for June 1847).

Let X y z denote, as usual, the three rectangular co-ordi-
nates of a point, and let us introduce two real functions of
these three co-ordinates, and of six arbitrary but real con-
stants, I mnV m' n', which functions shall be denoted by ic and
V, and shall be determined by the two following relations:

u{ll' -f mm' + nn') = Vx + fri'y + n'% ;

i^{ll' 4- mm' + nn^Y = [ly — mxf + [mz — nyf + [nx — hf ;

then the equation

u^^v'^=i\ (1.)

will denote (as received principles suffice to show) that the
curved surface which is the locus of the point x y zhtm ellip-
soid, having its centre at the origin of co-ordinates ; and con-
versely this equation u'^ + i^=l may represent any such ellip-

Sir W. Rowan Hamilton on Qjiaternions, 215

soid, by a suitable choice of the six real constants ImnV rri 11!,
At the same time the equation

will represent a system of two parallel planes, which touch
the ellipsoid at the extremities of the diameter denoted by the
equation

v = 0',

and this diameter will be the axis of revolution of a certain
circumscribed cylinder, namely of the cylinder denoted by
the equation

the equation of the plane of the ellipse of contact, along which
this circular cylinder envelopes the ellipsoid, being, in the
same notation,

M=0:

all which may be inferred from ordinary principles, and agrees
with what was remarked in the 29th article of this paper.

34. This being premised, let us next introduce three new-
constants, p, g, r, depending on the six former constants by
the three relations

2p = l + l'j 2q = m + m'f 2r=zti-\-n'.

We shall then have

I'x + m'j/ + n'z = 2 {px -\-qi^-\-rz)— [Ix -\-my-^ nz) ;

and the equation (1.) of the ellipsoid will become

{W -\-mm' -^nn'f
- {p + m^ + n^)(^x^ ^f + 2'^)

—^{lx-\- my + nz)[px -\-qy-{-rz)
+ ^{px-^qy-\-rzY

if we introduce three new variables, of, y\ s', depending on
the three old variables .r, y^ z, or rather on their ratios, and
on the three new constants p, q, r, by the conditions,

*' _ V _ ^' _ 2{px-{-qy + rz)
x~ y z x^-\-y^-\-z^

These three last equations give, by elimination of the two
ratios of ,r, y, z, the relation

^'2 ^ya ^ ^n = 2(/?y + qy' + rz') ;

the new variables a/, y\ z' are therefore co-ordinates of a new
point, which has for its locus a certain spheric surface, passing
through the centre of the ellipsoid ; and the same new point

216 Sir W. Rowan Hamilton on Quaternions.

is evidently contained on the radius vector drawn from that
centre of the ellipsoid to the point .r 3/ g', or on that radius
vector prolonged. We see, also, that the length of this radius
vector of the ellipsoid, or the distance of the point x y z from
the origin of the co-ordinates, is inversely proportional to the
distance of the new point a^ y' z' of the spheric surface from
the point / m n, which latter is a certain fixed point upon the
surface of the ellipsoid. This result gives already an easy
and elementary mode of generating the latter surface, which
may however be reduced to a still greater degree of simplicity
by continuing the analysis as follows.

35. Let the straight line which connects the two points
a' y z' and / m n be prolonged, if necessary, so as to cut the
same spheric surface again in another point a?" y ^": we shall
then have the equation

from which the new co-ordinates a/', y", 2" may be eliminated
by substituting the expressions

;r" = Z+^(^'-/), i/' = m + t{y'-m), z"=n + l{z'-n);

and th^ root that is equal to unity is then to be rejected, in
the resulting quadratic for t. Taking therefore for t the pro-
duct of the roots of that quadratic, we find

Z^ + m^ + n^ — 2(lp + mq-\- nr)
therefore also, by the last article,

X^ -\- 7f -{■ Z^

consequently

,2_ x^ + y^-\-z^

and finally,

{x"-l)^-\-{y"-m)^ + {is"-n)^=x'' + y^ + z\ . (2.)
Denoting by a, b, c, the three fixed points of which the
co-ordinates are respectively (0, 0, 0), (/, m, w), (p, q, r) ; and
by D, d', e, the three variable points of which the co-ordinates
are (y, y, s'), (:i", y, z"), (xy y^z)', abed' may be regarded
as a plane quadrilateral, of which the diagonals ae and bd'
intersect each other in a point d on a fixed spheric sur-
face, which has its centre at c, and passes through a and d';
so that one side d'a of the quadrilateral, adjacent to the fixed
side AB, is a chord of this fixed sphere. And the equation (2.)
expresses that the other side be of the same plane quadrilateral^
adjacent to the same Jixed side ab, is a chord of ajixed ellipsoid.

Sir W. Rowan Hamilton on Quaternions. 217

if the two diagonals ae, bd' of the quadrilateral be equally long ;
so that a general and characteristic property of the ellipsoid,
sufficient tor the construction of that surface, and t'or the in-
vestigation of all its properties, is included in the remarkably
simple and eminently geometrical formula

AE=BD'; ,(3.)

the locus of the point E being an ellipsoid, which passes
through B, and has its centre at A, when this condition is
satisfied.

This formula (3.)} which has already been printed in this
Magazine as the equation (10.) of article 30 of this paper, may
therefore be deduced, as above, from generally admitted prin-
ciples, by the Cartesian method of co-ordinates ; although it
had not been known to geometers, so far as the present writer
has hitherto been able to ascertain, until he was led to it, in
the summer of 1 846 *, by an entirely diiFerent method ; namely
by applying his calculus of quaternions to the discussion of
one of those new formsf for the equations of central surfaces
of the second order, which he had communicated to the Royal
Irish Academy in December 1845.

36. As an example (already alluded to in the 32nd article
of this paper) of the geometrical employment of the formula
(3.), or of the equality which it expresses as existing between
the lengths of the two diagonals of a certain plane quadrilateral
connected with that new construction of the ellipsoid to which
the writer was thus led by quaternions, let us now propose to
investigate geometrically, by the help of that equality of dia-
gonals, the difference of the squares of the reciprocals of the
greatest and least semi-diameters of any plane and diametral
section of an ellipsoid (with three unequal axes). Conceive
then that the ellipsoid, and the auxiliary sphere employed in
the above-mentioned construction, are both cut by a plane abV,
on which b' and c' are the orthogonal projections of the fixed
points B and c ; the auxiliary point d may thus be conceived
to move on the circumference of a circle, which passes through
A, and has its centre at c' ; and since AE, being equal in length

* See the Proceedings of the Royal Irish Academy,
t In reprinting one of those new forms, namely the following quater-
nion form of the equation of the ellipsoid:

a slight mistake of the press occurred at p. 459, vol. xxx. of this Magazine,
which however, with the assistance there given by the context, can scarcely
have embarrassed the reader. In the preceding page, for a hyperboloid of
one sheet, touching the same cylinder in the same sheet, should have been
printed, .... in the same ellipse.

218 Sir W. Rowan Hamilton on Quaternions.

to BD' (because these are the two equal diagonals of the qua-
drilateral in the construction), must vary inversely as BD (by
an elementary property of the sphere), we are to seek the
difference of the squares of the extreme values of BD, or of
B'D, because the square of the perpendicular BB' is constant
for the section. But the longest and shortest straight lines,
B'D], B'Dg, which can thus be drawn to the auxiliary circle
round C, from the fixed point B' in its plane, are those drawn
to the extremities of that diameter DiC'D2of this circle which
passes through or tends towards B' ; so that the four points
Di C Dg B' are on one straight line, and the difference of the
squares of B'Dj, B'Dg is equal to four times the rectangle
under B'C and C'D^, or under B'C and C'A. We see therefore
that the shortest and longest semi-diameters AEj, AEg of the
diametral section of the ellipsoid, are perpendicular lo each
other,because (by the construction above-mentioned)they coin-
cide in their directions respectively with the two supplementary
chords ADi, ADg of the section of the auxiliary sphere, and
an angle in a semicircle is a right angle; and at the same time
we see also that the difference of the squares of the reciprocals
of these two rectangular semlaxes of a diametral section of the
ellipsoid varies, in passing from one such section to another,
proportionally to the rectangle under the projections, B'C and
C'A, of the two fixed lines BC, CA, on the plane of the vari-
able section. The difference of the squares of these recipro-
cals of the semi-axes of a section therefore varies (as indeed it
is well-known to do) proportionally to the product of the sines
of the inclinations of the plane of the section to two fixed dia-
metral planes, which cut the ellipsoid in circles; and we see
that the normals to these two latter or cyclic planes have
precisely the directions of the sides BC, CA of the generating
triangle ABC, which has for its corners the three fixed points
employed in the foregoing construction : so that the auxiliary
and (liacentric sphere, employed in the same construction,
touches one of those two cyclic planes at the centre A of the
ellipsoid. If we take, as we are allowed tO do, the point B
external to this sphere, then the distance BC of this external
point B from the centre C of the sphere is (by the construc-
tion) the semisum of the greatest and least semiaxes of the
ellipsoid, while the radius CA of the sphere is the semidiffer-
ence of the same two semiaxes: and (by the same construc-
tion) these greatest and least semiaxes of the ellipsoid, or
their prolongations, intersect the surface of the same diacen-
tric sphere in points which are respectively situated on the
finite straight line BC itself, and on the prolongation of that
line. The remaining side AB of the same fixed or generating

Notices respecting New Books. 219

triangle ABC is a semidiameter of the ellipsoid, drawn in the
direction of the axis of one of the two circumscribed cylinders
of revolution ; a property which was mentioned in the 32nd
article, and which may be seen to hold good, not only froni
the recent analysis conducted by the Cartesian method, but
also and more simply from the geometrical consideration that
the constant rectangle under the two straight lines BD and
AE, in the construction, exceeds the double area of the triangle
ABE) and therefore exceeds the rectangle under the fixed
line ABand the perpendicular let fall thereon from the varia-
ble point E of the ellipsoid, except at the limit where the
angle ADB is right; which last condition determines a cir-
cular locus for D, and an elliptic locus for E, namely that
ellipse of contact along which a cylinder of revolution round
AB envelopes the ellipsoid, and which here presents itself as a
section of the cylinder by a plane. The radius of this cylinder
is equal to the line BG, if G be the point of intersection, di-
stinct from A, of the side AB of the generating triangle with
the surface of the diacentric sphere; which line BG is also
easily shown, on similar geometrical principles, as a conse-
quence of the same construction, to be equal to the common
radius of the two circular sections, or to the mean semiaxis
of the ellipsoid, which is perpendicular to the greatest and the
least. Hence also the side AB of the generating triangle is,
in length, a fourth proportional to the three semiaxes, that
is to the mean, the least, and the greatest, or to the mean, the
greatest, and the least, of the three principal and rectangular
semidiameters of the ellipsoid.

[To be continued.]

XXXVII. Notices respecting New Books.

Notice of a Memoir on Meteors of various sorts. By T.I. M. Forster,
F.R.A.S., &;c. Bruges, 1846.

EXPERIENCED in observing and in treating of these phseno-
mena, Dr. Forster refers his readers to his former communica-
tions of them, and to the numerous articles in the Royal Society's
Transactions, as well as in the Gentleman's and Philosophical Ma-
gazines.

He carefully examines the theory of phosphorescent jets of gas
rising unperceived while traversing the low and damp strata of the
atmosphere, but becoming ignited as soon as they reach a sufficiently
dry stratum. The ignition is then supposed to run down the column
of gas, and reveal the several bends it had been subjected to by
Various currents of wind. The occasional explosions may be ex-
plained by supposing the running fire to reach a spot overabounding
itl hydrogen, instances not unfrequent after heavy rains.

220 Notices respecting New Booh.

It was not till the 10th of August 1811 that the idea of their
periodicity occurred to Dr. Forster, when he and his father counted
some hundreds, and by their journal perceived their recurrence on
that same day. Indeed, in copying a curious old manuscript calen-
dar, he found the 10th and 18th of August called stellibvndce and
meterodes ; but he acknowledges their frequency at all times and
in all places.

Inclined to assign them a gaseous origin, our author has yet, in
deference'to the learned men who differed from him, endeavoured to
relate fairly the various arguments in favour of their several theories.

Aristotle regarded meteors as arising from exhalations denoting
an approaching change of weather. Theophrastus thought they pro-
gnosticated wind from the quarter towards which they rushed. And
Aratus agreed with him, especially if they left long lingering tails,
in which he was imitated by Virgil. Lucan in his Pharsalia rather
confounds meteors with the fixed stars. Homer compares the descent
of Minerva to the rush of a meteor.

Passing over the middle ages, when meteors were feared as indica-
tions of Divine anger, we find that in the seventeenth century electri-
city began to be suspected, and was supported by the highest names
of that sera. Then the magnificent meteor of the 18th of August 1783
brought out the elaborate paper by Dr. Blagden in the Philosophical
Transactions for the following year. As to their velocity, it varies
so much that this element cannot suffice to decide from what height
they fall. The meteor above alluded to moved at only six miles per
second when at about ninety miles above our heads. Cavallo esti-
mated its diameter at 3200 feet, and its elevation at 560 miles. Cer-
tainly the explosion not being heard for ten minutes after it was
seen is a sufficient proof of its distance. The general electric state
of the atmosphere that year over half the globe is well known, by
the remarks made in consequence of the violent earthquakes that
occurred.

In support of the theory that meteors are occasioned by the igni-
tion of columns of inflammable gas. Dr. Forster mentions the ignis
fatuus, and the flitting lights that are seen in May on cabbages.
Many naturalists regard meteors as one of the various phaenomena
attributable to electricity, and some expected to find that they chiefly
pointed to the magnetic pole.

Many roofs of thatch have been ignited by the fall of meteors
upon them, and this must be the explanation of towns recorded to
have been burnt by fire from heaven. The explosion of the meteor
of the 25th of September 1846, was heard a few seconds after it was
seen : but if, instead of the ambiguous term a few, spectators would
count slowly, they would aff^ord a much nearer approach to the true
time elapsed, especially if they would afterwards count at the same
rate when they can compare with a seconds watch, or with a clock.
(A.S.) The tail of that meteor was larger than usual, and lasted longer,
some persons stating fifty seconds, others some minutes. More
precise details are requisite. It was at first whitish, then purplish,
and lastly red, when it became curved, and faded in a serpentine

Notices respecting New Books. 221

form. This last phaenoraenon was observed in another instance
about twenty years since. Even the luminous arc of the 28th of
September 1828 might, our author thinks, be a still more dilatory
tail of a meteor that had shot across our hemisphere just before
sunset, and for that reason was not perceived. In July 1799
Dr. Forster's father saw a meteor cross the sky from south to
north, then return southward, and finally bend to the north-west.
Another peculiarity is that of rising in the sky instead of descending,
which has been reported as occurring sometimes near the equator,
where they are very numerous. And Dr. Forster himself saw a
whitish globe stationary for two seconds, and then turn a fine red.

A shower of small meteors is recorded to have occurred on the 25th
of April 1095 ; and Dr. Forster saw an approximation to this on a
bright winter night in 1832, inasmuch as the whole firmament was
in a glow from an immense number of very fine luminous tails nearly
parallel from E.N.E. to W.S.W. They might deserve the name
rather of streaks, no heads being visible. The duration of each might
not exceed a second, but the phaenomenon altogether lasted a quarter
of an hour and then ceased suddenly. And in November 1830 he
saw a similar multiplicity of little streaks, but crossed by others at
right angles. Another peculiarity was described by a clergyman
near Epping, that of seeing a meteor, after descending to the earth,
undergo a sort of reverberation by rising in an oblique direction,
and then break into sparks.

Among the numerous authors who have treated of this subject,
perhaps M. Quetelet's catalogue is the most complete, with the ex-
ception of his omitting the interesting meteor of 1783. M. Arago
and M. Biot have also treated the subject ably.

A copious journal of meteors has been kept in Dr. Forster's family
from 1767, but no periodicity was suspected till the 10th of August
1811 ; though then, on looking back through the journal, it was per-
ceived that there had been a great preponderance in the Novembers
ever since 1799, and in the Augusts from 1779. When employed a
few years after to construct perennial calendars. Dr. Forster indicated
a number of meteors as a phsenomenon to be expected on the 10th of
August.

This became confirmed in 1831 by other observers, and they added
the second period of the 13th of November. M. Queteletnow adds
April and December, while others suggest January, May, June and
July. He thinks their usual height in the atmosphere is from six-
teen to twenty leagues or more, though they are occasionally seen
slanting very near the ground. The most numerous sort, distin-
guished by the name of etoiles filantes, may revolve in trajectories
by swarms, forming a belt round the sun, which we have occasionally
to traverse. Then, owing to the earth's motion, these luminous
corpuscles would naturally, as they have been observed to do, appear
to " have their point of divergence towards /3 Camelop. in August,
and towards v Leonis in November, agreeing with our annual mo-
tion in the ecliptic." According to the known laws of optics, the
swarm would seem to separate in radii as we neared them, and.

222 Boyal Society,

owing to the compound velocities, seem to tend from N.E. to
S.W.

Although meteors differ very much from each other in some in-
stances, it is very difficult to classify them ; but an abundance of
them seems connected with a change of weather, and especially with
cirrostratus and cirrocumulus clouds. As to their direction, though
they sometimes converge towards one point, they rush at others
towards every point of the compass. He therefoi'e wavers only be-
tween an electric and a gaseous origin, — quoting electric experiments
referred to in England by the Abbe Bertholon, and gaseous ones by
Constable, as having produced excellent imitations.

In the terrible night of the 7th of July 1834, a crowd of nimbi
collected around Vesuvius about 9 o'clock, shooting their lightnings
down towards the mountain accompanied by rain and hail. The
lightning was sometimes bluish and sometimes reddish.

As to the periodicity of meteors. Dr. Forster finds that there are
decided changes in the electrometers also on the 10th of August and
13th of November ; and the greatest number he ever saw fell on the
10th of August 1811, just after a violent storm ; but when a storm
has happened some time before, the meteors are fewer at the two
periods observed. Also if one or more large meteors occur, there
are no small ones afterwards for a proportionate time, as if the atmo-
sphere had been cleared of the requisite material. Also it may be
remarked in general, that the winter and the higher latitudes are
least prolific of them.

Fiery balls do not often occur, but are very powerful. Thus the
one seen in France and in England the 17th of July 1771, must
have been at an elevation of fifty-four miles, and the report of its
explosion was not heard till two minutes after its occurrence, like
the rolling of thunder ; but the observatory windows at Paris were
" ebranlees." It appeared larger and brighter than the]), and its
swiftness was estimated at twenty-four miles per second.

From the quickly-increasing rarity of our atmosphere, Arbuthnot
thinks that at the height of sixty miles (the estimated height of tlie
meteor in 1718) the air is 30,000 times purer than on the level of
the sea. Yet Pringle estimated the height of the meteor of 1738 to
have been ninety miles. The diameter of some globes has been
estimated at 1^ mile.

XXXVIII. Proceedings of Learned Societies.

nOYAL SOCIETY.
[Continued from p. 770
June 17, " 13 ESEARCHES on the Function of the Intercostal
184<7. "-^ Muscles and on tlie Respiratory Movements, with
some remarks on Muscular Power, in Man." By John Hutchinson,
M.li.C.S. Communicated by Sir I3enjamin Brodie, Bart., F.R.S., &c.
The object of this paper is to demonstrate by models and dis-
sections the action of the intercostal muscles.

Royal Society, 223

After premising an account of the views of several eminent phy-
siologists, and in particular those promulgated by Haller, the author
shows that they resolve themselves into the general opinion that the
scalene or other muscles of the neck fix the first rib, in order to
enable the two sets of inlei'costal muscles to act either separately or
conjointly, as inspiratory or expiratory muscles. He then proceeds
to state the proofs that the intercostal muscles possess an action
which is independent of any other muscle, and also independent of
each other, so that any of the twelve ribs may be elevated or de-
pressed by them either separately or conjointly. He demonstrates
the nature of this action by means of models, producing oblique
tensions betweeti levers representing the ribs, and allowing of rota-
tion on their centres of motion ; and he shows that such tension in
the direction of the external intercostal muscles, elevates both the
levers until the' tension ceases, or the position of the bars by proxi-
mity obstruct each other. If the tension be exerted in a contrary
direction, as in the internal intercostal muscles, the bars are both
depressed. This movement was demonstrated by a model. It was
farther shown that two tensions decussating can, according to the
position of the fulcra, be made to act as associates or antagonists to
each other. Such motions are to be considered with reference to
the fulcra, bars with one fulcrum common to each having no such
action ; and the author accordingly draws the following conclu-
sions : —

1st. All the external intercostal muscles are true inspiratory mus-
cles, elevators of the ribs, and with this act they dilate the inter-
costal spaces, thus increasing the cavity of the chest.

2nd. The internal intercostal muscles have a double action; the
portions situated between the cartilages are associates in action
with the external layer, and act as elevators of the cartilages, while
the portion between the ribs are depressors, or antagonists of the
external layer, and are here true expiratory muscles ; with this they
decrease the intercostal spaces.

3rd. These muscles can elevate or depress the ribs independently
of any other muscle, fixing the first or last rib. Any one lamella,
or series of muscles, can, as required, independently perform in-
spiration or expiration at any one of the twenty-two intercofctal
spaces.

4th. In inspiration, the intercostal spaces increase, with a short-
ening of the muscle ; and in expiration, they decrease their perpen-
dicular distance, with a shortening of the muscle.

5th. All parallel intercostal muscles, acting with uniform force,
concur in the same effect, whether near the fulcrum or more distant
from it, and these muscles gain power with their increasing obliquity
as well as speed.

In the third part of the paper an account is given of the differ-
ence between the external thoracic space and the internal pulmonic
space. The respiratory movements are described in health and
disease, and it is shown that tiie chest is rarely enlarged at two
places at one and the 9ame time.

224 Royal Society.

In conclusion the author conceives that he has established the
following propositions : —

1st. Costal breathing may be distinguished from abdominal by
determining which part is first put in motion, and the kind of re-
spiration may be designated according to the name of such part.

2nd. Healthy costal breathing begins with the motion of a supe-
rior rib, which is followed by that of the lower ones in succession.

3rd. Ordinary respiration in men is abdominal, in women, costal ;
extraordinary breathing is the same in both sexes.

4th. Any of the ribs, from the twelfth to the first, may carry on
respiration.

5th. Diseased respiration is of various kinds ; the movements may
be symmetric or not symmetric, costal or abdominal; all or none of
the ribs may move ; the abdomen may or may not move ; the chest
may dilate in all its dimensions at one and the same time; costal and
abdominal breathing may alternate with one another ; costal motion
may be undulating or not ; and all these may be combined in one,
which the author terms " hesitating breathing ;" and lastly, the quan-
tity of air breathed is diminished when there exists pulmonary dis-
ease.

" On the Structure and Development of the Liver." By C.
Handfield Jones, M.B., Cantab. Communicated by Sir Benjamin
C. Brodie, Bart., F.R.S., &c.

The author gives a detailed description of the structure of the
liver in animals belonging to various classes of the animal kingdom.
He states that in the Bryozoon, a highly organized polype, it is clearly
of the follicular type ; and that in the Asterias, the function of the
liver is probably shared between the closed appendage of the stomach
and the terminal cteca of the large ramifying prolongations of the
digestive sac contained in the several rays. Among the Annulosa,
the earthworm presents an arrangement of the elements of the he-
patic organ, corresponding in simplicity with the general configura-
tion of the body, a single layer of large biliary cells being applied as
a kind of coating over the greater part of the intestinal canal. In
another member of the same class, the Leech, in which the digest-
ive cavity is much less simple, and presents a number of sacculi
on each side, these elements have a very diflferent disposition ; and
the secreting cells, although some remain isolated, for the most
part coalesce to form tubes, having a succession of dilatations
and constrictions, and finally uniting and opening into the intes-
tine. In Insects, the usual arrangement is that of long curved fila-
mentary tubes, which wind about the intestine ; these, in the meat
fly, are sacculated throughout the greater part of their course, till
they arrive quite close to the pylorus, where they open ; near their
origin they appear to consist of separate vesicles, which become
gradually fused together, but occasionally they are seen quite sepa-
rate. The basement membrane of the tubes is strongly marked,
and encloses a large quantity of granular matter of a yellowish tinge,
\rith secreting cells ; another portion of the liver consists of sepa-
rate cells lying in a granular blastema, which cells, in a later stage

Royal Society. 225

of development, are seen to be included in vesicles or short tubes
of homogeneous membrane, often coalescing and exhibiting a more
or less manifestly plexiform arrangement ; this portion of the
liver is regarded by Mr. Newport as really adipose tissue. The
author has termed it the Parenchymatous portion of the liver, on
account of its general appearance and mode of development, though
he has not been able to determine whether the tubes always origi-
nate from it. Among the Arachnida, the follicular type of arrange-
ment prevails; and the same is the case with the Crustacea, the folli-
cles in these last being distinctly visible to the naked eye. In Mol-
lusca also, we find the follicular arrangement universally to obtain ;
yet in certain cases the limiting membrane of the follicles cannot be
shown to exist, and the author therefore thinks that its importance
is probably not great, but that it serves chiefly to fulfil the me-
chanical function which its synonym " basement" indicates. The
quantity of retained secretion in the liver of molluscs seems clearly
to imply that the bile in them is not an excrementitious fluid ; it is
used slowly on account of the imperfect character of the respira-
tion.

In passing from the Invertebrata to the Vertebrate division of the
animal kingdom, and beginning with the class of Fishes, a great
change is inimediately manifest in the form and character of the
biliary organ ; it is now a gland of solid texture, to which the term
parenchymal is justly applied. Two portions may be distinguished
in it, namely, the secreting parenchyma, consisting of delicate cells,
or very often of nuclei, granular and elaborated matters in great
part, and the excreting ducts, which, though completely obscured
bj'- the surrounding bulky parenchyma, may yet be satisfactorily de-
monstrated, and traced often to their terminal extremities in the
following manner. If a branch of the hepatic duct be taken up in
the forceps, it may be dissected out without much difficulty from
the surrounding substance, which is very soft and yields readily to
gentle manipulation ; when a trunk is in this way removed and
placed under the microscope, a multitude of minute ramifications
are seen adhering to it ; among these not a few may be discovered,
which do not appear to have suffered injury ; some are occasionally
seen terminating by distinctly closed extremities ; more usually the
duct becomes very minute and gradually loses all definite structure,
appearing at last like a mere tract of granular matter ; in either
case there is no communication by continuity with the surrounding
parenchyma. Large yellow corpuscles, peculiar cells, and a consi-
derable quantity of free oily matter usually existing in the liver of
various fishes, seem generally to indicate a great superiority in the
amount of secretory over that of excretory action, and to betoken
clearly the feeble intensity of the aerating function.

In Reptiles, there is the same arrangement in the liver, namely,
a secreting parenchyma of cells and an apparatus of excretory ducts,
which have the same essential characters as those of fishes ; but
there exists very frequently in the parenchyma remarkable dark
corpuscles, which appear to be masses of retained biliary matter,

Phil. Mag. S. 3. Vol. 31. No. 207. Sept. 1847. Q

226 Royal Society.

the import of which, in the situation they occupy, is doubtless the
same as that of the similar masses existing in fishes.

In Birds, the parenchyma of the liver is remarkably free from oily
or retained biliary matters; it often consists almost wholly of free
nuclei and granular matter, with scarcely a single perfect cell ; the
excretory ducts often greatly resemble those of reptiles, sometimes
rather those of mammalia; the essential character is, however, always
the same, namely, that they terminate without forming any important
connexion with the jiarenchyma.

In Mammalia, the parenchyma of the liver consists usually of per-
fect cells, which are arranged often in linear series of considerable
length, radiating from the axis of each lobule ; these unite at variouf
points with each other, so as to present a more or less decidedly
plexiform appearance. Each lobule, as described by Mr. Kiernan,
is separated from the adjacent ones by the terminal twigs of the
portal vein, and to a greater or less extent by a " fissure," though in
most animals the lobules are continuous with each other both above
and below the fissure. The elaboi'ation of the secreted product
seems to be most completely effected in the cells adjoining the
margins of the lobules, which are often seen to contain a larger
quantity of biliary matter than those in the interior, and to be appa-
rently in the act of discharging it into the fissure; the margin of
the lobule then presents an irregular surface with large globules of
the secretion clustering together all over it. The capsule of Glisson
surrounding the vessels in the portal canals gives a fibrous invest-
ment to those surfaces of the lobules which are towards the canal ;
but when it has arrived in the fissures, it forms a continuous mem-
brane lining the surfaces of opposite lobules ; this membrane is often
truly homogeneous, and closely resembles the basement tissue : there
appears occasionally to be a delicate epithelium on its free surface ;
but this, as well as the membrane itself, is often absent, when the
margin of the lobules is in that condition which has just been de-
scribed and which may be termed active. The minute branches of
the hepatic duct as they approach their termination undergo a re-
markable alteration in their structure ; they lose their fibrous coat,
which blends itself with the membranous expansions of the capsule
of Glisson ; their basement inembrane becomes gi-adually indistinct,
and at last ceases to exist, and the epithelial particles no longer
retain their individuality, but appear to be reduced to mere nuclei,
set very close together in a faintly granular basis substance. The
mode of their termination is not uniformly the same; frequently they
present distinctly closed rounded extremities, between one and two
thousandths of an inch in diameter; at other times they seem to
cease gradually in the midst of fibrous tissue, the nuclei alone being
disposed for some little way in such a manner as to convey the idea
of a continuation of the duct. These ducts can seldom be dis-
cerned in the fissures, but have several times been seen in the
" spaces," where several fissures unite ; they do not form anything
like a plexus between the lobules. From the anatomical relation of
the ducts to the parenchyma, and from the circumstance that a

Royal Society. 227

distinct vessel conveying a different kind of blood is distributed to
the hepatic duct, as soon as the liver assumes the parenchymal form,
it seems probable that the mode in which the secreted bile is con-
veyed out of the organ, is by its permeating the coats of the minute
ducts in obedience to an endosmotic attraction, which takes place
between the bile in which the ducts may be said to be bathed, and
a denser (perhaps mucous) fluid formed in their interior. The
large quantity of oily matter frequently existing in a free state in
the secreting parenchyma of the liver, which must be regarded as a
product of secretory action, seems to suggest the idea, that a cer-
tain quantity of the biliary secretion may be directly absorbed into
the blood, and in this manner conveyed away from the organs, just
as occurs in the thyroid body, suprarenal capsules, and other
glands unprovided with efferent ducts.

With respect to the development of the liver, the author considers
the opinion of Reichart to be decidedly the correct one, namely,
that its formation commences by a cellular growth from the germi-
nal membrane, independently of any protrusion of the intestinal
canal. On the morning of the fifth day, the oesophagus and stomach
are clearly discernible, the liver lying between the heart, which is in
front, and the stomach which is behind ; it is manifestly a parenchy-
mal mass, and its border is quite distinct and separate from the digest-
ive canal ; at this period, the vitelline duct is wide, it does not open
into the abdominal cavity, but its canal is continued into an anterior
and posterior division, which are tubes of homogeneous membrane,
filled, like the duct, with opaque oily contents ; the anterior one
runs forwards, and forms behind the liver a terminal expanded
cavity, from which then passes one offset, which, gradually dilating,
opens into the stomach; a second, which runs in a direction up-
wards and backwards, and forms apparently a caecal prolongation ;
and a third and fourth, which are of smaller size, arise from the
anterior part of the cavity and run to the liver, though they cannot
be seen to ramify in its substance ; at a somewhat later period, these
offsets waste away, excepting the one which is continued into the
stomach, and then the mass of the liver is completely free and un-
connected with any part of the intestine. As the vitelline duct
contracts, the anterior and posterior prolongations of it become
fairly continuous and form a loop of intestine, the posterior division
being evidently destined to form the cloaca and lower part of the
canal. The final development of the hepatic duct takes place about
the ninth day by a growth proceeding from the liver itself, and
Consisting of exactly similar material ; this growth extends towards
the lower part of the loop of duodenum, which is now distinct, and
appears to blend with the coats of the intestine ; around it, at its
lower part, the structure of the pancreas is seen to be in process of
formation. The further progress of development of the hepatic
duct will, the author thinks, require to be carefully examined, but
the details he has given in this paper have satisfied him of the cor-
rectness of the statement that the structure of the liver is essentially
parenchymal.

Q2

[ 228 ]
XXXIX. Intelligence mid Miscellaneous Articles.

SUGGESTIONS FOR THE OBSERVATION OF THE ANNULAR
ECLIPSE, OCT. 9, 1847, MADE BY THE BRITISH ASSOCIATION
FOR THE ADVANCEMENT OF SCIENCE, OXFORD, JUNE 2(),

1847.

n[ "^HE following directions and suggestions, relative to the ensuing
-■- annular eclipse of the sun, which will take place Oct. 9, 1 847, are
proposed for the assistance of less-practised observers, or those who
may not have better information at hand, but who may nevertheless
render great service by noticing and recording, as well as circum-
stances permit, any of the various points herein alluded to.

I. As a general direction as to the limits within which the eclipse
can be seen annular in England and Ireland, if on any map a line be
drawn through Greenwich and Gloucester and produced, it will give
the northern limit at which the eclipse ceases to be annular.

A line parallel to the last, through Padstow in Cornwall on the
west, and Torbay on the east (which will extend across the channel
to Havre, &c., and passes just below Cape Clear on the west), will
be the line along which the eclipse is both annular and central.

The southern limit lies wholly below England.

II. As a rough guide to the time, the commencement of the an-
nulus will be nearly at 7** 23™ a.m. (civil reckoning) for the extreme
south-west of Ireland, at 7** 24"^ for a line through Land's End and
Milford Haven, at 7^* 2.5"" through the Isle of Wight and Reading,
at 7^* 25"" 50^ forWalmer (Greenwich mean time).

III. For the observations requisite, a telescope of very moderate
power is best. As the annulus will not last more than three or four
minutes, those unaccustomed to such observations should be cau^
tioned against attempting to observe all the phsenomena, or they
may thus run the risk of observing «OMe. If possible several observers
should combine for the purpose, and each agree to attend to one, or
some few of the phsenomena.

IV. To obviate some of the difficulties arising from the rapid
passage of the phsenomenon, the observer may be referred to Capt.
Smyth's Cycle (i. 141, 146), where some valuable practical hints
are thrown out for tranquillizing the observer's nerves in so transitory
a phsenomenon ; especially by previously making a careful drawing
of the spots (if any) existing on the sun's disc, which may be made
useful in marking and ascertaining the progress of the ecHpse.

V. With the view of correcting the moon's tabular north polar
distance and semidiameter, it is peculiarly desirable that observations
should be made along or near the line (passing through Greenwich
and Gloucester) on which the eclipse is barely annular. At some
of these the echpse will be completely annular, and here the follow-
ing observations should be made : —

The time of beginning of annularity and end of annularity should
be observed. As the duration only is required, a common watch
showing seconds will suffice for this purpose.

Intelligence and Miscellaneous Articles. 229

If possible, by means of a graduated pearl scale or other equivalent
means, the breadth of the narrowest part of the annulus should be
measured several times about the middle of the time of the annular
appearance, as well as it can be estimated.

At other places the eclipse will not be completely annular, and
here the principal object must be to make several measures of the
distance between the cusps about the time when that distance is
smallest. This measure may probably be made by means of a gra-
duated pearl scale, or by means of a divided object-glass applied in
front of the object-glass of the telescope, or by the use of a common
sextant.

VI. As to the particular points of physical interest to which at-
tention should be directed, they may be stated as follows : —

1 . It will be desirable in general to notice the fact of the appearance
of what are denominated " beads " and " threads " by the late Mr.
Baily and others, just before and after the completion of the annulus.

For details of older observations the observer should consult
Ast. Soc. Memoirs, i. 142-146, x. 10-17, 33-38.

The beads were observed by Mr. Baily, ib. x. 210, in 1842,
when they were not seen by Mr. Airy, ib. x. 218.

They were observed by Prof. Henderson at Edinburgh. Ast.
Soc. Notices, v. 186.

2. Whether in the neighbourhood of the cusp the limb either of
the sun or moon appears distorted ?

Whether the beads appear steady or waving, disappearing and
reappearing, &c. ?

See the observations of Mr. Caldecott at Trevandrum, Ast. Soc.
Notices, vi. 81.
Whether they present any peculiar changes when viewed through
differently coloured glasses, the observer alternating the colours,
which should be as dissimilar as possible, such as red and green ?
See Silliman's Journal, Jan. 1842.

3. Whether they are seen when the eclipse is projected on a screen?
In this way Prof. Chevallier saw none when others with coloured

glasses saw them. Ast. Soc. Notices, v. 186.

4. The drawing out of the beads into threads when very near
junction; and whether they waver and change, and the number of them?

See Ast. Soc. Mem., x. 15-17, 39 ; waving and changing, ib. x.
12, 13 ; not seen in 1842 by Mr. Baily, Notices, v. 210.

5. Whether before and after the formation of the threads the
moon's dark disc is elongated towards the point of contact ?

This was observed, ib. x. 29 ; and wavy motion in the limb,
ib. X. 12, 14, 30.

6. The beads are ascribed by some to lunar mountains : What
mountains exist at that part of the limb ?

See Ast. Soc. Mem., X. 9, 16, 30-36.

7. The exact intervals of time elapsed between the first and last
complete contact, and that of the first and last formation of beads or
other irregularities in or about the cusps, should be determined.
The difiference of the times being all that is wanted, a good ordinary
watch will be sufficient.

230 Intelligence and Miscellaneous Articles.

The remarkable fact of a recurrence of cusps observed by Mr.
Airy in 1842, and his explanation of it, should be attentively
coneidered. See Ast. Soc. Notices, v. 296.

8. If possible, accurate measures should be taken of the apparent
diameter of the dark disc of the moon upon the sun, which may be
expected to be greatly less than the truth, owing to the irradiation
of the sun's light.

9. It should be noticed whether any external luminous arch is
formed over the part between the cusps, a little before the first junc-
tion and after the final separation, and the colour of the light.

It was observed, and appeared brown to De Lisle (Phil. Trans.,
1748, 490), reddish in other cases (Ast. Soc. Mem., i. 144,
X. 37), Siadi purple in others (ib. x. 16).

ON THE PREPARATION AND COMPOSITION OF THE SALTS OF
ANTIMONY. BY M. E. PELIGOT.

Sulphates of Antimony. — When oxychloride of antimony (CI Sb^O-)
is treated with hot concentrated sulphuric acid, a salt is formed
which is deposited in acicular crystals, hydrochloric acid being at
the same time evolved. This salt, as well as another sulphate to
be described, can only be obtained in a dry state by long remaining
in vacuo, or in perfectly dry air upon porous plates of pipe clay.
These plates were heated to redness before the crystalline magma
was placed upon them, and they were left to cool in air deprived of
moisture. This method of drying yields products which usually
contain a slight excess of sulphuric acid. If however the points of
contact between the salt to be dried and the absorbent earth be re-
newed from time to time, and the absorption goes on for several
months, compounds of sufficient purity to remove all doubts of their
true composition may be obtained.

One hundred parts of the sulphate of antimony, obtained by com-
mon sulphuric acid and oxychloride of antimony, gave —

Sulphuric acid 51*9

Oxide of antimony (by carbonate of ammonia) .... 50*2
The composition of this salt is therefore —

4S03 2000 51-2

Sb^O* 1912 48-8

3912 100-0

Another specimen gave 53* 1 of sulphuric acid, and 44'3 of oxide of
antimony.

Another sulphate of antimony was obtained in the form of small
brilliant crystals, by treating sesquioxide of antimony with Nord-
hausen sulphuric acid. After remaining ten months on the dried
clay, it gave —

Sesquioxide of antimony .... 63"0 643

Sulphuric acid 37"1 35*0

The formula 2S03, Sb® O^ gives 65 "6 oxide of antimony and 34*4
sulphuric acid.

Mixtures of these salts in different proportions were also obtained ;

Intelligence and Miscellaneous Articles. 231

but no analysis indicated the existence of the compound 380^,
5Sb'-03, which, according to Berzelius, would be the neutral sul-
phate of antimony.

On treating the above-described salts with hot water, a subsalt is
obtained, the composition of which is represented by the formula —

Calculation. Experiments.

2Sb2 0' 3824 88-4 88-6

S03 '.. 500 11-6 11-4

4324 100-0 100-0

The analysis of two other specimens is correctly represented by
the formula 2Sb2 O^, SO^, 2H0.

Nitrate of Antimony. — This salt was obtained in the form of pearly
crystals by dissolving the oxide in cold fuming nitric acid, and adding
water to the solution. Its composition is 2Sb- O^, NO*.

Oxychlorides of Antimony. — Powder of Algaroth was prepared by
treating chloride of antimony with cold water. After some days the
mass became crystalline ; when well- washed its composition agreed
with the analyses which have served to fix the formula of this com-
pound. This formula is more simply replaced by CI Sb- 0'^

When the sesquichloride of antimony, or rather the sesquioxide
dissolved in a great excess of hydrochloric acid, is treated with hot
water, another oxychloride is obtained, which, on the cooling of the
liquor, precipitates in dense brilliant crystals. Its composition is
represented by the following formula : —

Calculation. Experiments.

CI 443 10-6 11-1 11-4

4Sb 3224 77-3 76-5 76-8

O' 500 12-1

4167 100-0

This compound consequently must here presented by the formula,
ClSb^O'^ + Sb^Os.

Tartrates of Antimony. — By allowing a syrupy solution of tartrate
of antimony, obtained by dissolving the oxide of the metal in tartaric
acid, to remain for a long time, large transparent crystals of tartrate
of antimony were obtained. The mother- water, after the separation
of the crystals, furnished more afterwards by spontaneous evapora-
tion.

This salt is very soluble in water. It is deliquescent in a moist
atmosphere. Its composition is represented by the following for-
mula : —

Calculation. Experiments.

, ^ .

C'6 1200 19-6 18-9 19-0

H'G 200 3-2 3-5 3*5

028 2800 46-0

Sb«03.... 1912 31-2 31-5

6112 100-0
At 320° F. this salt lost 23-1 per cent, of water.

232 Intelligence and Miscellaneous Articles.

On decomposing the formula as follows, the loss of twelve equi-
valents of water represents 22 per cent, of the weight of the salt —

2CsH2 08,Sb2 03, 12H0.

On pouring alcohol into a concentrated solution of the acidulous
tartrate of antimony, a precipitate is obtained which, when dried at
320° F., yielded 16*4 of carbon and TS of hydrogen. The compo-
sition of this salt is represented by the formula C^ H'^ O^, Sb^ O^, HO,
which requires 17*2 of carbon and 1 of hydrogen. The salt which
M. Peligot analysed contained a little more water than the quantity
required by this formula, but not enough to allow of the addition of
another equivalent.

Acidulous Tartrate of Antimony and Potash. — This salt was de-
scribed by M. Knapp, who obtained it by mixing solutions of tar-
taric acid and tartarized antimony. The salt which was analysed by
M. Peligot was in very regular crystals. It yielded —

Carbon 19-5 18-7

Hydrogen 2-7 2*7

Sesquioxide of antimony 3 TO

The formula C'^ H* O'^, Sb^ O^, KO, 8H0 represents its composi-
tion. It gives —

Carbon 19-1

Hydrogen 2*3

Sesquioxide of antimony . . 30"5

According to M. Knai)p it contains one equivalent less of water.

Oxalate of Antimony. — M. Peligot prepared this salt by four pro-
cesses : — 1st, by boiling in a solution of oxalic acid oxide of antimony
prepared from the chloride by carbonate of ammonia ; 2nd, by treat-
ing the powder of Algaroth with oxalic acid ; 3rd, by pouring hydro-
chloric acid into a hot solution of the double oxalate of potash and
antimony ; the oxalate of antimony precipitates in the state of a
crystalline powder ; 4th, by adding oxalic acid to a solution of the
same double salt.

The oxalates of antimony obtained by these processes are similar
in composition. The author attempted, but in vain, by varying the
proportions, to obtain other compounds of oxalic acid and oxide of
antimony. This salt is crystalline and insoluble in water. It is
decomposed by boiling water into oxalic acid, which dissolves, and
sesquioxide of antimony-
Its composition is represented by the following formula, : —
Calculated. Experiments.

, ^ .

C-* 300-0 10-2 10-1 10-6 10-6

06 600-0 20-6

Sb203 .... 1912-9 65-4 66-7 65-6

HO 112-5 3-8 3-8 4-5 40

2925-4 100-0

Double Oxalate of Potash and Antimony. — The preparation and
analysis of this salt are very difficult. The salt obtained by M. Pe-

Intelligence and Miscellaneous Articles. 233

ligot was crystallized in transparent prisms ; it is readily soluble,
and is decomposed by a large quantity of water.

The quantity of water in this salt appeared to vary from unknown
causes, but apparently dependent on the temperature at which the
salt crystallizes. The formula appeared to be 70^ O', Sb^ 03, 3K0,
6H0. This gives as the composition of 100 parts of the salt —

Carbon 13-9

Water 90

Oxide of antimony 25*7

Potash 23-5

M. Peligot obtained —

Carbon 13-7 14-3 14-4 14-0

Water 9-7 9-2 lO'l 8-9

Oxide of antimony 25-7 26-2 24-8

A7in. de Ch. et de Phys., Juillet 1847.

ACTION OF HYDROCHLORIC ACID IN THE FORMATION OF
OXALIC ACID.

M. Kopp states that the presence of hydrochloric acid in nitric acid
is peculiarly favourable to the formation of oxalic acid. The resins of
benzoin andTolu, treated with pure nitric acid, yield no oxalic acid ;
but with an impure acid it is obtained. Pure nitric acid occasions
the formation of terebic acid only, in acting upon oil of turpentine,
and to oxypicric acid, in oxidizing the gum-resins. By using nitric
acid containing much hydrochloric acid, oxalic acid only is obtained
under the same circumstances. — Ibid, Juillet 1847.

PROJECTION OF ALDEBAKAN ON THE MOON.
At the British Association in Oxford a question arose respecting
the apparent projection of Aldebaran on the disc of the moon in
occultations. Pi-of. Airy and Dr. Forster stated having seen this
phaenomenon, which Prof. Struve seemed disposed to attribute to
to some mal-adjustment of the telescopes. On looking back, how-
ever, to the Philosopliical Magazine, it will be found that this ap-
pearance has been three or four times recorded ; as well as some
other circumstances calculated to show that the light of different
stars is very differently refracted. See Phil. Mag. for April and
May 1824.

THE PUFF PARLIAMENTARY:— DISINFECTION.

The art of puffing has not yet exhausted its resources ; and a
Parliamentary Report well got up, printed at the expense of the
public, and from which extracts may go the round of the news-
papers, seems to be the last and boldest device for the purpose,
which however has been fearfully exposed in the Dublin Quarterly
Journal of Medical Science.

The Times newspaper in a leading article of the 20th of August,
felicitates itself on having " the pleasant task of giving what publi-
city it may to a discovery made by a French gentleman, M. Ledoyen,

SS4 hitelligence and Miscellaneous Articles.

a Parisian chemist, in concert, it would appear, with a Mr. F. C.
Calvert, who seems to have received his education as a chemist at
Paris, and who is now lecturer at the Royal Institution of Man-
chester. This discovery, which, under the auspices of Lord Mor-
peth, has been submitted to the most searching tests by Dr. South-
wood Smith, Mr. Toynbee and Mr. Grainger, promises fair to be
one of the greatest boons ever conferred on suffering humanity. The
discovery is nothing less than the means of disinfecting all foetid
animal substances and gases by a liquid which is very cheap, simple,
and can be applied by any person with the greatest facility.

" The three medical gentlemen appointed by Lord Morpeth to
inquire into the real value of M. Ledoyen's discovery, present us in
their report with a dismal catalogue of the offensive and dangerous
vapours from animal and vegetable substances which at all hours
infect the air we breathe, in a greater or less degree, accordingly
as we moreor less neglect their impure origins." "The Commissioners
state that they have tried the effect of this fluid, — 1, on substances
already in a state of decomposition ; 2, on substances undergoing
that process ; 3, on night soil ; 4, on impure air. In every instance
excepting the second these experiments have been attended with the
most miracidous result." *' It would almost seem that some mysterious
power had sent us M. Ledoyen and his discovery to compensate for
the shortcomings of the Premier and Lord Morpeth*."

So far The Times. — We now give a few extracts from the Dublin
Journal, and refer our readers to the article which it contains for
the details of the means by which these puffs have been procured,
and for a full account of the matter.

This boasted discovery professes to furnish "the means of disin-
fecting all foetid animal substances and gases by a liquid which is
very cheap, simple, and can be applied by any person with the
greatest facility. It disinfects night-soil, not destroying but in-
creasing vegetation, more particularly as regards agriculture, com-
pletely preventing the disease in potatoes when the land is manured
with disinfected night-soil. It disinfects hospital-wards of miasma ;
also cellars, water-closets, and buildings infected by impure gases.
It disinfects sailors suffering from fever on board of vessels ; it will
also disinfect ships at sea, and under quarantine. It disinfects
patients suffering with infectious disorders and wounds, also dead
bodies, so that they may be kept nearly a month ; also different
parts of the body can be kept for the purposes of dissection, for
coroners' inquests, &c."

* No wonder that competitors should have started up asserting their
claims to 80 wonderful a discovery. Mr. W. Maddick thus begins his letter
to the Editor of the Times; of whose judgement in matters of science he
seems to have a most exalted opinion:

" Siu,— All the world knows that a laudatory notice in your columns is a
very high honour; and as in your excellent leader of yesterday you have
highly eulogized Messrs. Ledoyen and Calvert for their alleged discovery,
I appeal with confidence to the proverbial justice ofThe Times, <kc. &c.

" I boldly claim originality in this matter, and challenge these gentlemen,
or any other," &c., &c.

Intelligence and Miscellaneous Articles. 235

"There is not a word of evidence in the document before us as
to the influence of this solution of nitrate of lead in curing, or ' dis-
infecting,' as they call it, by its vicinity, fever or other infectious
diseases. Of course no professional man (except Dr. Southwood
Smith) could bring himself to support such an absurdity as that
would amount to. With respect to the potatoe disease, Dr. Smith
has been even less guarded. He manured portions of bis garden
with his disinfected night-soil, and finds that potatoes grown on
these spots are finer than elsewhere. He says, ' I have this day
had specimens of them examined by Mr. Alfred Smee, who pro*
nounces them to be at present perfectly healthy.' What! not a
single Aphis vastator ! Oh, genius of humbug ! how numerous are
thy votaries! Truly, successful speculation constitutes the idolatry
of this age, and the wonder-workings of pseudo-science its super-
stition."

" Let us now briefly pass in review some of the evidence detailed in
this precious document — some of the ' Letters and Reports received
by the Chief Commissioner of Woods and Forests,'— set forward
in a parliamentary folio, gravely ordered to be printed by the
British senate, and consequently paid for by the country. Always
premising that we do not deny to this, in common with many other
chemical substances*, the power of destroying some unpleasant
odours, or, to deal more in the phraseology of the Report, stinks.
But against the disgraceful quackery with which this book abounds,
— a quackery not equalled by the most offensive and indecent ad-
vertisement,— and the humbug of presenting sucii a book to the
country, we loudly and strongly protest."

" We have already alluded to the circumstance that this imposture
has been attempted to be bolstered up by the testimony of night-
men, dissecting-room porters, ward-men, and other respectable
authorities of a similar kind. Some of the experiments made by
these intellectual and educated individuals may amuse our readers,
as they have doubtless enlivened the House of Commons. Speaking
of the contents of a privy —

" * William Fenwick did, as you gentlemen saw, taste it, and
William Dyer put some over his eyes without injuring them : if it
had not gone through your process, it would have blinded him ! ! '

*' We cannot however pursue a strain of levity when we come to
examine the part which a physician of repute has taken in this trans-
action. Dr. Southwood Smith, not content with bearing his share
in the fooleries of the Report already spoken of, volunteers his in-
dividual testimony as to the efficacy of the fluid in obviating con-
tagion among i\\2 medical and non-medical attendants on the sick.

" ' Whatever diflSculties,' he writes, 'your Lordship may have en-
countered in obtaining the necessai'y powers to make even any com-

* Sulphate of copper, nitrate of copper, chloride of copper, guper-nitrate
of bismuth, nitrate of lead, nitrate of silver, chloride of gold, protochloride
of tin, perchloride of tin, nitrate of mercury. This fluid has been examined
by Dr. Aldridge, and found to be a solution of nitrate of lead. Sir W.
Burnett has introduced the chloride of zinc for similar purposes in the
navy.

236 Intellisence and Miscellaneous Articles.

"is

mencement of a system of prevention by the removal of the causes
of fever, you have in your own hands, and have had for some months,
the sure and certain means of preventing the extension of fever to
the immediate attendants on the sick.'

" In the columns of newspapers, in the pages of journals, on the
covers of magazines, in the corners of railway guides, placarded on
dead walls and bankrupts' sliop-windows, dropped into the hat at
public meetings, thrust into the hand in streets, and forced upon the
attention at every turn, we thought all the modes of puffing quack
advertisements and indecent labels, eitlier in prose or rhyme, had
been exhausted : but we find that we were mistaken. A novelty in
this department has been introduced by Colonel Calvert ; and in the
pages of a parliamentary report* we see pufts as gross, and language
as indelicate, as any that disfigure the lowest newspapers."

We can only add an expression of our regret that an important
public cause, that of sanitary improvements, should have to encounter
prejudices raised against it from the exaggerations, misrepresenta-
tions, quackery, and jobbing which are too manifest in the conduct
of some of its advocates.

A GRANT OF 200/. TO MR. WILLIAM STURGEON.
We are glad to learn, from a communication dated Downing
Street, 12th August, from Colonel Grey, the private secretary of
Lord John Russell, that his Lordship has been pleased to grant the
sum of 200Z., from the Royal Bounty Fund, to Mr. William Stur-
geon of this town. Mr. Sturgeon was formerly lecturer on experi-
mental philosophy at the Hon. East India Company's Military Aca-
demy, Addiscombe ; and since his residence in Manchester, now
extending over a number of years, he has been superintendent of
the Victoria Gallery, delivering various courses of lectures there ;
and subsequently he filled the office of lecturer to the Manchester
Institute of Natural and Experimental Science. For a long series
of years Mr. Sturgeon has honourably distinguished himself by his
investigations and discoveries in the various branches of electrical
science, especially in electro- magnetism and thermo-electricity.

OBSERVATIONS ON CREATINE. BY M. HEINTZ.

About two years ago I described a peculiar substance which I had
discovered in the normal urine of man. From subsequent investi-
gations I find that this substance is identical with that which M.
Chevreul found in meat broth, to which he gave the name of crea-
tine, and the presence of which in the fresh muscular flesh of dif-
ferent animals has recently been shown by Liebig.

The most advantageous method of procuring the substance is that
subsequently pointed out by M. Pettenkofer ; it consists in adding

* That Parliamentary Reports are sometimes made vehicles of privileged
detraction and calumny the public are already aware. A late instance with
regard to the Greenwich Observatory has been exposed by the Astronomer
Royal,

Intelligence and Miscellaneous Articles. 237

to the alcoholic extract of the urine an alcoholic solution of chloride
of zinc ; in a short time a deposit is formed, which contains the
creatine in combination with the chloride of zinc, together with a
small quantity of phosphate of zinc. These two substances are sepa-
rated by boiling water, which dissolves the first, but is without action
upon the latter. The pure creatine is obtained from the aqueous
solution of its combination with chloride of zinc by precipitating the
zinc with hydrosulphate of ammonia ; after having evaporated the
filtered liquid as far as possible without a precipitate being formed
in the boiling solution, absolute alcohol is added to it, when the
creatine is immediately deposited in the form of small crystals,
resembling those obtained in operating upon the alcoholic solution
of the aqueous extract of meat.

After having washed these crystals with alcohol, I recrystallized
them from water. The elementary analysis of the pure crystals led
to the following formula, C^ H" N^ O* + 2HO, which is the same as
that advanced by M. Liebig.

When creatine enters into combination with chloride of zinc, it
parts with 2 atoms of water besides the water of crystallization, and
in exchange takes up 1 atom of this salt. This combination is repre-
sented by the following formula, C H^ N^ O^ + ClZn, and the atomic
weight of creatine is consequently 1412*5.

From the experiments of M. Liebig it results, that of all the
organs of the animal body it is only the muscles which yield crea-
tine. Now, as I have proved its presence in the urine of man and
animals, it appears placed beyond all doubt that this substance is
formed in the muscles, that it is absorbed by the lymphatics or
blood-vessels, and is finally secreted by the kidneys, like urea, &c.
We may therefore conclude that creatine should henceforth be
placed amongst the excrementitious substances ; and consequently it
is barely probable that it constitutes one of the most important ali-
mentary principles of meat broth, as M. Liebig is inclined to think.
Is it not rather one of the ultimate products of the chemical actions,
the presence of which we have great reason to suspect in the act of
muscular contraction? — Comptes Rendus, March 22, 1847.

THE NEW PLANET IRIS.

The following letter to The Times appeared on Wednesday,
Aug. 18th.

Sir, — In addition to the Berlin maps, which we have revised, and
in some instances corrected, ecliptical charts of stars down to the
tenth magnitude have been formed for some of the hours of right
ascension, which it is Mr. Bishop's intention to publish as soon as
they are completed. On the 13th of August I compared Wolfer's
map with the heavens, and was surprised to find an unmarked star
of 8"9 magnitude in a position which was examined on June 22 and
July 31 without any note being made. The mere existence of a
star in a position where before there was none visible, would not
have been sufficient to satisfy me as to its nature ; because during
an eight months' search I have^met with very many variable stars, — a

238 Intelligence and Miscellaneous Articles.

class which I believe to be far more numerous than is generally sup-
posed. But, on employing the wire micrometer we were enabled in
less than half an hour to establish its motion, and thus to convince
ourselves that I had been fortunate enough to discover a new mem-
ber of the planetary system. It may appear to many of your readers
rather bold to announce the existence of a new planet from the de-
tection of so small an amount of motion as 2 s. 5 in. R.A. ; but such
is the firm mounting of the large refracting telescope, and the per-
fection ofthe micrometers (for which wehaveto thank Mr. Dollond),
that a far smaller change would have been sufficient to convince us
as to the nature of the object in question. Mr. Bishop has fixed
upon Iris as an appropriate name for the new planet ; and we hope
that astronomers generally will join with us in its adoption. The
following are all the observations we have yet made : —

0. M. T. K.A. oflris.

h. m. s. h. m. s. o / //

Aug. 13, 9 39 46 19 57 30-38 13 37 21-5

~ 13, 10 37 24 19 57 28-41 13 27 27-6

— 14, 9 23 58 19 56 38-30 13 29 14-0

— 15, 9 0 39 19 55 4764 13 31 4-3

I remain. Sir, your most obedient Servant,

Mr. Bishop's Observatory, Regent's Park, J. R. Hind.

Aug. 17.

We have been favoured with the following additional information
by Mr. Hind : —

The planet was observed by Mr. Riimker at Hamburg, on Aug.
20, and by Prof. Gauss at Gottingen and Prof. Encke at Berlin, on
Aug. 21. M. Leverrier announced the discoveiy to the Paris Aca-
demy of Sciences on Aug. 16, giving at the same time a general
view of the various hypotheses which have been started respecting
the group of small planets. The orbit of Iris appears to be very
excentrical, and the period longer than that of any other asteroid ;
but further observations are required for the accurate determination
of the elements.

Prof. Schumacher's " Planeten-Circular" was despatched from
Altona on August 20, so that we may expect a general series of
meridian observations at the various European observatories during
the present apparition of the planet.

SUGGESTIONS FOR PROMOTING THE SCIENCE OF METEOROLOGY.

To the Editors of the Philosophical Magazine and Journal,
Gentlemen,
As I find the Meteorological Society is defunct, I beg leave to
suggest that in order that the science of meteorology may be im-
proved and promoted, and not left to chance, and in order that uni-
formity in the observations may be obtained, I propose that at the
several railway stations, the head clerk, or the cleverest man on the
premises, be supplied gratis with proper instruments, and that these
instruments should all be supplied by the same maker ; then will
they all start fair, upon certain data, which by the present system can-
not be done. And as my friend Mr. Luke Howard has suggested to

Meteorological Observations. 239

me, would not the electric telegraph be a capital means of transmit-
ting the intelligence of a thunder or hailstorm, or any change that
has taken place in any part of the kingdom where railways obtain,
and by that means unravel nature's secret with regard to meteoro-
logical phsenomena ?

But as those stations should be provided with every requisite for
taking all the necessary observations, so as to form a compendious
series of meteorological remarks, you will ask who is to furnish the
means ? In answer to that I would say, could not the Royal Society
do that, and might not the British Association take the concern
under their fostering care ? You will also say, would not the atten-
tion necessary to be paid to these observations lead to inattention in
respect to the trains ? I hope not ; and I believe, before long, such
improvements will be made in railways as to make it nearly a phy-
sical impossibility for accidents to occur.

Boston, July 3, 1847. Samuel Veall.

METEOROLOGICAL OBSERVATIONS FOR JULY 1847.

Chiswick. — July 1. Light clouds : fine: overcast. 2. Slight drizzle : cloudy.

3. Overcast : clear. 4. Very fine : clear : cloudy. 5. Sultry. 6. Very fine.

7, Overcast: slight shower. 8. Rain: cloudy: clear. 9. Cloudy and fine.

10. Overcast : clear. 1 1, 12. Very fine. 13. Sultry. 14 — 16. Excessively hot.

17. Thunder, lightning and heavy rain all the morning: fine: cloudy. 18. Cloudy.
19. Slight showers. 20. Overcast and fine. 21. Very fine. 22. Heavy clouds :
clear at night. 23, 24. Very fine. 25. Overcast. 26. Clear and fine. 27| 28.
Very fine. 29. Sultry. 30,31. Very fine.

Mean temperature of the month 65*''84

Mean temperature of July 1846 G5 '46

Mean temperature of July for the last twenty years ...... 63 '08

Average amount of rain in July 2"36 inches.

Boston. — July 1 — 3. Cloudy. 4, 5. Fine. 6. Fine : half-past 2 p.m. thermo-
meter 76°. 7. Fine : rain early this morning. 8. Cloudy : tremendous storm
of thunder, lightning and rain p.m. 9. Fine. 10, 11, Cloudy. 12. Fine : 4 p.m.
thermometer 81°. 13. Fine. 14. Cloudy. 15. Cloudy : 3 p.m. thermometer 74°.
16, 17. Cloudy. 18—21. Fine. 22. Rain. 23. Cloudy. 24—27. Fine.
28,29. Cloudy. 30,31. Fine.

Sandivick Manse, Orknei/.— July I. Cloudy. 2,3. Fog: fine. 4. Damp:
cloudy. 5. Cloudy : fog. 6, Fog. 7. Drops. 8. Rain : clear. 9. Bright :
fine. 10. Fog : bright : fine. 11. Bright : fine. 12. Clear : fine. 13. Damp:
cloudy. 14. Bright: showers. 15. Clear: fine. 16. Bright : fine. 17. Cloudy.

18. Rain. 19. Drizzle: damp. 20. Drizzle: cloudy. 21. Drizile : /og. 22.
Showers: rain. 23. Cloudy : showers. 24. Cloudy : fine. 25. Fine. 26. Bright:
drizzle. 27. Rain : cloudy. 28. Showers. 29. Showers : clear. 30. Bright :
showers. 31. Bright: rain.

Applegarlh Manse, Dumfries-shire. — July 1. Very fine : thunder. 2 — 4. Very
fine. 5. Very fine : mackerel sky and sultry p.m. 6. Very fine. 7,8. Heavy
showers : thunder. 9. Cloudy and threatening. 10. Rain. 11. Rain: fog p.m.
12. Fine, but cloudy. 13. Very fine: fog early a.m. 14. Heavy dew : very fine.
1.5. Very fine : shower and thunder. 16. Cool and breezy : thunder. 17. Very
fine : air elastic. 18. Very fine : drizzle p.m. 19,20. Very fine. 21. Fine, but
cloudy: shower and thunder. 22. ShoWers : refreshing, 23. Fair and fine.
24. Fair and fine, but dull. 25. Shower early A.M.: fine. 26. F'ine bracing air.
27. Cloudy : threatening : thunder. 28. Fair, but Cloudy. 29—31. Fair, but
cloudy : unsettled.

Mean temperature of the month 6l°"55

Mean temperature of July 1846 .' 59-20

Mean temperature of July for twenty-five years 58 '14

Average rain for twenty years 3*9i inches*

Q) 14

Uatt^jQ

'aiiqs

•uojsofi

•3ia!.«s!qo

^

•JJOIJipUBS

^M ^

■ajiqs
-sauj'uinQ

« ^

^" fe C & fe ^* I I fe

S S S . .J . . . S .. S S 6 .6 . S S a . . te' • >• fe s J s s

•wd I
•JiatMsiqj

oj oJ aJ lu „; g: g:* • S: ^' ^ ^ fe > ,; J 2 ,; g oj ^ jj ^ 2 <u . a," ^' ^' g: fc

« «

a,.-
c *

•ufd
?8

?6

IS >u

CI

wtei phIct »^!« F-!« Mte* --lei Mid ^Ici iHjci tH!«

t^vo t:^t^r^t--io*ovo r^iovo^o r~i>.>0'0 c^t>.« i>>vo*o*o r^c^toio^oo^o

•lU'B ^8

•uoisoa

n

0=2

•«!W

XBIM

c<5^cofOcoa\<Nt^t^-<*oir50M3>oooooocri'NooooiO'troioioci>-T)'cx)oo

-wg

cotN--O^C^ooc^<»QpO^O•7'l7-7<C^7"7«O^C^O^c~c»'7'Op0^t^<»(»*P'P
oooo^^0^o^6^c^c^^oooooooo^o^O^!3^^^0000^0^C^o^o^O^

cocof-'OC^cococoooo^a^o^•— <N';'•;'po^o^ap^ooopo^U5^^apr~t^
cooo6^0^CT^O^O^C^O^OC>OOC>OOO^O^c^C^0000^0^0^0^0^2>

QDQOOlO<s^o»•^lC^^oo^c^c^f^^ooor^roe^'-''-'oO'-^ooooco«^lco
S rS piS^S riS ^c» o<9 9-- — 0999c»<»c-.<7i— o^0^c^lO^Qpc^lqpcp
oCl00^6^c^l6^C^l0^6^000000000^o^o^o^oo^o^o^o^o^CT^o^<J^

§^§^2. olooiM *o Si» o^c^9•7' — — 9090^opl>•^pp90^g^9opg^c»op
0000^6^0^0^0^0^6^0^C>OOOCOOO^O^O^O^OOO^O^OO^O^O^O^

•tu*8 ?8
•uoisoa

JO sXbq

t>-ao

o o\o
O 0^c^

Qp r~Qp

C^ 0^ 0^
(M <N Ot

(O o

O '^

•^ 0^
00 vo
CM C^

6 6

CO CO

^ O 0\

CMC <N '

00 OO 0>'

. 00 -^

' Tf 00

o c^

t^CO

o op

6 CTi
COCN

oo""od^
CO CO
O 0^
6 a\
cod

lO !>• O
«3 «3 CSI
00 00 O

1 O CN

-^ to (N I

^O -^ lO ;
O O O I

o ao r~
qc CO -^
^^ ON o
(N CI CO

00 (yi

CI 7<

6 6

coco

r-oo

0^0

irjQO ys
o o o
6 6 6

CO CO CO

lOOO

COLO

o o

^ ci co'*in« rioo a.6 -^ 2 co^invd r^* O'S S gi c5d d d d d d oo
^ A (^ O

THE
LONDON, EDINBURGH and DUBLIN

PHILOSOPHICAL MAGAZINE

AND

JOURNAL OF SCIENCE.

[THIRD SERIES.]

OCrOBRR 1847.

XL. Fourth Memoir on Induction. By M. Elie Wart-
MANN, Professor of Natural Philosophy in the Academy of
Geneva *.

[With a Plate.]

[Continued from vol. xxx. p. 272.]

§ XIV. On the Commutators employed to render noltaic cur'

rents discontinuous.^ and to separate currents of induction,

116. TT is exactly a century since a remedy for various ail-
Jl ments was first sought in the electric fluid. The first
experiments were made at Geneva by Prof. Jallabert in IT^Tt*
At a later period, when the voltaic battery was invented, its
physiological effects were studied, and they are now employed
for the cure of various affections, such as obstinate ulcers J,
dumbness§, deafness II, blindness^, tic-doIoureux**,paralysisff,
&c. Lastly, since the discovery of magnetic and electric in-
duction in 1831 by Mr. Faraday, it has been found that the
induced currents, as well as the electrical discharges of the
Leyden jar, have an extremely short duration, and produce
greater shocks than batteries of a large number of elements.
The idea therefore has occurred of rendering the current of
the electromotor apparatus discontinuous, to approximate it

* Communicated by the Author.

\ Experiences sur I' Electricite, p. 127. 8vo. Geneva, 1748.

J Becquerel, Traiie de P/ii/sique,vo].'n. p. 638. Paris, 1844.

§ Namias, De alctini effetti delV elettnco sopra I Animate Economia, &c.,
p. 27. Venice, 1841.

11 Giornale per servire ai progressi delta Patologia et delta Terapeutica.
Gennajo, 1843, p. 108. Giornale delle Scienze Mediche di Torino, vol. iv.
p. 430.

^ Giornale per servire, &c.,'DecQmh&t 1841, p. 658. Biblioteca Italiana,
fascicolo 25, p. 12, &c.

** Zantedeschi, Trattaio delta Elettricitd, vol. ii. p. 525.

tt Giornale di Fisica, S^c. di Pavia, decade II. vol. vii. p. 284 ; and
vol. viii. p. 219. Annali delle Scienze del Regno Lombardo-Veneto, January
and February 1833, &c.

PhiL Mag. S. 8. Vol. 3L No. 208. Oct. 1 847. R

242 Prof. E. WaYtm&nn' sjburth Memoir on Induction.

to the cases of induced currents. Dr. NeefFof Frankfort on
the Maine, in 1835*, and M. Masson in the following yearf,
have made very conclusive experiments on this subject. An
instrument described by M. Poggendorff under the name of
inversorX, is intended to render the current of an ordinary
battery at the same time discontinuous and in an alternate
direction through a given conductor.

117. At the present day the employment of induced cur-
rents seems to become more and more general. In place of
the original magnets employed in the apparatus of Ritchie§,
Pohl||, Pixii^, Saxton**, Clarkeftj StorerJ:}:, and others, a
simple voltaic pair has been substituted, and an instrument has
been constructed, of small size, easy of transport, and producing
almost unlimited effects, called an electro-electric machine, or a
shock-machine. M. Bonijol constructs this machine with such
perfection that it has been generally adopted, and there is at
the present day scarcely an hospital where it is not found. It is
employed in the treatment of a multitude of nervous affections;
in that of amaurosis§§, in assisting parturition ||||, and as a dia-
gnostic to ascertain the state of vitality of the foetus.

118. I have had more than one opportunity of convincing
myself that many persons make use of the shock-machine
without understanding either its construction or its theory.
This maciiine, arranged on a different plan, might be rendered
both more intelligible in its mode of action and more useful to
the physicist and the physiologist. I will point out some of
the cases in which it may be employed, and afterwards the
arrangement applicable to each of them.

119. A voltaic current being given, it may be proposed —
1. To render it discontinuous, without changing its direc-
tion, in a conductor a ;

* Das BUtzrad, ein Apparat zu rasch abwechselnden gnhanischcn ScklieS'
sungen und Trennungen. Pogg. Ann., vol. xxxvi. p. 352, and vol. xlvi.
p. 104.

f Comptes Rendus de PAcad. des Sciences de Paris, vol. iv. p. 456.

X Pogg. Ann., vol. xlv. p. 372 and 385. § Phil. Trans., Oct. 1833,

II Pogg. A/m., vol. xxxiv. p. 185 and 500.

H Ann. de Ch. et de Phi/s., vol. 1. p. 322.

** Phil. Mag. N. S. vol. ix. p. 360. ft Ibid. p. 262.

XX Pogg. Ann , vol. Ixi. p. 417, 1844.

§§ Cunier, Dr., Annates d' Oculistique, vol. xii. and vol. xvi., where will
be found a memoir by Dr. Hcering On the Employment of the Electro-mag-
netic rotatory apparatus in Diseases of the Eyes.

nil See on this subject, P. Kerz, De ctcctro-viagvetisvii vi et iisu in arte
obstetricia. Bonn, 1846. — J. A. Schmidtmuller, Ilandbuch der medizinischen
GcburtsJiulfe. — T. Radford, Galvanism applied to the treatment of uterine
Haemorrhage. Manchester. — Von Kilian, Die Geburtslchre von Seiten der
Wissenschaft und Kunst, — Neue zeitschrift fiir Gcbtirtsfcunde, von H. Bursh,
d'Oiitrepont, &c., vol. xvi. No. 26, &c.

Prof. E. Wartmann* s^^ntrth Memoir on Induction, 243

2. To render it discontinuous, and in alternately contrary
directions.

This current being employed to react on a wire B, near the
conductor A, it may be required —

3. To isolate the direct currents, induced from the closing
of the circuit A ;

4. To isolate the inverted currents, induced on breaking
this circuit;

5. To emit these currents successively, giving them the
same direction ;

6. To emit them alternately in contrary directions, just as
they are produced directly.

It is known that there is a reaction of the induced currents
on the principal current. We may therefore desire —

7. To collect the totality of their reaction;

8. To avail ourselves only of the reaction of the direct in-
duced currents;

9. To avail ourselves only of that of the inverted induced
currents ;

10. To collect only the induction of the inductor on itself.
120. Physicists have studied the majority of these cases;

but the mechanical instruments which they have imagined and
described under the names of disj'unctor^, tachytrope\, rheo-
tropeX, gyrotrope^^ or co7nmutator\\, are scarcely applicable
except to one or other of the first two categories. The most
complete of these instruments, reinvented in Paris seven years
after having been described and employed in Germany, is
composed of four isolated wheels on the same axis, the outline
of which presents successively metallic and ivory arcs, against
which press conducting springs. The axis is set in motion by
means of a handle or tooth-wheel. Sometimes the interval of
the teeth is left void, and the spring in escaping determines
the opening of the circuit. Other commutators are formed
with needles arranged on isolated axes, in such a manner that
one is immersed in mercury at the instant when the other

* Dove, Magneto-elektrischcr Apparat zum Hervorbringen inducirter
Strome gleicher hitensitdt in von einander vollkommen getrennten Dr'dhten
Pogg. Ann., vol. xliii. p. 51 1 . ] 838.

"^ Dove, Ueber den Gegenstrom zu Avfang tend Ende eines prim'dren,
Pogg. Ann.y vol. Ivi. p. 251.

X Masson et Breguet, Memoirc sur ^induction. Ann. de C/i. et de Phys.t
vol. iv. p. 134. 1842.

§ Pogg. Ann., vol. xxxii. p. 539 ; and vol. xxxiv. p. 18S and 500.
1834-35.

II Jacobi, Sur P application de V Eleclro-magnetisme au monvement des via-
chines, § VII. Potsdam, 1835. Taylor's Scientific Memoirs, vol. i. p. 503,
4rcMvc'8 de PElectr., vol iii. p. 244.

R2

244- Prof. E. Wavtmann' sjburth Memoir on Induction.

comes out of it*. These different systems are complicated,
and subject to several inconveniences. The rheotrope, which
I shall proceed to describe, and which is especially applicable
to electro-electric machines, combines with the advantage of

* On the 18th of June, 1840, I communicated to the Society of Physics
and Natural History of Geneva an apparatus of this kind, the construction
of which presents no difficulty, and which is deposited in the Cabinet of
Physics in the Academy of Lausanne. The following is a description of it : —
" My commutator is composed of a pure copper stem a b (Plate II. fig. 1),
intersected in the middle by a piece of ivory c : the latter is hollowed into
the nut of a screw, in such a manner that the two halves of the stem screw
into it. Between these metallic extremities some sealing-wax is run, in
order to isolate them entirely. The cylinder thus formed is arranged ho-
rizontally, and each of its branches is furnished with symmetrical pieces at
equal distances. These pieces are two copper teeth ef, placed perpendi-
cularly and at a right angle on the axis ; then a copper circleg. Lastly, to
one of the extremities of the stem is fixed a pulley h, m the groove of
which there runs a ccrd i, which again passes over a lower pulley k, which
is much larger, vertical, and moveable by means of the handle m in one of
the supports of the apparatus.

"The six projecting pieces of copper dip into a glass vessel w (fig. 2),
placed on two small horizontal barso; it presents six isolated compart-
ments full of mercury. The extreme circles remain immersed in tliis
liquid during the entire rotation of the stem, the arrangement of the teeth
causing one to be immersed whilst the neighbouring one is not. It is
easy to regulate the quantity of mercury in the troughs so that the immer-
sion of the one may correspond exactly to the exit of the other.

" Supposing it be desired to emit into a rheometer the two induced cur-
rents, giving to them the same direction, it is sufficient to bring the extre-
mities of the wire in which the induction is produced in the extreme com-
partments reserved for copper circles. The ends of the wire of the mul-
tiplicator are tied to bars of copper connecting the troughs ef, ef, cor-
responding on the right and left of the isolator c to the needles fixed at
a right angle. So likewise on connecting the extremities of the rheometric
wire only with the troughs//, or with the troughs e e, it is evident that the
direct or inverted induced currents only may be collected.

" I have combined with this arrangement one which M. Bonijol has
employed in some of his apparatus. It consists (fig. 3) of planting one of
the ends of the stem in a flattened wooden cylinder ?•, on which a spring s
presses, passing into a circular cylinder t of hard wood, and the free extre-
mity of which u is placed by the rotation of r in contact with an amalga-
mated metallic capsule x, or is removed from it. Then, by connecting the
spring on one side, and the capsule on the other, with the wire which the
direct current of the pile traverses, we obtain by the simple rotation of the
stem any number of inductions.

"This apparatus enabled me to discover that the thermo- electric cur-
rents are capable of induction like the hydro-electric currents. I employ
a single bismuth-antimony pair, the solder of which is kept at 100° by steam.
The bismuth extremity is connected with the spring s, the antimony ex-
tremity with a wire covered with silk, which makes seventy turns on a
frame, and terminates at the capsule x. On the same frame is rolled an
isolated and finer copper wire which makes 1200 coils (110.), and both
ends of which terminate in the troughs p q. The induced circuit is closed
by a very delicate rheometer (5 a), which deviates five degrees and more,

Prof. E. Wartmann's/owr^i^ Memoir on Induction. 245

being more simple, and consequently less subject to derange-
ment, that of not requiring the employment of mercury, and
of serving to solve all the cases above stated.

121. H (Plate II. fig. 4) is a reel on which two insulated
wires are wound ; one the inductor A, by which the current
of the battery j9 n is made to pass; the other the induced B,
intended to become the seat of the currents of induction.
Three brass wheels r, 5, t, of the same diameter, are isolated
from one another on a common axis; their circumference pre-
sents an equal number of parts alternately of metal and wood.
Two metallic springs a, b are fixed against the wheel r, in
such a manner that the first leans against a conducting arc,
and the other against an insulating arc. The wheels s and t
are each pressed by two springs c d, ef^ similarly arranged.
The central metallic parts of the three wheels are in constant
communication with the springs ^, ^, i.

122. If it be desired to collect the voltaic current always in
the same direction after having rendered it discontinuous, it is
sufficient to connect the spring / with the pole p by a wire «,
and the other pole n with the spring^ by means of any con-
ducting wire different from the wires A and B wound upon
the reel. If it is wanted to obtain, as with the inversor, the
discontinuous current in directions alternately contrary, we
must join the springs c and e as well as the springs d and/j
and then connect the extremity of the conjunctive wire of the
battery with the spring h.

123. When it is desired to employ currents of induction,
the contact of the extremity / of the inducting wire with the
pole w is established permanently, and that of the extremity m
with the spring/*. Now, to isolate the direct currents induced
at the closing of the circuit A, we have only to connect the
ends x and y of the wire B respectively with the springs h and
g. — To isolate the inverted currents, we unite .r with h and y
with d. — To cause the direct and the inverted currents to pass
one after another in the same direction through the rheometer
G, for example, we connect the springs a and c with the end
s of the wire of the instrument, the springs b and d with the
end t, the extremity x with the spring h, and the extremity y
with the spring g. — To collect the induced currents alter-
nately in contrary directions, just as they are produced directly,

when it is traversed by direct and inverted currents in the same direction."
(See the Transactions of the Helvetic Society of Natural Sciences for 1840,
pp. 173, 195.)

Prof. Dove has demonstrated thermo-electric induction by a different
process. His researches were made at the same time as mine, and in an
independent manner. (See Pogg. Ann., vol. xlix, p. 97* 1840.)

246 Prof. E. "WaYlmann'sJourth Memoir on Induction.

we disconnect the extremities .2? and y of the wire B from the
springs of the rheotrope.

124>. Lastly, if we propose to employ the reaction of the
induced wire B on the inductor A, and that of the inductor
wound in a helix on itself, we substitute for the wire a the body
which is to be subjected to the effects of these reactions.
We then employ one of the four arrangements above described
(123.), according as we wish to obtain the totality of influence
of the two currents induced in the same direction, or in di-
rections alternately opposed, or again, the separate influence
of the direct or the inverted currents. The simple induction
of the inductor on itself is obtained with a reel with a single
wire in place of the conductor a, and the arrangement de-
scribed (122.).

125. It remains for me to give some details on the con-
struction of the rheotrope. The three metallic wheels r, 5, t
(fig. 5) present on their periphery twelve hollows filled in with
hard wood. These heterogeneous wheels have been worked
together by the lathe ; they are each C^'SO in diameter, and
0'^"Q6 in thickness. A metallic tooth of the middle wheel
s exactly corresponds to one isolating part of the extremes r
and t. They are placed on the same brass axis /cl, v/hich is
turned by a winch ti or a tooth-wheel. The spring / and the
wheel t are in metallic contact with the axis. The wheels r
and 5 are, on the contrary, each isolated from it by an ivory
ring covered externally with a brass cylinder. These two
cylinders bear the wheels, and are constantly pressed by
springs g^ /i, which embrace them on a semi-circumference.
The three springs^, ^, i terminate on the three heads ^>', h', i',
by means of which they can communicate together. Lastly, the
six springs a,b,c,cl,e,J'iire made of plates of hammered copper;
they are fixed to the base of the instrument by screws, ?',s',^',
the heads of which, similar tog', and pierced like them with
two holes, can receive the metallic wires intended to establish
a connexion between the different wheels. These springs are
cleft in order that the groove may facilitate the adjustment of
their length. Above they bear a screw (fig. 6) in the part
which has to rest on the circumference of the wheels ; the
opposite notch allows of regulating the elasticity of the spring
and the degree of friction. The play of these pieces may thus
be regulated with minute precision.

126. If it is not wanted to impart the same direction to the
two induced currents, the apparatus may be simplified by
giving it only two wheels. One is reserved to render the cur-
rent of the battery intermittent; the other is joined to the in-
duced wire; and according as there is coincidence or alter-

Prof. E. Wai'tmann'sj^?^rM Memoir on Induction. 247

nation in the closing of the two circuits, only either the direct
or the inverted currents are received. This double effect may
be obtained by changing the point of contact with one of the
springs, or by varying the position of one of the wheels on the
axis relatively to the other. Two wheels do not permit of
giving the same direction to the direct and the inverted cur-
rents ; because as it is evident that the induced circuit must
communicate with the two wheels when the principal current
is closed, a part of this current may proceed from the wire of
induction and modify the effect of the direct induced current.

127. Lastly, if it be desired to isolate only the inverted in-
duced currents, the rheotrope may be reduced to a single
wheel. It is sufficient for the proposed object to open the
induced circuit when the inducting circuit is closed, and vice
versa. But this arranijement would not be suited to isolate
the dn-ect induced currents, because it would be necessarj' to
close simultaneously the two circuits, and the voltaic current
would be propagated in the double channel presented to it.

128. It will be found convenient to mark letters on the dif-
ferent pieces ^, h\ i', r', s', t\ and to repeat them at the extre-
mities of the metallic conductors employed to connect these
pieces. These conductors will be fixed to the interior of the
lid of the case which contains the whole electro-electric ma-
chine; and a brief direction will indicate which ought to be
employed to produce the effects corresponding to the differ-
ent possible cases.

129. It is understood that the commutator with three or
with two wheels is applicable to all magneto-electric machines,
telegraphs, clocks, &,c., whose motive principle is the electri-
city of the magnet or of the battery.

§ XV. Employment of induced currents to restore sensation.

ISO. The cases of nervous weakness which have yielded to
a judicious application of electro-physiological shocks and
discharges are too well ascertained to admit of any question.
Since the marvellous effects of aether have been known, I have
proposed to several physicians the employment of the electro-
electric machine, or at least of intermittent currents of very
short duration, to obviate the dangers which the injection of
too strong a dose of this liquid, or a too prolonged inhalement
of it, might produce. I have made some experiments* with
a view to verify the accuracy of my expectations; and although
they are so few as to require to be repeated and varied, I shall

* In company with Dr. A. P. Prevost, and Mr. Schnetzler. I take this
opportunity of thanking these gentlemen for their zealous cooperation.

248 Prof. E. Wartmann^s fourth Memoir on Induction.

give them here, because similar results have recently been
announced by M. Ducros*.

131. The animals subjected to experiment were a rabbit
three months old, a chicken nine months old, and some frogs
of both sexes. They are all very sensitive to electric shocks.
The action of aether upon them is also very powerful, espe-
cially on the frogs, which should not be moistened with this
liquid.

132. The rabbitand the chicken appeared to have recovered
their sensation sooner under the influence of the shocks of
induction than by simple exposure to the air. In the frogs
no difference in this respect was remarked.

133. The aetherization was effected by plunging the animal
into a glass cylindrical vessel, in which boxes were arranged
furnished with sponges moistened with aether ; it was covered
with a piece of linen dipt in water. The internal atmo-
sphere was removed from time to time by removing the co-
vering.

134. The most remarkable case was presented by the
chicken. A quantity of aether, more than sufficient to produce
insensibility, was injected into its rectum. When it arrived
at this state, two or three shocks of the electro-electric appa-
ratus (110.) were passed from one wing to the opposite leg,
which shocks were effected by a Grove's pair; immediately the
eyes opened. On continuing the discharges in a very inter-
mittent manner, the animal was seen to struggle, to rise on
its feet, and then to fly to the end of the laboratory, relapsing
gradually into an insensible sleep under the influence of the
portion of injected aether which had not as yet produced its
effect.

135. The rabbit and the chicken were subjected to several
successive setherizations. The former, young and weak,
died six or seven hours after the fourth trial (injection). At
the end of fifteen hours its body was stiff, as if death had re-
sulted from natural causes. Its nerves exhibited the soften-
ing mentioned by some anatomists. The chicken, on the
contrary, survived, and even on the following day laid an egg
with a soft shell. It subsequently produced several others
perfectly healthy. It did not appear to feel the effects of the
shocks or injections to which it had been subjected. It ate
corn greedily, and the rabbit lettuce leaves, as soon as the
stupefaction produced by the aether had terminated.

136. Experiments were made on the frogs and the chicken ;
one while with the effect of the induced currents successively

* Comptes Rendus de l* Academic des Sciences de Paris, sitting of the
22nd of February 1 847, p. 286.

Prof. E. Wartm&nn* sjburth Memoir on Induction. 24-9

direct and inverted, at anotlier with inverted currents only,
employing the arrangement above described (127.)« There
w^as no perceptible difference between the two methods of
electrifying, even on circulating the inverted currents from
the feet to the wings, or vice versa.

§ XVI. Action of Induced Currents on Albumen.

137. Brande was the first who pointed out the coagulation
of albumen on the positive pole of the battery. M. Matteucci, in
treating of the physiological action of electric currents *, says,
that if the pole which was first positive be rendered negative,
the albumen is not seen to redissolve, and that consequently
an electric current may very well produce a cataract, but not
destroy it. On the other hand. Prof. Zantedeschi affirms that
he has seen the liquefaction of the albumen at the negative
polef. Repeated experiments have never shown me this re-
turn to the fluid state, and lead me to adopt entirely the con-
clusion of the celebrated physiologist of Pisa.

138. The coagulation of albumen does not present any re-
markable phase, when, under the immediate influence of a
battery, we substitute either direct or inverted induced cur-
rents, or the voltaic current rendered intermittent and strength-
ened by the reaction of the induction which it has engendered
in its own conductor and in the neighbouring conductor (124-.).
But the phaenomenon changes when the liquid is traversed by
induced currents in alternate directions.

139. Through the inducting wire A of an electro-electric
machine furnished with a bundle of iron wires, I passed the
current of five Grove's pairs of 0™'l square surface. The
extremities x and y of the induced wire B (fig. 7) terminated
in cups gg full of mercury. The circuit was closed by two
platina wires a, b of l"^"^ in diameter, one part immersed in the
cups, the other in the glass o full of the white of egg. The
latter immediately coagulated around each wire, especially
round that which communicated with the extremity of the
circuit B, from whence proceeded the inverted induced current,
and which corresponded to the positive termination of the
rheophorus A. At the end of a few minutes some bubbles of
gas appeared on the circumference of the coagulum. Some,
having increased in volume, rose lightly to the surface of the
viscous medium in which they were formed. The albumen,
riddled with holes, by which the gas escaped and continued

* Lezioni sopra ifenomenifisico-chimici del corpi viventi, p. 173. Pisa,
1844.

t Trattato del Magnetismo et delta ElettricUd, vol. ii. p. 511. Venice,
1845. %

250 Prof. E. Wartmann's^wr/// Memoir on Induction.

to be disengaged, turned black in several places: then a series
oi luminous spar/cles, and lastly real sparks of a bright yellow
glittered on the whole immersed part of the platina wire. At
the same time the induced wire B was heated around the reel,
the metallic pieces of the rheotrope rose in temperature, and
the upper sides of the glass, not filled whh the albumen, were
coated with aqueous vapour.

140. This remarkable phaenomenon is doubtless compli-
cated. The coruscations do not dart from one wire to the other
in the liquid ; they are seen along the wire. I thought at
first that the combustion (for it was such) only took place on
one of the electrodes (139.) ; but on repeating the experiment
many times, I saw it alternate on both of them according
as I reversed the poles of the battery, or present itself first
upon one wire, then upon the other, without the direction of
the current being changed ; or lastly appear upon only one
of them, whatever changes were made in the positions of the
rheophori and the extremities of the induced circuit. I attri-
bute this latter case, which only occurred when the surface of
the albumen was covered with a layer of aether, to the differ-
ence of the conditions of contact of the two platina electrodes
with the liquid : one, in fact, was then only covered with a
slight coagulum, whilst the other gave rise to a considerable
quantity of gas. These gases were collected on the aether in
a tube traversed by a [)latina wire cemented at its top. They
presented neither free carbonic acid, nor oxygen, nor hy-
drogen. I think that they were a mixture of oxide of carbon
and carburetted hydrogens.

HI. The albumen solidified around platina conductors
acquires the consistence of very soft glue; it is ductile, brown-
ish, even blackish, and diffuses a marked odour of burnt horn
or phosphoretted hydrogen. The platina does not take the
pulverulent appearance nor black colour which are communi-
cated to it by discontinuous alternate currents in other media;
it preserves its metallic appearance. With the assistance of
Prof. Marignac I analysed the coagulum; it contained no
trace of platina. There is therefore here no catalytic action.

142. These various remarks lead me to think that, in cir-
cumstances of imperfect conductibility of the albumen, and of
great power in the induced currents employed, the immersed
wires become heated when the coating of coagulum and of
gaseous bubbles has put a new obstacle to the passage of the
alternate currents (an obstacle rendered evident by the eleva-
tion of temperature of the external circuits), whence results a
true igneous decomposition and a burning, under the influence
of oxygen in a nascent state, of combustible elements exposed.

On eliminating the Signs in Star-Reductions, 251

l^S. Whatever value this opinion may have, it seems to
me that the decomposition of albumen by the passage of very
intense induced currents is a fact which deserves the serious
attention of physicians and physiologists. The presence of
this body in the blood, in urine, in the eye, in amniotic liquors,
&c., requires caution in the employment of violent alternate
currents.

14'1<. The appearances which I have described equally take
place in the albumen extracted from new-laid eggs, immersed
for some hours in the vapour of aether. They appear even
to be developed there more easily.

145. It is perhaps well to add, that the production of these
bright coruscations indifferendyon the two electrodes negatives
any explanation founded on a different polarity of the platina
wires, and all analogy with the pheenomena investigated by
MM. Gassiot*, Haref, and NeeH'|.

Geneva, June 18, 1840.

XLI. On eliminating the Sigtis in Star-Reductions.
By S. M. Drach, F.R.A.S.

To the Editors of the Philosophical Magazine and Journal,

Gentlemen,

THE subject of this paper was broached by the Astro-
nomer Royal in the Monthly Notices of the Royal Astro-
nomical Society for January 1847. 1 beg to propose the fol-
lowing extension, eliminating even the indices of the logs, em-
ployed.

Let A = E-P, B = F-Q, C = G-R, D = H-S;

a = e—pi &c. a' = ^—p' for decl., or =p' — ^ for N. P. Dist,
P, p, p'i &c. are numerical constants afterwards determined.

Corr. R.A. = SAa=2E^-2^P-SEp + 2Pp:
Corr. ^.P.D. = XAa' = ^Ee'-Xe'P-XEp' i-^Pp'.

Let P=28'75, Q = 30-5, R = 1'35, S = 20 (R. A. given in
time).

L Right ascension, ^? = 2% q = 2, r = 30*5, 5 = 2.

* Archives de V Electricite, vol. iii. p. 240.

\ Siliiman's American Journal, January 1841. I succeeded several
years ago in melting in an intermittent manner an iron wire of S"" diameter,
employed as a negative electrode on the surface of impure mercury in
which a copper wire bound to the positive pole is immersed. Twenty
Daniell's couples, or forty smaller Bunseu's, suffice for this experiment.

X Archives des Sciences Physiques et Naturelles, vol. i. p, 30.

252 On eliminating the Signs in Star-Redtictions.

< 1+ — sec 8 sin a +14'^ 53"" 14.s >-

+ 32^-937 \ 1 + — tan 8 sin a +T4M8m^s \

- SEp= - 308^-392 + 55^-422 { 1 + sin© +42° 32' 4"}
+ 1 -324(1+ sin 2© +55° 29'}

(J.)

+ 30'5(1— ^) + 21'254{l+ sin 53 +60° 30' 44"}
+ 0-2l7{l+ sin 2 £3 +235° 52'}.

Sum =2Art—2E^=— 387^*386 + periodical terms.
II. North Polar Distance, // = 2, (7' = 2, >'' = 30-5 5 — 2.

2Pp'— 2^P= — 74"'849 + 1 2"-476 cos 8 + 41"'912

(K.)

{1+ sin 8 sin a + 8^1 53°^ 14^}

+ 32"-937{l+ sin a +8^18'"49s}

>.

-XW

= _?^=_S

15

308"-392 + &c.

(J'.)
(K'.)

Sum =SAa — 2Ee=— 383"-241+ periodical terms.

Now if we add to (J.) and (J'.) the constant 180, and to
(K.) the constant 420 seconds, there will be only positive
quantities, and we shall have merely to subtract 10°^ or 10'
from the mean place ; the corrections being

E=28'75- 18-732 cos ©

<?=2+ 7^ cos a sec 8
lo

e' = 2 — •434COS 8 + sin a sin 8

1

y= 2 + — sin a sec 8

F=30'5 — 20-420 sin ©

y' = 2— cosasin 8
l0G = 13-5 + 10^-3-43sina&c. 0-lfl:=3-35706 + 0-1337sinatan8

0-lg'= 3-05 - 2-0055 cos a
1

H = 20 — 9-250 cosQ &c.

h-.

; 2 + :^ cos a tan 8
15

: 2 + sm «

It follows that the index of the first set in logarithms is
constantly unity, and that of the third set constantly zero,
permitting the omission of the latter. From 86° 10' S. dec. to
88° 50' N. dec. e,f, g, h will have their values range between
1 and 10, and their indices therefore always zero; these may
also be omitted. Now of the 8377 British Association Ca-

Mr. J. Brown on the Molybdate of Lead. 253

talogue stars, only sixteen fall out of this category in 1850;
viz.

Urs. Min. 2320, 4070, 4150, 4165, 6281, 6320, 6999, 7184,
Octans, 71, 2878, 5936, 5959, 6793, 7020, 7713, 8072;
a satisfactory result, as these polar segments = :j^yth=^^f7-ths
of the spherical surface.

With my constants, seven of those sixteen (2320, 4070,
4150, 5936, 6281, 6320, 7713) have all their R.A. coefficients
positive, the others have some negative. Indeed, near the
pole the annual change is so great, as to render greater con-
stants to include a dozen of these sixteen needless.

The saving in the above 8362 stars permits an additional
column, although five fig. logs, are required, giving a result as
far as O^-Ol or 0"-01.

App. R.A. = mean R.A. at epoch + yearly precession +
proper motion + ephemeral quantity.

+ Ee+F/+G^ + HA+ stellar quantity — lO"".
App. N.P.D. = similar quantity — 10',

which 10' might be already included in the mean place, as in
the planetary tables. The ephemeral quantity (depending on
the day of the year) is the same in R.A. time-seconds, or
N.P.D. space-sccojids.

Possibly these hints may be useful before reprinting the
British Association Catalogue.

S. M. Drach.

London, Sept. 2, 1847.

XLII. On the Molybdate of Lead. By Mr. John Brown*.

MOLYBDATE of lead was first analysed by Klaprothf,
who proceeded in the following manner : —
J 00 grains of the mineral finely pounded were treated with
dilute hydrochloric acid, and the whole of the silica was thus
separated. Upon cooling, the greater part of the chloride of
lead was deposited in fine crystals. The clear supernatant
liquor was then drawn off, and when sufficiently concentrated
the remaining chloride of lead was deposited. The whole of
the chloride was then carefully collected together, dried and
weighed. Its weight was 74*5 grs. From this the quantity
of oxide of lead was ascertained, which was 64*42 grs. Every
100 grains of molybdate of lead contain therefore 64*42 grs.
of oxide of lead. When the solution had thus been freed from
lead, it was concentrated by evaporation. Nitric acid was

* Read before the Philosophical Society of Glasgow, April 28, 1847, and
communicated by Dr. R. D. Thomson.

t Beitrage zur chemiscken Kenntniss der Minerat/corper, vol. i. 265.

^4 Mr. J. Brown on the Molybdate of Lead.

then added to the solution, which Immediately became of a
fine blue colour. When sufficiently concentrated, a quantity
of molybdic acid separated. The solution was then evapo-
rated to dryness, and the molybdic acid remained in the form
of a fine citron-yellow powder, which when completely dried
weighed 34<-25 grs.

The constituents therefore of 100 parts of the purest cry-
stals of Carinthian molybdate of lead are —

Oxide of lead . . 64<*42 59'59\ corrected from the

Molybdic acid . . 34*25 34--25J chloride.

As Klaproth did not know the true composition of chloride
of lead, the quantity of PbO given above is wrong. Calcula-
ting the quantity of oxide from the quantity of chloride which
he obtained, we get 59*59 per cent, of oxide of lead, which is
very near the theoretical quantity, or 60'57. But the great
error is in the molybdic acid. What Klaproth considered as
silica, was very probably molybdic acid, as that acid is not
entirely soluble in hydrochloric acid; and as he apparently
deducted this as impurity, he gets too little molybdic acid.
He also does not mention how he washed out the molybdic
acid from the chloride of lead. It could not well have been
done with water, for chloride of lead is soluble to a great ex-
tent. This is a great point of imperfection in the analysis.

II. This mineral was next subjected to a close examination
by Charles Hatchett, Esq., whose analysis is recorded in the
Philosophical Transactions (vol. xviii.), of which the following
is a summary : —

250 grs. of the ore, freed from as much impurity as pos-
sible, were put into a glass flask, and digested with sulphuric
acid for some time under a strong heat. When the solution
cooled, the clear liquor was drawn off", and the residual sul-
phate of lead washed by subsidence. This process was repeated
several times. The acid solutions were then filtered, and the
filtered liquid neutralized by caustic ammonia. After standing
for twenty-four hours a pale yellowish- coloured precipitate fell
down, which was collected on a filter, washed and dried : its
weight was then 4*20 grs. It had a yellowish colour, and
when dissolved in hydrochloric acid gave a blue precipitate
with yellow prussiate of potash.

Part of the clear blue solution, which was composed of sul-
phate and molybdate of ammonia, was then put into a retort
and evaporated down, the rest of the solution being added as
the liquid in the retort evaporated ; the whole was then dried
and strongly heated. In this manner all the sulphate of am-
monia was driven off, whilst the molybdate of ammonia was
decomposed into molybdic acid and ammonia, the former

Mr. J. Brown on the Molyhdate of Lead, 255

of which remained in the retort: the molybdic acid then
weighed 95 grs. The sulphate of lead formerly obtained was
then treated in the following manner ; — It was boiled with four
ounces of carbonate of soda in solution ; the powder was then
washed, and nitric acid much diluted was poured on it. The
whole dissolved except a small quantity of silica, which was
thrown on a filter : this when washed and dried weighed '7 gr.
The acid was then exactly neutralized with caustic potash,
which precipitated the lead as oxide : this, when washed and
dried, weighed H'6'00 grs.

The oxide of lead was then dissolved in nitric acid, and
sulphuric acid was added. After standing for some time the
solution was filtered and the filtered liquor saturated with
NH3 : after standing for some time a small quantity of per-
oxide of iron was precipitated, which when filtered and dried
weighed 1*0 gr. This, when added to the former quantity of
peroxide of iron, makes the quantity 5*2 grs., and the quantity
of oxide of lead 145 grs.

The composition of 250 grs. of molybdate of lead is there-
fore—

per cent,

Oxide of lead . .

. U5'0

58-00

Molybdic acid

95-0

38-00

Peroxide of iron .

5-2

2-08

Silica ....

•7

•28

24-5-9 98-36

If the iron and silica be subtracted as impurities, this ana-
lysis is very correct; but the method is very tedious and in-
convenient, and requires very great care.

III. The next person who turned his attention to this
mineral was Gobel*.

100 grs. of the mineral were digested with dilute hydro-
chloric acid with the assistance of heat : upon cooling, the lead
was deposited in the form of chloride. These crystals were
then collected together and dried ; the weight was then found
to be 72" 5 grs., which is equivalent to 59 grs. of oxide of lead.
The solution freed from lead was evaporated to dryness : when
perfectly dry a small quantity of nitric acid was added, and
the solution was again dried. It was then heated to redness
in a close vessel and weighed : its weight was thus found to
be 40'5 grs.

J 00 grains contain therefore —

♦ Schweigger's Journal fur Chemie und Phi/sik, vol, xxxvii, 71 ♦

256 Mr. J. Brown on the Molyhdate of Lead,

Oxide of lead . . 59'0 58*0 \ corrected from the

Molybdic acid . . 40*5 40*5 J chloride.

99-5 98-5

This method is essentially the same as that used by Klap-
roth. The result however is much nearer the truth. Gobel
however gets too much molybdic acid and too little oxide of
lead. This was probably owing to some of the chloride of lead
not being obtained, as it is soluble to a great extent in water
(1 in 152 of water)*; and the analyst does not state how he
washed the chloride of lead free from molybdic acid.

IV. The methods hitherto employed being liable to very
great objections, the molybdate of lead was analysed by another
method, which had proved successful in the hands of Mr.
William Parry last year in the Glasgow College laboratory.

26*84 grs. of the mineral finely pounded were boiled for a
considerable time with nitric acid and filtered. The unde-
composed mineral, along with a quantity of molybdic acid,
remained on the filter. This was then completely washed :
ammonia was then poured into the filter. The molybdic acid
was thus dissolved, and the insoluble matter remained on
the filter ; this was then washed, dried, ignited and weighed.
The weight of the insoluble matter in 26*84 grs. was 1*15 gr.

The solution containing the molybdate of ammonia was
then evaporated to dryness, and heated to redness in a close
vessel. The greater part of the molybdic acid was thus ob-
tained. Its weight was 6*76 grs.

The first washings from the molybdic acid and the insoluble
matter were then concentrated. Caustic ammonia was added
in order to neutralize the excess of acid, and afterwards sul-
phohydret of ammonia was added in excess. In this manner
the lead was precipitated in the form of sulphuret, while the
tersulphuret of molybdenum was redissolved in excess, giving
the solution a deep red colour. The sulphuret of lead was
then thrown on a filter, and washed with water containing
sulphohydret of ammonia. When completely washed, the
sulphuret of lead was dissolved in muriatic acid, and after
boiling for some time was filtered to get rid of the sulphur.
The filtered liquor was then concentrated, and the lead pre-
cipitated by means of o:falate of ammonia. The precipitated
oxalate of lead was then thrown on a filter, washed and dried.
By ignition the oxalate of lead was converted into the oxide ;

* I found in two experiments that 3963 grs. of water at 60° dissolved
26-2 grs. Pb CI = 1 in 151, and 4260 grs. HO dissolved 27-6 grs. Pb CI = 1
in 154 HO.

Mr. J. Brown on the Molybdate of Lead. 257

the quantity of which in 26*84< grs. was thus found to be 16*20
grs., which is equivalent to 60'35 per cent, of oxide of lead.

The next thing to be obtained was the rest of the molybdic
acid. This was contained in the washings from the sulphuret
of lead in the form of tersulphuret of molybdenum. When
the solution was sufficiently concentrated, it was made slightly
acid by means of nitric acid : a brownish-coloured precipitate
fell down, which is tersulphuret of molybdenum. This was
then thrown on a filter and washed. It was then dried at 212°
and weighed : its weight was 3*37 grs. From this and the
previous quantity of molybdic acid the quantity per cent, was
calculated, which was 39*30 grs.

According to this analysis, the composition of molybdate of
lead is —

Molybdic acid . . 39*30
Protoxide of lead . 60*35

99*65

V. In the course of the preceding analysis it was observed
that the sulphohydret of ammonia exercised a powerful solvent
action on the mineral itself. The following new method of
successfully analysing this mineral was therefore adopted.

23*0 grs., after being reduced to a very fine powder, were
digested with the aid of heat in sulphohydret of ammonia.
The solution became immediately of a deep red colour,
owing to the tersulphuret of molybdenum which was held in
solution by the sulphohydret of ammonia, while the lead was
precipitated as sulphuret, and fell to the bottom in the form
of a black powder. The clear supernatant liquor was then
drawn off, and a fresh portion of sulphohydret of ammonia
was added. This, after standing for some time, was thrown on
a filter, and washed with water containing sulphohydret of
ammonia. The tersulphuret of molybdenum passed through
in solution, while the sulphuret of lead remained on the filter.
When this was completely washed it was dissolved in dilute
muriatic acid, which took up the sulphuret of lead and left
the undecomposed matter along with the sulphur. These
were then thrown on a filter and washed: the whole jivas then
burnt. The sulphur was thus driven off^ while the insoluble
matter remained. The insoluble matter in 23 grs. amounted
to '24 gr., whilst in the former analysis it amounted to 1*15
in 23 grs.

When the washings from the sulphur were sufficiently con-
centrated, the lead was precipitated by means of ammonia and
oxalate of ammonia. The oxalate of lead was then thrown
on a filter and weighed. The quantity of oxide of lead in

Phil. Map. S. 3. Vol. 31 . No. 208. Oct. 1 847. S

258 Mr. J. Brown on the Molybdate of Lead.

22*76 grs. amounted to 13*71 grs., which is equivalent to 60*35
grs. per cent.

The next point was to precipitate the tersulphuret of mo-
lybdenum. This was done by making the sohition in sulpho-
hydret of ammonia slightly acid by means of muriatic acid.
The tersulphuret went down in the form of a brownish-coloured
precipitate. This was then thrown on a filter, dried, ignited
and weighed. The quantity in 22*76 grs. was thus found to
be 991 grs., which is equivalent to 39*13 per cent, of molyb-
dic acid.

The constituents therefore of molybdate of lead according
to this analysis are —

Molybdic acid . . 39*19
Lead protoxide . . 60*23

99*42
Phosphates and arseniates of lead were decomposed in the
same manner ; and it is evident this process would also answer
with antimoniates, vanadiates and seleniets.

Klaprotli.

Hatchett.

Gijbel.

Parry.

J, Brown.

Theory.

Molybdic acid , . .
Protoxide of lead
Peroxide of iron
Silica

34-25
59-59

38-00

58-00
2-08

•28

40-50
58-00

*40-40
59-60

39-88
59-56

39-30
60-35

*40-64
59-36

39-19
60-23

39-13

60-87

93-84

98-36

98-50

100-00

99-44

99-65

100-00

99-42

100-00

Note by Dr. R, D. Thomson.

Test for Arseniates^ Sfc. — I may notice a simple and quick
method of testing minerals containing arsenic in its various
forms, phosphates, molybdates, vanadiates, &c. A few grains
of the mineral to be examined are to be finely pulverized in an
agate mortar and introduced into a test-tube, and boiled with
bisulphohydret of ammonia for a few minutes. The mineral is
partially decomposed ; the sulphuret of lead precipitates, while
sulphuret of arsenic, &c. is dissolved by the excess of the re-
agent. The tube is then allowed to stand at rest, and the
supernatant liquor poured off or filtered. The excess of bi-
sulphohydret of ammonia being removed by evaporation, the
yellow sulphuret of arsenic precipitates. A molybdate is

• In these analyses the lead only was ascertained, and the deficiency
was taken as molybdic acid.

The two last analyses were made by means of sulphohydret of ammonia,
the tln-ee preceding by nitric acid.

On FossilCoXixmiiQS in the Carboniferous Strata near Wigan. 259

detected at once by the fine orange-red colour which the re-
agent assumes when it is heated in contact with that mineral.

A vanadiate gives a dark colour, but possessing less of the
red shade than the niolybdate. The liquor filtered from the
sulphuret of lead containing the vanadium in solution has a
green colour, becoming blue by the addition of hydrochloric
acid. Hence it appears that arsenic dissolved in bisulphohy-
dret of ammonia does not alter the colour of that reagent, while
the liquor gives a precipitate of orpiment by concentration.
Molybdenum and vanadium, on the other hand, render that
reagent I'eddish, and give brown precipitates by concentration.
The liquor filtered from the sulphuret of molybdenum is
colourless, or its hue is similar to that of the reagent, while
the liquor derived from the vanadium precipitate is green,

I have succeeded in decomposing a sufficient amount of
these minerals for quantitative analysis by the preceding prp^
cess when they have been carefully pounded and laevigated.
The process is particularly advantageous in the analysis of
molybdate of lead, where the use of nitric acid for dissolving
the mineral is objectionable in consequence of its tendency to
form the molybdate of molybdenum, and where hydrochloric
acid, by producing a chloride of lead, renders the employment
of an inconvenient quantity of water necessary. I have found
this process for testing very convenient where it was desirable
to use minute quantities of crystals, and where rapidity is an
object in view, as in examining a large collection of minerals
of the preceding description ; and I mention it for the sake of
those who may possess in their cabinets minerals of this nature
which they may desire to test, since it may be found a use-
ful adjunct to the blowpipe test.

The bisulphohydret affords a simple distinguishing test
between metallic arsenic and antimony, when spots have been
received on porcelain by Marsh's process. Arsenic dissolves
in the reagent, and leaves a yellow stain by evaporation. An-
timony dissolves and leaves an orange stain. For this expe-
riment it is convenient to use the inside of the cover of a
porcelain crucible.

X LI 1 1. 0}i Fossil CdXuxmies found standing in an erect position
in the Carboniferous Strata near Wigan, Lancashire. Bi/ E.

W. BiNNEY*.

THE fragmentary condition of the great bulk of fossil
plants found imbedded in the coal measures has led many
geologists to suppose that they had been drifted from adjoin-

* Read before the Literary and Philosophical Society of Manchester,
July 6, 1847, and communicated by the Author.

S2

260 Mr. E. W. Binney on Fossil Calamites in

ing lands, and had not grown in the position in which thej'
are now found. But although it is certain that plants which
have been drifted by water generally })resent a broken appear-
ance, it is equally true that plants grown upon the spots where
they are now found, having been laid low by the action of
currents of water, or weighed down and buried by the weight
of mud or silt that had fallen upon them, afford similar ap-
pearances, so that gi'eat care must be taken before we conclude
that a plant has not grown on the place where it is now found
merely because we find it in fragments.

A few years ago the whole of the fossil flora was generally
supposed to have been drifted. The first plant that was ex-
cepted from this rule and recovered its proper place was the
Stigmaria, whose long stringy rootlets prevented it from being
so conveniently drifted by currents as the advocates of the
drift hypothesis could desire; therefore it was allowed to have
grown where it is found.

When numbers of Sigillariae were found standing erect on
seams of coal, and their roots had not been traced to their
extremities, it was at first attempted to refer them to accident,
like the snags now found in the Mississippi and other rivers.
However, a more careful observation of these fossils, and the
great number in which they were found, at length induced
geologists to admit that they must have grown where they
are now met with. The discovery of the trees at Dixon Fold
on the Manchester and Bolton Railway by Mr. John Hawk-
shaw, F.G.S., and so ably described by the late Mr. Bowman,
F.G.S., in the first volume of the Transactions of the Man-
chester Geological Society, mainly contributed to establish
this view, which has been since clearly proved by the certainty
that Stigmaria? are the roots of Sigillariae, as the fossil trees
of St. Helens and Dukinfield testify.

As yet, however, Sigillaria was the only tree that to any
extent could be said to have been discovered m situ.

In the present communication, it is intended to show that
Calamites have been found standing erect on the places in
which they grew by the side of Sigillariae, and that the root-
lets of the former ver}' much resemble, if they are not identical
with, those of the latter plant.

The rootlets of Calamites have been very correctly figured
and described by Messrs. Lindley and Hutton in vol. i. pp. 78
and 79 of their Fossil P'lora of Great Britain ; but it is believed
by the writer that although numbers of erect Calamites have
been observed in the coal-measures, still none of them to his
knowledge have been described with their roots standing in
the position in which they had grown.

the Carbo7iiferous Strata near Wigan. 26 1

During an examination of the deep excavations through
the coal-measures made in forming the Bury and Liverpool
Railway in the vicinity of Wigan, I was so fortunate as to
discover on the 2 1st day of April last, in the Pemberton Hill
cutting about two miles west of Wigan, not only a whole forest
of Sigillariae standing erect with their roots just as they had
grown, but also many Calamites in a similar state of perfec-
tion.

The accompanying woodcut, fig. 1, representing a view of
the south side of the railway cutting, will show the position
in which the fossils occurred, although it is on an exaggerated
scale, and the characters of the trees are not given.

Fig. 1.

The excavation in which the fossils were met with is about
twenty-five feet deep, and consists chiefly of a light gray-
coloured silty clay known by the provincial name of" Warren,"
containing nodules of ironstone. This deposit is very similar
in composition to the strata in which the fossil trees at St.
Helens and Dukinfield before described were found. It lies
between two beds of coal each about two feet in thickness, and
occupies a position in the higher part of the middle division
of the Lancashire coal-field. The upper seam of coal is
covered, and in some places partly removed, by a deposit of
one or two yards in thickness of till. Near the bridge is seen
a flexure in the strata, as shown in the woodcut.

In Mr. Haliburton's section at Haigh (vol. vi. New Series,
of Manchester Memoirs, p. 437, Remarks on the Coal District
of South Lancashire, by James Heyvvood, Esq., F.R.S.), occur
the following strata : —

262 Mr. E. W. Binney on Fossil Calamites m

yards, ft. in.
Depth from the surface .... 10 0 0
Coal which burns to a white ash .10 6

Interval 800

Coal (Wigan yard coal) .... 0 2 6

Interval 16 0 0

Coal 026

Interval . 24< 0 0

Coal (Wigan four-feet coal) ... 1 1 0

Interval 32 0 0

Coal (Wigan seven- feet coal) ..210
The interval of eight yards is in my opinion the deposit in
which the fossil trees were met with.

In a distance of about fifty yards of the cutting, on my first
visit to the place, I observed full thirty upright stems of Sigil-
lariae, besides several flattened ones lying in a horizontal posi-
tion. These trees exhibited no evidence of their former struc-
ture, being mere casts, having their insides filled with a similar
material to the matrix in which they were found imbedded.
Their outsides consisted of a coating of bright coal of about a
quarter of an inch in thickness, and were ribbed and formed
as Sigillariae usually are. In diameter they varied from one
to three feet : their heights ranged from two to twelve feet;
but, with one exception of a stem with another lying directly
across it, none of them could be traced to their termination
upwards. Some of them rested with their stems on the top of
the lower seam of coal ; others had their roots midway between
the two seams ; and others again were found just under and
in the floor of the upper seam. Most of the trees, which on
exposure retained their coaly envelope, presented the irregu-
larly ribbed and furrowed appearance which the Dixon Fold
and St. Helen's trees exhibited, and which some geologists
contend are not sufficient to identify them with Sigillariae;
but six specimens were decorticated, and showed well-defined
scars and all the other characters of Sigillaria rcniforinis,
alternans and organum. All the uprigJd trees had roots of'
Stigmarice "doith their rootlets traversing the silty clay in all
directions.

Many stems of Calamites were found standing erect amongst
the last-described trees, some of which were traced four and
five feet in height without reaching their tops. These stems
varied in diameter from one to five inches : they showed no
structure internally, being mere casts filled with silty clay and
having a coaly envelope of about one-sixth of an inch in thick-
ness, which on being removed exposed the ribbed character

the Carboniferous Strata near Wigan. 263

and usual joints of this genus of plants. All those which
could be traced downwards exhibited rootlets proceeding from
the lower joints, less in size, but resembling those of Stigmarise.
One of the erect Calamites was traced for about two feet
upwards, and then at first sight appeared to terminate ; but
on more careful inspection it could be traced running in a
horizontal direction, but so much compressed as to remain
unseen without very close observation.

The erect stems both of undoubted Sigillariae and trees
which did not exhibit all the characters of Sigillariae as well
as those of Calamites, occurred in all parts of the deposit of
silty clay, from the top of the lower seam to the floor of the
upper one.

In the deposit where the trees occurred were found plants
of the genera Neuropteris, Pecopteris, Sphenopteris, Cyclo-
pteris, Odontopteris, Asterophyllites, Pinnularia, Lepidoden-
dron, Lepidophyllum, Lepidostrobus, Lycopodites, Spheno-
phyllum, &c.

Having thus given a hasty sketch of the locality where the
fossils occurred, and the fossils themselves as they appeared
to me on my first visit to the place, I shall proceed to describe
some erect stems of Calamites, which are intended to form
the chief subject of this communication. These trees were
not only seen by myself, but by Dr. J. Hooker and M. Jobert,
two well-known geologists ; and it is to the latter gentleman
that I am indebted for the drawings which accompany this
paper.

On the 22nd of May last, in company with the above-named
gentlemen, I again visited the Pemberton Hill cutting. Many
erect specimens of Calamites, both with and without roots, had
been seen on my previous visits to the place ; but the three
which it is now my intention to describe exhibited the lower
terminations, and more distinctly showed the rootlets than the
other specimens.

The three fossils marked Nos. 1, 2 and 3, in the rough
sketch before given, and No. 4, an individual examined by me
on a previous visit, occurred in the excavation on the south
side of the railway. They were all found standing in an erect
position about two yards distant from each other, having their
tops, as far as bared, two yards under the upper seam of coal.
They were each exposed from twenty inches to two feet, and
all presented the same external characters with regard to their
stems, joints, and rootlets, and most resembled the Calamites
approximatiis.

The description of No. 1 will serve for the other two.

This specimen appeared standing in the silty clay in a nearly

264

Mr. E. W. Binney on Fossil Calamites in

erect position, with the exception of a slight bend in its upper
part, as shown in the drawing. It was almost cylindrical, and
measured twenty-one inches from the base to its highest part,
which was exposed. Its greatest diameter, which occurred
near the top, was one and a half inches ; it then tapered
slightly towards the bottom, and terminated in a club-shaped
end. The exterior was covered with a coating of fine coal of
about one-eighth of an inch in thickness, which on being re-
moved exposed the usual ribs, furrows and joints, character-
istic of Calamites. The interior showed no trace of structure,
being composed of the same kind of silty clay as the matrix
in which the fossil was found.

The following is a sketch of No. 1 as it appeared in the cut-
ting, one-eighth the natural size of the fossil. The upper
part had been removed before we saw the specimen.

Fi'^ 2.

I'he joints or nodi were
ten in number, and occur-
red at irregular distances,
but nearer together at the
upper and lower extremi-
ties than in the middle of
the fossil.

At the joints small cir-
cular depressions were seen,
from which proceeded root-
lets. These could be traced
from eight to eleven inches
in length without reaching
their terminations. They
went down into the silty
clay, the higher ones ma-
king an angle of about 15°
with the horizon ; but the
angle gradually increased
as they went lower, until
they at last described an
angle of about 45°.

The rootlets appeared to have been originally cylindrical
and about one-eighth of an inch in diameter; but they were
now compressed, and their outsides covered with a thin coat-
ing of carbonaceous matter. On a careful removal of the
outside a delicate longitudinal stria could be perceived on the
rootlets; there also appeared something like a pith in their
middle.

Altogether the rootlets could not be well distinguished from
those of Stigmaria. They also appeared to come from the

the Carboniferous Strata near Wigan. 265

stems in something like quincuncial order, like the rootlets of
the last-named plant; but of this circumstance I cannot speak
with absolute certainty.

The specimen No. 4 differed from the other three only in
its base, which did not terminate in the same club-shaped
extremity which they did; but after the joints had gradually
approximated it turned inwards, and it could not then be seen
whether it ended or was inserted into some other stem.

In addition to the above there were many Calamites, both
in erect, inclined, and horizontal positions, but no leaves or
branches were satisfactorily traced to them.

In the course of his examination of upright stems of Sigil-
laria^ in the coal-measures, the writer has nearly always found
Calamites associated with them. At St. Helens they were
abundant, and their bases were found in contact with the
main roots of Sigillariae. One of the authors of the Fossil
Flora, Mr. Hutton, in describing the Burdiehouse fossils at
page 24, vol. iii. of that work, states as follows : " Amongst
vegetables, the characteristic fossils of this deposit are Lepi-
dostrobi, Lepidophyllites, Lepidodendra, and Filicites; the
rarity of Calamites, which occur but seldom, and of a dimi-
nutive size, and the almost entire absence of Stigmaria, are
very striking to those who are accustomed to view the fossil
groups usually presented by the beds of the carboniferous
formation ; whilst the profusion of Lepidostrobi and Lepido-
phyllites of various sizes and in various stages of growth asso-
ciated with the stems of Lepidodendra and those of no other
plant, is an additional argument for the opinion which has
always appeared highly probable, that they were the fruit,
leaves and stem of the same tribe of plants. Of Sigillaria, a
plant which in the flora of the carboniferous group generally
is of so much importance, we could not observe a trace."

In the course of his own observations, the writer has never
yet been able to meet with a stem of Sigillaria of so small a
size as six inches in diameter, or a Calamites of so large a size
as that. Doubtless there must have been young Sigillariae
whether or not there were large Calamites. Now what are
oung Sigillariae? This is a question which yet remains to
e answered.

It is now admitted that little is known about the true nature
of the genera Sigillaria and Calamites, except that they were
not the hollow succulent stems which they were once supposed
to be.

The rootlets of Calamites, as previously shown, if not actu-
ally identical with, at least very much resemble those of Sigil-
laria. In some specimens of this latter genus, especially those

I

266 Messrs. Frankland and Kolbe on the

of the species approximatus, figured and described in plate 216,
vol. iii. of the Fossil Flora, and the cruciatus^ figured in plate
19 of Brongniart's Histoire des Vegetaux Fossiles, their root-
lets are arranged in regular quincuncial order. In the largest
Calamites that to my knowledge has been figured, namely,
that called Gigas, plate 27 in Brongniart's work before alluded
to, the ribs and furrows begin to appear very like those of
Sigillaria, and the joints show indistinctly. 1 he termination
of the root of a Calnmites is exactly of the same form as the
terminal point of a Stigmaria, both being club-shaped.

I am not aware that up to the present time much, if any-
thing, is known of the structure of Calamites ; but if it should
resemble that of Sigillaria, it may tend to prove that Calamites
are but young Sigillaria.

In our observations it must not however be lost sight of,
that no central axis or pith has to my knowledge yet been
discovered in the stem of the Calamites like that found in Si-
gillaria. Both plants are proved to have had similar hahitatSy
and therefore it is very probable that they might have had
rootlets resembling each other without being the same plant.
Still, however, as Sigillaria was so long considered a separate
plant from Stigmaria, it is unphilosophical to take no notice
of the analogies of what are now considered distinct genera.
Although it will not by any means be safe to affirm that Sigil-
laria and Calamites are the same plant from their analogies,
still it is conceived that sufficient evidence has been adduced
in this paper to prove that the latter as well as the former
plants have generally grown on the places where they are
now found, and that the reason why one is so much more fre*
quently found in an erect position than the other, arises from
the circumstance of the stem of the one being much stronger
than that of the other. A deposit of mud on the branches
and leaves of the slender stem of a Calamites might weigh it
down and prostrate it, whilst the stout trunk of the Sigillaria
would resist such action and continue erect.

XLIV. Upon the Chemical Constitution of Metacetonic Acid,
and some other Bodies related to it. By E. Frankland,
Esq. and H. Kolbe, Ph.D.^

T^HE researches into the constitution of organic compounds
•*- certainly belong to the most interesting in chemistry.
But they are always attended with more or less danger, and
those who, leaving the safer road of experiment, plunge into
the depths of hypothesis, and build up theories apparently
* Communicated by the Chemical Society; having been read April 19,
1847.

Chemical Constitution of Metacetonic Acid. 267

ingenious, though often untenable, frequently stumble and
fall amongst a host of contradictions. It is a common error,
as experience teaches, into which young chemists are very apt
to fall, that, persuaded of the infallibility of their own views,
and blind to M-ell-founded objections, they endeavour to con-
vince by quick and ready argument rather than by solid rea-
soning, and consequently they cither offend others or feel
themselves offended w-hen contradicted.

When in the face of this danger we endeavour to advance
views concerning the rational composition of some organic
acids which do not accord with those generally received, we
do it with a certain degree of timidity, and with the most
strenuous endeavour to avoid those causes of error which have
been pointed out. It is far from our intention to give a de-
cided preference to the mode of viewing the subject here pro-
posed, or indeed in any way to force our own opinion, nor
is it unknown to us that even these views leave many facta
unexplained ; but we feel convinced that no detriment can
accrue to the progress of science by looking at subjects of
Buch importance with an impartial eye from all possible sides.

The starting-point of our experiments was the idea recently
expressed by Berzelius, that acetic acid might be considered
as a conjugate oxalic acid, as methyl-oxalic acid, Cg H^ Cg O3.
If this view of the subject, which explains so readily the con-
version of acetic acid into chloracetic acid, and the remark-
able reconversion of the latter into the former, and W'hich has
been further confirmed by the analogous relations of chloro-
carbohyposulphuric acid and methyl-hyposulphuric acid, is
correct, then the question arises whether it might not be ex-
tended to those other acids, so nearly related to acetic acid,
namely, formic acid, metacetonic acid, butyric acid, benzoic
acid, &c. We are of opinion that this question cannot a pri-
ori be answered in the negative ; on the contrary, it appeared
to us, after pursuing the subject further from that point of
view, that the manifold metamorphoses which the above
combinations undergo might be explained in a very simple
manner, and we consequently have submitted the question to
careful experimental examination ; and we believe that we
have gained a fact in support of the theory of conjugate com-
pounds in its application to the acids in question, by the ac-
tion which cyanide of ethyle exhibits with alkalies and acids.

When benzoic acid is supposed to consist of oxalic acid
with the carburetted hydrogen, Cj2 H5 (phenyle), as a con-
junct, then it is evident that benzoe-nitrile obtained by Feh-
ling* in distilling benzoate of ammonia, must be a cyanogen
* Liebig's Annalen, xlix. pi 91.

268 Messrs. Frankland and Kolbe on the

compound of the same carbo-hydrogen = Cj2H5Cy. This
mode of decomposition of phenyle-oxalic acid becomes thus
completely analogous to the well-known formation of cj'^ano-
gen by heating oxalate of ammonia, and to the formation of
hydrocyanic acid from the formiate of ammonia.

Benzoe-nitrile, viewed as cyanide of phenyle, would then,
together with the analogous body recently discovered by
Schlieper*, valero-nitrile (Cg Hg Cy, cyanide of valyle), be-
come allied to cyanide of ethyle; and as these bodies in contact
•with alkalies are so easily transformed into benzoic acid and
valerianic acid, it is to be presumed that cyanide of ethyle
under similar circumstances would become transformed into
ammonia and metacetonic acid.

We prepared for this purpose pure cyanide of ethyle, ac-
cording to the process of Pelouze, by the distillation of sul-
phovinate of potash with cyanide of potassium. The yellow-
coloured liquid which passes overt was mixed with water,
and separated again by chloride of sodium, dried over chlo-
ride of calcium, and lastly distilled in a bent tube freed from
air and hermetically sealed. Purified in this manner cyanide
of ethyle is a limpid colourless liquid, having an odour much
resembling that of the terrible cacodyle. The analysis of the
substance gave the following numbers: —

0*219 grm. gave 0*523 grm. of carbonic acid and 0'18G grm.
of water. Theory.

Carbon . . 65*19 6 65*45

Hydrogen . 9*46 5 9*09

Nitrogen. . 25-35 1 25*46

To settle the question started above, this cyanide of ethyle
was added drop by drop to a tolerably concentrated boiling
solution of caustic potash, and the product of distillation re-
turned to the retort as long as it retained any smell. During
this operation a considerable portion of ammonia was given
off. The alkaline residue distilled with sulphuric acid pro-
duced an acid liquid, which, neutralized with carbonate of
silver, baryta, or lead, gave the corresponding salts of those
bases. We had previously satisfied ourselves, by a carefully
conducted experiment, that no formic acid was present in
the acid solution.

The silver salt crystallizes from its aqueous solution in
small acicular prisms. It is sparingly soluble in water, and

• Annalen der Chemie, lix. p. 15.

t We found, in contradiction to Pelouze's statement, that cyanide of
ethyle is tolerably soluble in water; but when the solutionis saturated with
common salt, it again separates unchanged and comes to the surface.

Chemical Constitution of Metacetonic Acid. 269

the solution becomes blackened on boiling. The crystals dried
over sulphuric acid in vacuo had the composition of meta-
cetonate of silver.

I. 0*211 grm. gave, vjrhen burnt vi'ith oxide of copper,
0-153 grm. carbonic acid and 0-055 grm. water.

II. 0'167 grra. gave on careful ignition 0-100 grm. of me-
tallic silver.

Carbon . . .

19-77

6

19-90

Hydrogen . .

2-89

5

2-76

Oxygen . . .

13-06

3

13-27

Oxide of silver .

64-28

1

64-07

6

25-46

5

3-53

3

16-99

1

54-02

100-00 15 100-00

The salt of barytes is very soluble in water ; the solution of
the salt evaporated to dryness and dried for a long time at
212°, gave the following numbers : —

I. 0-339 grm. gave 0-311 grm. carbonic acid and 0-116
grm. water.

II. 0-258 grm. gave 0-211 grm. sulphate of baryta.

Theory.
Carbon . . . 24-98
Hydrogen . . 3-79
Oxygen . . . 17*58
Barytes . . . 53 65

100-00 100-00

The lead compound has the sweet taste of acetate of lead :
it does not appear to crystallize, but dries up to a tough
amorphous saline mass. A portion dried at 212° and in vacuo
gave exactly the quantity of oxide of lead, corresponding to
the formula PbO, Cg Hg O3, namely, —

0*446 grm. gave 0-282 grm, oxide of lead and 0*021 grm.
metallic lead, equivalent to 63*40 per cent, instead of 63*19.

From the above analyses, there can be no doubt that the
acid product of the action of caustic potash upon cyanide of
ethyle is metacetonic acid. The same result is produced by
weak sulphuric acid (1 part acid to 2 parts water). The silver
salt prepared from the acid product of distillation in this case
exhibited the properties described above, and an estimation
of the oxide of silver gave the follow-ing result : —

I. 0*150 grm. when ignited gave 0*090 grm. metallic silver,
corresponding to 64*30 per cent, oxide of silver ; the theo-
retical proportion being 64*07 per cent.

The mode of decomposition of cyanide of ethyle described
is therefore quite analogous to that of benzoe-nitrile and
valero-nitrile ; and if metacetonic acid is considered as ethyl-

2/0 On the Chemical Composition of Metacetonic Acid.

oxalic acid, with the formula C4 H5 Cg Og, the decomposition
may be expressed by the equation

Kl0^'^'V|KaaC,H,,C,03
3HO J '-^^"s-

The assumption will be rendered still more palpable, if,
as there is little doubt, at a future period we_^are able again to
produce cyanide of ethyle from metacetonate of ammonia.

If, on the contrary, metacetonic acid is considered as an
oxide of the radical Cg Hg, then it must be assumed, that
during the process of oxidation the two atoms of carbon, in the
cyanogen, amalgamate themselves, as it were, with ethyle, in
order to form the new radical metacetyle. Setting aside the
great improbability attached to such an fissumption, to which
there would be at present no analogous case, we feel justified
in giving preference to the former explanation, because it is
the more simple, and because it completely agrees with the
known transformation of cyanogen and water into oxalic
acid and ammonia ; and if, as one of us has found, valerianic
acid, placed in the circuit of the voltaic current, by the
assumption of 1 atom of oxygen, becomes converted into car-
bonic acid and the carbo-hydrogen Cg Hg, we consider that
our view has gained by that fact an additional support, and
that the acid in question contains that carbo-hydrogen in
conjugate combination with oxalic acid.

The foregoing observations lead to a great simplification of
our hypothesis in regard to organic radicals, inasmuch as
they do away with the necessity of supposing a specific ra-
dical for each acid belonging to an alcohol. The series of
radicals which are produced by the addition of one or more
equivalents of the carbo-hydrogen Cg Hg to 1 equivalent of
hydrogen, namely Cg H3, C4 H^, &c., the hydrated oxides
of which form the alcohols, are again found in accordance
with our view in the acids derived from them, conjoined
with oxalic acid. It may be conceived that the oxygen, in
converting alcohol or oxide of ethyle into acetic acid, first
acts upon the equivalent of carbo-hydrogen Cg Hg, which
alone distinguishes ethyle from methyle, and that the suc-
cessive products of oxidation of this body take up the re-
maining radical methyle into conjugate combination ; thus in
the formation of aldehyde and acetic acid,

20 J -\ no
c,n,oi_rc,ns,c,o^

30J-\2HO

Analysis of the Ashes of the Orange-Tree, 271

Moreover, it cannot be denied that our ideas concerning
the functions performed by compound radicals are very much
enlarged by these considerations. For when we find methyle
and ethyle combining like the electro-positive inorganic ele-
ments with the electro-negative non-metallic substances, the
property which they also exhibit of uniting with oxalic acid,
hyposulphuric acid, and with other, perhaps even neutral,
bodies to form conjugate compounds, evinces such an extensive
range of properties as is nowhere to be met with amongst the
more narrowly defined powers of combination of inorganic
substances ; and it is probable that nature, when she brings
forth the innumerable and manifold products of the organic
kingdom by a wonderful combination of those iew elements
which are at her disposal, may likewise make use of these sup-
posed extensive combining powers of the organic radicals, as
the simplest means of accomplishing her greatest works.

We beg to express our warmest thanks to Dr. Lyon Play-
fair for the use of his laboratory and apparatus in carrying
out the above investigation, and for the uniform kindness
which we as his assistants have experienced at his hands.

XLV. Analysis of the Ashes of the Orange-Tree (Citrus
aurantium). By Messrs. Thomas H. Rowney and Henry
How*.

T^OR the materials used in the following analyses we are
indebted to the kindness of Mr. Da Cumara, who had
sent it over for investigation from his plantations on the
island of St. Michel, being desirous to become acquainted
with the mineral constituents of the orange-tree, which forms
the principal wealth of his country. The analyses were per-
formed under the direction of Dr. Hofmann in the laboratory
of the Royal College of Chemistry.

To prepare the ashes in a fit state for analysis, the different
parts of the plant were heated in an inclined, open Hessian
crucible, until the carbon w"as consumedf. The ashes thus
obtained were mixed with a small quantity of oxide of mer-
cury and ignited a second time in a platinum capsule over a
spirit-lamp, in order to reproduce the sulphates, which in the
former process had been reduced to sulphides.

* Communicated by the Chemical Society; having been read April 6,
1847.

t To obtain the ash of the fruit, the oranges were cut into slices, and
after separation of the seed dried on the sand-bath in a covered porcelain
dish, and then burnt in a crucible.

272 Messrs. Rowney and How's Analysis of

The same quantity of ash served to determine the potash
and soda, sulphuric and phosphoric acids, perphosphate of
iron, lime and magnesia, silicic acid and accidental sand and
charcoal. For this purpose, Jthe hydrochloric acid solution
was evaporated to dryness, gently ignited and extracted with
hydrochloric acid. The solution thus obtained was divided
into different parts. The first portion served for the determi-
nation of the potash and soda.

For this purpose the acids, lime, magnesia, &c. were re-
moved by baryta, the excess of baryta by carbonate of am-
monia, and the ammoniacal salts by gentle ignition. The
residue, potash and soda, were estimated partly by separating
them by means of bichloride of platinum (analyses of the
ashes of the root and seed) and partly by the indirect method,
namely, by converting the mixed chlorides into sulphates,
weighing these and ascertaining the amount of sulphuric acid
by means of chloride of barium (analyses of the stem, leaves
and fruit).

In the second portion, sulphuric and phosphoric acids were
determined, the former as sulphate of baryta, the latter by
neutralizing the filtrate from the former with ammonia and
precipitating the phosphoric acid by means of sesquichloride
of iron and acetate of potash. This precipitate was dissolved
in hydrochloric acid, a sufficient quantity of tartaric acid was
added, and the phosphoric acid estimated in the form of py-
rophosphate of magnesia, by precipitating with ammonia,
chloride of ammonium and sulphate of magnesia. The latter
precipitate, frequently containing a small quantity of iron,
was redissolved in hydrochloric acid, and after the addition
of some tartaric acid reprecipitated by ammonia. A third
portion served for the estimation of perphosphate of iron, lime
and magnesia. For this purpose the liquid was neutralized
with ammonia, some acetate of potash was added, and the
solution strongly acidulated with acetic acid, in order to keep
the phosphate of lime, which might be precipitated, in solu-
tion; on heating perphosphate of iron subsides, from which
the sesquioxide of iron was calculated according to the for-
mula 2Fe2 0g + 3P05. From the filtrate the lime was preci-
pitated by means of oxalate of ammonia, and after the sepa-
ration of the lime, the magnesia by means of phosphate of
soda. Chlorine and carbonic acid were determined in sepa-
rate portions of the ash. In this manner the following expe-
rimental numbers were obtained : —

the Ashes of the Orange-Tree.

273

^

«<-""-<

(N >-i I— I

(N ^ r-H

eo<o IN

U5 .-c 00
C3 t** Ci

T^ l-H 1-1

05 >-i >-<

o o in

o »o «o «o eo
;oo CO (M »o GO
:ifs (N «p OS «p
i,co CO "^ OS ©^
" IN <y«

<N (N

s

00

CO >n
0010

(N

§^

(N '-<
NiN

(N

05 CO
OJ ^

i-i t^

eooq

M QO

C3S
IN

10 0

(N -*

(NiN

■— 1

ei5«o

toeo

OOS

10 »

1— 1

OS t>»
03 (N

00 50

coco

l-H

ao F-<

S?

o OS to >f5

CO to GO OS

OS OS 5p -^

T)< Ai lis 1^

IN COIN

°-^ So

P< SL -f= -13 .

s-a

S -^ u S o -rj
o w "C =^ "d 2

•3<g

^ .S -2 o

.2 &,<«

s a ®

So

60^ 0^0

o J3 o 3 o o .5

« 4) "o -^iS wis

a ^^

c

^^

^<s-=

-5 "-I S S
^,0 g ^

•ca

J3 £

GO'S

3 ta ♦;. .2

cd s c3 -5

Phil Mag, S. 3. Vol. 31. No. 208. Oct. 1847.

T

274

Messrs. Rowney and How's Analysis of

.tn in

CO

t>. © »n

• g(N rN '* 91

^

-*

©>>.©©

M

B

©CO l-H

© Hf 0

1

o6 -*©

&

©

©©© ©

• © >0 CO >-< 1— 1

©

-N

-<#»>.©© ©

2-<l< 00 ^ to <M

CO

«^

^ CO!M CO tf)

1.^ P>-* l>>03 CO CO

t^

CO

<M 0 © •-1 "* <M

g,© O t^ "* ©

l-H

"f

© -* «q ©© f-i

' ■ ■ (N *

• • • • 6 •

.ifl M5 . . . »>»

QO

^ „g

- 2 CO «>•

(N CO

CO

" fe?$

r-, CO

^

00

CO © © '"' ©

l>« CO

CO

'^

© >-i 10 ^ CO

1

©C. <N r^

©

©

©-^©6 '

^lo ic . . ro> ■*

CO

©

CO CO 10 If? © ©

© »n '^ (>( © c

-^

CO

"fe?^

CO ©

qp CO

©

CO CO 10 -^ CO "*

© <>J CO r-H CO »^

1

5

>{5

J^

N r-.

' Al ' ' ' '

_ <u

bD

CO

a

^. 2«0 ©

-H ©

p^

OS >>. i« (M

+j

f-H CO

«>»

©

■^ .-H -«t -^ ».

a

^ fc.t— f ©

IfJ »>»

'■t

© CO 0 (N M

<u

M

©6 ' " '-^ ©

©

©

© cb © ©©

u

1

'

>H

JJ

l3

.© o . . . ^ CO

>o

CO

10 COffI © © S

Ph

2i»©

• © CO

©

«>.

10 © 00 'Jt" CO CO

^Ir?

• © 00

CO

CO W <M CO W5 »

© CO i« <N eo<N

:^ to

6JJ

I— (

■ CO ■ ■ * ■

*b».

t>

«5

;§

- 2C0 t»

■>*< ©

(M

in

t>. ©00

S S^ ©

00 t^

«o

CO

CO © CO ©

,0

fe® ®

»>. CO

CO

«>.

© © -* ©

C+H

«

©© © ©

©

©

©<N ©6

a>

■ © © . . .9* <N

»>.

-^

CO ift WS 1(5 C

~ -5

SCO S<l

«>» tx

«>.

■^

10 CD «>. ITS CO •-

0

_; Sco^

l>» t^

«5

(M

COO CO © Tt< CC

42

gj©©

t* CO

CO

«>»

© 0.© © »7i C

■ c«s ■ ■ © ■

a

0

• 2^ 05 «0 • «

CO !>.

03

a S-«i< -H -^ : -^

<^^

CO

CO —< »>. 9»

fe© o <» • <N

-<!}<

IQ

10 (N -* ©

0

i

©©© ©

©

©

©<N© 6

y

P4

.i« ifl »rt © o .

«o

t>.

©©-*©© c

«

2© 5q QO 00 W5 •

-^

©

© 10 o «>. © »

>H

M S -rf —I >-H © CO : :

©

©

ITS 0 -«t CO © (>

' ^

gj© ^ CO 05 »? ; ;

ep

«P

© .-i ■<* © l-H C

1^

' ©1 ' ■ © ■

_ s

X

:

»_

s

CJ

S

3
O

'3

•c

0
a,

0
-a

4)

c

fcO

a

c
c

m4
4)

B

05 "^

CJ

ii

•^

rt

0

Sh

c^

^

s a

.S; -tJ

^

^

a

■3; 0

2-2

&
3

'S

(3

.2

^

4^

Sao

a

0

8,
a
•St

c
1

>

acid ....
arid chare
[ chloridei
iio-chloric
de of sod
1 sulphate
uric acid

is

1

osphate 0
nate of lii
hosphate
ide of silv
nic acid .

Silicic
Sand;
Mixed
Potass
Chlori
Mixed
Sulph

.a

a.T:

X 0 a, c 0 p

3
3

2"i

o,^ 0 0 ,g c

03

Ph

fL,OPHUU<

•— ^

the Ashes of the Orange-Tree. 275

Analysis of the Ashes of the Root. Composition directly
found.

Amount of ashes left by 100 parts of the root . . 4*48

I. II. Mean.

Potash .... 12-54 12*40 12*47

Soda 3-72 3-57 3*64

Lime 40' 16 40-31 40-23

Magnesia . . . 5*55 5-60 5-57

Sesquioxide of iron 0*83 0-82 0-83

Chloride of sodium 1*01 0-91 0*95

Phosphoric acid . 10*80 10-93 10-86

Sulphuric acid . . 4*61 4-76 4*68

Silicic acid ... 1*38 1*45 1*42

Carbonic acid . . 19-04 19-04 19-04

Sand and charcoal 0-42 0-63 0*53

100-06 100-37 100-22

Per-centage of the ash after deduction of the unessential

constituents, carbonic acid, sand and charcoal : —

Potash . . . . 15-43

Soda 4-52

Lime 49-89

Magnesia. . . . 6-91

Sesquioxide of iron 1-02

Chloride of sodium 1*18

Phosphoric acid . 13*47

Sulphuric acid . . 5*78

Silicic acid . . . 1*75
100*00

Analysis of the Ashes of the Stem.
Amount of ashes left by 100 parts of the stem . . 2*74

I.

II.

Mean.

Potash ....

9-66

9-73

9-69

Soda

2*61

2*47

2-54

Lime

45-46

45*96

45-71

Magnesia . . .

5*28

5-24

5-26

Sesquioxide of iron

0*48

0*48

0-48

Chloride of sodium

0-19

0-24

0-21

Phosphoric acid .

14*18

14-17

14-17

Sulphuric acid . .

3*90

3-79

3-84

Silicic acid . . .

0*92

1-14

103

Carbonic acid . .

16-51

16-50

16-50

Sand and charcoal

0-33

0-21 "

■ 0-27

99-52

99-93

99-70

T2

276 Messrs. Rowney and How*s Analysis of

Potash . . . . 11-69

Soda . . . . 3-07

Lime 55*13

Magnesia. . . . 6-34

Sesquioxide of iron 0-57

Chloride of sodium 0-25

Phosphoric acid . 17*09

Sulphuric acid . . 4-64

Silicic acid . . . ]-22
10000

Analysis of the Ashes of the Leaves.
Amount of ashes left by 100 parts of the leaves . . 13-73

I.

II.

Mean.

Potash ....

12-87

12-48

12-67

Soda

1-22

1-38

1-30

Lime

43-32

43-44

43-38

Magnesia . . .

4-49

4-30

4-39

Sesquioxide of iron

0-3G

0-44

0-40

Chloride of sodium

5-08

5-17

5-12

Phosphoric acid .

2-46

2-58

2-52

Sulphuric acid . .

3-35

3-47

3-41

Silicic acid . . .

3-67

3-78

3-72

Carbonic acid . .

23-22

22-97

23-09

Sand and charcoal

0-24

0-21

0-23

100-28 100-22 100-23

Potash .... 16-51

Soda 1-68

Lime 56-38

Magnesia. . . .

5-72

Sesquioxide of iron

0-52

Chloride of sodium

6-66

Phosphoric acid

3-27

Sulphuric acid . .

4-43

Silicic acid . . .

4-83

100-00

the AsJiea of the Orange-Tree. 277

Analysis of the Ashes of the Fruit.
Amount of ashes left by 100 parts of the fruit . . 3-94

I.

II,

Mean.

Potash ....

28-21

28-32

28-26

Soda

8-73

8-99

8-86

Lime

19-20

18-83

19-02

Magnesia . . .

6-39

6-14

6-26

Sesquioxide of iron

0-35

0-36

0-35

Chloride of sodium

2-93

309

302

Phosphoric acid .

8-55

8-64

8-59

Sulphuric acid . .

2-88

2-93

2-90

Silicic acid . . .

0-31

0-38

0-34

Carbonic acid . .

20-38

20-22

20-30

Sand and charcoal

1-69

1-62

1-65

99-62

99-52

99-55

Potash

36-42

Soda .

11-42
24-52

Lime .

.

Magnesia

• • . .

8-06

Sesquioxide of iron

0-46

Chloride of sodium

3-87

Phosphoric acid

11-07

Sulphuric

: acid . .

3-74

Silicic acid . . .

0-44

10000

Analysis of the Ashes of the Seed.
Amount of ashes left by 100 parts of the seed . . 3*30

Potash . , .
Soda *. . . .

Lime 16-59

Magnesia . . .
Sesquioxide of iron
Chloride of sodium
Phosphoric acid .
Sulphuric acid . .
Silicic acid . , .
Carbonic acid . .
Sand and charcoal

100-30 10002 100^

I.

II.

Mean.

35-22

35-29

35-26

0-77

0-84

0-81

16-59

16-65

16-62

7-87

7-51

7-69

0-68

0-72

0-70

0-77

0-67

0-72

20-33

20-39

20-36

4-46

4-48

4-47

1-02

0-96

0-99

6-83

6-83 ,

6-83

5-76

5-78

5-77

278 Sir W. Rowan Hamilton on Qjiaternions,

Potash .... 40-28

Soda 0-92

Lime 18-97

Magnesia . . . 8*74

Sesquioxide of iron 0'80

Chloride of sodium 0-82

Phosphoric acid . 23-24

Sulphuric acid . . 5-10

Silicic acid . . . 1-13—100-00
The preceding analyses furnish a new confirmation of the
fact first observed by De Saussure, namely, that the largest
amount of mineral constituents is deposited in those parts of
the plant in which the process of assimilation appears to be
most active. While the ash left by the root, stem, fruit and
seed did not exceed from 3 to 4 per cent., the leaves left
not less than 13 per cent, of fixed residue on incineration.

Regarding the composition of the different ashes, the great
amount of carbonic acid found in the ashes of the root, the
stem, and the fruit is at once obvious ; proving that not only
the fruit, but also the roots and stem, contain a large quan-
tity of organic acids.

From the composition of the ashes of the root, the stem,
and the leaves, the orange-tree belongs decidedly to the lime
plants. In these three ashes the joint amount of lime and
magnesia exceeds the quantity of the rest of the mineral con-
stituents. In the ashes of the fruit and seed, however, the
alkalies are as prevalent as they have been found in analogous
cases. The amount of phosphoric acid (23-24) in the ash of
the seed is considerable, as might be expected, still it is in-
ferior to the quantities (34'81 and 42-02) which Mr. Souchay
found on analysing the seeds of the citron [Citrus Medico)
and quince-trees [Pyrus Cydonia). Nevertheless the ash of
the orange-seed is very analogous in composition to the ashes
of the last-mentioned seeds, as may be easily seen on com-
paring their analysis *.

XLVI. On Qjiaternions ; or on a New System of Imaginaries
in Algebra. By Sir William Rowan Hamilton, LL.D.,
V. P.R.I, A. i F.R.A.S., Corresponding Member of the Insti-
tide of France,) Sfc., Andrews^ Professor of Astronomy in the
University of Dublin j aiid Royal Astronomer of Ireland,
[Continued from p. 219.]
37. 1)ESUMING now the quaternion form of the equa-
-*■*' tion of the ellipsoid,

(«p + H'-(/3f-p/3)^ = l> (1.)

• Liebig's Annals, liv. p. 343.

Sir W. Rowan Hamilton on Quaternions, 279

and making

'^+^=jz:;^9? «-/3=;2Z^> • • • (2-)

and

!t±£^,=Q, f^=Q', .... (8.)

the two linear factors of" the first member of the equation (1.)
become the two conjugate quaternions Q and Q', so that the
equation itself becomes

QQ'=1 (4.)

But by articles 19 and 20 (Phil. Mag. for July 1846), the
product of any two conjugate quaternions is equal to the square
of their common tensor; this common tensor of the two qua-
ternions Q and Q' is therefore equal to unity. Using, there-
fore, as in those articles, the letter T as the characteristic of
the operation o^ taking the teiisor of a quaternion, the fequaiion
of the ellipsoid reduces itself to the form

TQ = 1; (5.)

or, substituting for Q its expression (3.),

(6.)

which latter form might also have been obtained, by the sub-
stitutions (2.), from the equation (3.) of the 30th article (Phil.
Mag., June 1847), namely from the following*:

T(«p + pa + /3p-p/3)=l (7.)

38. In the geometrical construction or generation of the
ellipsoid, which was assigned in the preceding articles of this
paper (see the Numbers of the Philosophical Magazine for
June and September 1847), the significations of some of the
recent symbols are the following. The two constant vectors
« and X may be regarded as denoting, respectively, (in lengths
and in directions,) the two sides of the generating triangle
ABC, which are drawn from the centre c of the auxiliary and
diacentric sphere, to the fixed superficial point b of the ellip-
soid, and to the centre a of the same ellipsoid ; the third side
of the triangle, or the vector from a to b, being therefore de-
noted (in length and in direction) by » — x: while p is the
radius vector of the ellipsoid, drawn from the centre a to a

• See equation (35.) of the Abstract in the Proceedings of the Royal
Irish Academy for July 1846. The equation of the ellipsoid marked (1.)
in article 37 of the present paper, was communicated to the Academy in
December 1845, and is numbered (31.) in the Proceedings of that date.

280 Sir W. Rowan Hamilton on Quaternions.

variable point e of the surface ; so that the constant vector
I — X is, by the construction, a particular value of this variable
vector p. The vector from a to c, beiufr the opposite of that
from c to A, is denoted by — x; and if d be still the same
auxiliary point on the surface of the auxiliary sphere, which
was denoted by the same letter in the account already printed
of the construction, then the vector from c to d, which may
be regarded as being (in a sense to be hereafter more fully
considered) the reflexion of — x with i-espect to ^, is = —pxp-'^;
and consequently the vector from d to b is =j-j-pxp~^ The
lengths of the two straight lines bd, and ae, are therefore re-
spectively denoted by the two tensors, T(i + ^x^-') and Tp;
and the rectangle under those two lines is represented by the
product of these two tensors, that is by the tensor of the
product, or by T(»p4-px). But by the fundamental equality
of the lengths of the diagonals, ae, bd', of the plane quadri-
lateral ajbed' in the construction, this rectangle under bd and
ae is equal to the constant rectangle under bd and bd', that
is under the whole secant and its external part, or to the
square on the tangent from b, if the point b be supposed ex-
ternal to the auxiliary sphere, which has its centre at c, and
passes through d, d', and a. Thus T{ip + pK) is equal to
(Ti)^— (Tx)% or to x.^ — 1% which difference is here a positive
scalar, because it is supposed that cb is longer than ca, or that

Ti>Tx; (8.)

and the quaternion equation (6.) of the ellipsoid reproduces
itself, as a result of the geometrical construction, under the
slightly simplified form*

T(i^ + px) = x2-.2 (9.)

And to verify that this equation relative to p is satisfied (as
we have seen that it ought to be) by the particular value

P = '-x, (10.)

which corresponds to the particular position b of the variable
point E on the surface of the ellipsoid, we have only to observe
that, identically,

<(» — 1) + (< — x)x = »^ — »X + JX— x^

and that (by article 19) the tensor of a negative scalar is equal
to the positive opposite thereof.

39. The foregoing article contains a sufficiently simple

* See the Proceedings of the Royal Irish Academy for July 1846, equa-
tion (44.).

Sir W. Rowan Hamilton 07i Qiiaternions. 281

process for the retraiidation of the geometrical construction *
of the ellipsoid described in article 31, into the language of
the calculus of quaternions, from which the construction itself
had been originally derived, in the manner stated in the 30th
article of this paper. Yet it may not seem obvious to readers
unfamiliar with this calculus, why the expression — p>tp~* was
taken, in that foregoing article 38, as one denoting, in length
and in direction, that radius o/the auxiliary sphere which was
drawn from c to d; nor in what sense, and for what reason,
this expression —p)cp~^ has been said to represent the reflexion
of the vector — X with respect to p. As a perfectly clear answer
to each of these questions, or a distinct justification of each of
the assumptions or assertions thus referred to, may not only
be useful in connection with the present mode of considering
the ellipsoid, but also may throw light on other applications
of quaternions to the treatment of geometrical and physical
problems, we shall not think it an irrelevant digression to enter
here into some details respecting this expression— pxp~^, and
respecting the ways in which it may present itself in calcula-
tions such as the foregoing. Let us therefore now denote by
0- the vector, whatever it may be, from c to d in the construc-
tion (c being still the centre of the sphere) ; and let us pro-
pose to find an expression for this sought vector a; as a func-
tion of p and of X, by the principles of the calculus of quater-
nions.

40. For this purpose we have first the equation between
tensors,

Tcr=Tx; (11.)

which expresses that the two vectors o- and x are equally long,
as being both radii of one common auxiliary sphere, namely
those drawn from the centre c to the points d and a. And
secondly, we have the equation

V.(o- — x)/5 = 0, (12.)

where V is the characteristic of the operation of taking the
vector of a quaternion ; which equation expresses immediately
that the product of the two vectors cr — x and p is scalar, and

* The brevity and novelty of this rule for constructing that important
surface may perhaps justify the reprinting it here. It was as follows :
From a fixed point a on the surface of a sphere, draw a variable chord au;
let d' be the second point of intersection of the spheric surface with the
secant bd, drawn to the variable extremity d of this chord ad from a fixed
external point b; take the radius vector ae equal in length to bd', and in
direction either coincident with, or opposite to, the chord ad; the locus
of the point e, thus constructed, will be an ellipsoid, which will pass through
the point b (and will have its centre at a). See Proceedings of the Royal
Irish Academy for July 1846.

382 Sir W. Rowan Hamilton 07i Quaternions.

therefore that these two vector-factors are either exactly
similar or exactly opposite in direction ; since otherwise their
product would be a quaternion, having always a vector part,
although the scalar part of this quaternion-product (o-— x)p
might vanish, namely by the factors becoming perpendicular
to each other. Such being the immediate and general signi-
fication of the equation (12.), the justification of our establish-
ing it in the present question is derived from the consideration
that the radius vector p, drawn from the centre a to the sur-
face E of the ellipsoid, has, by the construction, a direction
either exactly similar or exactly opposite to the direction of
that guide-chord of the auxiliary sphere which is drawn from
A to D, that is, from the end of the radius denoted by x to the
end of the radius denoted by cr. For, that the chord so drawn
is properly denoted, in length and in direction, by the symbol
(T — X, follows from principles respecting addition and subtrac'
tion of directed lines, which are indeed essential, but are not
peculiar, to the geometrical applications of quaternions; had
occurred, in various ways, to several independent inquirers,
before quaternions {as 2)roducts or qnotieiits of' directed lilies
in space) were thought of; and are now extensively received.
41. The two equations (11.) and (12.) are evidently both
satisfied when we suppose (r = x; but because the point d is
in general different from a, we must endeavour to find another
value of the vector o-, distinct from x, which shall satisfy the
same two equations. Such a value, or expression, for this
sought vector <t may be found at once, so far as the equation
(12.) is concerned, by observing that, in virtue of this latter
equation, o- — x must bear some scalar ratio to p, or must be
equal to this vector p multiplied by some scalar coefficient x,
so that we may write

(rz=zx. + xp; (13.)

and then, on substituting this expression for cr in the former
equation (11.), we find that x must satisfy the condition

T(x+^p) = Tx, (14.)

in which this sought coefficient x is supposed to be some
scalar different from zero, that is, in other words, some posi-
tive or negative number. Squaring both members of this last
condition, and observing that by article 19 the square of the
tensor of a vector is equal to the negative of the square of that
vector, we find the new equation

-(x + jrp)2=-x2 (15.)

But also, generally, if x and p be vectors and x a scalar,

(x + a;p)^ = x^ + ir(xp -f px) -f- a;'^^^ ;

Sir W. Rowan Hamilton on Qtiaternions. 283

adding therefore x^ to both members of (15.), dividing by — .r,
and then eUminating x by (13.), which is done by merely
changing xp + xp'^ to a-p, we find the equation

<xp + pK = 0; (16.)

and finally

(T=- —pKp~'^'. (17.)

SO that the expression already assigned for the vector from c
to D, presents itself as the result of this analysis. And in fact
the tensor of this expression (17.) is equal to Tx, by the ge-
neral rule for the tensor of a product, or because (— pxp~')^
=pxp~Yxp~i=px^p"' = x'^, since x^ is a (negative) scalar; while
the product ((r — x)p, being = — (xp + px), is equal, by article
20, to an expression of scalar form.

42. Conversely if, in any investigation conducted on the
present principles, we meet with the expression —pxp~^, we
may perceive in the way just now mentioned, that it denotes
a vector of which the square is equal to that of x ; and that, if
X be subtracted from it, the remainder gives a scalar product
when it is multiplied into p : so that, if we denote this expres-
sion by cr, or establish the equation (17.), the equations (11.)
and (12.) will then be satisfied, and the vector o- will have the
same length as x, while the directions of <r — x and p will be
either exactly similar or exactly opposite to each other. We
may therefore be thus led to regard, subject to this condition
(17.) or (16.), the two vector-symbols o- and x as denoting, in
length and in direction, two radii of one common sphere, such
that the chord-line cr— x connecting their extremities has the
direction of the line p, or of that line reversed. Hence also,
by the elementary property of a plane isosceles triangle, we
may see that, under the same condition, the inclination of o-
to p is equal to the inclination of x to — p, or of — x to p ; in
such a manner that the bisector of the external vertical angle
of the isosceles triangle, or the bisector of the angle at the
centre of the sphere between the two radii o- and — x, is a new
radius parallel to p, because it is parallel to the base of the
triangle (acd), or to the chord (ad) just now mentioned.
And by conceiving a diameter of the sphere parallel to this
chord, or to p, and supposing — x to denote that reversed
radius which coincides in situation with the radius x, but is
drawn from the surface to the centre (that is, in the recent
construction, from a to c), while <r is still drawn from centre
to surface (from c to d), we may be led to regard <r, or — pxp~\
as the reflexion of — x with respect to the diameter parallel to
p, or simply with respect to p itseUi as was remarked in the
38lh article ; since the vector-symbols p, cr, &c. are supposed,

S84 Sir W. Rowan Hamilton 07i Quaternions,

in these calculations, to indicate indeed the lengths atid direc-
tions, but not the situatio?iSf of the straight lines which they
are employed to denote.

43. The same geometrical interpretation of the symbol
—pxp~^ may be obtained in several other ways, among which
we shall specify the following. Whatever the lengths and
directions of the two straight lines denoted hyp and x may be,
we may always conceive that the latter line, regarded as a
vector, is or may be decomposed, by two different projections,
into two partial or component vectors, x' and x", of which one
is parallel and the other is perpendicular to p ; so that they
satisfy respectively the equations of parallelism and perpen-
dicularity (see article 21), and that we have consequently,

x = x'-fx"; Y.x'p=0; S.x"p = 0; . . (18.)

where S is the characteristic of the operation of taking the
scalar of a quaternion. The equation of parallelism gives
px.' = x'p, and the equation of perpendicularity gives px"= —x"p ;
hence the proposed expression --pxp-^ resolves itself into the
two parts,

-px'p-^=:-x'/5p-»=-x'; \ ^^.

^pK"p-'=+x"pp-'=-}-x";J • • • ^ •-'
so that we have, upon the whole,

-pXp-'=-p{K' + K")p-'-->J-^K". . . (20.)

The part — x' of this last expression, which is parallel to p, is
the same as the corresponding part of — x; but the part +t",
perpendicular to p, is the same with the corresponding part
of + X, or is opposite to the corresponding part of — x ; we
may therefore be led by this process also to regard the expres-
sion (17.) as denoting the reflexion of the vector — x, with
respect to the vector p, legarded as a reflecting line ; and we
see that the direction of p, or that of —p, is exactly interme-
diate between the two directions of — x and —pxp~^f or be-
tween those of X and of pxp~'.

4-4-. The equation (9.) of the ellipsoid, in article 38, or the
equation (4.) in article 37, may be more fully written thus:

(,p+px)(p. + xp) = (x2-.T. . . . . (21.)

And to express that we propose to cut this surface by any
diametral plane, we may write the equation

OTp+pOT=0, (22.)

where ■bt denotes a vector to which that cutting plane is
perpendicular: thus, if in particular, we change ot to x, we
find, for the corresponding plane through the centre, the equa-
tion

Sir W. Rowan Hamilton on Qiiaterniotis. 285

xp + pK = Oy (23.)

which, when combined with (21.), gives

(x2 - «2)2 ={i-K)p.p{i-x)- (» - x)p2(, _ x) = (. - x) V,
that is,

/«2 _,2\2

(24.)

■=(^7

but this is the equation of a sphere concentric with the ellip-
soid ; therefore the diametral plane (23.) cuts the ellipsoid in
a circle^ or the plane itself is a ajcHc plane. We see also that
the vector x, as being perpendicular to this plane (23.), is one
of the cj/clic normals, or normals to planes of circular section;
which agrees with the construction, since we saw, in article 36,
that the auxiliary or diacentric sphere, with centre c, touches
one cyclic plane at the centre a of the ellipsoid. The same
construction shows that the other cyclic plane ought to be
perpendicular to the vector i ; and accordingly the equation

ip + pi = Q (25.)

represents this second cyclic plane; for, when combined with
the equation (21.) of the ellipsoid, it gives

and therefore conducts to the same equation (24.) of a con-
centric sphere as before; which sphere (24.) is thus seen to
contain the intersection of the ellipsoid (21.) with the plane
(25.), as well as that with the plane (23.). If we use the
form (9.), we have only to observe that whether we change
px to —xp, or ip to — f)», we are conducted in each case to the
following expression for the length of the radius vector of the
ellipsoid, which agrees with the equation (24.) :

Tf='i^) (^«-)

And because x^ — »^ denotes the square upon the tangent drawn
to the auxiliary sphere from the external point b, while
T(i — x) denotes the length of the side ba of the generating
triangle, we see by this easy calculation with quaternions, as
well as by the more purely geometrical reasoning which was
alluded to, and partly stated, in the 36th article, that the com-
mon radius of the two diametral and circular sections of the
ellipsoid is equal to the straight line which was there called
BG, and which had the direction of ba, while terminating, like
it, on the surface of the auxiliary sphere ; so that the two last
lines ba, and bg, were connected with that sphere and with
each other, in this or in the opposite order, as the whole se-

286 Sir W. Rowan Hamilton on Quaternions.

cant and the external part. In fact, as the point d, in the
construction approaches, in any direction, on tlie surface of the
auxiliary sphere, to a, the point d' approaches to g; and bd',
and therefore also ae, tends to become equal in length to Bci ;
while the direction of ae, being the same with that of ad, or
opposite thereto, tends to become tangential to the sphere, or
perpendicular to ac: the line bg is therefore equal to the
radius of that diametral and circular section of the ellipsoid
which is made by the plane that touches the auxiliary sphere
at A. And again, if we conceive the point d' to revolve on
the surface of the sphere from g to g again, in a plane per-
pendicular to Bc, then the lines ad and ae will revolve to-
gether in another plane parallel to that last mentioned, and
perpendicular likewise to bc ; while the length of ak will be
still equal to the same constant line bg as before : which line
is therefore found to be equal to the common radius of both
the diametral and circular sections of the ellipsoid, whether as
determined by the geometrical construction which the calculus
of quaternions suggested, or immediately by that calculus
itself.

^5. We may write the equation (21.) of the ellipsoid as
follows:

/(p) = l, ...... (27.)

if we introduce a scalar function^/of the variable vector p,
defined as follows :

[a^—i'^ffip) = {ip + px)(pi + xp) = ip^j -}- ipxp + pxpi + pK^p ;

or thus, in virtue of article 20,

(x2_,2^V(p) = (.2 4.x2)p2 + 2S..pxp. . . (28.)

Let p+T denote another vector from the centre to the sur-
face of the same ellipsoid ; we shall have, in like manner,

/(F + r) = l, (29.)

where

/(f-fr)=/(/») + 2S.VT+/(r), . . . (30.)

if we introduce a new vector symbol v, defined by the equation

{K^^i^fv={i^ + x^)p + ipK + xpi; . . . (31,)

because generally, for any two vectors p and t,

(p-f-T)' = p' + 2S.pr + T2, .... (32.)

and, for any four vectors, i, x, p, t,

S.iTXp = S.TXp»=S .xp»T=S .p<Tx; . . (33.)

which last principle, respecting certain transpositions of vector
symbols, as factors of a product under the sign S., shows,

Sir W. Rowan Hamilton on Qiiaternions. 287

when combined with the equations (27.)} (28.)> and (31.),
that we have also this simple relation :

S.vp = l (34.)

Subtracting (27.) from (29.), attending to (30.), changing t
to Tt. Ut, where U is, as in article 19, the characteristic of
the operation of talcing the verso?- of a quaternion (or of a
vector), and dividing by Tt, we find :

• 0 = -^:^^^:^^=2S.vUr + TT./(Ur). {35.)

This is a rigorous equation, connecting the length or the
tensor Tr, of any chord r of the ellipsoid, drawn from the
extremity of the semidiameter p, with the direction of that
chord T, or with the versor Ur ; it is therefore only a new
form of the equation of the ellipsoid itself, with the origin of
vectors removed from the centre to a point upon the surface.
If we now conceive the chord t to diminish in length, the
term TT./(Ur) of the right-hand member of this equation
(35.) tends to become =0, on account of the factor Tt; and
therefore the other term 2S . vUt of the same member must
tend to the same limit zero. In this way we arrive easily at
an equation expressing the ultimate law of the directions of the
evanescent chords of the ellipsoid, at the extremity of any given
or assumed semidiameter p ; which equation isO = 2S.vUT,
or simply,

0=S.VT, (36.)

if T be a tangential vector. The vector v is therefore perpen-
dicular to all such tangents, or infinitesimal chords of the
ellipsoid, at the extremity of the semidiameter p ; and conse-
quently it has the directiim of the normal to that surface, at
the extremity of that semidiameter. The tangent plane to the
same surface at the same point is represented by the equation
(34.), if we treat, therein, the normal vector v as constant, and
if we regard the symbol p as denoting, in the same equation
(34.), a variable vector, drawn from the centre of the ellipsoid
to any point upon that tangent plane. This equation (34.)
of the tangent plane may be written as follows:

S.v(p-v-')=0; (37.)

and under this form it shows easily that the symbol v"' repre-
sents, in length and in direction, the perpendicular let fall
from the origin of the vectors p, that is from the centre of the
ellipsoid, upon the plane which is thus represented by the
equation (34.) or (37.); so that the vector v itself, as deter-
mined by the equation (31.), may be called ihe vector ofproxi-

288 Sir W. Rowan Hamilton on Quaternions.

mity^ of the tangent plane of the ellipsoid, or of an element of
that surface, to the centre, at the end of that semidiameter p
from which v is deduced by that equation.

46. Conceive now that at the extremity of an infinitesimal
chord dp or t, we draw another normal to the ellipsoid ; the
expression for any arbitrary point on the former normal, that
is the symbol for the vector of this point, drawn from the
centre of the ellipsoid, or from the origin of the vectors p, is
of the form p + nv, where n is an arbitrary scalar ; and in like
manner the corresponding expression for an arbitrary point
on the latter and infinitely near normal, or for its vector from
.the same centre of the ellipsoid, is p + dp + (w + dw)(v + dv),
where d« is an arbitrary but infinitesimal scalar, and dv is the
differential of the vector of proximity v, which may be found
as a function of the differential dp by differentiating the equa-
tion (31.), which connects the two vectors v and p themselves.
In this manner we find, from (31.),

(x2-,2)2dv = (,2 + x2)dp + .dpx + xdp.; . . (38.)

and the condition required for the intersection of the two near
normals, or for the existence of a point common to both, is
expressed by the formula

p + dp + (w + d«)(v + dy) = p + wv; . . . (89.)

which may be more concisely written as follows :

dp + d.nv=0; (40.)

or thus :

dp + ndy + dnv = 0 (41.)

We can eliminate the two scalar coefficients, n and An, from
this last equation, according to the rules of the calculus of
quaternions, by the method exemplified in the 24th article of
this paper (Phil. Mag., August 1846), or by operating with
the characteristic S . vdv, because generally

S.yft2=o, S.vjttv = 0,

whatever vectors ft, and v may be ; so that here,

S . vdvwdv = 0, S . vdvdwv = 0.

• This name, "vector of proximity," was suggested to the writer by a
phraseology of Sir John Herschel's; and the equation (31.), of article 4.5,
which determines this vector for the ellipsoid, was one of a few equations
which were designed to have been exhibited to the British Association at
its meeting in 1846: but were accidentally forwarded at the last moment
to CoUingwood, instead of Southampton, and did not come to the hands
of the eminent philosopher just mentioned, until it was too late for him to
do more than return the paper, with some of those encouraging expressions
by which he delights to cheer, as opportunities present themselves, all per-
sons whom he conceives to be labouring usefully for science.

Sir W. Rowan Hamilton on Quaternions, 289

In this manner we find from (4<l.) the following very simple

formula: S.vdvd/3=0; (42.)

which is easily seen, on the same principles, to hold good, as
the quaternion form of the differential equation of the lines of
curvature on a curved surface generally ^ if v be still the vector
qfproximity of the siiperjicial dement of the curved surface to
the origin of the vectors p, which vector v is determined by

the general condition S.vdo = 0, (43.)

combined with the equation already written,

S.vp=l (S*.);
or simply if v be a normal vector, satisfying the condition (43.)
alone. Substituting, therefore, in the case of the ellipsoid,
the expression for dv given by (38.), and observing that
S . vdp^ = 0, we find that we may write the equation of the lines
of curvature for this particular surface as follows :

S.v(»d/5x+xdp«)dp = 0; .... (44.)

which equation, when treated by the rules of the present cal-
culus, admits of being in many ways symbolically transformed,
and may also, with little difficulty, be translated into geome-
trical enunciations.

47. Thus if we observe that, by article 20, <tx — xt< is a
scaiar form, whatever three vectors may be denoted by (, x, t;
and if we attend to the equation (43.), which expresses that
the normal v is perpendicular to the linear element, or infini-
tesimal chord, dp; we shall perceive that, for every direction
of that element, the following equation holds good:

S.v(»dpx— xdpj)dp = 0 (45.)

We have therefore, from (44.), for those 'particular directions
which belong to the lines of curvature, this simplified equation;

S.v»dpxdp = 0; (46.)

which may be still a little abridged, by writing instead of dp
the symbol t of a tangential vector, already used in (36.) ; for
thus we obtain the formula :

S.V.TXT = 0 . (47.)

We might also have observed that by the same article 20
(Phil. Mag., July 1846), jtx + jctj and therefore <dpx + xdpj is
a vector form, and that by article 26 (Phil. Mag., August
1846), three vector- factors under the characteristic S may be
in any manner transposed, with only a change (at most) in the
positive or negative sign of the resulting scalar ; from which
it would have followed, by a process exactly similar to the
foregoing, that the equation (44.) of the lines of curvature on
an ellipsoid may be thus written,

S.vdp.dpx = 0; (48.)

Phil, Mag. S. 3. Vol. 31. No. 208. Oct. 1847. U

290 SirW. Rowan Hamilton on Quaternions.

or, substituting for the linear element dp the tangential vector t,

S.vT«Tx = 0; (49.)

or finally, by the principles of the same 20th article,

VmK — XT<TV=0. (^O*)

48. Under this last form, it was one of a few equations
selected in September 1846, for the purpose of being exhibited
to the Mathematical Section of the British Association at
Southampton ; although it happened* that the paper con-
taining those equations did not reach its destination in time to
be so exhibited. The equations here marked (49.) and (50.)
were however published before the close of the year in which
that meeting was held, as part of the abstract of a communi-
cation which had been made to the Royal Irish Academy in
the summer of that year. (See the Proceedings of the Academy
for July 184G, equations (46.) and (47.).) From the some-
what discursive character of the present series of communica-
tions on Quaternions, and from the desire which the author
feels to render them, to some extent, complete within them-
selves, or at least intelligible to those mathematical readers
of the Philosophical Magazine who may be disposed to favour
him with their attention, to the degree which the novelty of
the conceptions and method may require, without its being
necessary for such readers to refer to other publications of his
own, he is induced, and believes himself to be authorized, to
copy here a few other equations from that short and hitherto
unpublished Southampton paper, and to annex to them an-
other formula which may be found in the Proceedings, already
cited, of the Royal Irish Academy : together with a more ex-
tensive formula, which he believes to be new.

49. Besides the equation of the ellipsoid,

(/p + px)(p.-f?cp) = (x^-i2)2(21.), art.44;
with the expression derived from it, for the vector of proxi-
mity of that surface to its centre,

(x2-,2)2y=(,2 + x2)^ + ,px-f xp. (31.), art. 45;
the equation for the lines of curvature on the ellipsoid,
VTiTx— xT<Tv=0 (50.), art. 47 ;

and the equation VT-fTv = 0, (51.)

which is a form of the relation S.vr = 0, that is of the equa-
tion (36.), article 45, of the present series of communications;
the author gave, in the paper which has been above referred
to, the following symbolic transformation, for the well-known
characteristic of operation.

(d^) + \i\y) "^ (d J '

* See the note to article 45.

Sir W. Rowan Hamilton on Quaternions. 291

which seems to him to open a wide and new field of analytical
research, connected with many important and difficult de-
partments of the mathematical study of nature.

A QUATERNION, Symbolically considered^ being (according
to the views originally proposed by the author in 184-3) an
algebraical quadrinomial of the form iso + ix+Jy + kz, where
wxi/z are any four real numbers (positive or negative or zero),
while ijk are three co-ordinate imaginary units, subject to the
fundamental laws of combination (see Phil. Mag. for July
1844-):

ij=k;Jk:=i; /ci=J; I . . . . (a.)

ji^—k; kj=—i\ ik——j;J

it results at once from these definitions, or laws of symbolic
combination, (a.), that if we introduce a new characteristic of
operation, < , defined with relation to these three symbols ijk,
and to the known operation of partial differentiation, performed
with respect to three independent but real variables xyz, as
follows :

id ;d ^d ,, .

'^ = T.+% + T.' (''•'

this new characteristic < mil have the negative of its symbolic
square expressed by the following formula :

of which it is clear that the applications to analytical physics
must be extensive in a high degree. In the paper* designed
for Southampton it was remarked, as an illustration, that this
result enables us to put the known thermological equation,

d^t> d?v d^u dv

db^+d^ + d^ + ^'dF"^'
under the new and more symbolic form,

(<=-r')-'" (^0

while < V denotes, in quantity and in direction, the flux of
heat, at the time t and at the point xy%.

50. In the Proceedings of the Royal Irish Academy for
July 1846, it will be found to have been noticed that the same
new characteristic < gives also this other general transforma-
tion, perhaps not less remarkable, nor having less extensive

* In that paper itself, the characteristic was written V ; bi't this more
common sign has been so often used with other meanings, that it seems desi-
rable to abstain from appropriating it to the new signification here proposed,

U2

292

Sir W. Rowan Hamilton on Quaternions.

consequences, and which presents itself under the form of a
quaternion :

, ^ /dt du i\v\

/dv du\ . (dt dv\ , (du
\d~v ~ Tz) +"^ Vd^^ ~ dI7 ■^'^ Vd^

+ 1

di\
d3//

(e.)

In fact the equations (a.) give generally (see art. 21 of the
present series),

( ix -{-jy + kz) {it +ju + kv)=:— {xt + yu-{- s:v)
+ i[yv—zu) +j{zt — xv) + 1i{xu —yt),

}•

(f.)

\{ xyztuv denote any six real numbers ; and the calculations by
which this is proved, show, still more generally, that the same
transformation must hold good, if each of the three symbols
iyj^ k, subject still to the equations (a.), be commutative in
arrangement, as a symbolic factor, with each of the three
other symbols .v, _?/, z ; even though the latter symbols, like
the former, should not be commutative in that way among
themselves ; and even if they should denote symbolical instead
of numerical multipliers, possessing still the distributive cha-
racter. We may therefore change the three symbols x,y, z,
respectively, to the three characteristics of partial differentia-
tion, -r-i -r-) -,- ; and thus the formula (e.) is seen to be in-
dx dy dz

eluded in the formula (f.). And if we then, in like manner,
change the three symbols /, ?/, v, regarded as factors, to

-T-i» -T-n -r-.i that is, to the characteristics of three partial dif-
dx' dy' d.s'

ferentiations performed with respect to three new and inde-
pendent variables x', y\ z\ we shall thereby change -p to

-T--r-;, and so obtain the formula:
dd^da?

^ d .d-d
^dx^^dy-^^'dz

)0-

d_ ._d_ _d^\

d^'"^'^d7/''^ d^7

\dxdotI ^

+

dy dj/'
dzdy

.(±± _i.A^

'^ \dxdy' dydx'J'

.{d d

'^'KdydF'

z)+J'

dy'

dzdz'J
./d_ d_
Vd^d^'

dx d^'

)

(g.)

On the Equation in Numbers A.r^+ B7/^+ C2^=D.27/2;. 293

which includes the formula (c), and is now for the first time
published.

This formula (g.) is, however, seen to be a very easy and
immediate consequence from the author's fundamental equa-
tions of IS-l-a, or from the relations (a.) of the foregoing article,
which admit of being concisely summed up in the following
continued equation:

t^=:f = k^ = ijk=-l (h.)

The geometrical interpretation of the equation S. vt«xt = 0 of
the lines of curvature on the ellipsoid, with some other appli-
cations of quaternions to that important surface, must be re-
served for future articles of the present series, of which some
will probably appear in an early number of this Magazine.
[To be continued.]

XLVII. On the Equation in Numbers Aa^ + By^ + Cz^^Dayz,
atid its associate si/stem of Equatio7is. By J. J. Sylvester,
Esq., M.A., F.R.S*

[Continued from p. 191.]

IN the last Number of this Magazine I gave an account of
a remarkable transformation to which the equation

is subject when certain conditions between the coefficients A,
B, C, D are satisfied ; which conditions I shall begin by ex-
pressing with more generality and precision than I was enabled
to do in my former communication.

1. Two of the quantities A, B, C are to be to one another
in the ratio of two cubes.

2. 27 ABC — D^ must contain no positive prime factor what-
ever of the form 6n + }. I erred in my former communication
in not excluding cubic factors of this form.

3. If 2"^ is the highest power of 2 which enters into ABC,
and 2" the highest power of 2 which enters into D, then either
m must be of the form 3«+ 1, or if not, then m must be greater
than 3w.

These three conditions being satisfied, the given equation
can always be transformed into another,

where A'u^ + WtP + Cxn^ = D'uvwy

A'B'C = ABC D' = D uvw = a. factor of z.

The consequence of this is, as stated in my former paper,
that wherever A, B, C, D, besides satisfying the conditions
above stated, are taken so as likewise to satisfy the condition, —
P, of ABC being equal to 23'»±', or 2°, of ABC being equal to
23m±ip3u±i^ provided in the first case that ABC is also of the

* Communicated by the Author.

294 Mr. J. J. Sylvester on the 'Equation in

form 9«+l, and in the second case ABC again of the same
form 9m ±1, but likewise D divisible by 9, j:? being in both
cases a prime, then the given equation will be generally inso-
luble. And I am now enabled to add that the only solution
of which it will in any case admit, is the solitary one found by
making two of the terms A.^■^, B?/^, Cz^ equal to one another ;
so that, for instance, if the given equation should be of the
form

^ +y + AB C . a^ = D^ys,

then the above conditions being satisfied, the one solitary so-
lution of which the equation can possibly admit, is J7= 1 ^=1,

Az^- 02 + 2 = 0,

which may or may not have possible roots. I call this a soli"
tary or singular solution, because it exists alone and no other
solution can be deduced from it ; whereas in general I shall
show that any one solution of the equation

A*^ + B?/3 + C^^=Da3/0

can be made to furnish an infinity of other solutions indepen-
dent of the one supposed given, /. e. not reducible thereto by
expelling a common factor from the new system of values of
x^y^ z deduced from the given system.

The following is the Theorem of Derivation in question :

Let

A«3 + B^3^C/=D«/3y.

Then if we write

F=A«3 G=B/33 H=Cy3,

and make

.r=F2G + G2H + H2F-3FGH
j/=FG2 + GH2 + HF2-3FGH

z=l{F3 + G3 + H3-3FGH},

or

= «/3y{F2 + G2+H2-FG-FH-GH},

we shall have

x^ -f- 2/3 -f. ABCs:^ = Yyxyz.

I am hence enabled to show that whenever x^ -^-y^ -\- kz^
= Djcyz is insoluble, there will be a whole family of allied
equations equally insoluble. For instance, because x^ + y^ + z^
= 0 is insoluble in integer numbers. I know likewise that

x^ +y^ + z^=x^y^ + x^z^ + y^^

a^ + y*^ + a« = .ry + a^^ — 2^/3^3
are each equally insoluble.

Numbers kx^ -\-'Qif -{■ Cz^ —T>xyz, 295

In fact

= t^ + x^ + li:^,

where u, v, w are rational integral functions ofx,i/, z.

Hence each of the factors must be incapable of becoming
zero*.

As a particular instance of my general theory of transfor-
mation and elevation, take the equation

x^ + y^ + 22^ = Mxi/z.
Then, with the exception of the singular or solitary solution
x=l .^=lj of which 1 take no account, I am able to affirm
that for all values of M between 7 and —6, both inclusive,
with the exception of M=— 2, the equation is insoluble in
integer numbers.

Take now the equation where M= — 2, viz.

x^+j/^ + 2z^+2xi/z = 0.
One particular solution of this is

x=l y— — \ z = l.
Another, which I shall call the second f, is

xszl j/=3 2=— 2.

From the first solution I can deduce in succession the follow-
ing:

j;=ll j/=5 z=— 7

a?=— 793269121 3/=11794900O 2;= — 1189735855
&c. &c. &c.

From the second,

^=—10085 j/ = 8921 2;=— 844-2
a? = &c. y=&c. z = hc.

As another example, take the equation
x^-\-Tl/^ + 6z^=6xyz.
One solution of the transformed equation
^3 ^ 2v^ + Str* = Qumso
is evidently

* It is however sufficiently evident from their intrinsic form, which may
be reduced to — (M^+SN^), that this impossibility exists for all the factors

except the first,
t See Postscript.

296 On the Equation in Numbers Aa;^ + By^+C2;*D=a^a.

Hence I can deduce an infinite series of solutions of the given
equation, of which the first in order of ascent will be

a;=5 y=1 z-=S.
Again, the lowest possible solution in integers of the equation

will be

A-=17 .?/=37 z—'-^i.
The equation

admits of the solutions

37=1 j/=2 z=- — 1

^'=—271 J/ = 919 2= -438.

I trust that my readers will do me the justice to believe that
I am in possession of a strict demonstration of all that has been
here advanced without proof. Certain of the writer's friends
on the continent have, in their comments upon one of his
former papers which appeared in this Magazine, complimented
his powers of divination at the expense of his judgement, in
rather gratuitously assuming that the author of the Theory of
Elimination was unprovided with the demonstrations, which he
was too inert or too beset with worldly cares and distractions
to present to the public in a sufficiently digested form. The
proof of whatever has been here advanced exists not merely
as a conception of the author's mind, but fairly drawn out in
writing, and in a form fit for publication.

P.S. It must not be supposed that the two primary or basic
solutions above given of the equation

a^ -]-if -\- 2r^ + 2x1/ z = 0,

viz. x = \ j/=— 1 z=l

x=l t/=3 2;= — 2

are independent of one another. The second may be derived
from the first, as 1 shall show in a future communication. In
fact there exist t/iree independent processes, by combining
which together, one particular solution may be made to give
rise to an infinite series of infinite series of infinite series
of correlated solutions, which it ma}' possibly be discovered
contain between them the general complete solution of the
equation

x^-\-if^-ir kz^=\)xyz, J. J. S.

26 Lincoln's Inn Fields,
Sept. 20, 1847.

[To be continued.]

[ 297 ]

XLVIII. Oil the Invention and First Introduction of Mr.
KcEnig's Printing Machine. By Richard Taylor, F.S.^.^c.

" As a step in the progress of civilization the Steam Press can only be
compared to the original discovery of Printing itself." — Times Newspaper,
j2cli/29, 1847, on the death of Mr. J. Walter.

MORE than a century after its introduction the first inven-
tion of the Art of Printing became a subject of long-con-
tinued controversy, remarkable for the insufficiency and fallacy
of the most confident assertions resting upon pretended tradi-
tions and unsupported conjectures. And, as Hadrian Junius
in 1575 first disputed the claims of Gutemberg after so long a
period had elapsed, so did Atkyns as late as 1664 first deny
the title of Caxton to the honour of having introduced the
art into our own country. Hence one of the writers in this
controversy remarks that " the Art of Printing, which has
given light to most other things, hides its own head in dark-
ness."

It will be our own fault if we allow any unfounded asser-
tions and pretensions to obtain currency with regard to an
improvement in the art, of which The Times newspaper has
said that "from the days of Faust and Gutemberg to the
present hour there has been only one great revolution in the
art of printing, and it occurred in the year 1814. Of that
revolution Mr. Walter was t lie prominent and leading agent.''

Now though I would on no account detract from the ge-
neral merits of the late Mr. Walter, as set forth in the Obi-
tuary and extended Memoir which appeared in The Times of
the 29th of July and 16th of September, yet I cannot allow
the representations which are made in these articles, as to
any share which he is alleged to have had in this important
invention, to pass without the most unqualified contradiction.

In the Obituary we read as follows : —

"But one achievement alone is sufficient to place Mr. Walter
high in that list which the world, as it grows older and wiser, will
more and more appreciate —

' Inventas aut qui vitam excohiere per artes,
Quique sui memores alios fecere merendc'

He first brought the steam-engine to the assistance of the public
press. Familiar as the discovery is now, there was a time when it
seemed fraught with difficulties as great as those which Fulton has
overcome on one element and Stephenson on another. To take off
5000 impressions in an hour was once as ridiculous a conception as
to paddle a ship fifteen miles against wind and tide, or to drag in
that time a train of carriages weighing 100 tons fifty miles. Mr.
Walter, who, without being a visionary, may be said to have thought
nothing impossible that was useful and good, was early resolved that
there should be no impossibility in printing by steam. It took a long
time in those days to strike oiF the 3000 or 4000 copies of The

298 Mr. R. Taylor on the Inveniion and First Introduction

Times. Mr. Walter could not brook the taedium of the manual
process. As early as the year 1804 an ingenious compositor, named
Thomas Martyn, had invented a self-acting machine for working the
press, and had produced a model which satisfied Mr. Walter of the
feasibility of the scheme. Being assisted by Mr. Walter with the
necessary funds, he made considerable progress towards the comple-
tion of his work."

" On the very eve of success he was doomed to bitter disappoint-
ment. He had exhausted his own funds in the attempt, and his
father, who had hitherto assisted him, became disheartened, and re-
fused him any further aid. The project was therefore for the time
abandoned." [Why abandoned, we may ask, if so feasible, and on
the very eve of success ?]

" Mr. Walter, however, was not the man to be deterred from what
he had once resolved to do. He gave his mind incessantly to the
subject, and courted aid from all quarters, with his usual munificence.
In the year 1814 he was induced by a clerical friend, in whose judge-
ment he confided, to make a fresh experiment ; and accordingly the
machinery of the amiable and ingenious Koenig, assisted by his young
friend Bauer, was introduced — not, indeed, at first, into The Times
office, but into the adjoining premises, such caution being thought
necessary from the threatened violence of the pressmen. Here the
work advanced, under the frequent inspection and advice of the
friend alluded to. At one period these two able mechanics sus-
pended their anxious toil, and left the premises in disgust. After
the lapse, however, of about three days, the same gentleman dis-
covered their retreat*, induced them to return, showed them to their
surprise their difficulty conquered, and the work still in progress."

Who would not infer from the above, that Mr. Walter,
having determined " to make a fresh experiment," in pur-
suance of those which he had long before abandoned (not-
withstanding his early resolution that there should be no im-
possibility in it), and " courting aid from all quarters with his
usual munificence," had been actually the person that enabled
Mr. Koenig to pursue his labours on Mr. Walter^s premises,
" under the inspection and advice of Mr. Walter's clerical
friend," and thus to produce his invention? Whereas, in
truth, Mr. Walter knew nothing of Mr. Koenig till after his
invention had been completed. He was merely the first
newspaper proprietor who purchased from the Patentees the
Printing Machines long before invented by Mr. Koenig. Of
these patentees I was one, and as I am now the sole survivor,
it devolves upon me to contradict any erroneous statements
and unfounded pretensions. I feel this to be the more ne-
cessary, as already the misstatements of The Times are cir-
culated, with additions and exaggerations, in other journals,

* To me this story appears not a little extraordinary : — the " discovery
of the retreat" of Messrs. K. and B. ! who were every day to be found su-
perintending our factory in Whitecross Street. — R. T.

of Mr. Koeuig's Printing Machine. 299

Thus, in an article in the Mechanics' Magazine for Sept. 18,
copied into the newspapers, I find the following passage : —

" No sooner were presses made of iron, than the idea occurred of
working them by steam ; and the first to welcome the new and happy
thought was the proprietor of a journal which stood in instant need
of some such powerful auxiliary to enable him to keep pace with a
circulation unexampled in the history of the press, and who, with-
out it, would most assuredly never have been able to attain to that
prodigious influence which for many years past has at once asto-
nished and awed the world. Koenig, the ingenious inventor of the
steam-press*, found in the proprietor of The Times his natural and
best possible patron. With the liberal aid of the late Mr. Walter,
he produced a machine of somewhat gigantic size, but nevertheless
possessing a completeness of design and purpose which cast all other
surface printing-presses into the shade."

And again —

" The steam-press has given occupation to many thousands, who,
but for its introduction, would have been standing idle, and who
ought, one and all, to bless the memory of Mr. Walter for enabling
the inventor to work out his ideas, and perfect his great and glorious
undertaking."

Now the whole of this is a fable. Mr. Walter was no
" natural and best possible patron " of Mr. Koenig's, — gave
him no " liberal aid in producing his machine," nor did any-
thing whatever to "enable him to work out his ideas." These
had all been worked out long before ; patents had been taken
out, a machine had been made, and was in operation on the
premises of the Patentees, before ever Mr. Walter, or any-
other newspaper proprietor, was applied to and invited to
adopt it. Mr. Perry of the Morning Chronicle declined,
alleging that he did not consider a newspaper worth so many
years' purchase as would equal the cost of machines. Mr.
Walter, "being a cautious man of the world," but enterprizing,
" it being," as his biographer says, " his habit in tiie game
of life never to throw away a chance," when he had fully sa-
tisfied himself by seeing that the invention was accomplished,
and in effective operation, consented to give an order for two
machines, for the cost of which he paid us a certain sum, and a
rental according to the number of copies printed ; and this rent
we received, until it was commuted for a sum agreed upon.

I do not mean to charge the writer in the Mechanics' Ma-
gazine with any intentional misrepresentation. He has evi-
dently been misled by the articles in The Times, which though
they do not directly assert all that he has inferred from them,
yet they imply as much. Thus a story gains in the telling,

* Mr. Kcenig's invention is very inappropiiately designated by the terms
" steam-press," and " the working of iron presses by steam." Its construc-
tion is wholly independent of the motive power employed.

SOO On the Invention of Mr. Koenig's Printing Machine.

till the most vague and unfounded suggestions, if uncontra-
dicted, are assumed as indisputable facts ; and it would be
recorded that if Koenig was the Gutemberg of the new dis-
covery, Walter was at least the Faust or Schoeffer of the
affair, or rather, both in one.

I am convinced that Mr. Walter, were he living, would
disclaim the pretensions that have been made in his name :
and indeed he has done so in the announcement which ap-
peared in The Times, Nov. 20, 1814, the day on which that
journal was first printed by the machines, and \yhich contains
the following passage : —

" That the completion of an invention of this kind, not the effect
of chance, but the result of mechanical combinations methodically
arranged in the mind of the artist, should he attended with many
obstructions and much delay may be readily admitted. Our share in
the event has indeed only been the application of the discovery, under
an agreement with the patentees, to our own particular business."

" The time for effecting the great revolution in the art of
printing," says Mr. Walter's biographer, " did not arrive till
the year 1814." Now it was in 1809 that, together with the
late Mr. George Woodfall, I joined Mr. Kcenig and Mr.
Bensley in taking out patents*, the machine being even then
so far advanced as to satisfy us as to the prospect of success,
and to enable us to have the specifications drawn up. Kcenig
had gone on with Bensley, to whom 1 had recommended him
some few years before, up to the year 1809, when the taking
of premises and the purchase of lathes, tools, &c., and the
employing of workmen, with the salaries of Mr. Koenig and
his able and excellent assistant Mr. Bauer, led Bensley to in-
vite us to a partnership in the undertaking. For several years
it occupied much of our time and attention, and cost us much
money (from which we had no return f ) and much anxiety.
Each experiment suggested some improvement, and one im-
provement led to others, so that additional patents had to be
taken out. But with Mr. Walter we had none of us any com-
munication, until, as I have before stated, the machine had
been completed and was at work on our own premises.

I have thought it right, under the circumstances, to put on

* One of the four patents bears date March 29, 1810 (See Phil, Mag.
vol. XXXV. 1st Series, p. 319). It was taken out in the name of Frederick
Koenig, and was assigned by articles of partnership to the firm of Bensley,
Koenig, Woodfall and Taylor.

•f Mr. Koenig left England, suddenly, in disgust at the treacherous con-
duct of Bensley, always shabby and overreaching, and whom he found to
be laying a scheme for defrauding bis partners in the patents of all the ad-
vantages to arise from them. Bensley, however, while he destroyed the
prospects of his partners, outwitted himself, and grasping at all, lost all,
becoming bankrupt in fortune as well as in character.

Cambridge Philosophical Society. SOI

record my own recollections as to the progress and introduc-
tion of this invention : and though they relate to transactions
which took place from thirty to forty years ago, I believe they
are in the main correct, and can be confirmed by documentary
evidence.

XLIX. Proceedings of Learned Societies,

CAMBRIDGE PHILOSOPHICAL SOCIETY.
[Continued from p. 143.]

ON the Partitions of Numbers, on Combinations, and on Permu-
tations. By Henry Warburton, M.P., F.R.S., F.G.S., Mem-
ber of the Senate of the University of London ; formerly of Trinity
College, A.M.

The use made by Waring of the Partitions of numbers in develo-
ping the power of a polynome, induced the author to seek for some
general and ready method of determining in how many different
ways a given number can be resolved into a given number of parts.
On his communicating the method described in article 5 of Section
L of this abstract, to Professor De Morgan, in the autumn of 1846,
that gentleman intimated a wish that the author would turn his
attention also to Combinations ; and such was the origin of the re-
searches which form the subject of the 2nd and 3rd sections.

I. Gn the Partitions of Numbers.

1. Let [N, p»j] denote how many different ways there are of re-
solving the integer N into ^J integral parts, none less than ij. Then

[N.p„] = [N±;,9,p„+^] (I.)

2. Such of the p-partitions of N as contain ij as a part, and no
part less than t], are obtained by resolving N— ij into p — 1 parts not
less than tj, and by adding ij, as a pXh part, to every such (p— 1)-
partition. That is,

[N,2.«]-[N,p]=[N-o,,i.-l]. . . . (IL)

3. In (IL), substitute ij + l, ij + 2, &c. successively for ij. The
sum of the results is

[N,^,]-[N,2?,+^+i]=Sj[N-,j-ri?,p-l]. . (III.)

Z tl

Inthisexpression, when 9=1* (—)—ij, the term [N,^„^^^.l]

vanishes, and the formula then becomes analogous to one published
anonymously by Professor De Morgan in a paper printed in the
fourth volume, p. 87, of the Cambridge Mathematical Journal.

4. In(IL), for [N,p„^i] substitute [N— pij.p,], and transpose

« I

f — jis employed to avoid the long phrase, "the integer nearest to

N
and not exceeding — ."

P

302 Cambridge Philosophical Society.

the terms. Then

[N,^„]-[N-,.j;-l] = CN-2)>j.p,]; . . (IV.)
*)

and this leads to

[N-i3,p-l]-[N-2ij,2?-2] = [N-pij,2j-l];

n ri vt

and that leads to the summation

[N.^„] = S^[N-;5,j.^.] (V.)

2

The lower limit of z in (V.) is made 0, in order that the formula
may comprehend the extreme case [0, Uj = 1, analogous to the ex-
treme case in Combinations.

5, After substituting 1 for ij, the author applies formula (IV.) to
determining in how many different ways N can be resolved into p
parts not less than 1. Let [N.^jJ be the term in a table of double
entry corresponding to column N, line p, in the table. From the
head, in line 0, of each of the columns 0, 1, 2, 3, &c., draw a diagonal,
advancing one column and one line at a time. Take these diagonals
one after another, and in each of them compute by formula (IV.) the
terms situate on lines 0, 1, 2, 3, &c., one by one in succession. If N
be the number at the head of the column from which any diagonal
takes its departure, there will be only N terms to compute on that
diagonal, the further terms being only repetitions of the term on the
line N. For the diagonal in question intersects line N in column
2N ; and, by formula V,

[2N,NJ = S^[N,r,]
z

= the sum of all the terms in column N. But, moreover,

[2N+y,Ni+y] = S?[N,r,]

z

as the same constant. The leading property of the table, indicated
by the formula

[N,p,]=S^[N-p,^.],
z
is, that the term [N, Pi] = the sum of all the terms in column N— j?,
from line 0 to line p inclusive. After the publication of the anony-
mous paper before referred to, Professor De Morgan discovered this
theorem also, but he did not announce it*.

II. On Combinations.

1. In ordinary Combinations, the combining elements are of differ-
ent kinds, and there is but one element of a kind : in the case here
considered, there are different kinds of elements, and there may be
many elements of a kind ; and more than one element of a kind may
enter into the same combination.

2. If II elements enter at a time into each combination, and the

* The author has recently discovered an equivalent formula in p. 264
of Euler's Int. in An. Infinitoruni ; but investigated by a totally different
method, and not applied as the author has applied it.

Cambridge Philosophical Society. 303

kinds are determinate in number, and their number is 5, let < > denote

how many diflferent combinations can then be formed: if the elements
are determinate in number, and their number is tr, let the number
of the combinations which can then be constructed, be denoted by
{m, (tJ. If (p (a:) be any function oi x, let D" (p(a;) denote the co-
efficient of a?" in that function developed according to the powers oix.
3. The same things as before being assumed, let a given set of
elements consist of a elements of the kind A, +/3 elements of the
kind B, + &c. Take the product, K, of the 5 geometrical progres-
sions,

•[H-Aa + A2a;2+ + A"a?"], [l+B^ + B2a;'+....+B'^a^],&c.

Then K will be of the form,

l+S[A]a? + S[A2 + AB]a;2+S[A3 + A«B + ABC]a:3+&c.,

and D«[K] will be of the form

S[A/'B^/C'-&c.],

the last expression being an aggregate of terms of the form A^'B'i'C'' ... ,
each containing a different combination of u of the given elements,
and their sum comprehending all the possible combinations of those
elements taken m at a time. Now, if A, B, C, &c. be each made
equal to 1, K will become

each of the terms A^.B*?. C''. &c. will become 1, and the number of
all the terms of the form A^B'i'C'" .. . which Dw[K] or S [A^B^C''...]
contains, that is to say, |m, (t} will be represented by D" [A] ; which

latter coefficient the author next proceeds to determine.
Now

« ^ . — 1— — -. &c. =[1— .-r ^ iri— ar^ ]..[1 — a;]

\—X \—X ^ -"- J L J

=[i---+']ci-»^+']...sTr^*-»i

u Ll"|i J

=[i-^"+'][i— ^+']...p^,s^[[«+ir-''V](vii.)

For brevity, write u^, a,^, /3i, &c. respectively, forM-l-l,a+l,/3 + l,
&o.; andalsowrite [l]for[l-^x]-^ [2]for[l-a;«i][l-a.]-*;
that is, for [1-.t"i] . [1]; [3] for [l-a;«i] [l-.r^i] [l-ar]"*;
that is, for [1— a;^i].C2], and so on. Then

D«[2]=D«[l]-D«-«i[l];
and

D«[3] =D«[2] -D«-^i[ 2] ;

* According to the factorial notation, here used by the author, «wl±i
represents s [s-4 1][5+2J. ..[*+_(«— 1)].

^VI.)

304 Cambridge Philosophical Society.

and

D"[4]=D«[3]-D"-5'i[3]; and soon; (VIII.)

and the developed product of the binomes,

[l-a;«i],Cl-aA],[i_a;yi],&c.;
that is to say,

1— a?*i+^«i+^i-a'"i+'^i+yi+ &c.

— &c. +&C.
when multiplied into the development of [1 —a;] ~*,
manifestly leads to the following formula :
D«M=D«[l]-S[D«-«i[l]]+S[D«-«>-^.[l]]l

- S [D^-^i-'^i-yiCl]] + &c. J

where, since the powers of a?, in (VI.) or (VII.) developed, are to be
all positive, no expression of the form

(«—«,), (m — ai— /3i), (m— ai— /5i— 7,), &c,
is to be negative. Then by giving to

D«[l],D"-«i[l],D«-«i-^i[l],&c. . . (IX.)
their respective values, we obtain the series of expressions :

where in all the kinds the elements are plural without limit ; a for-
mula given by Hirsch :

where the elements A are limited in number to a, but those of the
other (s— 1) kinds are plural without limit :

J p-iuL _|-„^_^j.-iii J-Lo

where, moreove«r, the elements B are limited in number to /3, but
those of the other (s— 2) kinds are plural without limit : and so for
the rest. The law of the terms being evident, they need not be
continued further.

Example of (IX,). Given one element of 1 kind, two elements of
a 2nd kind, three of a 3rd, and four of a 4th ; and let m=5. Then

1
{"•''} =1:2:3

-4.5.6
-3.4.5 + 1.2.3

^•^•S -2.3.4
-1.2.3

= 22.

Cambridge Philosophical Society.

305

4. If «=/5=y=&c., formula (IX.) becomes

{"■•■} =-i;4ip s<i;)[(-i)''^C»,-K]-"]. (X.)

Example of formula (X.) Given seven kinds of elements, and
three of each kind ; and let m=4. Then

^"'''>'"l.2.3^4.5.6[^-^-^-^-^-^Q~^-^-^-^-^-^-^]=

203.

5. If it is required to determine many, or all, of the terms of the
series {0,<r}, (l.o"}. {2, <r},..., {(T, cr}, formulas (VIII.) sug-
gest the following process for the determination of those terms.
An example will best explain the process.

Given 1 element of one kind, 2 elements of a second kind, and 3
of a third kind. How many combinations can be formed from these
elements, when taken 0, 1 , 2, 3, 4, 5, 6 at a time, respectively ?

Values of M

0

1

2

3

4

5

6

Coefficients of or" in[l — «•] ~^

1

3

6

1

10
3

15
6

21
10

28
15

Multiply by [1—^]; that is, subtract

Coefficients of «•« in \\-x"'\'\\-x\-^

Multiply by [1— x»J ; that is, subtract

1

3

5

7
1

9
3

11
5

13

7

Coefficients of ^» in[l-^2j [i _a;3] [i _^] -3
Mrltiply by [1— a?''] ; that is, subtract

1

3

5

6

6

1

6
3

6
5

Coefficients of ^x in[l— ^^j [i_^] [l-^^] t

[l-^]-3|

That is

1
{0,.}

3

5

{V}

6
{3,.}

5

{4,.}

3

{5,.}

1

6. Let a set of elements, S, such as we have been previously con-
sidering, consist of two similar sets, T and T', which do not contain
in common any elements of the same kind. If S consists of tr ele-
ments combining m at a time, and T consists of r elements combining
V at a time, T' will consist of {v—r) elements combining (u — v) at
a time. Consider u as constant, for the moment, and v as variable.
The author then shows that if by the process described in art. 5, the
whole series of terms {u, t'} and the whole series of terms iu—v,
c — rj, have been determined, we can thence determine the whole
series of terms {m, c} by means of the formula

{«,cr}=S^[{t;,r}.{M-v,(r-T}]; . . (XI.)

and of this he gives examples.

7. In formula (XI.) substitute (cr— «) for u ; and develope iu, <t\
and /<r — u,<t\ in the manner indicated by that formula. By com-
paring the 1st, 2nd, 3rd, &c. terms respectively of [u, cr} with
the last, last but one, last but two, &c. terms of icr—u, <r\, and vice

Phil. Mag. S. 3. Vol. 3 J . No. 208. Oct. 1 84.?. X

306 Cambridge Philosophical Society,

versd, the author shows that [u,or^ will be identical with /«■— m, tr},
provided {«.''} is identical with {t—v, r|, and provided also {u—v,
cr—t'^ is identical with /a— r— (m— jj), tr— rl. But this identity
actually exists when T consists of elements of one kind only, and
when T' also consists of elements of one kind only. For, in that
case, every term of the series |f, t'I and every term of the series
|m--u, cr— r} is equal to 1. Let the elements of the single kind
which T contains, be different from those of the single kind which
T' contains. Then the identity in question will exist, when S con-
sists of elements, finite in number, of two different kinds : conse-
quently, it exists also when T consists of elements, finite in number,
of two different kinds, and T' consists of elements, finite in number,
of one or two other kind or kinds ; that is, when S consists of ele-
ments, finite in number, of three or four different kinds. And there-
fore universally, in the case as well of finitely plural, as of singular
elements, the following law obtains :

{u,v) = [<T-u,(r) (XII.)

Hence it follows that in applying formulas (IX.) and (X.) to parti-
cular cases, the labour of computation will be shortened by substi-
tuting for the variable the lesser of the two numbers u and <r— m.

8. The author next considers how many different combinations
can be formed from a given set of elements, when every combination
is to be constructed in conformity with a given type ; in which type
there are m different kinds containing v elements each, m' other dif-
ferent kinds containing v' elements each, m" other different kinds
containing v" elements each, and so on ; and where, consequently,
in each combination, z, the number of kinds, is wj-{-Jw'-j-/w"+ &C' ;
and u, the number of elements, is mv + m'v' -^tri'v" + &c. The type
remaining constant, any combination conformable thereto may be
altered, either by changing the particular z kinds which are selected
out of the s given kinds ; or, the kinds remaining the same, by alter-
ing the distribution of the parts v,v,v, . . . (m)v', v', v', . . . {m*)v'' ,v" yv" ,
. . . {m") &c., among those kinds. When all the elements are plural
without limit, the changes of the former description will be repre-
sented by

1*11 '
and those of the latter description by

\m\\^ \m,'\\^ \iii."\\ , , . *

and their joint effect by the product

~pii~ ^ lm|l^ im'li, l»n"|l, * • • •* (XIU.)

But when the elements of all the given kinds are finite in number,
class these kinds, so that each kind in class 1 contains not fewer

Cambridge Philosophical Society. 307

than V elements ; each kind in class 2 contains fewer than v, but not
fewer than v' elements ; each kind in class 3 contains fewer than v',
but not fewer than v" elements ; and so on ; and so that the given
kinds may in this way be reduced, say, to t kinds containing v ele-
ments each +T' kinds containing v' elements each + T'' kinds con-
taining!;" elements each,&c. Thenlet^— ?n + T' = ^'; t' —m! + T'=zt";
and so on. The given kinds being thus ordered, since we are required
to select, 1st, m out of t kinds ; then, 2nd, m! out of t' kinds ; then,
3rd, m" out of t" kinds ; and so on ; the number of the different
combinations which can be constructed from those kinds in confor-
mity with the type, will be

f!ti.:^.::^,&c. . . . . (xivo

1»»|1 \m'\l J[to"I1 ' ^ '

If /y-f TV+TV. . &c. is reduced to a single term, t.v; then
formula (XIV.) becomes

t«l-i

\m\l^ 1»»'|1, l»»"|l,

&c (XV.)

Example of (XIV.). Given eight elements of 1 kind, seven of a
2nd kind, six of a 3rd, five of a 4th, four of a 5th, three of a 6th,
two of a 7th, and one element of an 8th kind, out of which it is re-
quired to construct combinations, each consisting of three kinds with
five elements each + two kinds with three elements each -)- one
kind with two elements. Of such combinations there can be formed

4»l-i 3^1-' 2»l-t _
is|i • i2|i • im

9. If it be required to determine how many different combinations
can be constructed, each containing u elements of z kinds, and the
given elements are all finite in number ; we must form all the differ-
ent S'-partitions of u ; and each of these partitions being regarded as
a type, we must determine, by formula (XIV.) or (XV.), how many
combinations correspond to each of these types ; and the total num-
ber required will be the sum of all these particular determinations.
But if the given elements may all be repeated without limit, it
follows from formula (XIII.), that the sum of all the particular de*
terminations may be represented by

Now

'\l'».i. l'"'li. l*""!'. &c./'

i lm|l^ lm'\\^ lm"\\ )

denotes how many different permutations can be formed, when, in
each different z-partition of u, the parts are permuted z together at
a time ; and the number of such permutations is

s-'[i]=s-2[«-i] == £!^r|Zl!!lL,

X2

308 Cambridge Philosophical Society.

Consequently the required sum is

,_j-, [„-i].-,i-

If in (XVI.) ^ varies from 0 to m— 1,

J 0 [^ pii ^ 2,_i|, J j„u'

this summation being a particular case of formula (XL). The result
agrees with D"[l] formula (IX.), art. 3.

10. When the given elements are all finite in number, we may
determine •[«, crj, by taking the sum of all the particular determina-
tions that may be obtained pursuant to art. 9, by giving to z the
successive values 0, 1,2, 3, &c. If w <: s, the upper limit of z is m,
and the number of types to be formed is [2m, m,]; which becomes
[2s, Sj], if M=s. If M > s, the upper limit of 2 is .9; and the number
of tj^pes to be formed is [w +5, sj . (See articles 4 and 5, Section I.)
But, if the repetition is finite, some of these partitions may fail to
yield combinations.

11. If the elements A, B, C, &c. represent different prime num-
bers, all the methods and theorems contained in this section will
apply, mutatis mutandis, to the composite numbers of whicli those
primes, or the powers of those primes, are divisors.

III. On Permutations.

1 . Let the given elements be of 5 different kinds. We can de-
termine in two known cases, by an explicit function of u, when the
elements are taken m at a time, in how many different ways they can
be permuted. The number of the permutations is denoted, when
there is but one element of a kind, by s'«l-i ; and when in all the
kinds the elements are plural without limit, by s". When the plu-
rality is finite, it is only in the particular case of all the elements
being permuted at a time, that there is a known formula to express
the number of their permutations.

2. Every combination constructed on a given type, u=.mv-\-m'v'
+ m"v"+ &c., will generate the same number of permutations,

]^ ^p

Mciinwi rii)'ii-im'n""in»»" . &c.

Therefore, if the number of the different combinations which can be
constructed out of the given elements in conformity with that type,
is represented by Q, Q X P will be the number of the permutations
corresponding to the type and to those elements. If the plurality
be without limit,

xP

JWlll^ 1»»'|1. l»i"|l^ gjg

will be the number of the permutations. If the given elements be
finite in number, as in formulas (XIV.) and (XV.), the number of

Cambridge Philosophical Society. 309

the permutations corresponding to those elements and to the type,
will be

JtoU * \m'\\ ' jm"!!

&c. X P.

Every different partition of u that may be formed within the limits
pointed out in art. 10, Section II., will give rise to a similar product,
QxP ; and the sum of all these particular products, S[Q x P], will
show how many different permutations can be formed from the given
elements, taken m at a time. The author illustrates this method of
computing the number of permutations, by examples.

3. Let P < " I denote how many different permutations can be

formed when u elements are taken at a time out of 5 kinds ; and P
{m, (TJ denote how many different permutations can be formed when
u elements are taken at a time out of c, a finite number of elements.
If all the elements may be repeated without limit,

=i>[i«i.[i+..+£i + ....^+. ...]*]•

Hence the author infers that, if the elements A are limited in num-
ber to a, while those of the other (s— 1) kinds are plural without
limit,

that if, moreover, the elements B are limited in number to /3, while
the other (s— 2) kinds are plural without limit,

p{^}=D»[l«l..<->[l+.+ ^+..^]

and so on, until finally, if all the elements are finite in number, and
the elements A, B, C, &c. are respectively limited, in point of num-
ber, to a, /3, 7, &c,.

fi -^'\

(XVII.)

4. Hence, if in all the s kinds the elements are dual, (XVII.)
becomes

• . . (XVIII.)

SIO Cambridge Philosophical Society.

This is the only addition which the author has been able to make
to the cases wherein PJ " , or P|m, o-} is expressed by an explicit

function of u, symmetrical in form.

Example. Let there be five kinds of elements, and two of each
kind. Let m=3.

5. The author gives the following theorem, which is precisely
analogous to that of art. 6, Sect. II., formula (XL), in Combina-
tions ; viz.

P{«,^}=S ;[^^'p(.,r}. ?{«-„, cr-r}]. (XIX.)

6. By a mode of proof precisely analogous to that employed in
art. 7, Sect. II., he shows that P{(r-1, o'}=P{(r, tr} ; that is to
say, that

l«l'.l/3|l.iy|l. &c.

denotes the number of permutations that can be formed with a ele-
ments A, (i elements B, &c. (where [a+/3 + y+ 3cc.] = (r), as well
when (T— I elements, as when c elements, are taken at a time.

Since correcting his paper for publication, the author has had his
attention called to the work of Bezout on Elimination (4to. Paris,
1779, p. 469), as containing a formula similar in structure to that
numbered VIII*. in the present abstract.

Bezout investigates the composition of a polynome function of 5
quantities. A, B, C, &c., consisting of terms which are of the form
ApE'iO', and of every dimension from 0 to m inclusive. Let [*]"
denote such a polynome, complete in all its terms, and N[]s]" the
number of its terms. Then, 1st,

and 2nd, the number of the terms in [«]« which are not divisible by
either A*, or B^, or C, &c., he expresses by

N W-N[«]«-« -f-N[s]«-*-/^- &c.
-NM«-^+ &c.
— &c.

Intelligence and Miscellaneous Articles. 311

He also observes (p. 89) that when A* B^, C^, &c. are the high-
est powers of A, B, C, &c. which a jjolynome, agreeing in other
respects with [s]", contains, the terms of such incomplete polynome
will agree in point of number with those terms in [s]" which are
not divisible by either A*+l, or B'^+^j or C>'+^ &c. The polynomes
from which Bezout proposes to eliminate certain terms, contain
terms of all dimensions from 0 to m inclusive. The terras which are
to remain after the others have been eliminated, and which are enu-
merated by means of the condition, that they are not divisible by
certain powers of A, or B, or C, &c., may be of all dimensions indis-
criminately from 0 to M inclusive. Bezout's object is exclusively
Elimination, and he makes no allusion to any other application of his
formulae.

The polynomes considered by the author, taken in their entirety,
agree in their general structure with those considered by Bezout ;
but the nature of the author's inquiries led him to confine his atten-
tion to the composition of those particular terms in a polynome
which were of the same dimension ; and to seek to express the
number of the terms, not of all dimensions indiscriminately, but of
each particular dimension separately. To show how it has hap-
pened that researches, very different at their point of departure,
have, as regards one point of investigation, ended in nearly similar
formulae, the author proceeds to deduce his formula (VIII*.) from
the investigations of Bezout. Such a deduction, he conceives, might
readily have been made by any one to whom it had occurred to make
it ; and the application of such a deduction, when once made, to pro-
blems in Combinations, would have been much too obvious to have
remained long unnoticed.

Expressions of the form above considered are regarded by Bezout
as of the nature of Differences ; and the truth of this view of the
subject may be shown in the following brief manner.

If (p(a;) generates \J/(m),[1 — a;«]f (a;) will generate \|/(m)— \|/(m— a),
which we may denote by A^vKm). Consequently \\—x^'\ [1 —x»'\
<p{x), that is to say,

[ -X )^(^)'

will generate A'^ g^p^u); and so on; the independent variable, ti,
undergoing, not uniform, but variable decrements, «, /3, y, &c.

L. Intelligence and Miscellaneous Articles.

ON THE AETIFICIAL PRODUCTION OF MINERALS, ANP ESPE-
CIALLY OF PRECIOUS STONES.

MEBELMEN states that the first results which he obtained
• related to minerals of the family ofSpinelles.
Tlie method adopted by the author to effect the crystallization of
these compounds, is based on the property which boracic acid pos-

312 Intelligence and Miscellaneous Articles.

sesses of dissolving metallic oxides in the dry way, and the volati-
lity of this acid at a high temperature. It occurred to him that by
dissolving alumina and magnesia, mixed in the proportions which
constitute spinelle, in fused boracic acid, and exposing the mixture
in open vessels to the high temperature of a porcelain furnace, that
the affinity of the alumina for the magnesia might cause the separa-
tion of a crystallized aluminate and the expulsion of the boracic acid.
The proportions employed were about one part of fused boracic
acid, and two parts of a mixture of alumina and magnesia, composed
so as to constitute the compound Al'-^ O^ MgO ; and from ^^ to
2-g-o of bichromate of potash were added to it. The ingredients,
well-mixed, were placed on platina foil, in a cup of porcelain, and
exposed to the highest temperature of the porcelain furnace of Sevres.
A product was obtained the surface of which was covered with cry-
stalline facets, and the interior contained cavities sprinkled with
crystals, the form of which was readily distinguishable with a glass.
These crystals were rose-red, transparent, scratched quartz readily,
and had the form of the regular octohedron without any modifica-
tion. They were completely infusible by the blowpipe. These
characters, combined with the composition of the crystals as deduced
from synthesis, appear to M. Ebelmen sufficiently conclusive as to
their identity with spinelle.

By substituting the equivalent of protoxide of manganese for
magnesia, a crystalline product was obtained in large laminae, ex-
hibiting the form of equilateral triangles or regular hexagons. The
author considers these as constituting the manganesian spinelle Al^O'
MnO, which has not hitherto been met with in the mineral kingdom.

Oxide of cobalt substituted for magnesia, equivalent for equiva-
lent, yielded crystals of a black-blue colour, in regular octohedrons.
They also scratched quartz, but not so readily as the two preceding.

In employing alumina and glucina in the proportions which con-
stitute cymophane AP O^ GIO, a mass covered with crystalline
asperities of great splendour was obtained. This product scratched
quartz and even topaz distinctly ; it therefore possessed hardness
comparable to that of natural crystallized cymophane.

Certain silicates, which are infusible by the heat of our furnaces,
appear also to be produced by the same process. Thus, on fusing
the elements of emerald with half their weight of boracic acid at the
same temperature as in the preceding experiments, a substance is
obtained which easily scratches quartz, and its surface presents a
great number of facets, the form of which is the regular liexagon.

The author proposes to continue these experiments, but at pre-
sent only states in addition, that it is possible to produce at tempe-
ratures lower than those obtainable in our furnaces, diaphanous
crystals, the hardness and external characters of which are analogous
to those of precious stones ; and he also concludes that many mi-
neral species may be formed at a lower temperature than that re-
quired for their fusion. — Comptes Rendus, August 16, 1847.

Intelligence and Miscellaneous Articles. 313

ANALYSIS OF THE GRAY COPPER FROM MOURAIA IN ALGERIA.

M. Ebelmen states that a copper mine, apparently of great im-
portance, has been for some time worked at the foot of the defile of
Mouraia in Algeria. The veins are composed principally of car-
bonate of iron and gray copper ; the latter sometimes occurring in
compact masses and sometimes in crystals, the prevailing form of
which appears to be a rhombic dodecahedron, but with numerous
modifications on the edges and angles.

The specimens received by M. Ebelmen for analysis contained
a great number of very brilliant small crystals of gray copper, on a
gangue composed of carbonate of iron and sulphate of barytes.
These specimens were digested for some time in warm dilute hy-
drochloric acid, which dissolved the carbonate of iron without alter-
ing the gray copper, the crystals of which were then readily de-
tached.

Qualitative experiments, conducted in the usual manner, showed
that the ore contained sulphur, arsenic, antimony, copper, iron and
zinc : lead, bismuth, and mercury were tried for, but not the smallest
quantity was found. No notable quantity of silver could be de-
tected ; and the fact that M. Berthier found 0*0008 in 1 part of the
ore, shows that the silver is very irregularly interspersed through
the veins.

For the quantitative analysis of this ore, M. Ebelmen employed,
with a slight modification, the method proposed by M. H. Rose ;
and taking the mean of several experiments, he obtained the fol-
lowing as the composition of this ore : —

Sulphur 27-2.5

Antimony 14'77

Arsenic 9*12

Copper 41-57

Iron 4-66

Zinc 2-24

99-61

If the analysis of this ore be compared with that of gray copper
from various localities, the greatest similarity will be found between
it and that from Sainte-Marie-aux Mines, which gave M. H. Rose —

Sulphur 26-83

Antimony 12*46

Arsenic 10*19

Copper 40-60

Iron 4*66

Zinc S'Q9

Silver 0-60

99-03
Annates des Mines, tome xi., p. 47.

S14 Intelligence and Miscellaneous Articles,

ANALYSIS OF KUPFERNICKEL.

M. Ebelmen states that this mineral comes from Ayer, in the val-
ley of Annivier (H' Valais). It possesses the usual characters of
Kupfernickel. It forms compact masses which are perfectly homo-
geneous, but exhibit no ti'aces of crystals ; the ore is mixed with
laminar carbonate of lime, which is easily separated by dilute hy-
drochloric acid. Its density is 7*39.

The analysis was effected by treating the purified mineral with
aqua regia. The sulphuric acid was precipitated by chloride of ba-
rium and the excess of barium by sulphuric acid. The arsenic acid
was converted into arsenious acid by means of ebullition with sul-
phurous acid, and the arsenious acid was precipitated by sulphuretted
hydrogen. The sulphuret of arsenic obtained was, after drying and
weighing, analysed by aqua regia to obtain the sulphur ; by heating
another portion in a current of hydrogen, a minute residue of anti-
mony was obtained. The liquor freed from sulphuret of arsenic was
concentrated along with nitric acid, and precipitated by excess of
ammonia; an abundant precipitate of peroxide of iron was formed,
which retained a little nickel, ns appeared from its colour.

It was redissolved on the filter by hydrochloric acid, and the
liquor was then treated cold with carbonate of bary tes. The peroxide
of iron only was precipitated ; the carbonate of barytes, with which
it was mixed, was readily separated. The liquor containing the
nickel was treated with sulphuric acid, and after filtration it was
added to the ammoniacal solution of the rest of the nickel ; this was
precipitated by excess of potash, and after drying and calcining, it
was weighed, and its quantity indicated that of the metallic nickel.

The ammoniacal liquor, afterwards treated with hydrosulphate of
ammonia, yielded a slight black precipitate, which, collected, calcined
and weighed, gave with borax the reaction of cobalt.

The results of the various experiments showed that the ore con-
sisted of —

Arsenic 54<'05

Antimony 0*05

Nickel 43-50

Cobalt 0-32

Iron 0'45

Sulphur 2*18

Gangue 0*20

100-75
Annales des Mines, tome xi. p. 5Q.

ON THE DEHYDRATION OF MONOHYDRATED SULPHURIC ACID.

M. Barreswil observes that anhydrous sulphuric acid has been
hitherto prepared by distilling protosulphate of iron or dry bisul-
phate of soda. These two processes produce an anhydrous salt
and sulphuric acid. The author states that he is not aware that an

Intelligence and Miscellaneous Articles. 315

attempt has ever been made to deprive concentrated sulphuric acid
of spec. grav. 1*848 of its water, without previously causing it to
enter into combination and form a salt. The same may also happen
when monohydrated sulphuric acid is employed in the preparation
of fluoboric and fluosilicic acids, which are considered as substances
having great affinity for water.

The reaction which M. Barreswil employs he considers as ex-
tremely simple. He mixes anhydrous phosphoric acid, with the
sulphuric acid of commerce, and leaves them in contact, and the
mixture is afterwards heated : the combination of the two acids pro-
duces an increase of temperature, and some acid vapours soon ap-
pear ; but this is prevented by proceeding cautiously, and keeping
the acids in a freezing mixture : by distillation anhydrous sulphuric
acid is disengaged, and vitreous hydrated phosphoric acid remains ;
the distillation is effected in the same way as the Saxon acid.

A circumstance which struck the author in this operation, is the
fact of the innocuousness of the mixture of monohydrated sulphuric
acid and anhydrous phosphoric acid, with respect to organic matters,
si;ch as paper and cotton, which are instantaneously destroyed by
the Saxon acid. The author considers this circumstance as a proof
that the sulphuric acid in the mixture is not anhydrous, but becomes
so when heat is applied.

Even if the reaction above described possesses interest in a theo-
retical point of view, M. Barreswil admits that as a manufacturing
process it is unimportant, and will hardly be regarded as a ready
method of obtaining anhydrous sulphuric acid. The high price of
phosphorus, and the difficulty of preparing anhydrous phosphoric
acid, ai-e obstacles to the employment of the process. — Comptes
Jiendus, Juillet 5, 1847.

OBSERVATIONS ON SILICA. BY M. DOVERI.

It results from the experiments detailed by the author —

1. 'I'hat the alkaline silicates, when decomposed by acids, and
particularly hydrochloric acid, deposit the greater part of the silica
which they contain if the acid in excess be added drop by drop ;
whereas the same quantity of acid added at once does not occasion
the precipitation of the smallest portion of silica.

2. That silica, once precipitated, does not redissolve in acids,
whatever may have been its origin, whether precipitated from an
alkaline silicate by an acid, or from fluoride of silicium by water.

3. That weak acids, as the carbonic, sulphiu'ous, boracic and the
vegetable acids, decompose the alkaline silicates at common tempe-
ratures, and precipitate the silica either as a jelly or in gelatinous
flocculi.

4. That very finely-divided silica, whether anhydrous or hydrated,
is capable of decomposing the aqueous solutions of the alkaline car-
bonates, and dissolving in the solution at a boiling heat.

5. That silica precipitated at common temperatures from a solu-

316 Intelligence and Miscellaneous Articles.

tion of an alkaline silicate or from fluoride of silicium, is a hydrate
of definite proportions, the composition of which may be represented
by the formula HO. Si O?. This hydrate, when heated to 212° F.,
loses one equivalent of water, and is converted into another com-
pound, HO, 2Si 03.

6. That when a solution of an alkaline silicate is treated with a
metallic solution, a precipitate is formed, which is a mixture of hy-
drate of silica and a metallic silicate ; the metallic silicate being
entirely dissolved by the mineral acids, while the free silica re-
mains undissolved.

7. That a limpid and very strong solution of silica in hydrochloric
acid may be obtained by dissolving in this acid silicate of copper,
and precipitating the copper by sulphuretted hydrogen.

8. That a solution of silica in hydrochloric acid, slowly evaporated
under the receiver of the air-pump, gives hydrate of silica (HO, Si O^)
perfectly crystallized in very small transparent needles, grouped
either in stars or tufts. — Comptes Rendus, Juillet 19, IS^?.

ON NITRIC MANNITE. BY M. SOBRERO.

Since the action of nitric acid on organic bodies has been
studied, a number of substances of great interest to science have
been discovered ; but the arts have hitherto acquired only fulraina-
ting-cotton, the fate of which is as yet uncertain. Whilst the ques-
tion as to cotton is under consideration, M. Sobrero announces to
the Academy another body which is fulminating in the highest de-
gree, resulting from the action of nitric acid upon mannite — the
nitric mannite, the composition of which has been already given by
MM. Flores Domonte and Menard.

Fulminating mannite possesses the property of detonating by the
stroke of a hammer with as much violence as fulminate of mer-
cury, and produces, during its decomposition, sufficient heat to in-
flame gunpowder. As soon as the author was acquainted with this
property, he set about to apply it, and prepared capsules with it
instead of detonating mercury for the discharge of fire-arms, and a
fowling-piece was discharged by it.

With respect to its use, the author has arrived at the following
conclusions : —

1 . Fulminating mannite must always be cheaper than fulminating
mercury.

2. It is more conveniently prepared, and does not expose the
workmen to the great danger which attends the manufacture of
fulminating mercury.

It must be cheaper than fulminating mercury, because the price
of manna is not very high ; because in the preparation of mannite
an uncrystallizable residue is obtained, mixed with a little mannite,
which may be employed in medicine and the veterinary art as a
purgative ; and because, according to the analyses of MM. Flores

Intelligence and Miscellaneous Articles. 317

Domonte and Menard, the mannite, in becoming nitric mannite,
must increase considerably in weight (from 100 to 225).

It is less dangerous in preparation and manipulation : in fact the
preparation is merely accompanied with the disengagement of some
vapour of nitric acid.

Fulminating mannite requires for detonation a violent blow be-
tween two hard bodies ; heat gradually applied to it fuses and after-
wards decomposes it, but without detonation. In fact it may be
placed on paper and touched with a red-hot coal, and fused without
detonation ; the paper on which it is put may be burnt, and it is
decomposed without detonation.

Lastly, fulminating mannite is decomposed by the blow of a ham-
mer, without, as far as appears, producing nitrous vapours. It
seems to be entirely decomposed into carbonic acid, water and azote ;
besides which it keeps indefinitely without undergoing decomposi-
tion.—-Co?np/e* Rendus, Juillet 19, 1847.

ON THE EXTRACTION OF SILVER.
BY MM. MALAGUTI AND DUROCHER.

From the numerous researches which the authors have made on
a large series of specimens from different parts of Europe, they have
inferred the general fact, that all metallic compounds which accom-
pany or are found near argentiferous minerals contain more or less
silver ; so that they deem it an established fact, that silver is pro-
bably one of the most widely-diffused metals in nature.

The researches of the authors have been made on sulphurets,
arseniurets, arsenio-sulphurets, some metallic oxides, and even native
metals. This fact being established, the mode in which tl>e silver
exists occupied their attention. As the subject appeared a difficult
one, it was simplified by inquiring in what state the silver existed
in galena, blende and pyrites, and they supposed it could exist only
in the native state, as chloride or sulplmret. Experiments appeared
to show that in these sulphurets the silver is not in the metallic
state ; and experiments still more numerous and decisive seemed
also to prove that the silver could not be in the state of chloride ;
and on this occasion they remarked a circumstance which has hitherto
escaped the observation of chemists : — Tliey found that all metallic
sulphurets, properly so called, and even some arseniurets, possess
the property of decomposing a certain quantity of chloride or bro-
mide of silver. This decomposition is effected more or less slowly
when contact is effected merely by water ; but it is produced much
more rapidly, and in some cases even instantaneously, when the
chloride or bromide of silver is in solution.

By comparative trials the authors succeeded in determining the
decomposing power of a great number of sulphurets and several
arseniurets. Thus —

318 Intelligence and Miscellaneous Articles.

100 of sulphuret of zinc decompose 3 of chloride of silver

100 .... cadmium ..14- ....

100 .... bismuth .. 2 ....

100 lead . . 5 ....

1 00 protosidphuret of tin "a • • • •

100 of bisulphuret of tin ..SO ....

100 protosulphuret of copper 360 ....

100 arseniuret of antimony 120 ....

100 .... cobalt 166 ....

In operating with natural sulphurets, the authors remarked very
considerable differences in their decomposing power. They attri-
bute these differences to the presence of small quantities of sulphu-
rets or arseniurets of very high decomposing power ; and they sup-
pose they may sometimes attach to the molecular condition of the
bodies. They found, for example, that a very pure and well-cry-
stallized blende from Kiinigsberg possessed decomposing power
equal to that of artificial sulphuret of zinc; while a blende equally
pure and as well crystallized, but coming from Radna, had a decom-
posing power which was twice as weak, and yet these two blendes
were of equal density.

The authors draw the following conclusions from the results of
their experiments : —

AH pure metallic sulphurets possess the power of decomposing,
under certain circumstances, a given quantity of chloride of silver,
and even of other insoluble chlorides. This power appears to be
modified in some cases by the molecular condition.

The decomposition of chloride of silver by sulphurets may be
effected, — 1st, by double decomposition ; 2nd, by reduction; 3rd,
by simultaneous reduction and double decomposition.

Natural sulphurets sometimes exhibit very high absorbent powers,
on account of the presence of minute quantities of foreign sulphurets
or arseniurets, acting by the reduction of the chloride of silver.

The decomposing action of sulphurets is exerted proportionally on
the bromide of silver, and it is but slightly appreciable on the iodide.

In these phaenomena the solvent exerts no influence ; for the
same results are obtained, except as to time, by simple contact aided
by water.

The general fact of the decomposition of insoluble chlorides by
sulphurets appears then to render it probable that, in natural sul-
phurets, the silver is in the state neither of chloride nor bromide.

Having then shown the improbability of the presence of metallic
silver or chloride in the natural argentiferous sulphurets, the authors
are of opinion that it must exist in the state of sulphuret ; but if
this conclusion were correct, how does it happen that blende, pyrites
and galena, do not yield silver to mercury? Is not the sulphuret
of silver almost as readily acted upon by mercury as metallic silver
itself? The authors propose shortly to communicate the second part
of this inquiry to the Academy. — Comptes Rendus, Juillet 26, 1 847.

Meteorological Observations. 319

VANADIATE OF LEAD AND COPPER.

M. Dufr^noy presented to the Academy, in the name of M. Do-
meyko, Professor of Chemistry and Mineralogy in the college of San
Yago, Chili, an account of this new mineral, which is composed of —

Oxide of lead 54-9

Oxide of copper 14'6

Vanadic acid 1 3*5

Arsenic acid 4*6

Phosphoric acid 0*6

Chloride of lead 0-3

88-5
Comptes Rendus, Mai 5, 1847.

METEOROLOGICAL OBSERVATIONS FOR AUG. 1847.
ChinvicL- — August 1, 2. Very fine : sultry. 3. Very fine : clear. 4. Very
fine : densely overcast. 5. Rain. 6. Overcast. 7. Very tine. 8. Very fine :
cloudy. 9. Cloudy: shower: clear. 10. Rain: showery. 11. Very fine.
12. Light clouds, with bright sun at intervals: clear at night. 13. Overcast:
very fine. 14, Very fine : cloudy. 15. Cloudy : clear : lightning at night. 16,
Rain. 17. Overcast. 18. Heavy rain. 19. Overcast: lightning at night.
20. Uniformly overcast : slight fog. 21. Slight fog : fine. 22. Overcast : rain :
cloudy. 23. Cloudy : rain. 24. Cloudy : clear at night. 25. Very fine. 26.
Overcast : very fine. 27, 28. Very fine. 29. Rain : very fine. 30. Very fine :
cloudy. 31. Very fine : clear at night.

Mean temperature of the month 62^*68

Mean temperature of Aug. 1846 64 •]6

Mean temperature of Aug. for the last twenty years 62 -32

Average amount of rain in Aug 2*41 inches,

Boston. — Aug. 1. Fine : 2 o'clock p.m. thermonn.eter 83°. 2. Fine : rain p.m.

3, 4. Fine. 5. Cloudy : rain p.m. 6. Fine. 7. Fine : rain p.m. 8. Fine.
9,10. Cloudy. 11. Cloudy : rain early A.M. 12. Cloudy. 13, 14. Pine. 15.
Cloudy. 16. Cloudy : rain a.m. and p.m. 17. Cloudy: rain p.m. 18, 19. Cloudy.
20— 25. Fine. 26. Cloudy. 27. Fine. 28. Rain. 29. Cloudy : rain early a.m. :
rain p.m. 30, 31. Cloudy.

Sandivick Manse, Orkiiey. — Aug. 1, 2. Brijiht : clear. 3. Bright: cloudy.

4. Cloudy : drops. 5. Bright : cloudy. 6. Cloudy : fine. 7. Rain : fine. 8.
Cloudy: rain. 9. Cloudy: fine. 10. Cloudy: rain. 11. Clear : showers.
12. Cloudy. 13. Clear : cloudy. 14. Cloudy : fine. 15. Bright : fine. 16,
17. Clear: fine. 18. Cloudy: fine. 19, 20. Cloudy. 21. Showers : rain. £2.
Cloudy : showers. 23. Clear : showers : cloudy. 24. Cloudy : rain. 25. Cloudy.
26. Cloudy : rain. 27. Cloudy : clear. 28, Bright : showers : clear. 29. Showers.
30. Rain : showers. 31. Bright: rain.

Apjtlegarlh Manse, Dumfries-shire, — Aug. 1. Fair, but cloudy. 2. Fair and
fine: shower early A.M. 3. One slight shower. 4. Rain early a.m. 5. Rain
nearly all day. 6. Frequent showers. 7. Heavy showers and sUn. 8. Rain.
9. Cloudy : cool : dry, 10. Heavy rain. 11. Fine a.m. : rain p.m. 12. Rain
nearly all day. 13, Fair and fine. 14. Very fine. 15, 16. Very fine : heavy dew.
17. Fine, though cloudy. 18. Very fine. 19. Still fine, but dull. 20. Heavy
showers. 21. Slight showers. 22, 23. Fine : clear. 24. Rain p.m. 25, 26. Fine,
though cloudy. 27. Fine, though cloudy : a few drops. 28. Fine, though cloudy :
one slight shower, 29. Fair and fine, 30. Fine : one slight shower. 31. Fine
harvest day.

Mean temperature of the month 57°'15

Mean temperature of Aug. 1846 61 '2

Mean temperature of Aug. for twenty-five years 57 '14

Average rain for twenty years .,, 3*16 iOCbes,

-saiajiutiQ

•ao)8og[

•3lO!.ttsm3

•JlOIiipUBS

'Xauiijo

•ajiqs
-saujiuna

S^

CO o\ n O) r^ n •->
o •^"-< c — o -^

I'l i S S"ll c'l S:' » I* I "• « B? ^' fe' I i i c ^' S ^*l i fe S:* &'

So QUO S? pa> y^

^* ^' ^' «* i i £ i i i ii^foii] ^ ^ >• ^ ^ i i i ^ & i % i

1 ^* §"« » c 5:

u u u u

o § ■•"■«

" f6

H« ^i« "-tte? •HJN

co*o icr>-oo o^io c>-oi u:)0^o^r>.oo t^ON^ooooio loto-^^o^o o ot^-ioiO'^

3 <"

'am

•Ul'Bfg

•uojsoa

•uiH

•XBK

-wg

■^s

•iioisoa

•q^uOH
JO sXbq

*^ to

«0^t-0<^lO^^O^O^OlOO^^Ol

r^oooo^o■^o•^u^«^ot^ooc^•^^^'7<oc^o^^p^o>ooo^ooo — criir^'P
0^6^6^C^0^6^CT^^^C^^3^a^CT^O O O O O O 6^6^0^6^0 ^^0 O O O O^CTM3^

I:^oooo^^•^"o■^^Ol^^lOoo*o--lcoc^(^Io^^^■^c>•l^~oo^op•7<o^poqoao
0^o^cy^O^C?^C^C^O^O^O^O^C^O O O O O O O 0^0^0^0^0 O O 0^0 O 0^0%

-<lCQOOO^cocr)•^^^«■oo■^ooo^u:)OOoo^oo^OlCc-»<^^(^l^c^o^oo^o
0^0^6^6^6^0^6^c^C^l6^0^0^0000C>OOO^o^o^OOOOOOOO^O^

0*000 r^t^o t^o coo ir^-^oo •^co i>-oo cot^-^rot^oo ooo OMnt^'tcyioo

■^ (N O

CTNCX3 O

— "^-lO ro
-^O O <M
r^ ip uo r-
0^ 0^ 0^ 0^
Ol C< <N <N

o -- 00
o in o

(N 00
t^QO

o o

Oslo CO o

00 t^io -^

^ Of •-' o

O <N >0 lO

rtoo — a\
o o o 00

■^^O 00 I
t^(0 00 •

CI T-

6 6

I CO CO

t^lO — c

O OMJ^

00 •^o>
^ -^o
ooo
6 6 6

CO CO CO

lO O CO CO
■^ 'S' CO t^

^ VO -" '

CO c^oo

t^OO 00

coio es 00

UO — i <M 00
<M COOl O

6 6 6 6

I CO CO CO CO

■^O ^ !>•

O T* 7H 9^
6 6 6 6^

CO CO CO (N

— coo

— vo 10

OM>-0

■^ CO r^
«M I— oi
— — <N

cy\o

d CO

6 6
coco

in 10 o>

"5}<«5 o
— 00

6 6 6

CO CO CO

i-JeJco-^m<oc^oodo— "<Nco'^iniot^ooo\o — c^co•^ln^Ol>•Qoo^o —
— >.. vj-4^ — — rHi-ipH — ^rH — — <NCN<N<M<NOI<N<NC»C<COC0

w A « O

LONDON, EDINBURGH and DUBLIN

PHILOSOPHICAL MAGAZINE

AND

JOURNAL OF SCIENCE.

[THIRD SERIES.]

NOVEMBER 1847.

lA. Researches on the Voltaic Arc, and 07i the iiifluence which
Magnetism exerts both on this Arc and on bodies transmitting
interrupted Electric Currents. By M. Auguste De la Rive,
Professor in the Academy of Geneva, Foreign Member of
the Royal Society, Corresponding Member of the Academy
of Sciences at Paris, 8cc.^

rilHE luminous voltaic arc occurring between two conduct-
J- ing bodies, each communicating with one of the poles of
the pile, is not merely one of the most brilliant phaenomena in
physics, but, from the numerous aspects under which it may
be regarded, it is also one of the most important.

As a source of light, this phaenomenon, when exhibited in
a vacuum, enables us to examine what influence this particular
origin of the light employed may have in various optical ex-
periments. Compared with the solar light, the light of the
voltaic arc presents some curious differences and also resem-
blances. If, on the one hand, we find in it the seven coloured
rays of the spectrum, on the other the black streaks are re-
placed by brilliant ones, and these are differently interspaced.
In this field of inquiry, much, or rather all, yet remains to be
investigated.

As a source of heat, the voltaic arc enables us to study the
fusion and solidification of even the most refractory bodies in
vacuo, and consequently under circumstances exempting them
from oxidizing action and other chemical influences, which
usually result from the application of a high temperature in
atmospheric air. It likewise allows us to determine the effects
produced upon bodies at a high temperature, by various gases
or vapours, distinct from those which enter into the composi-
tion of atmospheric air, and at different degrees of density.

As an electro-chemical power, the voltaic arc may be ap-

* From the Philosophical Transactions for 1847, part i. ; having been
received by the Royal Society Nov. 26, 1846, and read Jan. 7, 1847.
Phil. Mag. S. 3. Vol. 31. No. 209. Nov. 184.7. Y

322 M. De la Rive's Researches on the Voltaic Arc.

plied so as to submit to the electrolizing action of the electric
current gaseous media, which, from some experiments already
made, appear capable of decomposition by this process.

As a mechanical power, the voltaic arc, by bringing bodies
into a state of minute division, and impressing upon them, in
this state, a tendency to motion, places them in a favourable con-
dition for the study of their molecular constitution, and of the
relations which connect this constitution with electricity and
magnetism. The struggle that takes place between cohesion
and the expansive force of the electric current, the reduction of
matter to the molecular state, and the form and nature of the
deposits resulting therefrom, are so many phaenomena capable
of throwing light on the obscure subject of molecular physics.

The few preceding remarks suffice to give some idea of the
extent of an investigation embracing the whole range of
experimental research on the voltaic arc under its various
aspects, which I am far from pretending to have attempted.
I shall confine myself at present to a few details, and especially
to such as exhibit the action of magnetism on the voltaic arc,
and on those bodies which transmit interrupted currents. I
shall begin by describing some particular phaenomena which
I observed during my study of the voltaic arc under various
circumstances, while employing different substances as elec-
trodes, both in the air and in a vacuum ; I shall then proceed
to examine the action of a powerful electro-magnet on this
voltaic arc, and I shall conclude by describing some remark-
able experiments also illustrating the influence of magnetism
on conducting bodies, of whatever nature, traversed by inter-
rupted currents.

§ 1. Some Phccnomena coyicerning the Voltaic Arc.

Davy was the first who produced the phsenomenon of the
voltaic arc with two points of charcoal. More recently,
Messrs. Grove-l^ and Danlellf employed with success the
points of different metals, and arrived at interesting results:
I also published some experiments I made on the voltaic arc J
in 184-1. Subsequently, MM. Fizeau and Foucault observed
some remarkable facts of the same kind on the occasion of an
investigation into the intensity of the light emitted by charcoal
in the experinient of Davy §. The researches made up to the
present time, have already led to many results, of which I
shall consider only the most important.

1. That the voltaic arc may be produced, a pile of greater
tension is required than that which is necessary for the ordi-

* Bibl. Univ. June 1 840, i. 27. p. 387- 1' ^rck. deV Elect, torn. i. p. AQZ.
X Arch, de V Elect, torn. i. p. 262. § Ibid. toni. iv. p. 311.

M. De la Rive's Researches on the Voltaic Arc. 323

nary calorific and electro-chemical phaenomena. The neces-
sity of this condition proves the great resistance presented to
the passage of the electric current by the minutely divided
matter, whatever it may be, which connects the two poles.

2. The luminous arc cannot exist, unless contact be pre-
viously made between the electrodes, and unless these, or at
least one of them, be terminated at the point of contact by
points fine enough to produce in them an increase of tempera-
ture. When this increased temperature is once produced, we
may, by separating the electrodes gradually and with pre-
caution from each other, obtain the luminous arc, the length
of which will depend on the intensity of the pile. Daniell
discovered the important fact, which was confirmed by M.
Van Breda in a very recent investigation inserted in the
Comptes Rendus de V Academie^^, that without contact having
taken place, the luminous arc may be produced between two
electrodes placed very near together, by causing the discharge
of a Leyden jar to pass between them : this is owing to the
discharge being always attended by the transference of highly
diffused matter, which closes the circuit during the instant of
time necessary for the formation of the arc.

3. Tiie enormous elevation of temperature which accom-
panies the production of the luminous arc, is also manifested
in the electrodes, especially in the positive ones, which become
much more strongly healed than the negative.

4. Matter is thus transported from the positive electrode to
the negative, a fact which may be verified with electrodes of
all kinds, but particularly with those of charcoal.

5. The various phaenomena presented by the voltaic arc,
are modified to a greater or less extent by the nature of the
electrodes and by that of the surrounding medium. Thus
Mr. Grove adduces facts from which it appears that the pre-
sence of oxygen is necessary in most cases to produce a very
luminous and brilliant arc. It results also from his experi-
ments, as well as those of other philosophers, that when two
different substances are made use of for the electrodes, it is
not a matter of indifference which of the two is placed at the
positive pole.

I now proceed to my own researches. I commenced by
studying the production of a luminous arc between a plate and
a point of the same material in air, and in vacuo. By means
of a micrometer screw I was able to make the point recede
from the plate very gradually, and judge of their mutual di-
stance with great precision. The limit of distance beyond
which the luminous arc ceases to appear, is constant for the

[* See also p. 638 of the December Number of this Journal for 1846.J

Y2

324- M. De la Rive's Researches on the Voltaic Arc.

same plate and the same point: when, however, the plate
communicates with the positive pole, it is in general double
that which it is when the point communicates with the same
pole. But in proportion as the strength of the pile is greater,
the difference is so much the smaller.

With respect to the absolute amount of this distance, it is
very variable, depending on the strength of the pile, on the
nature and molecular state of the electrodes, and on the time
occupied in the experiment. Thus, with a Grove battery
composed of fifty pairs of plates sixteen square inches in sur-
face, it is two or three times greater than with a pile of seventy
elements of two or three square inches. With metals easily
fused or oxidized, as zinc and iron, it is much greater than
with platinum or silver. The duration of the phtenomenon
influences the result, inasmuch as the high temperature of the
electrodes allows them to be drawn asunder to a greater di-
stance without breaking the arc. The same effect may be
produced by heating them artificially, by means of a spirit-
lamp. It is evident from what I have said that the length of
the luminous arc has a relation to the greater or less facility
which the substances composing the electrodes possess of being
segregated, a facility which may depend upon their tempera-
ture diminishing their cohesion, upon their tendency to oxi-
dize (which produces the same effect), upon their molecular
state, and lastly upon their peculiar nature. Carbon derives
from its molecular constitution, which renders it so friable,
the property of being one of the substances which produces
the longest luminous arc.

The deposits of the transported matter, form upon the plate,
when it is negative and the point positive, a species of very
regular ring, the centre of which is the projection of the point
upon the plate. This takes place equally, whether the plate be
vertical or horizontal, plainly indicating a determinate direc-
tion in the transfer of the substance from die positive to the
negative electrode; in the air and with metallic electrodes,
the deposits always consist of the oxidized dust of the meial,
of which the positive electrode is composed.

I shall here enter into some details. A plate and a point
of platinum have been used as electrodes in a vacuum, in air
and in hydrogen. In a vacuum with a Grove battery of fifty
pairs of plates, which had previously been used, I had only a
very feeble effect, and particularly when the plate served as
the positive electrode. The point was hardly removed a mil-
limetre* from the plate when the arc broke ; to re-establish it,
it became necessary to renew the contact between the point
* 1 millimetre = 0-03937 inch.— Trans,

M. De la Rive's Researches on the Voltaic Arc. 325

and the plate, by touching another point of the plate, the first
point which was touched appearing to have undergone such
a modification as to prevent the re-formation of the arc. The
same effect is produced when the experiments are made in the
air, but it ceases when the power of the battery is increased :
this is probably due to an augmentation of cohesion consequent
on the increase of temperature in that part of the plate which
acts as the positive electrode. Besides, when the experiment
is made in air, the voltaic arc is more marked and of greater
length than when it is made in vacuo, at least if the battery be
weak; for when the battery is powerful, composed, for ex-
ample, of fifty pairs of plates freshly charged, it appeared to me
that the contrary obtained. 1 did not, however, perceive any
great difference ; but the vacuum in which I experimented was
far from being perfect; it was that of a pneumatic pump, en-
closing therefore highly rarefied air.

In the latter case, that is to say, with the pile composed of
fifty pairs strongly charged, and in highly rarefied air, a bluish
spot, perfectly circular and presenting the appearance of a
coloured ring of Nobili, was formed on the plate of platinum
when it served as the positive electrode. The same spot ap-
peared in atmospheric air, but its diameter was one-half less,
and its colours much less vivid. In hydrogen, no coloured
spot was formed ; its formation is therefore evidently the result
of the oxidation of the platinum at a high temperature when
acting as a positive electrode in the ordinary atmosphere, and
still more so, perhaps, in rarefied air*. When the same plate
of platinum was made use of as a negative electrode, the point
being positive, it became covered with a white circular spot,
formed of a vast number of minute grains of platinum, which,
having been raised to a high temperature, remained adhering
to the surface. The white spot, like the blue one, was much
larger in rarefied air than in a vacuum. If the experiment be
prolonged for a minute or two when the plate is negative, the
rod of platinum terminating in a point,'which is positive, soon
becomes highly incandescent ; its end is fused and falls on the
plate in the form of a perfectly spherical globule. When the
plate is positive and the point negative, the latter is less heated,
and does not become fused ; but the plate, unless it be very
thick, is liable to be perforated : besides, as may easily be

* This effect may possibly have been owing to the action of the oxygen
brought by the voltaic current into that particular state which Schonbein
first described under the name oi ozone. Indeed, in this state the oxygen
may attack those metals which are supposed to be inoxidizable ; and M.
Marignac and I have shown that this may be effected by causing a succes-
sion of electric discharges to pass through the oxygen, even when very dry,
with which the phaenomenou of the voltaic arc has a great resemblance.

326 M. De la Rive's Researches on the Voltaic Arc.

imagined, the phsenomenon lasts much longer in the latter
Ciise. The light is less brilliant, but it is accompanied by a
reflexion of a superb blue, which may be seen when the ex-
periment is made in the interior of a bell, whether the air
be rarefied or not. This blue reflexion is observed on the side
of the bell, and is to be seen whatever may be the nature of
the electrodes, or the colour of the light to which these give
rise in the centre of the bell ; only when this central light is
very brilliant, it becomes slightly paler by the effect of contrast.

I substituted for the platinum point a point of coke, but
the plate of platinum remained ; this being positive and the
point negative, I obtained a luminous arc more than double
the length of the arc produced by the point of platinum.
With respect to the arc, instead of its being a cone of light,
having its base on the plate and its apex at the point, as was
the case when the latter was platinum, it was composed of a
multitude of luminous jets diverging from different points of
the plate, and tending to various parts of the point of coke.
This fact shows clearly the influence that may be exercised by
the negative electrode, the function of which is very far from
being a merely passive one. Let me add, that although the
strength of the pile was precisely ihe same as when the point
was of platinum, not only was the luminous arc much longer
with the point of coke, but the heat developed in the plate of
platinum was so much greater that it was soon melted and
perforated. The coke being positive and the plate negative,
the length of the arc was less than in the preceding case, and
particularly so in air, where it was sensibly less than in a
vacuum. The heat generated was however still very great,
the point of coke becoming quickly incandescent throughout.
I ought to add, that with the point of coke, the luminous arc
was so brilliant that the blue light which I have mentioned
almost entirely disappeared, which was not the case with any
other kind of point.

Leaving the plate of platinum, I adjusted a zinc point. The
effects were most brilliant, but of short duration, the point
speedily melting. In common air, a deposit of white oxide
was precipitated upon the platinum plate; in highly rarefied
air (the vacuum of an air-pump), a black deposit was formed :
in both cases it communicated with the positive pole. An
iron point being substituted for that of zinc, equally produced
in common air a brownish-red deposit of oxide of iron, and in
rarefied air a deposit of black oxide.

I call the attention of chemists to these two facts, as well as
that of the oxidation of the platinum at a high temperature in
rarefied air. They appear to prove the influence which the
state of greater pr iess density of the surrounding oxygen may

M. De la Rive's Researches on the Voltaic Arc, 327

exert on the phsenomenon of oxidation and on the nature of
the oxide formed. A plate and a point of soft iron were used
as positive and negative electrodes, both in a vacuum and in
the atmosphere ; the same results appeared with a plate and
a point of silver, a plate and a point of copper, and a plate
and point of argentane*. The blue light was perceived in
all the experiments ; coloured circles were likewise seen on all
the plates when they had acted as positive electrodes in rare-
fied air. The silver and copper plates presented in this case
very decided cavities, caused by the passage of the matter
from the positive to the negative pole. The points became
incandescent throughout when they served as positive elec-
trodes; whereas when negative, they were heated only at their
extremities. The copper point when positive became isolating
at its extremity, and it was necessary to excite it by friction
in order to renew the experiment. This circumstance is pro-
bably attributable to the formation of a thin film of oxide.
The point and plate of copper gave out a luminous arc of a
beautiful green light, which contrasted in a remarkable manner
with the blue reflexion visible in this, as in the other experi-
ments. Mercury was likewise employed, both as a positive
and negative electrode. In a vacuum as well as in atmospheric
air, the luminous effect was most brilliant. The mercury was
excessively agitated, rising up in the form of a cone when it
was positive, and sinking considerably below the positive point
when it was negative. The quantity of vapour thrown off by
the mercury during this experiment filled the bell so quickly
that it was not easy to observe the details.

I shall terminate this section by stating a fact which appears
to me to be important; it is the influence which the nature of
the metallic points forming the electrodes exercises on the
temperature which they acquire in relation to the production
of the voltaic arc. If the two points are of the same metal,
both platinum, or both silver, the positive one alone becomes
incandescent throughout its whole length. If the silver point
be positive and that of the platinum negative, the latter be-
comes incandescent, and the silver one is much less heated.
Thus, when the voltaic arc is formed, the circuit must be re-
garded as completed, and then it is those parts of the circuit
which present the greatest resistance to the current which
become the hottest ; at first it is that portion forming the arc
itself, and then, in the rest of the circuit, the metal which is
the worst conductor. But if the conductors be of the same
material on both sides of the arc, or if there be only a slight

* An alloy of copper and nickel : also known by the names oi pachfong
and melchior.

328 M. De la Rive's Researches on the Voltaic Arc.

difference of conductibility between them, then the develop-
ment of heat, instead of being uniform, as it might appear it
ought to be, is much greater on the positive side. This im-
portant fact evidently proves that this portion of the circuit
has to resist a much more energetic action than that which the
other side experiences ; a fact which is confirmed by the mo-
lecular segregation accompanying this action at the positive
electrode. This want of resemblance in the phaenomena pre-
sented by the two electrodes, although placed in conditions
entirely synmietrical, deserves to be taken into serious consi-
deration, for it may throw light upon the nature of the electric
current, and upon the link which unites it with the molecular
state of the bodies through which it is transmitted.

§ 2. Influence ()f Magnetism on the Voltaic Arc.

Davy was the first who observed that a powerful magnet
acts upon the voltaic arc as upon a moveable conductor, tra-
versed by an electric current ; it attracts and repels it, and
this repulsion and attraction manifests itself by a change in
the form of the arc. Even the action of the magnet may, as
I have found, break the arc by too great an attraction or re-
pulsion exerted upon it, causing the communication which the
transmitted particles establish between the electrodes to cease.

The action which I have just mentioned is not the only one
which magnetism exerts on the voltaic arc. I have already
stated the curious fact, that if two points of soft iron acting as
electrodes, be both placed within a helix formed of thick cop-
per wire of several coils, the voltaic arc developed between
the two points of iron ceases the moment a strong current is
passed through the wire of the helices, and reappears if this
current be arrested before the points have become cold. The
arc cannot be formed between the two iron points when they
are magnetized, whether by the action of the helices, or by that
of a powerful magnet, unless they be brought much nearer to
one another, and the appearance of the phaenomenon is then
entirely different. The transported particles appear to dis-
entxage themselves with difficulty from the positive electrode,
sparks fly with noise in all directions, while in the former case
it was a vivid light without sparks, and without noise, accom-
panied by the transfer of a liquid mass, and this appeared to
be effected with the greatest case. It is of little moment with
respect to the result of the experiment, whether the two rods
of magnetized iron present to that part of their extremities
between which the luminous arc springs, the same magnetic
poles or different poles.

The positive electrode of iron, when it is strongly magnet-

M. De la Rive's Researches on the Voltaic Arc. 329

ized, produces, the moment that the voUaic arc is formed be-
tween it and a negative electrode of whatever nature, a very
intense noise, analogous to the sharp hissing sound of steam
issuinjif from a locomotive engine. This noise ceases simul-
taneously with the magnetization.

For the purpose of better analysing these different phaenc
mena, I placed an electro-magnet of large dimensions and great
power in such a manner as to enable me to place on each of
its poles, or between them, different metals destined to form
one of the electrodes of the pile, while one point of the same
metal, or another substance, acted as the other electrode. I
have aUke employed as electrodes, placing them in the same
circumstances, two points of the same metal or of different
metals. The following are the results which I have obtained.
A plate of platinum was placed on one of the poles of the
electro-magnet, and a point of the same metal was placed ver-
tically above it; the voltaic arc was produced between the
plate and the point, the plate being positive and the point
negative. As soon as the electro- magnet was charged, a sharp
hissing was heard ; it became necessary to bring the point of
the plate nearer to enable the arc to continue, and the bluish
circular spot which the platinum plate presented, became
larger than when the experiment was made beyond the influ-
ence of the electro-magnet. The plate was made negative,
and the point positive; the effect was then totally different;
the luminous arc no longer maintained its vertical direction
when the electro-magnet was charged, but took an oblique
direction, as if it had been projected outwards towards the
margin of the plate; it was broken incessantly, each time ac-
companied by a sharp and sudden noise, similar to the dis-
charge of a Leyden jar. The direction in which the luminous
arc is projected, depends upon the direction of the current
producing it, as likewise on the position of the plate on one
or other of the two poles, or between the poles of the electro-
magnet. A plate and a point of silver, a plate and a point of
copper, and generally a plate and a point of any other metal,
provided it be not metal too easily fused, present the same
phaenomena.

Copper, and still more silver, present a remarkable pecu-
liarity. Plates of these two metals retain on their surfaces the
impression of the action that took place in the experiments
just described. Thus, when the plate is positive, that portion
of its surface lying beneath the negative point presents a spot
in the form of a helix; as if the metal melted in this locality
had undergone a gyratory motion around a centre, at the same
time that it was uplifted in the shape of a cone towards the

330 M. De la Rive's Researches oti the Voltaic Jrc.

point. Moreover, the curve of the helix is fringed throughout
by minute ramifications, precisely similar to the tufts which
mark the passage of positive electricity in a Leyden jar. When
the plate is negative and the point positive, the marks are
totally different, being merely a simple point, or rather a circle
of a very small diameter, whence proceeds a line more or less
curved, forming a kind of tail to the comet, of which the small
circle might be the nucleus: the direction of this tail depends
upon the direction in which the luminous arc has been pro-
jected.

When, instead of a plate and a point, two points are used
for electrodes, it is evident that no visible trace of this phseno-
menon can be obtained ; but both the sharp hissing and the
detonations may be produced, which latter are sometimes so
loud as to bear a resemblance to distant discharges of mus-
ketry. For this the electro-magnet must be very powerful,
and the current which produces the arc very intense. I had
observed that when 1 took for a positive electrode a point of
platinum, and for a negative electrode a point of copper, and
placed them between the two poles of the electro-magnet, the
production of the voltaic arc between the two poles was ac-
companied by a sharp hissing noise ; whereas in the opposite
case, the copper being positive, and the platinum negative,
the detonations were heard, attended by a frequent breaking
of the arc. On examining this phasnomenon more closely, I
perceived that the fact I have just meTitioned was due to the
platinum becoming heated much more rapidly than the copper
when they were employed as electrodes in producing the vol-
taic arc ; and I have satisfied myself that in order to obtain
the hissing sounds, it is necessary that the positive electrode
should be at a sufficiently high temperature to experience a
commencement of liquefaction ; for without this condition,
only a series of detonations are heard. The hissing would be
the result of the easy and continuous transport of matter more
or less liquefied from the positive electrode, whilst the deto-
nations would be the effect of the resistance opposed by the
same matter to the disintegration of its particles when it is not
sufficiently heated. Numerous experiments made with metal-
lic points^ whether of the same or different natures, as silver,
iron, brass, as also platinum and copper, some of which be-
come heated sooner than others under the same circumstances,
have quite confirmed me in this view of the subject. It is
merely necessary to be careful, in order to produce the hissing
noise, to maintain as much as possible the continuity of the
arc when once the positive electrode becomes incandescent;
while, on the other hand, to obtain the detonations, one of the

M. De la Rive's Researches on the Voltaic Arc. 331

electrodes must be held in the hand, and then the arc fre-
quently made and broken without waiting till the metallic
points acquire too high a temperature.

It remains now to be considered why the influence of pow-^
erful magnetism, such as that exerted by the electro-magnet,
is necessary for the production of these sounds, which are not
heard in the ordinary experiment of the voltaic arc. This can
arise only from the change which the magnet produces in the
molecular constitution of the matter of the electrode, or rather
in the highly diffused matter which forms the voltaic arc.
This action is besides shown by the shortening of the arc, and
by the remarkable difference which it presents in its appear-
ance; it is therefore not surprising that it should also be capable
of producing a phasnomenon such as sound, which essentially
depends on the variations in the molecular state of bodies.
This view of the subject appears to me to deserve very par-
ticular attention : the results at which I have arrived, in pur-
suing it more closely^ form the subject of the following section.

§ 3. Irifluence of the permanent actioji of Magnetism on con-
ducting bodies traversed hy interrupted electric currents.

Faraday's brilliant discovery of the action exerted by mag-
netism on a ray of polarized light, -when that ray traverses a
transparent body submitted to the action of a powerful electro-
magnet, had no sooner been announced by its illustrious
author, than the majority of philosophers saw in it a proof
that magnetism, when at a high degree of intensity, has power
to modify the molecular constitution of all bodies. They
consequently attributed the phsenomenon observed by Faraday,
not to the direct action of the electro-magnet on the polarized
ray, but to the modification effected by this action on the mo-
lecular constitution of the substance traversed by the ray. I
was of this opinion, and communicated it to Mr. Faraday,
who alludes to it in his memoir. Desirous, however, of found-
ing this opinion on facts of a different kind, I asked myself if
it were not possible to find in the electric current, an agent
capable of performing the same function for opake conducting
bodies that polarized light does for transparent ones. I had
stated in my paper on the sound emitted by iron wires tra-
versed by interrupted electric currents, that the nature as well
as the intensity of the sounds were singularly modified by
the molecular .slate of the wire submitted to the experiment.
I had particularly mentioned the influence of temper and an-
nealing, of greater or less tension, and of temperature. 1 had
shown that iron wire, when under the influence of an action
which renders it magnetic, does not emit the same sound as

332 M. De la Rive's Researches on the Voltaic Arc.

when it is in its natural state. Finally, by modifying, through
the agency of heat, the molecular arrangement of some metals,
such as platinum and brass, I had succeeded in obtaining from
them, during the passage of the interrupted current, sounds,
which, though feeble, were yet distinct.

The preceding reflexions tended to confirm me in my
opinion, that sounds produced under die influence of mag-
netism in the experiments on the voltaic arc, are owing to a
molecular modification effected by the action of the magnet,
and the more so inasmuch as the voltaic arc may be regarded
as produced by a succession of interrupted currents, following
each other with extreme rapidity, rather than by a perfectly
continuous current. I accordingly took bars of other metals
besides iron, as of tin, zinc, lead, bismuth, &c. I placed them
on the poles of the electro-magnet, and 1 caused an interrupted
current from a Grove's battery of from five to ten pairs to
traverse them. They emitted no sound as long as the electro-
magnet was not magnetized ; but as soon as it was, sounds were
very distinctly heard, composed of a series of strokes corre-
sponding to the interruptions of the current, and analogous to
that produced by a toothed wheel. The bars were eighteen
inches long, and from nine to ten lines square. Rods of
copper, platinum, and silver produced a similar effect ; a rod
of iron did not emit a much louder sound under the influence
of the magnet than it did when not exposed to this action.

What appeared to me most remarkable was, to find lead, a
body so little elastic, yield a sound as powerful as those pro-
ceeding from the other metals, when placed under the same
circumstances. The position of metallic bars with respect to
the poles of the electro-magnet did not in any way modify the
result of the experiment ; they might be placed axially, that
is to say, in the direction of the poles, or equatorially, that is,
across the poles; the effect remained the same, being merely
weakened as the distance between the bar and the poles in-
creased. In order to hear the sound distinctly, when not very
powerful, it was sufficient to establish a communication be-
tween the metallic bar and the ear by means of a wooden rod.
In this manner the sound was not unfrequently heard prolonged
some seconds, though growing constantly feebler, until it ceased
entirely, after the source of magnetism had been withdrawn
from the electro-magnet. Mr. Faraday has remarked an
analogous fact in the action of the transparent medium on the
polarized ray, an action which does not cease immediately
with the magnetism of the electro-magnet. Is this prolonga-
tion owing to the magnetization of the electro-magnet not
ceasing in a sudden manner ; or to its return to its primitive

M. De la Rive's Researches on the Voltaic Arc. 333

molecular state not taking place instantaneously in the sub-
stance submitted to its action ? This question 1 am unable to
decide. I incline, however, rather to the latter of these ex-
planations, seeing that the effect is not equally perceptible in
all bodies, and that it is, for example, more sensible in a bar
of bismuth than in one of copper.

It is needless to remark that the calorific action of the cur-
rent could not have any influence on the production of the
phaenomenon, since there could have been no development of
heat, on account of the dimension of the bars compared with
the force of the current. Besides, if the expansion arising
from the heating of the body traversed by interrupted currents
had caused the sound, the effect would have been produced
equally, whether the bar had been under the influence of the
magnet or not. This last remark applies equally to the fol-
lowing experiments, as to the preceding.

The intensity of the sound appears to depend much lesson
the nature of the substance submitted to the experiment, than
on its form, its volume, and its mass. Tubes of platinum, of
copper, and of zinc, emitted more marked sounds than massive
cylinders of the same metals. I wound a leaden wire in the
form of a helix round a cylinder of wood ; I did the same with
a very fine platinum wire, and also with copper, zinc, and tin
wires, taking care to place the coils of the helices so far apart
that each should be isolated. Placed like bars and tubes,
whether in the direction of, or across the poles of the electro-
magnet, these helices emitted very powerful sounds when, the
electro-magnet being charged, they were traversed by the in-
terrupted current. It was particularly surprising to hear the
lead wire emit so strong a sound. A helix constructed with
copper wire, covered with silk, and composed of several coils
wound round each other, emitted a very intense sound ; it
also emitted one, but much feebler, under the action of the
electro- magnet.

It is almost needless to remark, that in all these experiments
an ordinary magnet produces the same effect as an electro-
magnet. But what is more interesting, is to replace the action
of the electro-magnet by that of a helix traversed by a strong
continuous current, in the axis of which helix is placed the
bar, the tube or the coiled wire, through which the interrupted
current is transmitted. Experiments have shown me that in
this case the results are the same ; the intensity of the sounds
is not very different, especially when tubes and wires coiled as
helices are used.

If, between the exterior helix and the metal submitted to
the action, a tube of soft iron is placed, the effect is a little

334 M. De la Rive's Researches on the Voltaic Arc.

heightened : it is neither increased nor lessened when the tube
is of' copper, only in this case another sound is heard which
seems to proceed from the copper tube. This tube, however,
is not traversed by a current, but it is probably acted upon
by the currents of induction, which the interrupted currents
traversing the conductor placed in the axis of the helix pro-
duce in it. I constructed a double helix formed of two thick
copper wires covered with silk and coiled, each forming several
circumvolutions, the one exterior to the other. In making a
continuous current pass through the exterior wire, and an
interrupted current through the interior one, I heard a re-
markably intense sound. In the reverse case, the sound ex-
isted, but was much weaker. This fact is evidently connected
with the known property of helices traversed by electric cur-
rents exercising scarcely any magnetic influence exteriorly,
whilst in the interior this action is very energetic.

Metals and solid bodies are not the only substances which
produce the phaenomena I have just described; all conducting
bodies, whatever may be their state or their nature, appear to
be capable of producing them. Thus, I have observed them
with pieces of charcoal of all kinds and shape. Mercury also
produces them in a very marked manner. I have inclosed
mercury in a tube of glass an inch in diameter, and ten inches
long : the tube was completely full and closed with care, so
that the mercury could have no motion. As soon as it was
traversed by an interrupted current, transmitted by means of
two platinum wires, and the electro-magnet or the helix was
made to act upon it, a sound- was heard remarkable for its
intensity. When the mercury was placed in an open trough,
instead of being inclosed in a tube, it likewise produced a
sound, and in addition there was seen on its surface an agita-
tion or vibratory motion, very different from the gyratory
motion observed by Davy, which appears under the influence
of the poles of a magnet when traversed by a continuous cur-
rent.

Dilute sulphuric acid, and what is even better, salt water,
were successively put in a capsule of platinum placed on one of
the poles of an electro-magnet. A point of platinum immersed
in the liquid, served, together with the capsule, to send an
interrupted current through it. A sound was again heard,
but less distinct, on account of the noise produced by the
disengagement of the gas : still it was so clear that no doubt
could be entertained of its existence.

It may perhaps be thought that in the experiments I have
just described, the sounds are produced by the mechanical
action of attraction or repulsion exerted by the electro-magnet

M. De la Rive's Researches on the Voltaic Arc. SS5

on the substance traversed by an interrupted current, and
that, consequently, magnetism has no more share in the phae-
nomenon than a finger might be supposed to have, when
pressing on a sonorous cord. The simple description of the
experiments shows this interpretation to be inadmissible. In
the first place, the sound is the same wit'fe the wires in a helix,
whether these wires be stretched or not, or whether they be
of lead, platinum, or brass. Besides, how could this account
for the sound produced in large masses, especially in liquids,
such as mercury, and for the fact, that the position of the
conductor traversed by the interrupted current with regard to
the poles of the electro-magnet does not exert any influence
on the phsenomenon? Further, it must be remarked that the
sound in question is not a musical sound, such as would be
produced by a string or mass made to vibrate by a cause acting
exteriorly at its surface; it is a series of sounds corresponding
exactly to the alternations of the passage of the current; like
a species of collision of the particles amongst themselves.
Thus, the phsenomenon is molecular; and it leads to the de-
monstration of two important principles.

The first principle is, that the passage of the electric current
modifies, even in solid bodies, the arrangement of the par-
ticles ; a principle which I have already deduced from the ex-
periments contained in my preceding memoir on this subject.
The second principle is, that the action of magnetism, under
whatever form it may be exerted, modifies alike the molecular
constitution of all bodies, and that this modification lasts as
long as the cause producing it endures, and only ceases with
it. What is the nature of these two modifications ? This is
what we must endeavour to investigate and to ascertain. I pur-
pose to engage in this inquiry, and indeed I have already made
some attempts of which it would, however, be premature to
give any account. I shall confine myself at present to a single
remark, which does not appear to me to be devoid of interest :
it is, that the influence of magnetism on all conducting bodies
seems to impress on them, as long as it lasts, a molecular
constitution similar to that which iron, and generally all bodies
susceptible of magnetism possess naturally ; for it developes
in them the property of producing, when traversed by inter-
rupted currents, sounds identical with those emitted also by
iron and other magnetic bodies when transmitting these cur-
rents, but produced in these last without requiring the action
of a magnet.

[ 336 ]

LII. Analyses of the Ashes of Rough Brown Sugar ajid Mo-
lasses. By Thomas Richardson*.

TOURING some inquiries which I had occasion to make in
■*-"^ the manufacture of an artificial manure for the sugar-
cane, it was necessary to know the composition of the ash of
coarse brown sugar and molasses as imported into this country.
The results may be interesting to some of the readers of your
Journal.

I. Rough Brown Sugar.

20648 grs. in its ordinary state left 2*74' grs. ash =1*33
per cent. 1 43*05 grs. of ash furnished 18*16 grs. silica, and
4*75 grs. carbonic acid.

28-61 grs. of ash furnished 8-43 grs. SOg BaO =2*89 grs.
sulphuric acid.

28'61 grs. of ash furnished 0'19 gr. oxide of copper.

28*61 grs. of ash furnished 1*95 gr. peroxide of iron.

28-61 grs. of ash furnished 7*00 grs. COg CaO =3-92 grs.
lime.

28*61 grs. of ash furnished 7*81 grs. P052MgO =2*86 grs.
magnesia.

16*12 grs. of ash furnished 7*36 grs. CI2 Ag = 1*96 gr.
chlorine.

28*61 grs. of ash furnished 28*61 grs. chlorides of alkaHes,
and this yielded 33*88 grs. of the double chloride of platinum
and potassium =10*34 grs. chloride of potassium = 6*53 grs.
potash, leaving 4*16 grs. chloride of sodium = 2*20 grs. soda.
The ash also contained a trace of oxide of manganese.

The result of the analysis is therefore —

Potash 22*84

Soda 7-69

Lime 13-69

Magnesia 1000

Peroxide of iron . . . . 6*11
Oxide of copper .... '66
Oxide of manganese . . . trace

Sulphuric acid 10-12

Silica 12*68 .

Carbonic acid 2*32

Chlorine 12-20

98-31
* Communicated by the Author.

analyses of the Ashes of Rough Brown Sugar and Molasses. 337

Omitting the carbonic acid and combining the chlorine with
the sodium and potassium, we obtain the following composition
in 100 parts: —

Potash 19*42

Lime 14.-67

Magnesia 10'72

Peroxide of iron .... 6'55
Oxide of copper .... '71
Oxide of manganese . . . trace
Chloride of potassium . . . 8'03
Chloride of sodium . . . 15*46

Sulphuric acid 10'85

Silica 13-59

lOO'OO

II. Molasses.

Great dijfficulty was experienced in incinerating the bulky
charcoal mass left by boiling down the molasses. Part of the
oxide of iron and sulphuric acid were decomposed ; and this
accounts for the excess in the analysis, as these substances
were obviously in part twice estimated. 477'77 grs. left 17*21
grs. ash =3*60 per cent.

64*27 grs. of ash furnished 6*46 grs. carbonic acid.

64*27 grs. of ash furnished 8*65 grs. charcoal, containing
'55 gr. peroxide of iron and *26 gr. lime.

64*27 grs. of ash furnished 1*02 gr. silica.

21*423 grs. of ash furnished "28 gr. peroxide of iron.

21*423 grs. of ash furnished 3*83 grs. C02CaO = 2-14grs.
lime.

21 -423 grs. of ash furnished 5*34 grs. PO5 2MgO= 1 -95 gr.
magnesia.

21*423 grs. of ash furnished chlorides of alkalies 14*91 grs.,
and this gave chlorides of platinum and potassium 32*88 grs.
= Cl2 K 10*34 grs., leaving ClgNa 4*57 grs., = potash 6*53 and
soda 2*42 grs.

12-62 grs. of ash furnished 2*38 grs. SOgBaO =*818 gr.
sulphuric acid.

18*405 grs. of ash furnished 9*92 grs. Clg Ag=2*453 grs.
chlorine.

The ash also contained traces of oxides of copper and man-
ganese.

Collecting these results we obtain the following composition :
Phil, Mag, S. 3. Vol. 31 . No. 209. Nov, 1 847. Z

SS8 Prof. Loom Is o« the determination of differences

Potash 30-50

Soda 11-30

Lime 10*42

Magnesia 9-13

Peroxide of iron .... 2-15

Oxide of copper .... trace

Oxide of manganese . . . trace

Sulphuric acid ..... 6-48

Chlorine 13-33

Carbonic acid 10-04

Silica 1-58

Charcoal 11-78

106-71

Omitting the charcoal and carbonic acid, and combining
the chlorine as before, we have as follows : —

Potash ....

Lime

Magnesia . . .
Peroxide of iron .
Oxide of copper .
Oxide of manganese
Chloride of potassium
Chloride of sodium
Sulphuric acid . .
Silica ....

36-23
12-72
11-14

2-62
trace
trace

1-58
25-87

7-91

1-93

100-00

The molasses and sugar came from the same manufactory,
and were made from the same sugar-cane.

Lin. Letter from Professor Loomis of the New York Uni-
versity/ to Lieut. -Colonel Sabine, Foreign Secretary/ of the
Royal Society^ on the determination of differences of Longi-
tude made in the United States hy means of the Electric
Telegraph, and on projected observations for investigating
the Laws of the great North American Storms.

Dear Sir, New York University, Aug. 2, 1847.

I HAVE been for some time engaged upon a work in which
)'0U may perhaps feel some interest, — it is the exact de-
termination of the difference of longitude between New York,
Philadelphia and Washington, by means of the magnetic

of Longitude hy means of the Electric Telegraph. 339

telegraph. Morse's magnetic telegraph has been in operation
between these places for a considerable time, and Prof. Bache
proposed to use the line for the transmission of signals for the
comparison of our local times for the use of the coast survey.
Accordingly I erected a temporary observatory last season as
near to this city as I conveniently could, and set up a transit
instrument and clock. A wire was then carried from my ob-
servatory to the telegraph-office, thus connecting me with the
regular Philadelphia line. A wire was also carried from the
Philadelphia telegraph-office to the High School Observatory
in Philadelphia, and another wire was carried from the Wash-
ington telegraph-office to the National Observatory. Thus
three observatories, at New York, Philadelphia, and Wash-
ington, were in telegraphic communication ; and having de-
termined our local times by astronomical observations, we
only needed some signal which could be heard simultaneously
at the three places. This signal was affiarded by the click of
a magnet in the usual mode of telegraphic communication.
Our plan of operation is as follows : — At ten in the evening,
when the usual business of the telegraph company is concluded,
our three observatories are put in communication with each
other. After corresponding with each other long enough to
ascertain that everything is in good order. New York com-
mences giving clock signals. At the commencement of a
minute by my clock I strike the key of my register, and a click
is heard simultaneously at New York, Philadelphia, and
Washington. The three observers record the time each by
his own clock. At the expiration of 10^ I give a similar sig-
nal, and all three record; after another 10® I do the same,
and so on to twenty signals. Having waited one minute,
Philadelphia repeats the same series of signals, and all record
the time. We then wait another minute, and Washington
repeats the same signals. Thus we have obtained sixty com-
parisons of our clocks, which will give our difference of lon-
gitude with as great accuracy as we can determine our local
times.

In our first experiments we met with a great many disap-
pointments, as might have been anticipated from the novelty
and delicacy of the undertaking ; but we have triumphed
over them all. Onjive different nights we have transmitted
good signals back and forth, and we propose to continue the
comparisons until a further degree of accuracy is not to be
expected. The errors of our clocks have not yet been rigo-
rously computed, and we have not obtained final results ; but
we have made sufficient comparisons to know that the results
of different nights agree remarkably well with each other. I

Z2

34:0 On the determination o/^ differences of Longitude.

think the extreme discrepancy of different nights' work will
amount to only a small fraction of a second. It appears to
me that this mode of determining differences of longitude
must supersede every other method between places which are
connected by a telegraphic wire. The observations can be
repeated indefinitely, and I think the longitude can be deter-
mined with a precision fully equal to that of the local times.
I presume the same cannot be said of any other method yet
practised. I have not heard of this method being tried in
any part of Europe, though the application is very obvious.
Can you inform me of any such trials ?

In my former correspondence with you, and in my printed
papers, I have more than once alluded to the importance of a
combined movement in this country for meteorological obser-
vations. I am happy to say that the prospect of such a com-
bination is brightening. You are probably aware that the
Smithsonian Institution has been organized, and Prof. Joseph
Henry has been placed at the head of it. The plan of orga-
nization is to appropriate c§^l 5,000 a-year to the promotion
o^ original researches. Prof. Henry is disposed to include in
this plan a grand meteorological campaign, to continue for
three years, — to cover the entire area of the United States
with the greatest possible number of observers : we want 300,
and I think they could be obtained. I have been drawing up
a paper for Prof. Henry which will be placed before the
Smithsonian Institution, and also before Congress the coming
winter. I think the plan will be carried into execution, pro-
vided we can obtain the co operation of the British Govern-
ment. You know from the papers which I have sent you that
our great storms frequently extend far to the northward of
the United States. When the centre of a storm travels along
the valley of the St. Lawrence, its margin often extends to the
Gulf of Mexico. Observations spread over the entire United
States would frequently include only half the area of a violent
winter storm ; and this is the class of storms from which most
is to be expected, because their phaenomena are most strongly
developed. Unless therefore we could obtain simultaneous
observations from the British possessions on the north of us,
we should feel that our observations were deprived of more
than half their value. Will you not see if the I3ritish Govern-
ment and the Hudson's Bay Company cannot be induced to
co-operate with us ? What I propose is, that at every go-
vernment station a register should be kept for a period of one,
two, or three years. I should hope 100 such stations could
be procured. The first cost of the instruments would not be

On the Algebraic Equation of the Fifth Degree, 341

great, and the expense of observing probably nothing at all.
If your government will co-operate, I think the Smithsonian
Institution will undertake the organization for the United
States.

With much respect I remain,

Yours truly,

Elias Loomis.

LI v. On the Algebraic Equation of the Fifth Degree.
By the Rev. Brice Bronwin*.

I

T appears that the resolution of equations of the fifth and
higher degrees into factors, one of which is of the second
degree, depends upon the solution of the proposed equation
itself. This circumstance appears to me deserving of notice,
as it seems to indicate the impossibility of solving such equa-
tions in finite terms. Suppose

x^+Ka;'^-\-Bx'^+Cx-\-Ti={pir^-\-ax'^-\-bx + c)'\ ,^.

{x^-ax+f) = 0. J

Multiplying the two factors, and comparing the result with
the first member, we find

Eliminating b and c from these, we have

2fl/^-(a3 + A« + B)/+D=0
/3-(a2+A)/2 + C/+Da=0.
From these we easily deduce

/2 _ (3^2 _^ A)/+ a'l + Aa2 + Ba + C = 0

(«3 + A« - B}/2 - (2Ca - D)/- 2Da^ = 0.

Eliminating /2 by 2a/^=(fl3 + Aa + B)/— D, we shall have
two equations, in whichywill be only of the first degree;
and then, by eliminating it from these, there results an equa-
tion in a of the tenth degree ; and it is obvious that/, c, and
b will be determined from a by simple equations.

Now let a„ ag, &c. be the roots of the equation in a, and
Oi'i, a^g, &c. those of (1.); then, since x'^—ax-\-f=-0 must con-
tain two of the last, we shall have

^5 = ^2"T""^3» <'^6"^^'*^2'^"'^4' ^7 ^=^ "^a "^ "^S* ^8 ^^ "^S "I" "^4' ( * \ ')
^=^3 + '2?6j «]o = a?4 + T5. -J

* Communicated by the Author.

342 The Rev. B. Bronwin on the Algebraic

To which we may add,

0 = ^1 +.r2 + a?3 + ^4 + ^5-

By eliminating .Tp Xc^, &c. from these, we find

3a5=— 2a4— 2cf3 + a2 + «iJ 3^6=— 2a4 + ag— SflTg + ^u
3fl7=«4— 2^3— 2a2 + «iJ 3«8= — 2a4 + fl'3 + «2~2^i>
3a9=C4— 2«3 + fl2~2ai, 3^10=^4+ ^3~2a2—2ai,

which may be verified by putting for a„ «2, &c. their values
in a-'i, .^2, &c. Therefore six of the roots «i, ag, &c. are linear
functions of the remaining four, and the equation in a of the
tenth degree is reducible to one of the fourth.

We also find

^l=-3 (^l + '^2 + «3 + «4)» •^•2=-3-(2«l — «2-^3 — «4)> I
^3=3- (-«l + 2Cf2-«3-«4)» •^4=3-(-«l-% + 2^3-«4)> K^O

^^5= -(~a,-fl2-«3+2«4)-

Now let the reduced equation in a be

a'^ + mc^-\-tia'^+pa + r=0, .... (4'.)
the roots of which are «i, «2» ^3> ^4? ^"^ therefore

— »2 = 2(«x), « = S(a?i«2)> — J^=^(^i«2^3)j r—a^acfi^a^.
Consequently, — m = 3a?i by (3.),

w = 6a?2 + 2.ri (372 + ^3 + ^4 + ^5) + '^[Xy':^ — ^x\ + '^{XyX^

because S(^i) = 0;

—7; = ^x\ + 2.r2 (a;2 + .^3 + X4 + 075) + ^i2 (a^i^-g) + 2 (a^i^a-^s)

2 1

= 2^3 + ^iSCj^-ja^g) -\-X{x^x^^ — — — n^— — mk—V>\

r = x\-\- x''^{x^x^ + 2(^13^2^3.^4) = — »i4 + _ ^^2^ ^ c^

Hence w, 7?, and /- are given in terms of 'm, and m=^ —Zx^ can
only be found by solving (1.); or the resolution of the pro-
posed into factors, one of which is of the second degree, de-
pends upon the solution of the proposed itself.

We may introduce fifth roots if we please; thus, let

A'*+^^3-f^X^ + ^X + /=0, .... (5.)

Equation of the Fifth Degree, 343

the roots being

Xj = rt6, A2=a«, A3=a5, X^-a\.

Here we shall find, as before, putting for a^^ Cg* &c. then*
values in w^, Xc^ &c., that

■\-5x^{xl^x\ + x\ + x*^+xl + x\ + xl + xlz=z~27x\

We may find h, k, and / in terms ofg, as we found n, p, and
r in terms of ?w ; and as S(^j), 2(a:^), &c. are known functions
of A, B, &c., we shall have g, h, &c. functions of x^. The
determination of these therefore may be said to depend upon
the solution of the given equation. If otherwise found, as
they may be by finding the equation on which g depends, it
must be by an equation of the fifth degree not reducible ; for
the five values of Xy, .r^, &c. being distinct, there will be as
many distinct values of ^.

It may be observed that if we make A any other integer
function of a, not passing the fifth degree, we shall still have
an ultimate equation to solve of the same degree.

To give two very simple examples of the equation in g, let
x^ + Kx+B — O.
Then

2:K) = 0, 2:(a;?)=0, 2(^t)=-4A, 2(^J) = -5B;
and

g=27:rf + 20A.ri4-5B.

Eliminating x^ between this and a:J + Aa*j + B=0, we hav

{g + 22B)5 + 74A^(^ + 22 B) - l^M'B = 0.
Again, leta?^ + Aa;^ + B = 0. In this case

and

g=z'21x\ + iiOAx\+5^.

Eliminate -Tj from this and x\-{- Ax\->r^ = 0, and there results

{g + 22B)* + 3W(g + 22B) - 3*A^^B = 0.

By making g + 22B = i; in the first of these examples, and
^-l-22B = t;^ in the second, the equations in v are similar to
those in .r, and are no way in a more solvable form.
Let us now take the equation of the sixth degree,

a^ + ha^ +^x^ + Cx^ + Dx + ¥.=: {af^ ^ aa^ -{■bx'^ + ex + d)

{x^—ax+f)=0.

844 The Rev. B. Bronwin on the Algebraic

There are fifteen ways in which this may be done, and con-
sequently the equation in a will be of the fifteenth degree. As
before,

&c., and

0 = a?! + ^2 + a73 + 0^4 + a?5 + OTg.

If we eliminate x^, Xc^ &c. from these sixteen equations, we
shall have ten resulting equations between a^, «2» &C'j which
will give fl!g, ^7, &c. in terms of the first five of these quantities.
The equation of the fifteenth degree is therefore reducible to
one of the fifth, or

a^ 4- ma!^ + nc^ -i-pa^ + qa + r=0,
where

— m=ai4-«2 + «3 + «4 + «5=4)A'i.

The determination of m then will be the same thing as solving
the given equation of the sixth degree. And it is easy to see
that we shall arrive at results precisely the same in equations
of a still higher degree.

If we resolve the given equation into the factors x^-^ax^
+ bx + c and x^—ax'^+Jlr+g, we shall have

^j — — •*'| ~T" <^2 "1 ill^g, fltg =S i^j -j- ^2 -j- iJ?^, OCC,

and the equation in a will be of the twentieth degree. But
since 0^= — «i, «i2= ~^'2.i &c., the equation in a^ will be only
of the tenth degree. The reduced equation however, whether
we find by it a or a^, will be of a higher degree than the fifth.

Let us now return to (1.), or the equation of the fifth de-
gree, in order to find Lagrange's final equation of the sixth
degree.

Make

^, = Q\ + Q\ + fl*3 + a*4, x^ = aS^i + u^\ + et^a + uH\,

Whence we find

5Q\=x^ + u'^x^ + ^% + y'*^4 + 8%

59^2 = -^1 + «% + ^% + y^^4 + ^-^5
59^3 = a?i + a^^g + fi% + /^4 + 1%
5d*4 = 0?! + (SMJj + /ScVg + yx^ + dx^f

where 1, a, /3, 7, 8 are the five roots of unity. If we make

^ = 01.% y^ct^, 8=«% we have

Equation of the Fifth Degree, 34f5

5S*2 = J^i + a^^2 + '^^'^a + *'*'^4 + "^^
Sd\ = a?i + «*^2 + «*^3 + <''^4 + ^^^,
56*4 = iTi + a^3 + ^^^3 + *^^4 + "''^S'

Let 5i, figj 635 ^4 be the roots of

fl4 + Mg3^Nfl2 + Pfl + Q=0.
Then

-M=0, +624-63+64;

to find which I employ

(m + u + to + 3 + ^)5 = 2(«^) + 5X Ku) + 1 02(w3u^) + 202 (M^tw)
+ 302(w^u^w) + 602(m^«W2) + l^Owmzt,

collecting the terms separately, and reducing by means of

l+ot + a%+a3 + a^ = 0, 0:1 + 5-2 + ^3 + ^4 + ^-5=0,

and also

X\ - X^X^ = - x\{Xc^ + 0^3 + ^4 + ^bli

a^lx^x^ + x^x^) = - x\xlx^ - x\xlx^ - x\x\x^ - x\xlx^
—x\[XciX^X^ + Xc^XgXr^ + Xc^X^X^ + x^x^x^, &c.

We thus find

5^(3i + fl2 + S3 + M = 192«)-102(o;^)X«)-202(^^^2^3)
— 1 302(^^4^3) ~ 4-OS (^?a;2a:'3^4) + 4i^0x^Xc^QX^x^ + 250

+ 5;^o-«or, + aX^i + ^K^4 + ^^X) •

The first six terms of the second member are all given,
being symmetrical functions of a7j, a?2j &c. Let their sum be
R; then, putting — M for 5)(6i), the above will be

-5^M-R=s260(a?2ar2a74 + x2a?2a?j + ).

Or if we make

5^M + R_

250 ~'^*
it will become

-<^=x\xlx^^x\xlx^-\- '

(6.)

Now if we make a?i, x^ change places in the second member of
the last, then a'^, ajg, and o-g* ^3> &c., we shall find that it has
six different values, as stated by Lagrange. Thus

846 Capt. J. H. Lefroy o« a great Magnetic

-<^^-x\x\x^^x\xlx^^

-(p„=a;2^X + ^I-^I^i +

— <^^=-X^X^X^-\-X^Xi^^ +

-(^^-x\x\x^ + x\x\x^ +

(pg — c*^./.5X_^-t-</:.^.X3<*g-^

By adding these six equations, the sum of the second mem-
bers will be a symmetrical function of iPj, &c., and we easily
find — 2(fi)=S(.2;'Ja;2), a given quantity. Thus the coefficient
of the second term of the equation of the sixth degree, of which
the roots are fj, (^5, &c., is known, and the other coefficients
may be found by means of it. It does not appear that there
is any other relation between (pj, f^? &c. ; and therefore it
would seem that the equation of the sixth degree is not redu-
cible. But if any one thinks that there may exist such a rela-
tion as *

fl=/(^2)»

y* denoting a rational function, he may, from what precedes,
make the trial. Success however seems so hopeless, that it is
pity that time and talent should be wasted upon it.

Gunthwaite Hall, near Barnsley,
October 2, 1847.

LV. Letter from Capt. J. H. Lefroy, B.A.^ Director of
the Magnetic Observatory of Toronto in Canada, to Lieut.-
Colonel Sabine, R.A., on a great Magnetic Disturbance on
the 2Uh of September 1847.

Observatory, Toronto,
My dear Colonel, September 24, 1847.

THIS day has been distinguished by a greater disturbance
than any we have had yet. The observed range of De-
clination was 4° 2' ; and I have little doubt that the actual
range was greater, as the non-commissioned officer on duty,
when he found that the movement was beyond the scale of
the Observatory declinometer, lost time in sending for me,
instead of at once lighting the lamp of the transportable one,
and following it up on that. The observed range of horizontal
force was over 600 divisions, or 0*052 of the horizontal force !
The day has been raw and cloudy, with occasional rain, so that
if an aurora existed, it could not have been seen. The disturb-
ance seems to have begun between 2liiand 22'' Gottingen time
on the 23rd, as the observation at 22^ was decidedly unusual;
but extra observations did not commence until 23^ 20°*. The
extreme disturbance began about 0'' 35°* on the 24th, when
both thelargedeclinometer and large bifilar went off" their scales.

Disturbance on the 24!th of September 184-7. 34-7

At this time I was called, and we began to observe the trans-
portable declinometer and bifilar. The last also went oiF the
scale. The lowest reading of the former was 692*5 at 1^ 0"^
Gott., and the highest 1 126*0 at l^' 45"^: this gives a range of
3° 36'*7; but at a subsequent period (5^ 0"^ Gott.) a reading of
1177*2 was obtained, thus giving the enormous range of
4° 2'"3*. I did not take a reading of your compass; but
looking hastily at it, I perceived that during the great shock
it was ranging more than 3° 20' from its usual position. As
both bifilar scales were exceeded, we can only say that the
range of that element between 0^ and l"* Gott. exceeded 600
divisions, or 0*052 of its whole amount, on the testimony of
two instruments ; a fact which cannot, 1 think, but make it a
most interesting question, what is the nature of a force sub-
ject to such immense variations, and how can they occur with-
out affecting or being affected by the other physical agents in
the globe? This disturbance was attended by a great degree
of motion in the magnets, a peculiar mechanical agitation,
which they only exhibit on rare occasions ; it lasted, more or
less, down to 12^ Gott. As the results have not been made
up, I cannot state precisely the range of inclination, but per-
haps may do so before I close this.

After some little trouble, I think we have got Dr. Robin-
son's Anemometer into beautiful working order. If the prin-
ciple on which the velocity is estimated is correct, as we must
feel confident it must be, I think it has a great superiority over
any other instrument of the kind yet invented. The facility
and precision with which the velocity is measured, and the
beautiful manner in which sudden changes are shown, together
with the large scale on which directions are marked, make it
a pleasure to use it, and make Osier's instrument look quite
clumsy beside it; it is a most elegant instrument, and will give
diurnal curves of velocity with a precision we have never
attained before. I found on careful examination that Osier's
anemometer, which has been up seven years, was much the
worse for wear, and not in a condition to give a satisfactory
comparison with the other ; we have therefore, with a good
deal of difficulty, taken it down. I have put it into the hands
of an engineer here, and he is to refit all the essential parts,
particularly the shoulder and collar of the vane, which were
worn, and made the vane unsteady : we shall then be able to
compare pressures and velocities.

Believe me, my dear Colonel,

Faithfully yours,

J. H. Lefroy.

• I think our greatest range before this was only 2° 23'; this occurred
last April.

[ 348 ]

LVI. On the Decomposition of Valerianic Acid by the Vol-
taic Current. By H. KolbEj Ph.D.*

THE very remarkable changes which a series of organic
compounds undergoes by means of the voltaic current,
have induced me to make that mode of decomposition the
subject of a thorough investigation. As however the nume-
rous difficulties which present themselves in researches of
this nature, and the immense extent of the field which opens
before us, do not admit of the results being communicated
in a complete and connected form, I beg to lay before the
Chemical Society a short preliminary notice of the changes
which valerianic acid undergoes when exposed to the oxidizing
action of the voltaic current, reserving a more complete de-
scription of the products obtained till the investigation shall
have been brought to a close.

When the voltaic current, excited by six pairs of Bunsen's
carbo-zinc battery, is permitted to act on a concentrated neu-
tral solution of valerianate of potash in the cold, two plates
of platinum forming the electrodes, a brisk evolution of gas
takes place simultaneously from both ; the gases evolved
consist of hydrogen, carbonic acid and a new carbo-hydrogen,
but contain no traces of oxygen gas as long as the solution
of valerianate of potash does not become too much exhausted.
At the same time a light oily liquid separates at the surface,
having an agreeable Eethereal odour, and the alkaline solution
ultimately consists chiefly of carbonate and bicarbonate of
potash, the latter of which generally separates during the
operation in a crystalline form.

The neutral ethereal oil is a mixture of two compounds ;
the one containing oxygen, the other perfectly free from it.
By the action of an alcoholic solution of potash the former is
decomposed, and the latter can then, by means of water, be
separated unchanged. In the pure state it exists in the form
of a light colourless sethereal oil, possessing an agreeable
aromatic smell. It is insoluble in water, but soluble in al-
cohol and aether; it boils at 108° C. without decomposition,
and has the composition Cg Hg. Oxygen and iodine are with-
out action upon it, but chlorine, bromine, and fuming nitric
acid form with it products of substitution.

The oil containing oxygen, which in the first instance was
found mixed with this substance, I have not yet been able to
obtain in a pure state ; but several circumstances render it
more than probable that it is formed by the union of vale-
rianic acid with the oxide of the above carbo-hydrogen. An

• Communicated by the Chemical Society} having been read April 19,
1847.

Dr. Kolbe on the Decomposition of Valerianic Acid. 349

alcoholic solution of potash treated with it is found to con-
tain as a product of decomposition a considerable amount of
valerianate of potash. But on account of the small quantity
of material which has been at my disi)0sal, I have not suc-
ceeded in separating the alcohol Cg Hjq Og, which must have
been formed at the same time.

The gaseous carbo-hydrogen, which is evolved with the hy-
drogen^ is a substance analogous to defiant gas ; it is cha-
racterized by a peculiar aethereal smell, and has a specific
gravity double that of olefiant gas. It unites with chlorine
even in the dark, forming a heavy oily liquid, having a marked
similarity to chlorelayl, and is generally composed of a mix-
ture of several products of substitution. Its rational com-
position is expressed by the formula Cg Hg. The changes
which valerianic acid undergoes, in accordance with the fore-
going experiments, are capable of a very simple explanation,
if we consider that acid as a conjugated combination of the
carburetted hydrogen, or the radical Cg H^ with oxalic acid,
in a similar manner to the new view taken of the constitution
of acetic acid. For whilst by the addition of one atom of
oxygen oxalic acid becomes converted into carbonic acid, this
radical is set free ; but a portion of it unites w ith the excess
of oxygen to form an oxide, and this enters into combination
with a portion of undecomposed valerianic acid, giving rise
to a new aether, Cg Hg O + Cg H9 Cg O3.

Another portion of the radical is probably decomposed at
the moment of its formation, in consequence of the conco-
mitant evolution of heat into hydrogen and the gaseous carbo-
hydrogen Cg Hg. This latter view is supported by the fact,
that if the temperature of the solution of valerianate of potash
exceeds a certain point during the decomposition, not a single
drop more of the aitherial oil is produced.

The following formula will throw light on this decompo-
sition : —

o/-lc,

KO + Cg Ho C2 0^\ _ /KO + 2CO2

^8 "9-

Both butyric and acetic acids are acted on in a similar
manner to valerianic acid ; the products of decomposition of
acetic acid are all gaseous, and appear to contain oxide of
methyle. Butyric acid gives in addition to the gaseous com-
pounds a volatile oil composed of Cg H7.

The minute description of this product will form the sub-
ject of a future memoir.

The foregoing investigation has been carried out during
the late session in the laboratory of Dr. Lyon Playfair, as
whose assistant I have been engaged during that time j and I

350 Mr. Adieus Experiments with Galvanic Couples

cannot allow this opportunity to pass by without thanking him
for the kindness and liberality which he has shown in placing
his laboratory at my disposal, in leaving so much of my time
on my own hands, and in rendering me every assistance in
his power.

LVII. An Account of Experiments with Galvanic Couples
immersed in pure water and in oxygenated water. By Mr.
Richard Adie*.

TN the years 1845 and 1846, I published in the Edinburgh
-*- Philosophical Journal two series of experiments, made
with a view to prove that the action of the water battery was
maintained by absorbing oxygen from the atmosphere. Some
of these experimentsf show that it is the oxygen only that is
drawn from the atmosphere, and that the presence of the
other component parts is unnecessary. But there was one
given to show that zinc and copper elements placed in a her-
metically sealed tube along with pure water did not act, there
being no flocculent deposit of oxide of zinc, which is formed
in abundance when a minute aperture admits the atmosphere
to the contents of the tube. After a lapse of two years, I ex-
amined an arrangement of this kind which had been her-
metically sealed since December 1844 ; there was no apparent
change, the water was transparent, and the metals bright.
I had scarcely put the tube down when it burst with violence ;
this fact immediately satisfied me that the water battery must
have a true decomposition of water action when it acts on
zinc associated with copper or any other metal less oxidizable
than the copper, independent of the much more extensive
effect due to atmospheric oxygen. It is from a desire to trace
by experiment the double action of this battery that I respect-
fully submit for the consideration of the Society the following
results : —

In fig. \, aaa represents six pieces of zinc soldered at c c
to a corresponding number of pieces of copper bb h, arranged
alternately as in the figure, and insulated from one another
by strands of thread, dddd. These were placed inside a
flint-glass test-tube, which was after their insertion drawn off
at the blowpipe to a capillary point. The tube was now filled
with pure water, and to dislodge the air from among the
fibres of the thread, the water was repeatedly boiled, closing
and re-opening the capillary point at each boiling. When
the air was well-removed, the tube was hermetically sealed,

* Communicated by the Chemical Society; having been read April 19,
1847.
f Edinburgh New Phil. Journal, vol. xxxviii. p. Q^, and vol. xl.

immersed in pure water and in oxygenated water. 351

the water at the time of closing being Fig. 1.

near the boiling temperature. On ^ ^

cooling, the space left vacant by the

contraction of the fluid was estimated

to be T^fth of a cubic inch ; the super- d

ficies of each plate ^th of a superficial
inch. From previous trials, I knew that
when the above arrangement had a
communication with the atmosphere, a
flocculent deposit of the protoxide of
zinc was soon perceived, which steadily
increased. With the same hermetically
sealed there was no such deposit ; nei-
ther was there any perceptible change,
until the bursting of the vessel after
two years revealed another action of
the battery. Judging from the thick-
ness of the broken glass, I endea-
voured at the time to make an ap-
proximate estimate of the volume of
the gas generated, which in the vacant
space of y yth of a cubic inch, where it
could lodge, produced pressure suffi-
cient to burst the vessel : the result of
my estimate gave less than a cubic inch
of gas measured at the usual atmo-
spheric pressure ; for the development
of which six zinc surfaces of ^th of a U 11

superficial inch each had been two
years in action. In a repetition of this
experiment, with zinc filings in lieu of plates, a small quan-
tity of gas was collected, and proved to be hydrogen.

Afterwards examining the inner surfaces of the firagments
of the glass, the surface of the plates, and the fibres of the
thread with a powerful lens, I found all of them covered with
minute transparent crystals ; the largest crop of these was on
a copper surface opposite a spot on one of the zinc plates, to
w hich nearly the whole of the corrosion of the metal appeared
to have been confined. The red ground of the copper sur-
face showed them most distinctly. On heating the copper
the crystals parted with water of crystallization, and became
circular white spots, very much resembling the protoxide of
zinc.

My friend Mr. Waldie examined the thread; his process
was, incinerating, dissolving the ash in hydrochloric acid,
adding excess of potash, filtering to separate a trace of oxide

352 Mr. Adie's Experiments with Galvanic Couples

of iron, and treating the filtered liquid with hydro sulphuret
of ammonia, which gave a yellowish white precipitate. This
result proves that the minute transparent crystals among the
fibres of the thread contained protoxide of zinc.

On a former occasion I employed either the air-pump or
ebullition to deprive water used in exciting voltaic couples, of
absorbed air, I gave preference to the method of boiling the
water in the battery cell, as the more severe test, for showing
how far a battery's action depended on oxygen from the atmo-
sphere. The proof which appeared to me to furnish satisfac-
tory evidence of the assistance given by absorbed oxygen, was
when the indication in the galvanometer fell near to zero by
prolonged boiling, and rose again when water holding dis-
solved air was thrown into the cell. According to this test,
a zinc and platinum couple lose much of their action when
excited by pure water boiled for near two hours. The galva-
nometer needles always indicated a slight action, however long
the boiling was prolonged ; but as 1 found when care was
taken to have an atmosphere of steam resting on the surface
of the boiling water the action of the couple was at its lowest,
I was led to think that what remained might be due to oxy-
gen from the atmosphere, which it was impossible to remove
perfectly. The experiment given above renders this view no
longer tenable ; for if zinc and copper elements can at ordinary
temperatures slowly generate gas, it must follow that all the
elements less oxidizable than copper will at boiling tempera-
tures possess, when associated with zinc, a voltaic action
independent of oxygen from the atmosphere.

To try the effect of a zinc and copper couple "excited by
pure boiling M^ater, I attached a pair of plates to a more sensi-
tive galvanometer than I had hitherto used : the plates were
placed in a Florence flask and covered to a depth of two
inches with pure water previously distilled in glass vessels ;
there was only a small orifice in the cork at the top of the
flask for a steam escape, in order to preserve the boiling sur-
face from the atmosphere.

Previous to boiling, the galvanometer needle stood at 50°

Indication the moment boiling was about to begin , 70°
after long boiling 20^^

A similar experiment with iron and copper elements : —

Indication previous to boiling 20°

at boiling 46°

after long boiling 7°

In this experiment the indication rose on cooling to 30°,
and afterwards fell back.

When silver or platinum was substituted for the copper the

immersed in pure water and in oxygenated water. 353

t»esults were in the same order, giving the highest action at
the time the water is parting with dissolved air, and lowest
when the water is thoroughly boiled. Where zinc is the
positive element the action falls considerably, as the boiled
water cools before it has time to re-absorb air. With a little
common salt added to the water of a zinc and platinum couple,
ebullition serves greatly to exalt the action, for the arrange-
ment is no longer dependent on oxygen from the atmosphere.

These experiments, in extension of those I formerly sub-
mitted to the public through the Edinburgh Philosophical
Journal, do not militate against the general conclusion then
drawn, that the water battery supported its action by absorb-
ing oxygen from the atmosphere ; they only show that there
is in addition a minute degree of action when two metallic
elements are excited by pure water.

Perhaps the experiments of the most importance for deter-
mining the theory of the action of gas absorbing galvanic
couples, are those where one metal only is excited by oxy-
genated water ; to illustrate this action I made the following
experiments : —

Two slips of zinc cut side by side from the same sheet were
placed in a running brook, the one opposed to a rapid part of
the current, the other in a still place at the edge. Connecting
these in the usual manner with the galvanometer, there was
a permanent deflection of 25° ; and on changing the respective
places of the plates in the stream without disturbing their
attachments to the galvanometer, the needles immediately
passed to the opposite side of the card ; in both cases the
piece of zinc in the current acted as a negative or platinode
plate. With both plates in still water and a tube filled with
oxygen inverted over one, the effect was the same. It is the
greater supply of oxygen to the plate in the current which con-
verts it into a negative or platinode. A cell containing two
small silver wires and the cyanide of silver solution used for
electro-plating was attached in place of the galvanometer,
when, after a lapse of two hours, metallic silver was seen pre-
cipitated in a minute quantity on the silver wire connected
with the piece of zinc in still water.

Two plates of iron were placed in the stream, under like
conditions to the zinc ; after two hours metallic silver was
distinctly seen precipitated on the silver wire connected with
the iron plate in still water.

The fact here shown, of two similar pieces of iron giving
rise to a galvanic current capable of precipitating metallic
silver, appears to me to be important, for it proves that the
electricity in passing through the water intervening between

Phil, Mas. S. 3. Vol. 31 . No. 209. 'Nov. 184.7. 2 A

354 Mr. Adie's Experiments with Galvanic Couples

these two plates, either decomposes it with the aid of oxygeil
in solution, or that oxygenated water forms a binary com-
pound, capable of acting as an electrolyte.

The fact of iron and oxygen uniting together at ordinary
temperatures when moisture is present, is well known. It is
the office performed by the water during this union, wherein
lies the true ground of the theory of gas-absorbing batteries.
A single plate of iron exposed to water and oxygen gas, has
local differences on its surface which act in the same way as
if the iron had been in two halves and placed in a stream in
the manner described : the oxidation of the iron is developing
a voltaic current which passes through the fluid from one
point of the plate to another, either by a process of decompo-
sition and re-composition of water, or by the decomposition
of the compound formed by the solution of the gas in water.

The first of these views can only be supported by holding
that the solution of oxygen so changes the affinities, that iron
with its aid can at ordinary temperatures decompose water.
I see no evidence sufficient to give probability to this hypo-
thesis, while, if the second supposition be admitted to meet
all the facts shown by experiment, it will establish the exist-
ence of an electrolyte more easily decomposed than water,
and as universal in nature ; and account for the very reduced
action of zinc and copper elements excited by pure water
freed from absorbed air or from oxygen gas, the active prin-
ciple derived from the air.

I may here take occasion to add, that a saturated solution
of carbonate of potash and soda in an open cylindrical vessel
has so shut out the oxygen of the atmosphere from some
pieces of iron immersed in it, that now, after two years and
four months immersion, there is no rust on the surface of the
iron.

The experiments with two similar pieces of zinc or of iron
placed in a running stream, as already described, were per-
formed during the cold weather of winter, with the tempera-
ture varying from 32° F. to 42°. On the return of a little
warmer weather I recommenced the experiments with iron
plates, from a wish to try if two similar pieces of iron could
be made to develope a voltaic current of the same electro-
motive force as that derived from a platinum and iron couple
excited by still water.

A piece of iron- wire M'as cut into two equal lengths; each
of those was bent into the form of a flat spiral (fig. 2), and a
copper wire well-varnished was soldered to the iron at A,
for connecting the plate with a small decomposing apparatus
in the usual manner.

immersed in pure water and in oocygenated water. 355

A pair of iron plates thus formed was
taken to the banks of a small stream in
Cheshire, called the Grange brook ; one
plate was placed in a current of mode-
rate velocity, where the water poured
through the interstices of the coils ; the
other plate was dropped into still water
in a convenient place at the edge. In
both plates the solderings at A pro-
jected above the surface and were kept
dry.

i When copper wire poles were placed
in a decomposing cell filled with sulphate
of copper solution, and connected with
the galvanic couple formed of two pieces
of iron (fig. 2), distinct evidence of the precipitation of me-
tallic copper on the wire connected with the plate in still
water, was observed after an hour's action, temperature 45°.
One of the iron plates was now removed to a cell filled with
water, and associated with a platinum plate, the arrangements
for precipitating metallic copper remaining as before. With
the temperature at 42°, the depositing of the metal did not
proceed so actively as it had done with an iron plate in a cur-
rent of water for a platinode.

When the decomposing cell was filled with a solution of Sul-
phate of zinc, and zinc wire poles supplied, after three hours'
action, temperature 46°, the wire in connexion with the iron
plate in still water showed, with the aid of a lens, a distinct
deposit of metallic zinc. Repeating this experiment with an
iron and platinum couple in still water, the metallic deposit of
zinc was again obtained, temperature 46°, the rate of action in
both experiments being, as near as I could judge, the same.

The inference from these results is, that a piece of bright
iron placed in a current of water performs the office of a piece
of platinum, as well as the latter metal does when excited by
still water.

The quantities of metal precipitated during two or three
hours' action of these oxygen absorbing batteries is in no
case sufficient to give results by weight. I have tried experi-
ments of one week each, but the changes in the level of the
stream and other sources of derangement, made me prefer
trials of two or three hours each, where 'there is no difficulty
in detecting any decided change in the rate of action.

The Grange brook is supplied with water almost wholly by
the drainage of a rather poor clay soil, reposing on the new
red sandstone formation of the district*

2 A 2

S56 Dr. R. Hare on Improvements in the

The plate in the centre of the stream shows less rusting
than the one at the edge in still water ; but judging from the
analogous case of the copper sheathing of a ship, it should
waste away the fastest, the particles of peroxide of iron as
they are formed being removed by the force of the stream,
while the voltaic current developed during this action only
circulates to some other point of the same plate, or belongs
to what is called local action.

From the above results, the benefit to be obtained in a con-
stant battery by making the negative plate rotate, should be
apparent ; but to prevent waste it would still be necessary to
employ one of the more costly metals, which are not liable to
oxidation.

In concluding these experiments, I may again notice, that
a tube of oxygen suspended over a plate of iron in still water
has the same effect as the current of the stream, converting
the oxygenated plate into a platinode. The carbonic acid
present in all surface water may by some be thought to per-
form an essential part in the ordinary rusting of iron. But
where every care is taken to exclude this gas from a tube
filled with oxygen, a small quantity of water, and a piece of
iron, the oxidation of the iron proceeds with rapidity, accom-
panied by changes which appear to me to preclude the idea
that even a trace of carbonic acid can be essential. The oxy-
gen gas disappears ; at first an abundant formation of red or
peroxide of iron is seen ; then, after the supply of oxygen has
decreased, the green-coloured protoxide is gradually formed.
These two oxides afterwards begin slowly to unite, and form
the well-known black or magnetic oxide. In an experiment of
this kind every trace of the red and green-coloured oxides had
disappeared at the end of three months from the time of
closing the tube, and there remained only an inky precipitate,
which was proved to be the black oxide of iron.

LVIII. On certain Improvements in the Construction and Sup-
ply of the Hydro-Oxygen Blowpipe^ by 'mhich Platinum
may he fused in the large txay. By Robert Hare, M.D.*

/^N my return from Europe in 1836, I was very much in
^^ want of a piece of platinum of a certain weight, while
many more scraps than were adequate to form such a piece
were in my possession. This induced new efforts to extend
the power of my blowpipe; and after many experiments, I
succeeded so as to fuse twenty-eight ounces of platinum into
one mass.

* Conr,jnunicated by the Author.

Hydro-Oxygen Bloisopij^e for the Fusion of Platinum. 357

Although small lumps of platinum had been fused by many
operators with the hydro-oxygen blowpipe as well as myself,
it had not, up to the year 1837) been found sufficiently com-
petent to enable artists to resort to this process. I am in-
formed by Mr. Saxton, that some efforts which were made
while he was in London were so little successful, that the pro-
ject was abandoned. There was an impression that the metal
was rendered less malleable when fused upon charcoal, as in
the experiments alluded to. This is contradicted by my ex-
periments, agreeably to which fused platinum is as malleable
as the best specimens obtained by the Wollaston process,
and is less liable to flake. Dr. Ure, on seeing specimens of
platinum which I had elaborated and fused in the form of wire,
of leaf, ingots and plate, said that there was no one in Europe
who could fuse platinum in such masses. He also informed
me that it had been found so difficult to weld platinum, that
no resort was had to that process. In this I concur, having
had the welding tried by a skilful smith, both with a forge
heat, and with a heat given by the hydro-oxygen blowpipe.
An incorporation of two ingots was effected on their being
hammered together, when heated nearly to fusion ; but on
hammering the resulting mass cold, a separation took place
along the joint by which the ingots were united.

The difficulty seems to arise from the rapidity with which
the platinum becomes refrigerated. It seems to have a less
capacity for heat than iron ; and, not burning in the air as
iron does, has not the benefit of the heat acquired by iron
from its own combustion with atmospheric oxygen.

Lately, by means of the instrument and process which it is
my object here to describe, I have been enabled to obtain
malleable platinum directly from the ore, by the continued
application of the flame. From some specimens of platinum
I have procured as much as ninety per cent, of malleable
metal. The malleability is not inferior to that of the best
specimens obtained by reducing it to the state of sponge,
through the agency of aqua regia and sal-ammoniac. There
is however a greater liability to tarnish, arising probably from
the presence of a minute portion of palladium.

Of the fusion of iridium and rhodium, I have already given
an account in the Bulletin of the American Philosophical
Society, which was subsequently embodied in an article in
this Journal for August 1847.

It remains now to give an account of the apparatus employed
in the fusion of platina on a large scale.

Fig. 1 represents the association of fifteen jet-pipes of plati-
num with one large pipe B D at their upper ends, so that

358 Dr. R. Hare on Improvements in the

their bores communicate, by means of an appropriate brass
casting, with that oFthe large pipe, the joints secured by hard
solder. Their lower extremities are made to protrude about
half an inch from a box A, of cast brass, their junctures, with
the appropriate perforations severally made for them, being
secured by silver solder. They come out obliquely in a line
along one corner of the box, an interval of about a quarter of
an inch alternating with each orifice. By means of flanges,
the brass box is secured to a conical frustum of copper (fig. 2),

so as to form the bottom thereof, while the pipe, extending
above the copper case, is screwed to a hollow cylinder of brass
Aj fig. 3, provided with two nozzles and gallows-screws g g,
for the attachment of appropriate hollow knobs, to which pipes
^re soldered, proceeding from the reservoirs of oxygen and
hydrogen. Cocks are interposed by which to regulate the
emission of the gases in due proportion,

In connecting the pipes conveying the gases with the brass

Hydro-Oxygen Blowpipe for the Fusion of Platinum. 359

cylinder A, fig. 3, care should be taken to attach that con-
veying oxygen to the upper nozzle, while the other, conveying
hydrogen, should be attached to the lower nozzle ; since by
these means their great difference in density tends to promote
admixture, which evidently it must be advantageous to effect.

The object of surrounding the jet-pipes with water by means
of the copper box*, is to secure them against being heated to
such a degree as to cause the flame to retrocede and burn
within them, so as finally to explode within the cylinder A,
gg, fig. 3. It is preferable to add ice or snow to the water,
in order to prevent undue heat.

Fig, 4 represents a moveable platform A, of cast iron, wholly
supported upon the point of the iron lever D B, which is
curved towards the extremity under the platform, so as to
point upwards, and to enter a small central conical cavity
made for its reception. The lever is supported by a universal
joint upon the fulcrum C, so that by means of a sliding weight
at one end, the platform and its appurtenances are counter-
poised at the other. The platform is kept in a horizontal
position by the cannon-ball, supported in a sort of iron stirrup
terminating in a ring, in which the ball is placed. Upon the
platform is situated an iron pan with a handle holding the
brick, on a cavity in which, as already mentioned, the metal
is supported. The apparatus being duly prepared and con-
nected with the supply-pipes, the hydrogen is first allowed to
escape and then the oxygen, until the ignition has attained
apparently a maximum. The accomplishment of this object
may of course require the adjustment of either cock several
times, especially where there is any decline in the pressure
either of the one or the other gas in its appropriate reservoir.

By means of the handles of the lever and of the pan, the
operator is enabled to bring the metal into the position most
favourable for the influence of the iieat, while his hands and
face are sufficiently remote to render the process supportable.
In fusing any quantity, not being more than four ounces, the
platform may be dispensed with, the handle of the pan being
held in one hand of the operator, while by the other the cocks
may be adjusted.

When the blowpipe of fifteen jets, or any larger, may be

* Since the engraving was made, I have preferred to use water-tight
boxes, with gallows-screws and nozzles, situated one near the bottom on
one side, the other on the opposite side near the top. By means of the
lower nozzle, a pipe is attached, communicating with a head of cold water,
the other being so situated as to carry the water into a waste pipe or large
tub : a circulation may be kept up during the whole time that the opera^
tion is going on,

As a support, a brick kaolin is used, having an oblong ellipsoidal depres-
sion on the upper face for the reception of the metal to he fused.

JI

360 Dr. R. Hare on Improvements in the

employed, and the platform is necessarily resorted to, the
cocks must be adjusted by an assistant.

Fig. 5 repre-
sents a cask made
of boiler iron,
three-sixteenths of
an inch thick, so
as to resist an
enormous pres-
sure. The joints
are secured by
riveting, as in
constructing high
pressure boilers.

This cask com-
municates with
the hydrant pipes,
so called, by which
our city is sup-
plied with water,
of which the pres-
sure varies from a
half to more than
two atmospheres,
say from seven to
thirty pounds per
square inch, ac-
cording to the
number and bore
of the cocks from
which the water
may be flowing
at the time for
the consumption
of the community.
Hence experi-
ments, while using
this head, are best
made towards bed-time, or between that time and sunrise.
The vessel is filled with water by opening a cock F on one side
of the pipe C, and allowing the air to escape through the valve-
cock B. Being thus supplied, the cock F closed, and a commu-
nication with a bell-glass, into which oxygen is proceeding from
a generating apparatus, being made by means of a flexible
leaden tube, on opening the valve- cock B and the cock E, the
water will run out and be replaced by gas from the bell. This
process being continued till the iron cask is sufficiently supplied

Hydro-Oxygen Blowpipe Jbr the Fusion of Platinum, 361

with gas, the cock E must be shut. Whenever the gas is
wanted for the supply of the blowpipe, it is only necessary to
establish a communication between the valve-cock B and the
upper gallows-screw (fig. 3) of the cylinder A, and to open the
cock F, so as to admit the water to press upon the gas, the
efflux being regulated by B, or preferable by a cock of the
ordinary construction, one of which kind should be interposed
at a convenient position between the valve-cock B and cylin-
der A.

T represents a glass tube, which, by due communication
with the interior, shows the height of the water, and conse-
quently the quantity of gas in the vessel.

G H represents a gauging apparatus, consisting of a cast-
iron flask, of about a half pint in content, and a glass tube of
about a quarter of an inch in bore, which should be at least
five feet in height. The tube is secured air-tight into the neck
of the flask, so as to reach nearly to the bottom within. The
flask is nearly full of mer-
cury. Under these circum-
stances, when a communi-
cation is made by a leaden
pipe between the cavity of
the flask and that of the
reservoir, an equilibrium
of pressure resulting, the
extent of the pressure is
indicated by the rise of the
mercury in the tube.

In order to generate hy-
drogen for the supply of a
reservoir like that repre-
sented by the preceding
figure, I have employed the
vessel represented by fig. 7.
This vessel, by means of a
suitable aperture, suscep-
tible of being closed by a
screw-plug, is half-filled
with diluted sulphuric acid.
Being furnished with a tray
of sheet copper D, punc-
tured like a coal-sieve, and
supported by a copper sli-
ding-rod E, strips of zinc
are introduced in quantity equal to the capacity of the tray.
The sliding-rod passes through a stufling-box F, at the top of
the reservoir, so that the operator may, by lowering or raising

E

362 Dr. R. Hare o?i Improvements in the

the tray, regulate or suspend the reaction between the zinc
and its solvent, accordingly as the supply of hydrogen is to be
produced, suspended, increased, or diminished.

The communication with the reservoir is open and regulated
by means of a cock P, furnished with a gallows- screw G, for
the attachment of a leaden pipe, as above described, in the
process for supplying the reservoir with oxygen.

Another apparatus for producing a supply of hydrogen is
represented in fig. 6. It consists of two similar vessels of

Hydro-Oxygen Blcmpipefor the Fusion of Platinum. 363

boiler iron, each capable of holding forty gallons. They are
lined internally with copper, being situated upon a wooden
frame, so that the bottom of one is two-thirds as high as the
top of the other. The upper portions of these vessels com-
municate by a leaden pipe B, of about half an inch bore, fur-
nished with a cock, while the lower portions communicate by
another leaden pipe of a bore of one and a half inch.

The upper vessel is surmounted by a globular copper vessel,
of about twelve inches in diameter, which, from its construc-
tion, renders it possible to introduce an additional supply of
concentrated acid, while the apparatus is in operation, with-
out reducing the pressure within the reservoir, by permitting
the excess above the pressure of the atmosphere to escape.
This object is accomplished as follows : —

The valve at the end of the rod attached to the lever L
being kept shut by the catch M, the screw-plug H removed,
the acid is introduced through the aperture thus opened. In
the next place, the plug being replaced, and the valve depressed
by means of the lever and rod, so as no longer to close the
opening which it had occupied, the acid descends from the
chamber into the cavity of the vessel beneath it. The valve
is of course restored to its previous position as soon as the acid
has effected its descent.

The lowermost vessel is furnished with a perforated copper
tray, supported by a copper sliding rod, in a way quite ana-
logous to that already described in the case of the copper
reservoir. It is also supplied with zinc and its solvent in like
manner, being made half-full of the diluted sulphuric acid.
Of course, on contact being produced between the zinc and
its solvent, the generation of hydrogen will take place. So
long as the communication between the upper portions of the
two vessels is open, the gas will extend itself into both, occu-
pying the whole of the upper vessel, and that half of the lower
one which is unoccupied by the liquid. But if in this way
the pressure reaches to two atmospheres, as indicated by the
gauge*, on shutting the communication through the pipe B,
the pressure in the inferior vessel will augment, that in the
superior vessel remaining as before ; but the liquid will con-
sequently begin to pass out of the inferior vessel through the
pipe A, and thus may lessen the contact between the acid and
zinc, and finally suspend it altogether. Meanwhile the gas in
the upper vessel being condensed to nearly half its previous
bulk, the pressure, will be nearly four atmospheres. It will,

* I have used for a gauge an instrument like G, fig. 5, the tube being
Sbout two feet in length, and sealed at the upper end.

364 Dr. R. Hare on Improvements in the

in fact, always be nearly double that which existed before the
pipe B was closed.

In order that nearly the whole of the acid shall be expelled
from the inferior vessel, the tray must be depressed till it
touches the bottom of that vessel.

The pressure being four atmospheres at commencement, as
soon as, by means of a pipe attached to the valve-cock N, an
escape of gas is allowed, the acid is forced again upon the
zinc, and thus prevents a decline of pressure to any extent
sufficient to interfere with the process.

The gases may be used from a receiver in which they exist,
in due proportion, safely by the following means : —

Two safety-tubes are to be made, not by Hemming's pro-
cess exactly, but as follows :

A copper tube, silver soldered, of which the metal is about
the eighth of an inch in thickness, is stuffed with the finest
copper wire, great care being taken to have the filaments
straight and parallel. The tube is then to be subjected to
the wire-drawing apparatus, so as to compress the tube on its
contents until the draught becomes so hard, as that it cannot
be pushed further without annealing. The stuffed tube thus
made is to be cut into segments, in lengths about equal to the
diameter, by a fine saw. The surfaces of the sections are to be
filed gently with a smooth file. By these means they appear
to the naked eye like the superficies of a solid metallic cylinder.
Brass caps being fitted on these sections, they are to be inter-
posed by soldering, at the distance of a foot or more, into the
pipe for supplying the jet. Under these circumstances, the
posterior section becoming hot, may allow the flame to retro-
cede ; but the anterior section being beyond the reach of any
possible combustion and remaining cold, will not allow of the
retrocession ; and as soon as the flame passes the first section,
the operator, being warned, will of course close the cock, and
subject the posterior section to refrigeration before proceeding
again.

But this plan of operating may be rendered still more se-
cure by interposing a mercury bottle, or other suitable iron
vessel, half-full of oil of turpentine, between the reservoir and
safety tubes, as in the arrangement of a Woulfe's bottle. A
leaden pipe proceeding from the reservoir is, by a gallows-
screw, attached to an iron tube which descends into the bottle,
so that its orifice may be near the bottom. The leaden
pipe communicating through the safety tubes with the jet-
pipe, is attached to the neck of the bottle. Thus the gaseous
mixture has to bubble through the oil of turpentine in order
to proceed through the safety tubes to the jet-pipe. If, while

Hydro-Oxygen Blowpipe for the Fusion of Platinum. 365

this process is going on, the flame should, by retrocession,
reach the cavity of the bottle, exploding in contact with the
turpentine, a compound is formed, which is, per se, inexplosive
from the excess of carbonaceous matter. Meanwhile the
shock, acting on the surface of the oil, drives it into the bore
of the iron tube, and thus, both by its chemical and mecha-
nical influence, renders it utterly impossible that the flame
should reach the cavity of the reservoir.

Apparatus for the Fusion of Iridium or Rhodium or masses of
Platinum less than five ounces in weight.

For the fusion of either iridium or rhodium or masses of
platinum not exceeding the weight of half an ounce, an instru-
ment with three jets has been employed, the bore of each jet-
pipe being such as not to admit a wire larger than the ^^^d
of an inch in diameter. The flame produced by these means
was quite sufficient to envelope the mass to which it was ap-
plied.

In fusing any lumps or congeries of platinum, not exceed-
ing five ounces, an instrument has been used capable of giving
seven jets of gas, issuing of course from as many pipes. Of
these pipes, six protrude through the brass casting forming
the bottom of the copper case constituting the refrigerator, so
as to be equidistant from each other upon a circumference of
three-fourths of an inch in diameter, the seventh protruding
from the centre. The bores of these jets are such as not to
admit a wire larger than ^^2"^ of an inch in thickness. Those
of the larger instruments, represented by the accompanying
engravings, were such as to admit wires of J^^h of an inch in
thickness.

The jet-pipes may be made by the following process : — A
thin strip of sheet metal, somewhat wider than the length of
the circumference required in the proposed pipe, after being
roughly turned about a wire so as to form an imperfect tube,
is drawn through several suitable holes in a steel plate, as in
the wire-drawer's process. Under this treatment the strip
becomes converted into a hollow wire ; the edges of the strip
being brought into contact reciprocally, so as to leave only an
almost imperceptible crevice. Having drawn one strip of
platina in this way, another strip sufficiently wide nearly to
inclose it is to be drawn over that first drawn, care being taken
to have the crevices left at the meeting of the edges on con-
trary sides. The compound hollow wire or tube thus fabri-
cated, is finally to be drawn upon a steel wire of the diameter
of the requisite bore.

The following method of making jet-pipes, though more

366 Dr. J. W. Griffith on the Composition of

difficult, is preferable, as there is less liability of the water of
the refrigerator leaking into the bore.

Select a very sound and malleable cylinder of platina, of
about three-eighths of an inch in thickness, perforate it by
drilling in a lathe, so that the perforation may be concentric
with the axis. A drill between one-sixteenth and one-eighth
of an inch in diameter may be employed. In the next place,
the cylinder may be elongated by the wire-drawing process
until the proper reduction of metallic thickness is effected, the
diameter of the bore being prevented from undergoing an
undue diminution by the timely introduction of a steel wire.

Of course the metal must be annealed as often as it har-
dens, by drawing. For this purpose a much higher tempe-
rature is necessary in the case of platinum, than in that of
either copper, silver, or gold.

The annealing is best performed by the hydro-oxygen
flame. If charcoal be used, the greatest care must be taken
to have the fireplace clean.

Agreeably to a trial made last spring, palladium may be
used as a solder for platinum ; and as it is nearly as difficult
to fuse as this metal, it is of course for that purpose preferable
to gold, where great heat is to be resisted. No doubt by em-
ploying palladium to solder the exterior juncture of the double
drawn tubes above mentioned, they might answer as well
nearly as when constructed of solid platinum.

This idea has been verified by a successful trial : and,
moreover, silver has been successfully employed to solder the
portions of the tubes, protected from heating by being within
the cavity occupied by water. The portions which protrude
beyond the brass box (see fig. 1) may be left unsoldered.

LIX. Oti the Coviposition of the Bile of the Sheep. By J.'W,
Griffith, M.D.^ F.L.S., Physician to the Finshury Dis-
pensary^.

T^HE following analysis was made with the view of compa-
-*■ ring the composition of this fluid with that of the biliary
secretion in other animals ; the conditions under which the
analyses of the bile in them were performed have therefore
been observed as closely as possible.

The bile in a perfectly fresh state was evaporated to dry-
ness in a water-bath, the residue powdered and exhausted
with alcohol of 840 specific gravity, the solution filtered, and
the alcohol distilled off at 212° F. ; the dry residue was next
powdered, dissolved in absolute alcohol, the solution filtered
* Communicated by the Author.

the Bile of the Sheep. 367

and digested with animal charcoal : when decolorized, it was
ao-ain filtered, the alcohol distilled off, and the residue pow-
dered, exhausted with aether, and perfectly dried at 212° F.

In this state it was almost white, having a slight tinge of
buff.

The ash was prepared in a muffle at a low red heat.

I. 3-02 grains of the dried bile gave 0-305 ash =10-09 per
cent. ; 4--83gave 0-48 ash =9-93 per cent.; 4-41, charred and
washed in the manner proposed by Rose, gave 045 ash = 10*20
per cent.

The amount of chloride of sodium present in the prepared
bile was very small ; thus —

II. The 0-48 of ash from the 4-83 bile (1.) yielded 0-04 chlo-
ride of silver =0*43 per cent, of chloride of sodium ; 6*83 of
the bile gave 0'05 chloride of silver =038 per cent, of the
chloride of sodium ; the soda remaining, determined as sul-
phate, amounted to 0*99 =6-32 per cent.

III. 4-045 bile burnt with chromate of lead, yielded 8*94
carbonic acid and 3-275 water, = carbon 60-07, and hydrogen
8-97 per cent.

IV. 4*005 bile gave 8*845 carbonic acid and 3*20 water,
= carbon 60*22, and hydrogen 8*87 per cent.

V. 3'62 gave 2*29 ammonio-chloride of platinum = 3*97
nitrogen per cent.

Hence

I. II.

Carbon 60*07 6022

Hydrogen 8*97 8*87

Nitrogen 3*97

Oxygen ..... 20*29

Soda 6-32

Chloride of sodium . . 0'38 0*43

100*00

In all the specimens of the ash of bile which I have ex-
amined, on solution in water and the addition of nitrate of silver,
the yellow colour resulting from the formation of the tribasic
phosphate of silver was distinctly perceptible in admixture
with the white colour of the chloride. The yellow precipitate
was dissolved by a drop of nitric acid. Whether the phos-
phate thus indicated arises from the solubility of the phosphate
of soda existing in the bile prior to the separation of the mu-
cus in an alcoholic solution of bile, or to the oxidation by the
heat of a certain amount of phosphorus existing in the elec-
tro-negative constituent of this fluid, and its subsequent com-
bination with the soda, I have not determined.

368 Notice respecting the Meteor of September 25, 1846.

By comparing the above results with those obtained by
Kemp, Theyer and Schlosser, &c., from the analysis of the
bile of the ox, the two fluids are seen to exhibit the same
composition.

The nature of the true constitution of the bile is still a mat-
ter of doubt; the opinion that it was a compound of an elec-
tro-negative substance (bilic acid) with the base soda seemed
latterly to have been almost established. If however the ex-
periments of Mulder, which have recently been published,
should be confirmed, no dependence can be placed upon
direct analysis, since from the moment of the secretion of that
fluid it begins to undergo decomposition : even on drying at
212° F. ammonia is evolved, and the bile ceases to be perfectly
soluble in water; and all fresh bile contains ammonia. Should
these results be proved correct, the analysis of this fluid must
be conducted in a different way from that which has been
ordinarily adopted. On dissolving some purified fresh sheep's
bile in alcohol, adding a drop of muriatic acid, then a little
chloride of platinum, and setting the mixture aside, I obtained
a precipitate of the ammonio-chloride, the crystals of which
were perfectly distinct under the microscope. This appears
to give support to Mulder's statement that ammonia is present
in the bile.

9 St. John's Square, August 1847.

LX. Notice respecting the Meteor qf September 25, 1846.
By the Rev. J. Slatter.

To the Editors of the Philosophical Magazine and Journal.

Rose Hill, near Oxford,

Gentlemen, October 2.3, 1847.

A S I do not generally see your Publication, I was quite
■^^^ unaware of any accurate notice having been put on
record of a large meteor which appeared one night in the end
of September 1846. During the late meeting of the British
Association at Oxford, a conversation arose, from which I
learnt that Sir John Lubbock had observed it also, and made
a communication to your Magazine respecting it*.

I saw it myself in lat. 51° 43' 50" N., and long. 1° 13' 45" W.
It passed from E. to N.E. at an altitude at first of about 50°,
declining somewhat towards the end of its course, but not
more to my notion than would be caused by perspective, sup-
posing its path to have been on a meridian line, and parallel
to a tangent at the earth's surface. The night was very cloudy,
but there were many openings between the clouds. The body
* In the January Number for this year, p. 4.

Mr. J. Glaislier on the Aurora Borealis. 369

of the meteor was visible at these points, and appeared round,
and certainly not less than 15' in diameter, — I should say
double that measure. I was in some degree enabled to judge
by estimating, after it had passed, the size of the gaps in the
clouds where it was fully visible. The light was very great,
enabling me to see surrounding objects as plainly as during a
vivid flash of lightning, and lasted about two seconds.

Now to compare my observation with the diagram and
notice sent you last year by Sir J. Lubbock, I conclude he
must have seen the meteor just before its disappearance; in
which case, the course being very much foreshortened, it
would occupy the portion of the heavens which he has indi-
cated by a blurred mark of his pencil. On this hypothesis it
must have passed about 8° or 10° from the zenith of his place
of observation, which I suppose to be in longitude 0° 4-'*5 W.,
lat. 51° 20' N.

1 consider then that the meteor at the end of the phoeno-
menon bore N. by 10° W. at Sir J. Lubbock's station at
an altitude of about 40°; at my station at the same instant it
bore N.E. at an altitude of 45°. From these data, I calculate
its height to have been sixty-one miles nearly.

But taking its course as upon a meridian line, and the esti-
mated altitude when due east of me, I make its height about
fifty-six miles. Considering the roughness of the data, I re-
gard this degree of accordance, proceeding upon two inde-
pendent methods, as tolerably satisfactory. Then, if my esti-
mate be at all correct, it had a diameter of at least 700 yards,
and its velocity was thirty-six miles in a second.
I remain, Gentlemen,

Your obedient servant,

John Slatter.

LXL On the Aurora Borealis, as it isoas seen on Su7iday
eve?iijig, October 24, 1847, at Blackheath. By James
Glaisher, Esq., of the Royal Observatory, Greenwich,

THIS day having been remarkable for one of the most
brilliant displays of Aurora Borealis which it has ever
been my good fortune to witness, it has occurred to me that a
notice of its principal phases, so far as they fell under my own
observation, may not be unacceptable to your readers.

The barometer reading during the day previous had de-
clined rapidly, and during this day it had increased as rapidly.
The day had been for the most part overcast, and light rain
had fallen occasionally ; towards evening the sky became per-

Phil, Mag. S. 3. Vol. 31. No. 209. Nov. 1847. 2 B

370 Mr. J. Glaisher on the Aurora Borealis

fectly cloudless; the night was beautiful, and the full moon
shone with unusual brilliancy.

At about 6^^ SO'" p.m. a bright red streamer was seen to
spring up from the N.W.

At 6'' 40"^ another streamer was seen in the N.W., and at
the same instant one sprung up from the N. ; both of which
were of a beautiful red.

At 6'' 56™ a less brilliant streamer was seen in theN.W., and
withiu three minutes after this time, several faint streamers
were seen in the N., N.N.W. and N.W,

From 7^' to 7'^ 12™ a few streamers were seen, and after this
time no trace of the Aurora could be seen for some time.

Between 7'^ 30'" and 9^' 40*" there were occasional streamers,
both white and red, appearing between the N.W. and the
E.N.E.

At 9'' 55^ a splendid column of red light appeared in the
N.W., whose base was about 5° in breadth. This pyramid
exhibited all the lints of the most brilliant sunset, and appear-
ed to be composed of streamers whose colours shaded from
the most intense crimson into the ruddiest and most brilliant
orange, which orange parts again contrasted with the ruddy
hue of the next portion, forming b}' means of contrast upon
contrast an endless gradation of shade and colour, — a truly
sublime and gorgeous appearance. About this time, the fur-
nace glow which pervaded this appearance increased in in-
tensity, and had all the appearance of the reflexion from an
immense conflagration ; in the mean time the orange colour
entirely disappeared, and gave place to an uniform deep crim-
son, increasing, as before stated, in intensity, and apparently in
denseness.

At lO'i 0"^ the same appearance continued as above; but in
addition to it, there was a collection of vertical columns of
light from 2° to 3° in breadth; and from the E.N.E, there
was a column similar in form and colour to the one in the
N.W., with the exception of being less brilliant. These two
red columns formed the east and west boundaries of the fan-
like appearance of the whole mass, all the columns of which
converged to a point a ^qw degrees S. of the zenith.

The columnar appearances situated between the red columns
were of the most silvery light, shaded with a most delicate and
pure gray; theywere perpetually glancing and shifting upwards
and downwards; the lower parts of each column would suddenly
glance into the place of the upper portion of the same column,
whilst the upper portion would shoot higher towards the zenith,
and then both together suddenly descend. This vibrating mo-
tion was simultaneous in all the columns, excepting the splendid

of October 2% 1847. 371

red portions at either termination, which remained immove-
able, though it rather appeared, that as the central silvery
light fluctuated, now bright, now dim, these rosy extremities
fluctuated in direct opposition, their rosy hue becoming fainter
and inclining to a neutral tint in proportion to the increase
of the silvery brightness. The whole variation of appearance
somewhat resembled the reflexion cast upon a wall by a Gothic
casement lighted from within by some fitful and inconstant
light. Towards 10^ 12"™ a considerable diminution in the
brilliancy of the light, fleecy, silvery columns took place; the
regular and casement-like appearance disappeared by degrees
and assumed more of the character of the extremities, although
they still continued their fitful, glancing and radiating motion.
During these appearances two or three milk-white, cloud-like
masses came up from the N.W. and slowly moved towards
the S.E. ; each of these masses seemed to have a kind of pulsa-
tion within themselves.

At 10^ 19"^ little could be seen of the Aurora, excepting
the red column in the N.W. ; this still retained much bril-
liancy, though all else seemed merged into the sky, when at
times, like the bursting of a firework, a stream would spring
up from this column, white and brilliant, except at their upper
portions, which were tinged with rose colour.

About this time, the moon, which had been shining upon
a cloudless sky, was suddenly surrounded with a splendid
corona, exhibiting concentric circles, first of a neutral tint,
next of violet, then green, and the outermost red ; the ex-
ternal boundary of the latter passed nearly midway between
the moon and the planet Mars ; this appearance continued at
its extreme brilliancy a short time only, but more dimly it
continued for a long time.

From \Q)^ SO"" to U** 0% with the exception of an occa-
sional streamer, there was no appearance of the Aurora; and
at times no aurora at all was visible.

At 1 1^ 14™, to this time no arch-formation had been seen,
or bank of vapour; a bright arch however was supposed to
have formed at about this time, but, if so, it continued a short
time only.

Shortly after 11^ IS"^ a faint stream or column of white light
was seen in the N.N.E.,and a splendid red patch of light, nearly
in the east, was seen, which grew very bright, and the phee-
nomenon at midnight exhibited an appearance as beautiful as
any of those that had preceded it. An arch appeared extend-
ing from the N.W. to the S.E. ; from this arch very bright
and flickering pencils of light darted out, both upwards and
downwards.

2B2

872 Royal Society.

At 12^ 30"^ the streams were frequent; the arch now ex-
tended from the N. by W. to tiie E. by N., and at every part
of this arch an occasional streamer, with its taper-like form,
sprung up; and this appearance continued till after \'S^.

I did not observe any halo around the moon at any time,
and the Aurora, with the exception of the beautiful white
clouds, was confined to the northern hemisphere.

On Friday the 22nd, and on Saturday the 23rd, the mag-
netic instruments at the Royal Observatory were greatly
disturbed, as they were during the auroral appearances on
the 24th ultimo*.

Many of the preceding observations were made by an
assistant at my residence, as my own attention was almost
completely occupied by observations of the magnetical instru-
ments; so much so, that I was oblijied to neijlect some of its
finest appearances, but which I believe were pretty vvell ob-
served as above described. The watch by which the times
were taken was compared at about midnight, so that the se-
veral times are true Greenwich mean solar times.

James Glaisher.
Blackheath, Oct. 26, 1847.

P.S. An Engraving of its appearance, as seen at about 10^,
will appear in the Illustrated London News of Oct. 31.

LXII. Proceedings of Learned Societies.

ROYAL SOCIETY.
[Continued from p. 227.]
June 17, "/^N the Solution of Linear Differential Equations." By
1847. ^^ Charles James Hargreave, Esq., B.L., F.R.S., Pro-
fessor of Jurisprudence in University College, London.

1. By the aid of two simple theorems expressing the laws under
which the operations of differentiation combine with operations de-
noted by factors, functions of the independent variable, the author
arrives at a principle extensively applicable to the solution of equa-
tions, which may be stated as follows : — " if any linear equation
<p{x,'D).u=X have for its solution M=\|/(a?,D).X, this solution being
so written that the operations included under the function ^ are not
performed or suppressed, then (p(D,— a?).M=X has for its solution
M=i//(D,— a?).X." The solution thus obtained may not be, and often
is not, interpretable, at least in finite terms ; but if by any trans-
formation a meaning can be attached to this form, it will be found
to represent a true result.

An important solution immediately deducible from this principle
is given by Mr. Boole in the Philosophical Magazine for February

• See the weekly reports of the weather supplied by the Astronomer
Royal to the Registrar-General.

Royal Society, 373

1847, and is extensively employed in the present paper. It is imme-
diately obtained by making the conversion above proposed in the
general equation of the first order and its solution.

2. By the use of this theorem and the general theorems above
referred to, the solution of the equation

D'M + 2Q.DM + ^c^ + Q2 + Q'— ^ ^^ ^ju = P,

is found in the form

7,f=x'»£-/Q'^(D24-c2)»'-i{a;->(D2+c2)-'»(a;-('»-i) .g/Q*'.P)} ;

of which various particular cases and transformations are given and
discussed ; including the well-known forms

D^M + — Dm + c": M = P,
X —

^ ,^ / m(m—\)\ ^

and extensions of these forms.

The application of the process to equations of the third and higher
orders gives rise to solutions of analogous forms ; and in particular
the equation

is solved in the form

bnZ^ + bn-iZ"-^ + ..._b„ A , B

where — s—. „_. ,- — t 1 « + ...

a»2;»-f-a„_iz" '-)-... a„ x—a x—p ;

and by the application of the theorems first referred to, a still more
general form is solved.

The solutions above-mentioned are subject to the important re-
striction that m, A, B, &c. (denoting the number of times that the
operations are to be repeated) must be integer ; but in the subse-
quent part of the paper, a mode is suggested of instantaneously con-
verting these solutions into definite integrals not affected by the re-
striction.

3. The interchange of symbols above suggested frequently renders
available forms of solution which otherwise would not be interpret-
able in finite terms. The ojjeration (fD)^ is not intelligible if m
be a fraction ; but if by any legitimate process this be changed into
the factor (<p{—x))% the restriction ceases to operate. By the ap-

874 'Royal Society.

plication of this principle, solutions of a simple character are ob-
tained for (b being integer),

(a;2+c2)D2M— 2wDM+ft(2a— 6+ 1)m=P,

d^u_b{b + \)

dt^ cosH ""• '

<px,'D'^u + ^x.Du + ('<p'x-^0''x)u='P.

4. The advantages of the forms above given in this particular,
that the number and order of the operations in the solution are ex-
pressed geyierally, and not by a series of substitutions involving
changes of the variable as in the ordinary mode of solving Riccati's
equation, appear more clearly in the application to partial linear
diflPerential equations. Thus, the equation

d^u In d^u
dx'^ X dxdy

which may be solved by m successive substitutions, receives its
solution in the general form

which exhibits at a glance all the successive processes to be per-
formed upon "i/ix^y) in order to arrive at the result. It v/ill be ob-
served that the process e*^'^' performed upon ■i^y denotes •i^{y-\-<^x).
Among other results worthy of notice on this branch of the subject
may be noticed the solution of

d'^u a (du du\ a(a — l)^m(m—l)_
d^+^q[d^ + d^J+ (p+qy =?'(/''^)

(solved by Euler in a series when there is no second term) ; viz.

j^_^m-«CD2_D'2)™-i|a:-'(D'^— D''^)-™{a?«-'"+i.\J/(a;,?/)}| ;

t|/ being determined from <p by the equations -^jcic 4:^; and the solu-
tion of

(a„x+bn)^^ + (an-ix+b„-0^-^^-^+.. + (a,x^bo)~=,p(x,y)

which is readily deduced from the solution of the corresponding
form in ordinary equations.

5. The character of most of the solutions may be described as
follows : they consist in the performance (repeated m times) of ope-

Royal Society. S75

rations of the form ^D upon the second side X; multiplication by
the factor x-^ ; and the performance (repeated m — \ times) of the
inverse operation (^D)-'; and it will be seen that, in all cases
where X=0, it is sufficient to perform the direct operation ^D a
single time.

It is a remarkable phenomenon connected with the solutions last
mentioned, that they are instantaneously convertible into definite
integrals by changing (pD into <pz, multiplying by e*-^, changing x~'^
into D'-' (D' denoting differentiation with regard to z), and assign-
ing proper limits for the integral. In this manner definite integrals
are immediately found for

D«M+-=0,

X

D»M+a?.M=0,

(a«a; + ft«)D'*M + . . + {UfjK + io) w = 0,

and other forms.,

6. The application of the principle above stated to equations of
finite differences gives solutions for the equations

{anX-\-bn)Ugj,n+ + (aia? + &i)war+i + («o^ + ^oV^=Q*>

and where the number of operations to be performed is denoted by
a fraction, solutions are found in the form of definite integrals.
The solution of the first when Q^=0 is

tly 0
+y/3(a»^;'»+. •a.t' + ao)" V''(^-'^)M^-^)^2 • • -^^-^^^

+ &c. ;

and that of the second is somewhat similar.

From some investigations effected by interchanging the symbols
X and D in the solution of the general linear equation in finite dif-
ferences of the first order, it would seem that definite summations
may be used to represent the solutions of certain forms of equations.
Thus a partial solution of

is c2(r;2!)»f*« from «= — a to z==0.

7. In attempting the solution of some equations by means of suc-
cessive operations, not consisting exclusively of D combined witli

376 Cambridge Philosophical Society.

constants, but involving also functions of x, the only result which
appeared to the author worthy of notice is the solution of

D%+6Dm+c2m-w(w+ 1)3;^= X ;

from a particular case of which, the general solution of Laplace's
equation,

may be found in the simple form

with a similar function using — V^ — l for v — 1.

CAMBRIDGE PHILOSOPHICAL SOCIETY.
[Continued from p. 311.]

On the Symbolical Equation of Vibratory Motion of an Elastic
Medium, whether Crystallized or Uncrystallized. By the Rev. M.
O'Brien, late Fellow of Caius College, Professor of Natural Philo-
sophy and Astronomy in King's College, London.

The object of the author in this paper is twofold : first, to show
that the equations of vibratory motion of a crystallized or uncrj^stal-
lized medium may be obtained in their most general form, and very
simply, without maldng any assumption as to the nature of the mo-
lecular forces ; and secondly, to exemplify the use of the symbolical
method and notation explained in two papers read before the Society
during the present academical year.

First, with regard to the method of obtaining the equations of
vibratory motion.

This method consists in representing the disarrangement (or state
of relative displacement) of the medium in the vicinity of the jioint
X1/Z by the equation

i. dv X' , dv r. , dv ^ I d-v ^ ,, , d"v j i> , n -o

ov= --dx+ — 01/ + — dz-\ ?x^+ dxdy + &c. — &c.

dx dy dz 2 dx'^ dxdy

(where i' = ^a + ij/3 + ^y, ^15^ denoting, as usual, the displacements
at the point xyz, and ajSy being the direction units of the three
coordinate axes), and in finding the ivhole force brought into play at
the point xyz (in consequence of this disarrangement) by the symbo-
lical addition of the different forces brought into play by the several
terms of Bv, each considered separately. It is easy to see that these
different forces may be found with great facility, without assuming
anything respecting the constitution of the medium more than this,
that it possesses direct and lateral elasticity. By direct elasticity we
mean that elasticity in virtue of which direct or normal vibrations

Camh'idge Philosophical Society. 377

take place ; and by lateral, that in virtue of which lateral or transverse
vibrations take place.

The forces due to the several terms of h are obtained by means
of the following simple considerations.

Let AB be any line in a perfectly uniform medium, and conceive
the medium to be divided into elementary slices by planes perpen-
dicular to AB ; let OM{=x) be the distance of any slice PP' from
any particular point O of AB, and suppose this slice to suffer a dis-
placement equal to — cx'^ (c being a constant) in the direction OAB,

and the other slices to be similarly displaced. Then it is evident
that the medium suffers by these displacements a uniformly increasing
expansion in the direction OB, and a uniformly increasing condensa-
tion in the direction OA ; the rate of increase both of the expansion
and condensation being c. Now in all known substances, whether
solid, fluid, or gaseous, a disarrangement of this kind would bring
into play on the slice O a force along the line AB proportional to
the rate of increase c, i. e. a force Ac, A being a constant depending
upon what we may call the direct elasticity of the substance.

Again, suppose that the slice PP' receives a displacement — cx'^

in the direction OC perpendicular to AB, and the other slices similar
displacements. Then the line AB will become curved into a para-
bola A'OB', and all the lines of the medium parallel to AB will be

similarly curved, the radius of curvature being equal to — and per-
pendicular to AB. Now in all known substances* a disarrangement
of this kind would bring into play upon the slice O a force in the
direction OC proportional to the curvature c, i. e. a force Be depend-
ing upon what we may call the lateral elasticity of the substance.

Lastly, suppose that M.V^=y, and that the point P of the medium
receives a displacement cxy parallel to AB, and the other points
similar displacements. Then the slice PP' will, in consequence of
this kind of displacement, turn through an angle tan~'(ca;) into the
dotted position, and the other slices will suffer similar rotations,
those on the other side of O, such as QQ', turning the opposite way.
Now it is easy to see that a disarrangement of this kind produces a
uniformly increasing expansion in the direction OC, and a uniformly
increasing condensation in the direction OC, the rate of increase
both of the expansion and condensation being c. But the expansion
and condensation here described are quite different from that pre-
viously noticed ; since it is produced, not by displacements parallel
to C'C, but by lateral displacements, i. e. perpendicular to C'C. On
this account all that we can assert without further investigation is,
that the force brought into play upon an element at O by this dis-
arrangement acts along the line C'C, and is proportional to c, i. e.
equal to Cc, where C is some constant evidently depending in some
way both upon the direct and lateral elasticity of the medium.

* Fluids and gases possess lateral elasticity as well as solids, only in a
comparatively feeble degree.

378 Cambridge Philosophical Society,

There is however a very simple way of finding the precise value
of the force brought into jilay by a disarrangement of this kind ; for
if we turn the axes of x and y in the plane of the paper through an
angle of 45°, it appears that this disarrangement is nothing but a
combination of the two kinds of disarrangement previously noticed ;
and from this it immediately follows, in the. case of an uncrystallized
medium, that the force brought into play at O is (A— B)c ; in other
words, the coefficient C, which must be multiplied into c, in order
to give the force brought into play by the disarrangement cxy, is
equal to the coefficient of direct elasticity (A) minus the coefficient
of lateral elasticity (B).

In the case of a crystallized medium, it may be shown that six
relations, corresponding to the relation C=A— B, are most probably
true, and are essential to Fresnel's theory of transverse vibrations ;
that is to say, the medium is capable of propagating waves of trans-
verse vibrations if these six conditions hold, but otherwise it is not.

In employing the above considerations to determine the equations
of vibratory motion, the directions AB and C'C are always taken so
as to coincide with some two of the three coordinate axes ; and it is
this circumstance that makes the method peculiarly applicable to
crystallized media. Indeed, if it were necessary to take the lines
AB and C'C in any directions but those of the axes of symmetry,
the above considerations would not apply without considerable mo-
dification.

The equations of vibratory motion obtained by this method for an
uncrystallized medium, are the well-known equations involving the
two constants A and B. The equations obtained for a crystallized
medium are perfectly free from any restriction of any kind, are appli-
cable to all kinds of substance, whether we suppose its structure to
be analogous to that of a solid fluid or gas, and hold for all kinds of
disarrangement, whether consisting of normal or transverse displace-
ments, or both.

When we introduce the six relations between the constants above
alluded to, and moreover assume that the vibrations constituting a
polarized ray are in the plane of polarization, we arrive at Professor
MacCuUagh's equations*. If, on the contrary, we suppose the vi-
brations to be perpendicular to the plane of polarization, we arrive
at equations which agree exactly with Fresnel's theory in every par-
ticular-j-.

If we introduce these six relations into the equations for crystal-
lized media deduced from M. Cauchy's hypothesis, that the mole-
cular forces act along the lines joining the different particles of the
medium, it will be found that these equations are immediately re-
duced to the equations for an uncrystallized medium. From this it
follows that M. Cauchy's hypothesis cannot be applied to any but
uncrystallized media. In fact, it may be easily proved that if the

* Given in a paper read to the Royal Irish Academy, Dec. 9, 1839,
page 14.

t On this subject see a paper by the late Mr. Greene in the seventh
volume of the Cambridge Transactions, p. 121.

Cambridge Philosophical Society. 379

equations derived from this hypothesis be true, a crystallized medium
is incapable of propagating transverse vibrations.

Secondly, respecting the use of the symbolical method and notation
above alluded to.

The application of the symbolical method and notation to the subject
of vibratory motion is very remarkable, and leads to equations of
great simplicity. In the case of an uncrystallized medium, the three
ordinary equations of motion are included in the single symbolica-
tion equation

dt^ \d9!'' dy'' dz^i ^ \ dx '^ dy ' dz ) \dx ^ dy ^ dz )

If we employ the notation Am'.m, £md 9,ssurae the symbol 50 to re-
present the operation

d I /oj^ , d
dx dy dz '
the equation of nogtion becomes

^ =B(AlD.3D)v+(A-B)-i-AlD.v ;

or, by using the notation Dm'.w also, it may be put in the form
^ = { AIDAlD.-B(DlD.)°-}y.

The symbol 3D written before any quantity U which is a function
of xyz, has a very remarkable signification ; the direction unit of the
symbol 3DU is that direction perpendicular to which there is no va-
riation of U at the point xyz, and the numerical magnitude of 3DU is
the rate of variation of U, when we pass from point to point in that
direction.

The symbols AiP.u and DiD.v have also remarkable significations.
AlD.u is a numerical quantity representing the degree of expansion,
or what is called the rarefaction of the medium at the point xyz.
DtD.u represents, in magnitude, the degree oi lateral disarrangement
of the medium at the point xyz, and, in direction, the axis ^bout which
that displacement takes place.

These two symbols may be found separately by the integration of
an equation of the form

d''\J _^ (d^V rfnj rf2U\
dt^- \c?j?2 dy'^ dz^J'

When the six conditions above alluded to are introduced, the
equation of motion for a crystallized medium becomes

380 Royal Astronomical Society.

where Aj A^ A3 are the three coefficients of direct elasticity with
reference to the three axes of symmetry, and B, B/ B^ B2' B3 MJ the
six coefficients of lateral elasticity with reference to the same axes.
If the vibrations be transverse, this equation is reducible to the
form

^ = -(DID.) V?a+&^oj/3 + c^?7)

= — (DlD.)^(a -aAa + i^jS A/3 + c^^y)v,

assuming the vibrations of a polarized ray to be perpendicular to
the plane of polarization.

The well-known condition that a plane polarized ray may be
transmissible without subdivision, and the velocity of propagation
may be immediately deduced from this equation.

If we assume the vibrations of a polarized ray to be in the plane
of polarization, the equation becomes

1!H = _DlD.(a^aAa+62/3A/3 + c2yA7)D5D.u.

This includes Professor MacCullagh's three equations.

ROYAL ASTRONOMICAL SOCIETY.
[Continued from p. 146.]

May 14, 1847. — Extract of a letter from Mr. Adams, with new
Elements of Neptune.

" The following elements of Neptune have been obtained by taking
into account Prof. Challis's observations made since the reappear-
ance. * * * The elements are now sufficiently correct to enable me
to approximate to the perturbations of Neptune by the action of
Uranus, in order to compare more accurately the ancient observa-
tions of 1795 with those .... made recently. I have used the old
observations, supposing the elements not to have changed. I hope
immediately to set about a new solution of the perturbations of
Uranus, starting with a very approximate value of the mean distance.
* * * I do not think, with Professor Pierce, that the near commen-
surability of the mean motions will interfere seriously with the re-
sults obtained by the treatment of perturbations ; but it will be in-
teresting to see how nearly the real elements can be obtained by
means of the perturbations."

Elements of the Orbit of Neptune.

Mean longitude, Jan. 1, 1847, G. M. T... 328 13 54-5 T

Longitude of perihelion (on the orbit)... 11 13 41-5 LM. Eq. 1847-0

Longitude of ascending node 130 5 39-0 J

Inclination to ecliptic 1 47 I'S

Mean daily motion 213774

Semi-axis major 302026

Eccentricity of orbit 0-0083835

On the communication of Mr. Adams's paper, the Astronomer

Royal Astronomical Society. 881

Royal gave orally a continuation of the history of Neptune, embra-
cing the ])rincipal points that have been ascertained since his com-
munication of Nov. 14, 1846. The planet having been actually dis-
covered in the heavens by means of certain predicted elements, the
fair presumption was that those elements were very approximately
correct. Adopting these elements, therefore, Mr. Hind examined
Lalande's and other observations, with the hope of finding some
former observation of the planet as a star now missing, but satisfied
himself that there was none. In the meantime, the continuation of
the observations of the planet in the last months of 1846, and the
comparison of them with Professor Challis's early observations of
August, led to some unexpected conclusions. It was found that,
though one^/«ee of the planet might be very well represented by M.
Le Verrier's or Mr. Adams's elements, yet the apparent movement of
the planet could not be represented within several minutes. Elements
were then investigated from the observations themselves (without
any reference to the preceding deductions from the perturbations of
Uranus) by Mr. Adams in England (see Monthly Notices for March,
p. 244), and by Professor Pierce and Mr. Sears C. Walker in Ame-
rica. Attention is particularly due to the former of these investiga-
tions, in which are exhibited, not only the results for the different
elements, but also for the probable error of each. The most import-
ant conclusion was, that the planet certainly moved in a much smaller
orbit, and probably in an orbit of much smaller eccentricity, than
that indicated by the calculations of perturbation. With elements
thus roughly corrected, the orbit was again traced back through the
ancient observations ; and it was found by Dr. Petersen of Altona,
and Mr. Sears C. Walker, that a star observed by Lalande on May
10, 1795, and now missing from the heavens, was very probably the
planet. The observation however vi'as marked doubtful in Lalande's
printed volume : and to this circumstance is probably due a most
remarkable discovery. The manuscripts of Lalande's observations
were some years ago transferred by his representatives to the obser-
vatory of Paris. To examine into the presumption of doubt in the
observation, the astronomers of the Observatory of Paris referred to
the originals, and there they found that the observation of May 10,
1795, was entered without any expression of doubt at the time;
that an observation of May 8, 1795, was omitted in the printed vo-
lume ; that it was omitted solely because it could not be reconciled
with the observation of May 10 ; and that, upon reducing both pro-
perly, they exhibit most distinctly the retrograde motion of a planet
nearly parallel to the plane of the ecliptic, the right ascension and
the polar distance having both changed in the proper proportion.
It seems now inconceivable to us that an astronomer, having his atten-
tion strongly called to the difference between the two days' results,
should rather assume that there were in the observations two inde-
pendent errors (one of right ascension and one of polar distance),
than that the body observed was really a planet. With the place of
the planet at an epoch so distant, its elements are ascertained with
great accuracy.

S82 Royal Astronomical Society.

It is remarkable that the missing star, to which allusion has been
made, is actually entered as an observed star in the Berlin Star-Map ;
and this circumstance prevented Mr. Adams from tracing the new
orbit of the planet so soon as he would otherwise have done. This
insertion of an unobserved star can be accounted for only on the
supposition that the star had been taken by the observer in his work-
ing-catalogue as a zero-star, and had then been inserted as a matter
of course.

The mean distance of Neptune from the sun now appears, instead
of 38, to be something near 30 ; and its periodic time, instead of
220 years, to be nearly 166. It is certainly a most curious thing
(in which much is owing to chance) that elements, now known to
be extremely erroneous, should have accounted for the perturbations
of Uranus through 150 years with such accuracy, and should also
have given the planet's place, for the particular year in which the
attention of astronomers was first strongly directed to it, with such
precision. It remains to be seen whether the new elements of Nep-
tune will, with any possible mass, explain the perturbations of
Uranus. In any case, Bode's law, on the assumption of which the
original investigations of M. Le Verrier and Mr. Adams entirely de-
pended, fails completely.

Calcul detaille d'une In^galit6 Nouvelle a Longue Periode, qui
existe dans la Longitude moyenne de la Lune. By M. Hansen.

The author states that he has lately made known to some astro-
nomers a discovery of two inequalities in the motion of the moon,
whose periods are respectively nearly 273 and 239 years. Denoting
by g, g' , g" the geocentric mean anomaly of the moon, and the helio-
centric mean anomalies of the earth and Venus, these inequalities are^—
27"-4x sm(-g-l6g'+l8g" + 35° 20'-2)
-f-23"-2 X sin (8i^"-13^'-f-315° 30') ;
of which the first depends on a new argument, while the second
depends on the argument of an equation of long period in the motion
of the earth, discovered by Mr. Airy.

As the calculation of those parts of the coefficients which depend
on the product of the square and cube of the sun's disturbing force
by the disturbing force of Venus is extremely laborious, and is more-
over connected with other unpublished calculations of other inequa-
lities of the moon, it does not appear possible to publish it at pre-
sent. Indeed M. Hansen does not consider himself able yet to
answer for their perfect correctness, though he has the strongest
reason to believe that they are very nearly correct. The present
paper therefore includes only the calculation of that part of the co-
efficient of the first inequality which depends on the first power of
the disturbing force.

It appears difficult to abstract very completely the remaindef of
this paper, but the following indications will enable a person ac-
quainted with the developments of physical astronomy to follow the
whole process.

The perturbing function CI for the moon as disturbed by Venus
being formed, it will be found that it may be expanded in a rapidly

lloyal Astronomical Society. 383

converging series of fractions, whose numerators contain successive
powers of r, the moon's radius vector, and whose denominators
contain different powers of the same multinomial (which, when ec-
centricities and inclinations are omitted, is a trinomial) that occurs
in computing the perturbations of the earth by Venus. Upon ex-
panding any of these fractions with trinomial denominator, there
occur terras depending on \Q(j" — \Gg\\lg" —ilg' ,axidi \Sg" —\%g' :
then, upon introducing the inclinations and eccentricities, the first
(among other combinations) will be multiplied by sin^ \ inclin. x
cos 2g"—'2 V (where v is the difference of longitude of node and pe-
rihelion of Venus), and also (in other terms) by e"'^ cos 2 g' ; the
second by e". e' cosg" + g' ; and the third by e'^ cos 2 g'. Each of
these combinations produces terms whose argument is IS g''—16 g'.
Then upon multiplying these terms by a power of r, since the ex-
pression for any power of r contains e. cos g, the product will contain
terms depending on IS g''— 16 g'~g. The coefficient necessarily
contains one of the following products of three small quantities :
p. sin^ ^ inclin., e . e"^, e.e'.e', c.e''^ (of which the first is the most im-
portant), and it is therefore extremely small; but the resulting
perturbation is made important by the excessive smallness of the
divisor introduced in integration. It is well known that the divisor

in this case will be proportional to ( 18 -^ — 16 -^ ^ | ; and,

*^ ^ \ dt dt dt)

tin"

taking for -—-, &
dt

to a Julian year.

dn"

taking for -^, &c., the value in sexagesimal seconds corresponding
dt

^ =2106G41"-3
dt

^=1295977-4
dt

*^ = 17179167'4
, dt

whence 18^ -16^' -^ =4747"-7,
dt dt dt

a quantity very small in comparison with -^.

dt

In this manner the greatest part of the term in question is pro-
duced. Other parts arise from the circumstance that, the dimen-
sions of the moon's orbit being slightly altered, the perturbing force
of the sun upon the moon is not the same as it would otherwise be.

M. Hansen remarks that this term is remarkable as depending
upon higher multiples of the anomalies than have ever before been
considered, and as having the longest period in proportion to the
periodic time of the disturbed body that is yet known.

The term depending on S g" —\^ g' arises mainly from the cir-
cumstance, that, the earth's motion in its orbit being different from
what it would have been without the perturbation by Venus, the
disturbing force of the sun upon the moon is not the same as if that
perturbation had not existed.

SS* lioyal Astronomical Society.

M. Hansen states that he has examined several inequalities of
long period in the moon's motion which hitherto have escajjed notice,
but that in no other instance does the coefficient amount to 1".

In concluding the account of this remarkable discovery, it is gra-
tifying to add that it explains almost precisely the observed inequa-
lity in the moon's mean motion, which for the last fifty years has
troubled physical astronomers.

After the reading by the Secretary of a portion of this paper, the
Astronomer Royal gave an oral explanation of its general subject in
the following manner : —

The disturbing effect of Venus upon the moon is not the whole
attraction of Venus upon the moon, but the difference of the two
attractions, of Venus upon the moon and of Venus upon the earth.
Thus, when the moon is between the earth and Venus, the attrac-
tion of Venus upon the moon is stronger than that of Venus upon
the earth, and therefore it tends to pull Venus away from the earth.
When the moon is more distant from Venus than the earth is, the
attraction of Venus on the earth is the stronger, and tends to ])ull
it away from the moon, which, in regar'd to the disturbance of the
relative jilaces of the earth and moon, is the same thing as pulling
the moon away from the earth. In both these positions, therefore,
the disturbing force of Venus tends to pull the moon away from the
earth. When the earth and the moon are equally distant from Venus,
the attractions of Venus upon the two are equal, but not in parallel
lines ; the attractions tend to draw them along the sides of a wedge
whose point is at Venus, and, therefore, to diminish the distance
between them, or to push the moon towards the earth.

Inasmuch as, in one pair of positions of the earth and moon, the
disturbing force of Venus tends to increase the distance between
them, and in another pair of positions it tends to diminish that di-
stance, it is important to ascertain which of these disturbances is the
greater. Suppose the distance of the moon from the earth to be
Y^ part of the distance of the earth from Venus. Then, when the
moon is between the earth and Venus, its distance from Venus is
yW of ^^^ whole ; the force upon it is '^^y* ^^ ^^^^ upon the earth ;
the excess of this (or the disturbing force tending to pull the moon
away from the earth) is -gV^' °^' nearly -^^ of that on the earth. In
like manner, when the moon is further from Venus than the earth
is, its distance from Venus is -Lg-^ of the earth's distance ; the force
upon it is ^g^g° of that upon the earth ; the defect of this (or the
disturbing force tending to pull the earth away from the moon) is
, gg' . or nearly -A- of that on the earth. But when the earth and

10201' J 5 1 , -,.

the moon are at equal distances from Venus, the proportion of then-
relative approach (as produced by the action of Venus) to the whole
effect of Venus upon them, is evidently represented by the inclina-
tion of the two lines drawn from them to Venus, or is the same as
the proportion of the distance of the moon from the earth, to the
distance of the earth from Venus, and is therefore yi^^ of the whole.
Thus the force tending to pull the moon from the earth at one time
is about double the force tending to push the moon towards the earth
at another time ; and therefore, upon the whole, the tendency of the

Royal Astronomical Society. 585

disturbing force of Venus is to pull the moon from the earth. To
arrive at this conclusion, we have considered only four points of the
moon's orbit : in other points the effects of the perturbation are
more complicated ; but they do not alter this general conclusion.

The same remark applies to the disturbing effect of Venus upon
the moon when at a given point of its orbit, provided the nature of
that point be such that at different times it is in all possible posi-
tions relative to Venus. For instance, the moon's apogee is (in
consequence of the motion of the line of apses, and of the relative
motions of the earth and Venus) sometimes between the earth and
Venus, sometimes more distant from Venus than the earth is, some-
times 90° to the right, sometimes 90° to the left. We may assert
therefore that, upon the whole, the disturbing force of Venus upon
the moon, when she is in apogee, tends to draw her away from the
earth. The same may be predicated when the moon is in perigee.

Next, it is important to ascertain how the disturbing force de-
pends upon the moon's distance from the earth. For this purpose,
instead of supposing, as before, that the moon's distance is y^-g- part
of the distance of the earth from Venus, let us suppose it y^ part of
that distance. Then when the moon is between the earth and Venus,
the force upon the moon is VWx" °^ ^^^^ upon the earth, and there-
fore the excess, or the disturbing force, is ^qj, or nearly -^ of the
whole force upon the earth. In the former assumed instance it was
■gL. Thus, upon doubling the moon's distance from the earth, the
disturbing force is doubled. And similarly for other distances of
the moon from the earth, the disturbing force (in similar positions
with regard to Venus) is proportional to the moon's distance. Thus,
when the moon is at apogee, in a given position with regard to Venus,
the disturbing force is greater than when the moon is in perigee in
the same position. And, upon the whole, in all possible relative
positions of the moon and Venus, the action of Venus pulls away
the moon from the earth, more when she is in apogee than when she
is in perigee.

Now we may consider the general effect of these forces upon the
dimensions of the moon's orbit. So long as the force which draws
the moon towards the earth is always the same at the same distance,
the moon will continue to describe an orbit of the same dimensions
over and over again. But if at any time the force directed towards
the earth suddenly grows smaller, the moon will immediately rush off
in an orbit which, on the opposite side, is larger. If the force
towards the earth gradually grows smaller, the dimensions of the
orbit will gradually increase. And the periodic time in the orbit
described at every successive revolution will undergo the change
corresponding to the change of dimensions (that is, to the change of
major axis) of the orbit, and will therefore become continually
greater and greater.

These are the changes which produce the most serious disturbance

in the apparent place of the moon. If a force, after acting for a long

time, produce a small change in the eccentricity of the moon's orbit,

the effect on the moon's place is simply the amount of the corre-

Phil. Mag. S. 3. Vol. 31. No. 209. Nov. 1 847. 2 C

386 Royal Astronomical Society,

spending change in the equation of the centre, and cannot possibly
exceed that amount. But if the force have been for a long time
gradually altering the major axis, and consequently the periodic
time in the moon's orbit, then during the whole of that time the
moon has been performing her revolutions quicker or slovi^er than we
expected, and therefore at the end of that time she is in advance
or in retard of her expected place by an amount equal to the accu-
mulation of all the advances or retards in all the revolutions through
which the change has been going on. The planetary inequalities of
long period are all of this kind. The major axis here plays the same
part as the pendulum of a clock. If a small force acting for a year
pushed the seconds-hand forwards by an inch, the clock would be
merely a few seconds wrong ; but if in the same time it shortened
the pendulum by an inch, the clock would have gained fifty hours ;
and if the time occupied by the change had been greater, the dis-
turbance in the clock indication would have been proportionably
greater.

In order then to find inequalities of long period in the motion of
the moon produced by Venus, we must seek for some alternate in-
crease and decrease, occupying a very long period, in the force by
which Venus draws the moon from the earth.

No such slow increase and decrease have been found in the general
force by which Venus disturbs the moon.

The next point of inquiry is, whether a combination of the changes
in the force of Venus with the changes in the position of the moon
in its orbit can produce a force, which, for a very long time together,
gradually increases the force drawing the moon from the earth, and
then for an equal time gradually diminishes that force.

A force which acts in opposite ways, nearly on opposite sides of
the moon's orbit (pulling the moon from the earth on one side and
pushing it towards the earth on the other side), may produce this
effect, provided the period of the change in the nature of the force
(from pulling to pushing) correspond nearly, but not exactly, with
the time in which the moon moves from apogee to perigee. For
(as we have seen) the effect of a certain force of Venus is to produce
a greater disturbing force on the moon at apogee than at perigee ;
and this force, or a change in this force, will, at apogee, produce a
greater effect on the dimensions of the moon's orbit than at perigee,
both because the disturbing force is actually greater, and because it
acts on the moon when the moon's velocity is smaller. Therefore,
if a pulling force, gradually increasing in magnitude, act on the
moon at apogee, it will gradually increase the dimensions of the
moon's orbit : if a corresponding pushing force act at perigee, it
will gradually diminish the dimensions of the moon's orbit; but the
former prevails, and the orbit will gradually increase in size. If
after a time the pulling force at apogee gradually diminish, and at
length become a pushing force, while the pushing force at perigee
gradually diminishes, and at length becomes a pulling force, then
the orbit will gradually diminish in size. And this change of forces
would be produced by such a modification in Venus's force, as that

Royal Astronomical Society. 387

of which we have spoken, namely, a force which acts in opposite
ways on opposite sides of the moon's orbit, and in which the period
in the change of the nature of the force coincides nearly, but not
exactly, with the time in which the moon moves from apogee to pe-
rigee ; for then the pulHng force at apogee will after a long time be
changed to a pushing force, and the pushing force at perigee will in
the same time be changed to a pulling force. If, for instance, the
change in the disturbing forces of Venus (from pushing to pulling)
occupied fourteen days exactly, and if the moon's motion from apo-
gee to perigee occupied fourteen days and five minutes, then in 4032
anomalistic semi-revolutions of the moon (which would bring her
from apogee to apogee), there would have been 4033 changes of the
force (which would change it from pulling to pushing), and there-
fore in this time, and no sooner, a complete pulling force at apogee
would be changed to a complete pushing force at apogee.

It is necessary now to point out how such a modification of the
force of Venus can be found.

The only disturbing forces which are yet completely brought under
the management of mathematicians are of two kinds ; a constant
force (always pushing or always pulling with the same amount of
force), and a force alternately pushing and pulling, having equal
periods and equal maximum magnitudes in each state. The latter
of tliese, if projected graphically, with the time for abscissa, is re-
presented by the ordinates of a line of sines : algebraically, it is ex-
pressed by tt.cos (^bt + c).

Now, while the relative positions of the earth and Venus change,
the disturbing force on the moon (estimated by the force which, on
the whole, it exerts to pull the moon from the earth) undergoes very
great changes. When Venus is nearest to the earth, this force is
about 250 times as great as when Venus is furthest from the earth.
It declines very rapidly from its greatest magnitude. If therefore
we represent the disturbing force from one conjunction to the next
by a curve, this curve will be very high at the beginning and end,
and very near the line of abscissa at the middle, and through the
greater part of its extent.

The separation of this foi-ce into a number of different forces, fol-
lowing the two laws mentioned above, is effected by a process sug-
gested and facilitated by algebra, but in which, nevertheless, every
step has its physical meaning. It may be stated at once, that this
remark applies universally to the algebraical operations of physical
mathematics. As a simple instance, we may refer to the equation
(a + bysz a3 -J- 3 a"^ b +3 ab'^ + b^, which probably was suggested by
algebra ; but which may be illustrated by taking a cube, whose side
is a + b, and (by three saw-cuts) cutting it into eight pieces, when
the single piece representing a^, the three pieces each representing
a- b, the three pieces each representing ab^, and the single piece
representing b^, will be found. And there is perhaps no better dis-
cipline for the mind than thus tracing the evidence of the truth of
algebra, especially in its more profound processes.

The separation, then, of the force of Venus goes on by the fol-
lowing steps : —

2C2

888 Royal Astronomical Society.

1st. A constant pulling force, equal to the mean value of the
force.

2nd. A force pulling when Venus is in conjunction, pushing at
the time intermediate to the conjunctions, and pulling when Venus
is in conjunction again ; thus going completely through its changes
once between conjunction and conjunction.

3rd. A force pulling when Venus is in conjunction, then pushing,
&c., going through its changes twice.

4th. A force pulling when Venus is in conjunction, then pushing,
&c., going through its changes thrice.

In this manner the forces go on, continually diminishing in magni-
tude. When we arrive at the 18th, the force is extremely small.

The algebraical expression for the collection of these terms, put-
ting fl for the difference of mean longitude of the earth and Venus, is

A-i-B . cos HC . cos 2 9 + D . cos 3 9 + &c,

This is on the supposition that the orbits of the two planets are
circular and in the same plane. But, in consequence of their eccen-
tricities and inclinations, the forces of any one system alternately
pushing and pulling (Nos. 2, or 3, or 4, &c.) will not have the same
maximum magnitude throughout. But each can, in all cases, be
expressed by the combination of three such forces, in each of which
the maximum forces are equal throughout. Thus, if we combine a
large force, going through its changes twenty times in a certain
period, with a small force going through its changes nineteen times
in the same period, and another small force going through its
changes twenty-one times in the same period, then it will be found
that both the small forces increase the large force (whether in its
pulling or in its pushing state) near the beginning and the end of
the time; that both diminish the large force near the middle of the
time ; and that the two small ones destroy each other at a quarter
and three-quarters of the time. The elFect of this combination is
therefore precisely such as is spoken of above.

Thus, then, for the complete expression of the force, we are
driven to an infinite number offerees following the law of alternately
pulling and pushing, but with very great variety of magnitudes of
force and of periodic time. The greatest portion of these produce
no sensible effect ; some because (though their magnitudes are large)
they act for so short time in one way, or their periods are so little
related to the periods of any movement of the moon, that their effects
never accumulate ; others because their magnitudes are small, and
there is no unusual circumstance favourable to their increase.

But there is one of these forces which, in the algebraical expres-
sion, depends on 18 x mean longitude of Venus — 16x mean lon-
gitude of the earth, whose coefficient is exceedingly small, but which
goes through all its changes, from pulling to pulling again, in the
time,

27'^ 13»» T'^SS^'G;

or from pulling to pushing, in the time

13d \^^ 33™ 47«-8.

Intelligence and Miscellaneous Articles. 389

Now, the anomalistic revolution of the moon, from apogee to apogee
again, is performed in the time

27d i3h i8«» 32«'3;
or from apogee to perigee, in the time

IS'i 18^^39"^ 16«-1.

Here we have a real instance, exactly corresponding to the case which,
for the sake of explanation, we assumed a short time back, and the
results are truly such as were there described. During about 4000
half-revolutions of the moon, or 2000 revolutions, the pulling force
at apogee is gradually diminishing till it becomes a pushing force,
and during about 2000 more revolutions, the pushing force at apogee
is gradually diminishing till it becomes again a pulling force ; the
opposite changes going on in the force at perigee : and thus, for
reasons fully explained before, the moon's orbit is gradually con-
tracting during 2000 revolutions, and gradually expanding during
2000 revolutions more. And although the change in the size of the
orbit is totally insensible in observation (for, according to a rough
calculation, the utmost accumulation of change in the major axis of
the moon's orbit is only ten feet, sometimes in increase and some-
times in decrease), yet the consequent alteration in its periodic time,
continued through so many revolutions, is sufficient to cause the
irregularity in question. The inequality in longitude, as measured
on the moon's orbit, exceeds thirty miles, sometimes in advance, and
sometimes in retard.

For a complete understanding of this matter, it must carefully be
borne in mind that the force at the apogee, which has been described
as a pushing force through 136 years, is not absolutely a pushing
force through every month of that time, but that (in consequence of
the motion of the moon's line of apses) if we take any period of nine
or ten years, the moon's apogee will in that time have passed through
every position with regard to Venus, and therefore, upon the whole,
during that period of nine or ten years, the force at apogee will have
been a pushing force. In like manner, in another period of 136
years, if we take any period of nine or ten years, upon the whole,
during that period of nine or ten years, the force at apogee will have
been a pulling force.

The general cause of the inequality depending on the argument
8p"— 13^', has been sufficiently stated in one of the last paragraphs
of the abstract of M. Hansen's paper.

LXIII. Intelligence and Miscellaneous Articles.

ON THE GELATINOUS SUBSTANCES OF VEGETABLES.

MFREMY, in a memoir read before the Academy of Sciences,
• has arrived at the following conclusions : —
1st. There exists in vegetables, along with cellulose, a substance
which is insoluble in water, alcohol and aether, which the author

890 Intelligence and Miscellaneous Articles.

n&mes peciose, and which, by the action of the weakest acids, is con-
verted into pectin. Diluted acids produce this effect only at the
temperature of ebullition ; and acetic acid, which, as is well known,
does not act upon starch, is also without action on pectose. Pectose
cannot be confounded with cellulose, for the latter, as was ascer-
tained by M. Payen, gives no traces of pectin when treated with
acids. M. Fremy's experiments confirm those of M. Payen.

2nd. The author has found in the greater number of fruits and
roots, an amorphous substance, comparable to ferments, and espe-
cially to diastase : the gelatinous substances contained in vegetables
experience by its action a series of isomeric transformations. This
substance M. Fremy calls pectase ; in acting upon the gelatinous sub-
stances it gives rise to the different phsenomena which constitute
pectic fermentation.

3rd. The acids which are employed to convert pectose into pectin,
may, according to their nature and proportion, form different sub-
stances, each of which possesses well-delined distinctive properties.
Thus, when the acid is very weak, pectin, properly so called, is ob-
tained, which does not render acetate of lead turbid. If the acid be
more concentrated, or if the ebullition has been longer continued,
the substance formed precipitates the neutral acetate of lead ; this
substance the author calls parapectin; and lastly, by employing a
powerful acid, a third substance may be formed, which is distin-
guished by the name of metapectin ; this is feebly acid to coloured
test-papers, and precipitates chloride of barium ; the other com-
pounds are neutral.

4th. If a small quantity of pectase be added to a solution of pec-
tin, and the temperature be kept at about 86° F., the pectin is soon
observed to change into a gelatinous, consistent substance. This
curious transformation, which explains the production of vegetable
jellies, may be effected without the contact of the air ; there are
formed in this case two acids ; one is new, and termed pectosic acid,
and the other is pectic acid. Pectosic acid, which might be con-
founded with pectic acid, is immediately distinguishable from it by
its perfect solubility in boiling water. In the reaction of pectase
on pectin, pectosic acid is first produced, and is afterwards changed
into pectic acid by the prolonged action of the pectase. The free
alkalies or their carbonates are capable of converting in the cold,
pectin at first into pectosates and afterwards into pectates.

The phsenomena now described are so easy of observation, accord-
ing to M. Fremy, and characterize pectin so distinctly, that he finds
it difficult to imagine how in later times pectin has been confounded
with gums, mucilages, aad especially with pectic acid, which is
insoluble in water.

The author has particularly examined pectic acid, and is of opinion
that he has overcome the difficulties attendant upon its analysis, and
especially the determination of its equivalent. He has also found
that pectic acid, heated to 392° F., loses water and carbonic acid,
and a new pyrogenous acid, which he calls pyropectic acid, is pro-
duced.

InteUiaence and Miscellaneotts Articles. 391

"is

Pectic acid possesses the singular property of dissolving in con-
siderable quantity in neutral or acid salts ; it then forms com-
pounds precipitable as jellies by alcohol ; these precipitates are
often mixed with pectin, render it gelatinous, and prevent by their
presence the recognition, by means of elementary analysis, of the
simple relations which connect pectin with the other gelatinous
bodies.

5th. The gelatinous bodies may undergo a last period of trans-
formation, and be changed into two very soluble and energetic acids.
It is sufficient to boil pectic acid in water for a certain time to con-
vert it into an acid, called by the author parapectic acid, and
into another acid termed metapectic acid. The parapectic and meta-
pectic acids are also formed during the action of acids or alkalies on
pectin or pectic acid : the pectates may by long boiling be con-
verted into metapectates. These two acids are readily distinguished
from each other ; for the first precipitates barytes water, and the
second does not ; they decompose the double tartrate of copper and
potash, as glucose does. To be certain that this property was not
derived from the presence of sugar, the author had recourse to a
polarizing apparatus and the action of yest. Guided by the advice
of M. Biot, M. Fremy found that the parapectic and metapectic acids
effected no rotary action on polarized light, and that the presence of
yest produced no traces of fermentation.

6th. After having examined all the properties of the gelatinous
bodies, and found that by employing very weak agents, comparable
to those which exist in vegetables, their acidity might be suc-
cessively developed, and from neutral bodies, which they originally
were, they might be transformed into energetic acids, the author
examined whether, during the act of vegetation, gelatinous substances
did not undergo changes comparable to those which he had pro-
duced artificially. On following for two years, with this intention,
the modifications which are effected in fruits during their matura-
tion, M. Fremy found that the gelatinous bodies which occur in
them could pass through the different intermediate states which he
has described ; thus green fruits contain abundance of pectose. As
maturation advances the pectose is changed into pectin ; and when
the fruits are perfectly ripe, the pectin is frequently completely con-
verted into metapectic acid. The modifications examined in this
memoir are then precisely those which occur during the maturation
of fruits.

The author found in the numerous analyses which he performed
that the composition of the gelatinous bodies could not be repre-
sented by carbon and water, and consequently that they were far
removed from neutral bodies, properly so called. As experiment
always indicates a larger quantity of hydrogen than really exists in
organic bodies, the author states that he cannot attribute the differ-
ence which he has obtained to an error of analysis.

The table presented to the Academy shows that all gelatinous
substances, similar to those which are derived from starch, are iso-
meric, or at least they differ only by the elements of water. This

Intellige7ice and Miscellaneous Articles.

result might be foreseen ; for when a mixture of pectase and pectin
is put into a bottle, and it is hermetically sealed, the pectin is suc-
cessively converted into pectosic, pectic, parapectic, and metapectic
acids, without forming any secondary product.

The capacities of saturation given in the following table prove that
the acidity of the gelatinous bodies increases in proportion as their
equivalent diminishes. Thus parapectin, the equivalent of which is
very heavy, forms a neutral salt with lead which contains 10 per
cent, of oxide, and does not redden tincture of litmus ; and meta-
pectic acid, the equivalent of which is very light, produces a salt of
lead which contains 67 of oxide, and its acidity resembles that of
malic or citric acid.

Names of the gela- Composition of the Composition of the

tinous substances. gelatinous substances. salts of lead.

Pectose

Pectin C6+ H« O^e, 8H0

Parapectin C^" H^o O^^ 8H0 C^ H^o 0^\ 7H0, PbO

Metapectin ...... C^*W 0^\ 8H0 C«i H<o O^^ 6H0 2PbO

Pectosic acid .... C32 H"-" O^^, 3H0 C^"- W O^s, HO, 2PbO

Pectic acid C^^ H^o O^s, 2H0 C^^ W° 0'«, 2PbO

Parapectic acid . . C^-^ H^s O^i, 2H0 C^* H'^ O"-', 2PbO

Metapectic acid . . C^ H^ 0^, 2H0 C^ H^ O?, 2PbO

M. Fremy states that the formula of pectic acid, which he has
here adopted, gives in 100 parts exactly the same quantities as de-
termined by M. Regnault, and as published by himself in his first
memoir on gelatinous bodies.

The author concludes that he has succeeded in proving that vege-
tables contain a neutral insoluble substance, which is convertible
during vegetation into an energetic acid. — Comptes Rendus, Juin 14,
1847.

PREPARATION OF PROTOXIDE OF TIN.

M. Roth gives the following process for preparing the red prot-
oxide of tin : — The white hydrate is to be prepared, and after being
well-washed it is to be digested at 133" F. in a solution of prot-
acetate of tin, with a slight excess of acid, and of specific gravity
about r06. The protoxide is then converted into hard heavy grains,
wliich yield a greenish-brown powder ; these grains inflame when
heated, and readily blacken in the sunshine. They behave with
reagents like common protoxide. — Journ. de Ph. ct de Ch., Aout
1847.

iON THE PRESENCE OF ARSENIC, COPPER AND TIN, IN THE
MINERAL WATERS OF BAVARIA.

According to the experiments of M. Buchner, Jun., the brown-
ish-yellow ochrey deposit of the springs of Ragoczy and of Pandour,
at Kissingen, contain only doubtful traces of copper ; but they con-

Intelligence and Miscellaneous Articles, 893

tain sufficient quantities of arsenic to admit of the extraction of tho
\netal.

The reddish-brown ochre of the ferruginous spring of Briickenaii
contains mere traces of arsenic, but there is much copper. Tin has
been discovered in the ochres of Kissingen and of Briickenau. Ex-
periments performed to ascertain the presence of arsenic and copper
in the brownish-yellow ochre of the ferruginous waters of Kellberg-
were not followed by any positive results. — Journ. de Ph. et de Ch.,
Aout 1847.

SOLUBILITY OF COMMON SALT IN ALCOHOL.

M. Wagner has determined the degree of solubility of chloride of

sodium in alcohol of different densities and at various temperatures.

The results are that —

o
Alcohol of 75 per cent, dissolves at 57*20 F, 0'661 part of salt.

75 .... 59'45 0-700 ....

75 100'40 0-736

75 160-70 1-033

95-5 .... 59-0 0-174

96-5 171-05 0-171 -

Ihid,

ON SOME IMPROVED FORMS OF CHEMICAL APPARATUS.
BY THOMAS TAYLOR, ESQ.

Among the many advantages possessed by the Chemical Society,
it appears to me not the least, that it affords to its members a ready
mode of communicating to one another many of those little practical
facts and modes of operating, which, although perhaps not of suffi-
cient importance to merit distinct notice in the scientific journals,
are nevertheless of considerable value to those engaged in the prose-
cution of the science. In furtherance of this view I will therefore
describe some new forms of apparatus which I have myself been in
the habit of using for some time past.

The first of these is a mode of closing the mouths of gas-bottles,
or indeed of any wide-mouthed vessel into which tubes are to pass,
as in Woolf 's apparatus, gas generators, &c. To effect this the top
of the bottle is first to be slightly ground, so as to procure a level
surface, a piece of sheet caoutchouc is then laid upon it, and this is
covered by a disc of wood of the same size as the top of the bottle,
and from a quarter to half an inch in thickness. The wooden cover is
held in its place by means of a small double clamp of brass or of
varnished sheet iron, which passes across the cover, and the ends of
which are bent under the rim of the bottle, against which they are
pressed by a screw fixed in the centre of the clamp. By turning
the screw the caoutchouc is sufficiently compressed to render the
joint perfectly air-tight. The tubes intended to pass into and out

8M

Intelligence and Miscellaneous Articles.

of the bottle are cemented into the wooden cover, usually on one
side of the clamp ; and they pass of course through corresponding
holes in the caoutchouc. By making these holes somewhat smaller
than the diameter of the tubes, the caoutchouc contracts so closely
around them, that not only is any liquid which might be accidentally
thrown up effectually prevented from getting between the caoutchouc
and the wooden cover, but the necessity of cementing the tube into
the cover may be even dispensed with. This method is so effectual
and easily arranged, that I am quite convinced it will supersede the
use of corks in the preparation of all gases which only require the
application of a moderate heat and do not act upon caoutchouc.
Ground glass plates might of course be substituted where caoutchouc
is inapplicable, or a sheet of ground glass might be cemented upon
the lower part of the wooden cover ; but these modes would be rather
expensive, and the cases in which they would be required are not
very numerous. In small bottles the use of a clamp is not essential,
as sufficient pressure may be obtained by inserting two wedges of
wood beneath a string tied around the neck, and over the top of the
bottle.

J\ .. Fig- 1-

ANTiAAA/ivfAA

Fig. 1. A clamp of sheet iron having a small centre of brass B, in which the
screw C works. D disc of wood. E sheet of caoutchouc. F glass bottle. G 11
glass tubes.

I will next direct the attention of the Members to a new mode of
cupelling, or rather to a new form of muffle. Cupellation is an ope-
ration not often performed by amateurs, chiefly I believe on account
of the difficulty in doing it unless provided with furnaces built ex-
pressly for the purpose. The following plan I have found to afford
most accurate results, while it may be performed in almost any fur-
nace : — The mouths of two black lead crucibles of the same size are
to be ground flat, so that when applied one to the other they may
stand quite steady. An oblong or semicircular notch is to be cut
out of the mouth of one of the crucibles, and a hole is also to be

Intelligence and Miscellarieous Articles.

395

drilled through its bottom. This crucible when placed upon the top
of the other constitutes the muffle, and of course resembles in shape
a skittle. To cupel with this apparatus, the lower crucible is nearly
filled with clean sand, set upon the bars of the grate in the centre of
the furnace, and brought to a low red heat. The cupel containing
the lead and the alloy is then placed upon the sand and immediately
covered by the other crucible, taking care that the notch in its side
shall be opposite to, and correspond with the furnace-door; more
fuel is added, during which it is well to cover the hole in the top
of the muffle with a crucible lid, in order to prevent the admission
of dirt. When the muffle has become throughout of a bright red
heat, the furnace-door is thrown open, and the ignited fuel gently
moved aside, so as to permit a view of the side opening in the
muffle. The current of air which is thus established through the
muffle instantly causes rapid oxidation of the lead, and this may be
regulated at pleasure by closing the door more or less. If from
the fuel falling down any difficulty should be experienced in main-
taining a free passage for the air, a portion of a porcelain tube or a
gun-barrel may be passed through the furnace-door to within an
inch of the muffle ; but this proceeding is generally rendered quite
unnecessary by taking care to place some large pieces of coke im-
mediately around the door of the furnace.

In many cases it will be found advantageous to convert the lower
crucible itself into the cupel by first half-filling it with sand and
then ramming in pounded bone-earth. I have found the above me-
thod to possess the following advantages : — In the first place, the
crucibles may be maintained at a much higher temperature than can
be readily obtained when the ordinary mufile is used, while the de-
gree of heat and the quantity of air admitted may be regulated with
the greatest nicety. Secondly, owing to the greater draught of air,
the oxidation of the lead is more quickly effected ; and lastly, by
looking through an opening in the furnace cover, the operation may
be watched from first to last.

Fig. 2.

L_.

Fig. 2. a B black lead crucibles. C the upper opening. D the lower opening.
E the cupel. The dotted semicircle represents the position of the furnace-door.

9M

Intelli<rence and Miscellaneous Articles.

Improved Form of Messrs, WiWs and Farrentrapp's Apparatus.

The only inconvenience I have found in the process proposed by
Drs. Will and Varrentrapp for the estimation of nitrogen in organic
bodies, is the liability of the liquid in the condenser being thrown
back into the combustion-tube by sudden absorption taking place,
or fi'om too violent an evolution of the gases, part of it being ejected
from the other extremity of the condenser. So well-aware were its
authors of this inconvenience, that they recommend in the analyses
of substances rich in nitrogen the introduction of sugar, or some
other body abounding in carbon, into the combustion-tube, I have
found that the necessity of this addition, which is of course open to
many objections, may be entirely avoided by using a condenser
nearly three times as large as that generally employed, and by sur-
mounting each of the bulbs with another bulb of about half its ca-
pacity. The opening between the bulbs should be very wide, they
being run into one another in the same manner as in the lower bulbs
of Liebig's potash apparatus. With a condenser of this description,
the large bulbs being 1| inch in diameter and about 4 inches aj)art,
I have never experienced the least accident, nor am I compelled to
pay that constant attention to the progress of the combustion which
Drs. Will's and Varrentrapp's condenser usually requires.

Fig. 3.

Mr. Taylor also exhibited a small instrument for holding Da-
guerreotype plates during the process of washing off. It consisted
of two pieces of brass or plaited wire fitted into a wooden handle.
One of the wires is bent into the form of an acute triangle, its base
being slightly turned up, so as to form a ledge for the silver plate
to rest upon. The other wire is placed between the sides of the
triangle curved, so as to form a spring, which rests upon the top
of the plate, and keeps it in its place. By inserting the fore-finger
in the loop of the spring, the plate may be shaken violently without
becoming dislodged.

Fig. 4.

From the Proceedings of the Chemical Society.

Intelligence and Miscellaneous Articles. 997

PREPARATION AND COMPOSITION OF LIGNIN.

MM. Poumarede and Figuer state as a test of the purity of lignin,
that when immersed in concentrated sulphuric acid it is not ren-
dered black. In order to procure it in this state, a piece of wood
is to be transversely rasped, and the raspings are to be immersed in
soap ley for twenty-four hours ; the mixture is then to be diluted
with once or twice its weight of water and poured off; the insoluble
residue is to be largely washed with water, treated with a slight ex-
cess of dilute hydrochloric acid, and again washed with water. After
this the ligneous fibre is to be treated with great excess of a solution
of common salt ; the digestion is to be continued with occasional
stirring for two or three days, a fresh portion of the solution being
once used ; this being poured off, the fibrous matter is to be treated
with a weak alkaline solution till it comes away colourless ; it is to
be again washed, and the remaining alkali is to be saturated by slight
excess of hydrochloric acid, and after again washing with distilled
water till litmus is not reddened, the product, placed on a sieve, is
to be dried either in the sun or a stove.

The lignin thus obtained, after being washed with alcohol and
aether, is not coloured by concentrated sulphuric acid, and is to be
considered as absolutely pure. It is white and silky, and possesses
the organic structure of the wood from which it has been obtained ;
and the authors consider themselves authorized to conclude that in
analysing this substance, they operate on the vegetable skeleton such
as it exists in plants.

The authors find that the results of their analyses differ but very
little from those obtained by M. Payen ; they nevertheless deem it
necessary to state them as satisfactorily proving the agreement which
exists between the various kinds of lignin of very ditterent origin.

Lignin of the poplar, dried at 288° F. ; mean of three experi-
ments : —

Carbon 43-88

Hydrogen 6'23

Oxygen 49*89

100-00

Lignin of the beech, dried at 288° F. : —

Carbon ...i >••>>.. . 43-85

Hydrogen 6*22

Oxygen 49-93

100-00

Blotting-paper treated with acids, alkalies, water, und alcohol,
dried at 288° F. :—

I. XL

Carbon 43-87 43*84

Hydrogen 6-12 6*22

Oxygen 50-01 49*94

100-00 100-00

398 Intelligence and Miscellaneous Articles.

Cotton treated only with boiling water, hydrochloric acid, and
dilute solution of potash, cold : —

I. II.

Carbon 43-46 43*10

Hydrogen .... G'38 6-43

Oxygen 50-lG 50-45

100-00 99-98
Flax, treated like cotton : —

I. II.

Carbon 43-92 43'33

Hydrogen 6-01 6 41

Oxygen 50*07 50*26

100-00 100-00

Papyrin. — In employing sulphuric acid to determine the purity of
lignin, the authors have discovered a new substance which consti-
tutes a very curious modification of ligneous tissues. It results from
the first action of sulphuric acid on lignin, and is the product which
arises before its conversion into dextrin.

Let blotting-paper be immersed for not more than half a minute
in concentrated sulphuric acid, and then be immediately washed with
a large quantity of water to prevent the action of the acid ; and if
it be then immersed for a few moments into water containing a few
drops of ammonia, a substance is obtained which possesses all the
physical characters of an animal membrane. When moistened with
water, it has the soft and greasy feel of animal membrane softened
in water ; when dried it has the appearance and the toughness of
parchment, and when glazed it has considerable transparency.

This substance, which the authors call j^apijr'm, is identical in
composition with lignin. It was found to yield—

I. II. III.

Carbon 43*30 43*89 44*44

Hydrogen 6*28 6-27 6-23

Oxygen 50-42 49*84 49*33

100*00 100-00 100-00

Journ. de Ph. et de Ch., Aout 1847.

SOLUBILITY OF CHLORIDE OF SILVER IN HYDROCHLORIC ACID.

M. Pierre states that concentrated hydrochloric acid is capable of
dissolving j-i-^dth of its weight of chloride of silver ; when it has
been diluted with twice its weight of water, it is capable of retain-
ing more than -^i^dth of its weight.

M. Gerhardt observes that this fact is important, and says he
had previously stated it ; and it appears to him to be the cause of
the difference of the numbers obtained by MM. Berzelius and Ma-
rignac as to the theoretical number expressing the atomic weight of
chlorine according to Dr. Prout's law of multiples. — Ibid. Sept. 1847.

Meteorological Observations. 399

Daubeny on Active and Extinct Volcanos.

Professor Daubeny of Oxford has in the press, and nearly ready
for publication, a new and much-enlarged edition of his Description
of Active and Extinct Volcanos.

The present Edition will be found to contain nearly twice the
amount of matter included in the preceding one, embracing not only
such new facts and observations with respect to volcanos as have
been brought to light since its first appearance in 1826, but likewise
the allied phsenomena of Earthquakes and Thermal Springs, as well
as a fuller discussion of the theories connected with those subjects.

METEOROLOGICAL OBSERVATIONS FOR SEPT. 1847.

Chiswick. — September 1. Clear: cloudy: clear. 2. Cloudy: boisterous. 3.
Cold rain : overcast. 4. Fine. 5. Clear : shower : clear. 6. Very fine. 7. Clear
and cold : cloudy : rain at night. 8. Rain. 9. Very fine. 10. Overcast : very
fine. II, 12. Very fine. 13. Densely overcast: rain. 14. Very fine : slight
shower: clear and cold at night. 15. Fine: boisterous, with rain at night.
16. Boisterous. 17. Rain. 18. Cloudy, with very clear intervals. 19. Cloudy:
heavy rain at night. 20. Fine: slight showers. 21. Rain. 22. Cloudy: fine.

23. Cloudy and mild. 24. Foggy : very fine. 25, 26. Fine. 27. Frosty : clear :
very fine : clear and frosty at night. 28. Slight fog ; overcast. 29. Slight fog :
very fine. 30. Dry haze : overcast.

Mean temperature of the month 53°'40

Mean temperature of Sept. 1846 60-79

Mean temperature of Sept. for the last twenty years 52 '77

Average amount of rain in Sept 2*73 inches.

Boston. — Sept. 1. Fine. 2. Windy. 3. Cloudy : rain p.m. 4. Fine. 5. Fine:
rain P.M. 6,7. Fine. 8. Cloudy. 9 — 11. Fine. 12. Windy. 13. Rain:
rain A.M. and P.M. 14. Fine. 15. Fine: rain p.m. 16. Fine: stormy from
10 A.M. 17. Cloudy. 18 — 20. Fine. 21. Fine: rain p.m. 22. Cloudy : rain
A.M. 23. Cloudy. 24—28. Fine. 29. Cloudy. SO. Fine.

Sandwich Manse, Orkney. — Sept. 1,2. Showers. 3. Bright: showers: sleet.
4 — 6. Showers. 7, 8. Cloudy : showers. 9. Drizzle : showers. 10. Cloudy.

11. Cloudy: rain. 12. Showers. 13. Cloudy : clear. 14. Cloudy. 15,16.
Bright : rain. 17. Cloudy : showers. 18. Showers. 19. Clear : showers : sleet.
20. Showers: rain: cloudy. 21. Bright: fine. 22. Damp: rain. 23. Showers.

24. Showers : cloudy. 25. Rain : clear. 26. Bright : clear. 27, 28. Clear.
29. Clear: aurora. 30. Clear.

Applegarlh Manse, Dumfries-shire. — Sept. 1. Sharp showers and high wind.
2. Clear and fine harvest day. 3. Rain. 4. Fine clear sharp weather. 5. Fine
harvest day. 6. Clear and bracing. 7. Rain, though not heavy. 8. Fair, but
cloudy. 9. Close rain. 10. Fine : some drops p.m. 11. Fair a.m. : rain p.m.

12. Fair, but threatening. 13. Fine. 14. Bracing day : flying showers. 15.
Fine a.m. : heavy rain p.m. 16. Rain and high wind. 17. Few drops of rain.
18. Fair, but dull. 19. Frequent showers. 20. A few drops. 21. Rain p.m.
22, 23. Showery. 24. Fair and fine. 25. Slight drizzle. 26. Very fine day.
27. Very fine day: frost a.m. 28, 29. Very fine days : no frost. 30. Fair, but cold.

Mean temperature of the month 50°"9

Mean temperature of Sept. 1846 59*6

Mean temperature of Sept. for 25 years 53 '2

Mean rain in Sept. for 20 years , 3*13 inches.

m

•^loiAipaBg

■uoisog

•5[3!.vis!qo

•5{0!.»pu«g

■ajiqs
-saiJjuinci

■uoisog

•ui'd t

0:5?

•lU'B

16

Sic

•uojsoa

•ii!H

•XEJtt

•*•£

^§

■uoisoa

JO SiCBQ

^c«

^ & Bt & ^ g

a a a a a §
a G a c a ^

e P P ? « c8

a a

M ^ ^ ^ C

> ^ ? c' ^ ri 2 B:" >■ ^" ^' & & ^ 3 ^- ^' ^' i ■! i ^^i^tci^o cJ
c a ^ m S p *■

fe • • J te J ^ ^ J ^ • -B • > f- -• t: -• t; J fe ^* 5= i.-

u u u u o

^ ^

S S S cu S

i 5^' i i i ?■ ^' ^* ^' ^ ^ i i ^ ^ ^i^i^i^^i^'^i-^'^ i

o^O^O^<3^0^6^0^0^CT^O^'i) o^c^c^O^O^c^o^o^O^O^O^O^O^O O O O C O

<Nioi>.i:— loooiyi— ^oOlOOO^t^ooa^-^lr5rt'*<Mr^Olooo•^oc'5c*:
a^6^a^6^0^6^0^l3^a^6^6^co C^O^O^aO 0^0^0^<3^^^0^0^0^0^00 O o O

lr5ooo^^^ooc^^o'^^<^^oom-'t^c^■^cr5lcco — QOtooooejoiooofMioco

d — OlO(Nl0^o■'tlomoo--lOO^oot>■<^l^o■^loaoc^^^oc^lOc»(^^l^co
lr:l:-^t:^^-~l>.^:^t^^l>.l:^o^o^^p^o^ocpe^c^^py^ap^^^po«7•7'<NC^(N

C^o^0^c^l6^0^0^0^6^6^0^o^0^o^O^QO o^o^O^O^CT^O^O^O CT\0 O O O O

^OOO0^0^O^■^O<»O^»t00lt^COC^O(»C0>CO^0L0<~0l:^^0CM-^|-l
ncoip-rPcocoipTfipip^p cocorO'^C^O O rj<Ol *o iprr^O ipoqp O^Qpqp

C<0<M(MOI<M<MC»0«CN<MC^(N(N01C«CNCNC4<NO«<MOIO)C»<M<NCN(Ne«

ICO 00

00 o^O^

a6 00

e> -^ ■^

CO 00 op
O^ C?s 0^
(N !N <N

0^ 00 -^ -^ t^
i-^ r^ IN 00 t^

oD CO (> o^ a\

c^io o

lOOO —

o o o

00 lO

00 r^

c^oo
lO o
00 1--

o^oo tc

CO CO ■<*

LO « —

CO lO
(N CO

o a^

— 1 -*

— 1 C^

O O t-^ lO C^

COU5 cyvco — <

— <N d Ol -H

§2

t^OO

— 00

F- 00

C^GO

?M CO t--

t^ 1*00
LO CO O

o^o^-<

00 00

o o

O~i00
O\00

t-» lO i-o o —
CO C^-d — ' •«*
(M «N CO CO (N

9

O

T H E
LONDON, EDINBURGH and DUBLIN

PHILOSOPHICAL MAGAZINE
JOURNAL OF SCIENCE.

[THIRD SERIES.]

DECEMBER 1847.

LXIV. On the Diamagnetic conditions of Flame and Gases.
By Michael Faraday, F.R.S., Foreign Associate of the
Academy of Sciences, Sec,

To Richard Taylor, Esq.

Royal Institution,

My dear Sir, Nov. 8, 1847.

I LATELY received a paper from Professor Zantedeschi,
published by him, and containing an accomit of the dis-
covery, by P. Bancalari, of the magnetism (diamagnetism) of
flame, and of the further experiments of Zantedeschi, by which
he confirms the result, and shows that flame is repelled from
the axial line joining two magnetic poles. I send you the
paper that you may, if you estimate its importance as highly
as I do, reprint it in the Philosophical Magazine ; and I send
also with it these further experimental confirmations and ex-
tensions of my own. As M. Zantedeschi has published his
results, I have felt myself at liberty to work on the subject,
which of course interested me very closely. Probably what I
may describe will only come in confirmation of that which has
been done already in Italy or elsewhere ; and if so, I hope to
stand excused ; for a second witness to an important fact is by
no means superfluous, and may in the present case help to
induce others to enter actively into the new line of investigation
presented by diamagnetic bodies generally.

I soon verified the chief result of the diamagnetic affection
of flame, and scarcely know how I could have failed to observe
the effect years ago. As I suppose I have obtained much
more striking evidence than that referred to in Zantedeschi's
paper, I will describe the shape and arrangement of the essen-
tial parts of my apparatus. The electro-magnet used was the
powerful one described in the Experimental Researches
(2247 *.). The two terminal pieces of iron forming the virtual
magnetic poles were each r7 inch square and six inches long;
• Page 398 of this Journal for May 1846.

Phil. Mag. S. 3. Vol. 31 . No. 210. Dec. 1847. 2 D

402 Dr. Faraday on ihe Diamagnetic conditions

but the ends were shaped to a form approaching that of a cone,
of which the sides have an angle of about 100°, and the axis
of which is horizontal and in the upper surface of the pieces
of iron. The apex of each end was rounded ; nearly a tenth
of an inch of the cone being in this way removed. When
these terminations are brought near to each other, they give
a powerful effect in the magnetic field, and the axial line of
magnetic force is of course horizontal, and on a level nearly
with the upper surface of the bars. I have found this form
exceedingly advantageous in a great variety of experiments.

When the flame of a wax taper was held near the axial
line, but on one side or the other, about one-third of the flame
rising above the level of the upper surface of the poles, as
soon as the magnetic force was on, the flame was affected ;
and receded from the axial line, moving equatorially, until it
took an inclined position, as if a gentle wind was causing its
deflection from the upright position ; an effect which ceased
the instant the magnetism was removed.

The effect was not instantaneous, but rose graduallv to a
maximum. It ceased very quickly when the magnetism was
removed. The progressive increase is due to the gradual pro-
duction of currents in the air about the magnetic field, which
tend to be, and are, formed on the assumption of the magnetic
conditions, in the presence of the flame.

When the flame was placed so as to rise truly across the
magnetic axis, the effect of the magnetism was to compress
the flame between the points of the poles, making it recede in
the direction of the axial line from the poles towards the middle
transverse plane, and also to shorten the top of the flame. At
the same time the top and sides of the compressed part burnt
more vividly, because of two streams of air which set in from
the poles on each side directly against the flame, and then
passed out with it in the equatorial direction. But there was
at the same time a repulsion or recession of the parts of the
flame from the axial line ; for those portions which were below
did not ascend so quickly as before, and in ascending they
also passed off in an inclined and equatorial direction.

On raising the flame a little more, the effect of the magnetic
force was to increase the intensity of the results just described,
and the flame actually became of a fish-tail shape, disposed
across the magnetic axis.

If the flame was raised until about two-thirds of it were
above the level of the axial line, and the poles approached so
near to each other (about 0*3 of an inch) that they began to
cool and cqmpress the part of the flame at the axial line, yet
without interfering with its rising freely between them ; then,

of 'Flame and Gases. 403

on rendering the magnet active, the flame became more and
more compressed and shortened ; and as the effects proceeded
to a maximum, the top at last descended, and the flame no
more rose between the magnetic poles, but spread out right
and left on each side of the axial line, producing a double
flame with two long tongues. This flame was very bright
along the upper extended forked edge, being there invigorated
by a current of air which descended from between the poles on
to the flame at this part, and in fact drove it away in the equa-
torial direction.

When the magnet was thrown out of action, the flame re-
sumed its ordinary upright form between the poles, at once;
being depressed and redivided again by the renewal of the
magnetic action.

When a small flame, only about one-third of an inch high,
was placed between the poles, the magnetic force instantly
flattened it into an equatorial disc.

If a ball of cotton about the size of a nut be bound up by
wire, soaked in asther and inflamed, it will give a flame six
or seven inches high. This large flame rises freely and natu-
rally between the poles ; but as soon as the magnet is rendered
active, it divides and passes off in two flames, the one on one
side, and the other on the other side of the axial line.

Such therefore is the general and very striking effect which
may be produced on a flame by magnetic action, the import-
ant discovery of which we owe to P. Bancalari.

1 verified the results obtained by M. Zantedeschi with dif-
ferent flames, and found that those produced by alcohol, tether,
coal-gas, hydrogen, sulphur, phosphorus, and camphor were all
affected in the same manner, though not apparently with equal
strength. The brightest flames appeared to be most affected.

The chief results may be shown in a manner in some re-
spects still more striking and instructive than those obtained
with flame, by using a smoking taper. A taper made of wax,
coloured green by verdigris, if suffered to burn upright for a
minute and then blown out, will usually leave a wick with a
spark of fire on the top. The subdued combustion will how-
ever still go on, even for an hour or more, sending up a thin
dense stream of smoke, which, in a quiet atmosphere, will rise
vertically for six or eight inches; and in a moving atmosphere
will show every change of its motion, both as to direction and
intensity. When the taper is held beneath the poles, so that
the stream of smoke passes a little on one side of the axial
line, the stream is scarcely affected by the power of the mag-
net, the taper being three or four inches below the poles ; but
if the taper be raised, so that the coal is not more than an inch

2 D 2

404 Dr. Faraday on the Diamagnetic conditions

below the axial line, the stream of smoke is much more
affected, being bent outwards ; and if it be brought still higher,
there is a point at which the smoke leaves the taper-wick even
in a horizontal direction, to go equatorially. If the taper be
held so that the smoke-stream passes through the axial line,
and then the distances be varied as before, there is little or
no sensible effect when the wick is four inches below : but
being raised, as soon as the warm part of the stream is between
the poles, it lends to divide; and when the ignited wick is
about an inch below the axial line, the smoke rises vertically
in one column until about two-thirds of that distance is passed
over, and then it divides, going right and left, leaving the space
between the poles clear. As the taper is slowly raised, the
division of the smoke descends, taking place lower down, until
it occurs upon the wick, at the distance of O'-i or 0*5 of an
inch below the axial line. If the taper be raised still more,
the magnetic effect is so great, as not only to divide the stream,
but to make it descend on each side of the ignited wick, pro-
ducing a form resembling that of the letter W; and at the
same time the top of the burning wick is greatly brightened
by the stream of air that is impelled downwards upon it. In
these experiments the magnetic poles should be about 0*25 of
an inch apart.

A burning piece of amadou, or the end of a splinter of wood,
produced the same effect.

By means of a small spark and stream of smoke, I have even
rendered the power of an ordinary magnet, in affecting them,
evident. The magnet was a good one, and the poles were
close to each other and conical in form.

Before leaving this description of the general phsenomenon
and proceeding to a consideration of the principles of mag-
netic action concei-ned in it, I may say that a single pole of
the magnet produces similar effects upon flame and smoke,
but that they are much less striking and observable.

Thougli the effect be so manifest in a flame, it is not, at
first sight, evident what is the chief cause or causes of the
result. The heat of the flame is the most apparent and pro-
bable condition ; but there are other circumstances which may
be equally or more influential. Chemical action is going on
at the time : — solid matter, which is known to be diamagnetic,
exists in several of the flames used : and a great difference
exists between the matter of the flame and *the surrounding
air. Now any or all of these circumstances of temperature,
chemical action, solidity of part of the matter, and differential
composition in respect to the surrounding air, may concur in
producing or influencing the result.

of Flame and Gases. 405

I placed the wires of an electrometer, and also of a galvano-
meter, in various parts of the affected flame, but could not
procure any indications of the evolution of electricity by any
action on the instruments.

I examined the neighbourhood of the axial line as to the
existence of any current in the air when there was no flame or
heat there, using the visible fumes produced when little pellets
of paper dipped in strong solutions of ammonia and muriatic
acid were held near each other; and though I found that a
stream of such smoke was feebly affected by the magnetic
power, yet I was satisfied there was no current or motion in
the common air, as such, between the poles. The smoke
itself was feebly diamagnetic; due, I believe, to the solid par-
ticles in it.

But when flame or a glowing taper is used, strong currents
are, under favourable circumstances, produced in the air. If
the flame be between the poles, these currents take their course
along the surface of the poles, which they leave at the opposite
faces connected by the axial line, and passing parallel to the
axial line, impinge on the opposite sides of the flame ; and
feeding the flame, they make part of it, and proceed out equa-
torially. If the flame be driven asunder by the force of these
currents and retreat, the currents follow it; and so, when the
flame is forked, the air which is between the poles forms a
current which sets from the poles downwards and sideways
towards the flame. I do not mean that the air in evety case
travels along the surface of the poles or along the axial lines,
or even from between the poles ; for in the case of the glowing
taper, held half an inch or so beneath the axial line, it is the
cool air which is next nearest to the taper, and (generally)
between the taper and the axial line, that falls with most force
upon it. In fact the movements of the parts of the air and
flame are due to a differential action. We shall see presently
that the air is diamagnetic as well as flame or hot smoke ; i. e.
that both tend, according to the general law which I have ex-
pressed in the Experimental Researches (2267, &c.), to move
from stronger to weaker places of magnetic force, but that
hot air and flame are more so than cold or cooler air : so, when
flame and air, or air at different temperatures, exist at the
same time within a space under the influence of magnetic
forces, differing in intensity of action, the hotter particles will
tend to pass from stronger to weaker places of action, to be
replaced by the colder particles ; the former therefore will have
the effect of being repelled ; and the currents that are set up
are produced by this action, combined with the mechanical
force or current possessed by the flame in its ordinary relation
to the atmosphere.

406 Dr. Faraday on the Diamagnetic conditions

It will be evident to you that I have considered flame only
as a particular case of a general law. It is a most important
and beautiful one, and it has given us the discovery of dia-
magnetism in gaseous bodies: but it is a complicated one,
as I shall now proceed to show, by analysing some of its
conditions and separating their effects.

For the purpose of examining the effect of heat alone in
conducing to the diamagnetic condition of flame, a small helix
of fine platina wire was attached to two stronger wires of cop-
per, so that the helix could be placed in any given position as
regarded the magnetic poles, and at the same time be ignited
at pleasure by a voltaic battery. In this manner it was substi-
tuted for the burning taper, and gave a beautiful highly-heated
current of air, unchanged in its chemical condition. When
the helix was placed directly under the axial line, the hot air
rose up between the poles freely, being rendered evident above
by a thermometer, or by burning the finger, or even scorching
paper; but as soon as the magnet was rendered active, the
hot air divided into a double stream, and was found ascending
on the two sides of the axial line; but a descending current
was formed between the poles, flowing downwards towards the
helix and the hot air, which rose and passed off sideways
from it.

It is therefore perfectly manifest that hot air is diamagnetic
in relation to, or more diamagnetic than, cold air ; and, from
this fact I concluded, that, by cooling the air below the natural
temperature, I should cause it to approach the magnetic axis,
or appear to be magnetic in relation to ordinary air. I had
a little apparatus made, in which a vertical tube delivering
air was passed through a vessel containing a frigorific mix-
ture; the latter being so clothed with flannel that the ex-
ternal air should not be cooled, and so invade the whole of
the magnetic field. The central current of cold air was di-
rected downwards a little on one side of the axial line, and
falling into a tube containing a delicate air-thermometer, there
showed its effect. On rendering the magnet active, this effect
however ceased, and the thermometer rose ; but on bringing
the latter under the axial line it again fell, showing that the
cold current of air had been drawn inwards or attracted to-
wards the axial line, i. e. had been rendered magnetic in rela-
tion to air at common temperatures, or less diamagnetic than
it. The lower temperature was 0° F. The effect was but
small; still it was distinct.

The effect of heat upon air, in so greatly increasing its dia-
magnetic condition, is very remarkable. It is not, I think, at
all probable that the mere effect of expanding the air is the
cause of the change in its condition, because one would be led

of Flame and Gases. 407

to expect that a certain bulk of expanded air would be less
sensible in its diamagnetic effects than an equal bulk of denser
air ; just as one would anticipate that a vacuum would present
no magnetic or diamagnetic effects whatever, but be at the
zero point between the two classes of bodies (Experimental
Researches, 2423, Sl^l-). It is certainly true, that if the air
were a body belonging to the magnetic class, then its expan-
sion, being equivalent to dilution, would make it seem dia-
magnetic in relation to ordinary air (Experimental Researches,
2367, 2438) ; but that, I think, is not likely to be the case,
as will be seen by the results described further on in reference
to oxygen and nitrogen.

If the power conferred by heat is a direct consequence, and
proportionate to the temperature, then it gives a very remark-
able character to gases and vapours, which, as we shall see
hereafter, possess it in common. In my former experiments
(Experimental Researches, 2359, 2397) 1 heated various dia-
magnetic bodies, but could not perceive that their degree of
magnetic force was at all increased or affected by the tempe-
rature given to them. 1 have again submitted small cylinders
of copper and silver to the action of a single pole, at common
temperatures and at a red heat, with the same result. If
there was any effect of increased temperature, it was that of
a very slight increase in the diamagnetic force, but I am not
sure of the result. At present, therefore, the gaseous and va-
porous bodies seem to be strikingly distinguished by the power-
ful effect which heat has in increasing their diamagnetic con-
dition.

As all the experiments, whether on flame, smoke, or air,
seemed to show that air had a distinct magnetic relation, which,
though highly affected by temperature, still belonged to it at
all temperatures ; so it was a probable conclusion that other
gaseous or vaporous bodies would be diamagnetic or mag-
netic, and that they would differ from each other even at com-
mon or equal temperatures. I proceeded therefore to examine
them, delivering streams of each into the air, in the first in-
stance, by fit apparatus and arrangements, and examining the
course taken by these streams in passing across the magnetic
field, the magnetic force being either induced or not at the
time.

In delivering the .various streams, I sometimes introduced
the gases into a globe with a mouth and also a tubular spout,
and then poured the gas out of the spout, upwards or down-
wards, according as it was lighter or heavier than air. At
other times, as with muriatic acid or ammonia, .1 delivered the
streams from the mouth of the retort. But as it is very im-

408 Dr. Faraday on the Diamagnetic conditions

portant not to deluge the magnetic field with a quantity of in-
visible gas, I devised the following arrangement, which an-
swered well for all the gases not soluble in water. A WoulPs
bottle was chosen having three apertures at the top, a, h and c ;
a wide tube was fixed into aperture a, descending within the
bottle to the bottom, and being open above and below ; by
this any water could be poured into the bottle and employed
to displace the gas previously within it. Aperture b was closed
by a stopper. Aperture c had an external tube, with a stop-
cock fixed in it to conduct the gas to any place desired. To
expel the gas and send it forward, a cistern of water was placed
above the bottle, and its cock so plugged by a splinter of
wood, that when full open it delivered only twelve cubic inches
of fluid in a minute. This stream of water being directed into
aperture a, and the cock of tube c open, twelve cubic inches
of any gas within the Woulf's bottle was delivered in a minute
of time ; and this I found an excellent proportion for our mag-
net and apparatus.

With respect to the delivery of this gas at the magnetic

poles, a piece of glass tube bent into this shape j was held

by a clamp on the stage of the magnet, so that it could easily
be slipped backward and forward, or to one side, and so its
vertical part be placed anywhere below the axial line. The
apertuie at this end was about the one-eighth of an inch in-
ternal diameter. In the horizontal part near the angle was
placed a piece of bibulous paper, moistened with strong solu-
tion of muriatic acid (when necessary). The horizontal part
of the tube was connected and disconnected in a moment, when
necessary, with the tube c of the gas-bottle, by a short piece
of vulcanized rubber tube. If the gas to be employed as a
stream were heavier than the surrounding medium, then the
glass tube, instead of having the form delineated above, was
so bent as to deliver its stream downwards and over the axial
line. In this manner currents of different gases could be de-
livered, perfectly steady and under perfect command.

The next point was to detect and trace the course of these
streams. A little ammonia vapour, delivered near the mag-
netic field, did this in some degree, but was not satisfactory ;
for, in the first place, the little cloud of muriate of ammonia
particles formed, is itself diamagnetic ; and further, the tran-
quil condition of the air in the magnetic field was then too
much disturbed. Catch-tubes were therefore arranged, con-
sisting of tubes of thin glass about the size and length of a
finger, open at both ends, and fixed upon little stands so that
they could be adjusted either over or under the magnetic poles
at pleasure. When they were over the polesy I generally had

of Flame and Gases. 409

three at once; one over the axial line and one at each side.
When they were under the poles, the lower end was turned
up a little for the purpose of facilitating observation there.

The gas delivered at the poles, as already described, con-
tained a little muriatic acid (obtained from the solution in the
paper), but not enough to render it visible. To make it ma-
nifest up which catch-tube it passed, a little piece of bibulous
paper, folded and bound round and suspended by a copper
wire, was dipped in the solution of ammonia and hung in each
of the tubes. It was then evident at once, by the visible fume
formed at the top of one of the tubes, whether the gas delivered
below passed up the one or the other tube, and which : and
yet the gas was perfectly clear and transparent as it passed by
the place of magnetic action.

In addition to these arrangements, I built up a sheltering
chamber about the magnetic poles and field, to preserve the
air undisturbed. This was about six inches long by four inches
in width and height, and was easily made of thin plates of
mica, which were put together or taken down in a moment.
The chamber was frequently left more or less open at the top
or bottom for the escape of gases, or the place of the catch-
tubes. Its advantages were very great.

yJir. — In the first place air was sent in under these arrange-
ments, the stream being directed by the axial line. It made
itself visible in the catch-tube above by the smoke produced;
but whether the magnet was active or not, its course was the
same; showing that, so far, the apparatus worked well, and
did not of itself cause any erroneous indications.

Nitrogen. — This gas was sent from below upwards, and
passed directly by the axial line into the catch-tube above;
but when the magnet was made active, the stream was affected,
and though not stopped in the middle catch-tube, part ap-
peared in the side tubes. The jet was then arranged a little
on one side of the axial line, so that, without the magnetic
action, it still ascended and went up the middle catch-tube:
then, when the magnetic action was brought on, it was clearly
affected, and a great portion of it was sent to the side catch-
tube. The nitrogen was, in fact, manifestly diamagnetic in
relation to common air, when both were at the same tem-
perature; but as four-fifths of the atmosphere consists of ni-
trogen, it seemed very evident, from the result, that nitrogen
and oxygen must be very different from each other in their
magnetic relations.

Oxygen. — A stream of oxygen was sent down through air
between the poles. When there was no magnetic action it
descended vertically, and when the magnetic action was on it

410 Dr. Faraday on the Diamagnetic conditions

appeared to do the same ; at all events it did not pass off
equatorially. But as there was reason, from the above expe-
riments with nitrogen, to expect that oxygen would appear,
not diamagnetic but magnetic in air; so the place of the
stream was changed and made to be on one side of the
axial line. In this case it fell perfectly well at first into a
catch-tube placed beneath ; but as soon as the magnet was
rendered active, the stream was deflected, being drawn towards
the axial line, and fell into another catch-tube placed there to
receive it. So oxygen appears to be magnetic in common air.
Whether it be really so, or only less diamagnetic than air (a
mixture of oxygen and nitrogen), we shall be better able to
consider hereafter.

Hydrogen. — This gas proved to be clearly and even strongly
diamagnetic ; for notwithstanding the powerful ascensive force
which its stream has in the atmosphere, because of its small
specific gravity, still it was well deflected and sent equatorially.
Considering the lightness of the gas, one might have expected
that it would have been drawn towards the axial line, as a
stream of rarefied air (if it could exist) would be. Its dia-
magnetic state, therefore, shows in a striking point of view,
that gases, like solids, have peculiar and distinctive degrees of
diamagnetic force.

Carbonic acid. — This gas made a beautiful experiment.
The stream was delivered downwards a little on one side o
the axial line; a catch-tube was placed a little further out, so
that the stream should fall clear of it as long as there was no
activity in the magnet. But on rendering the magnet efficient,
the stream left its vertical direction, passed equatorially, and
fell into the catch-tube; and by looking horizontally, could
be seen flowing out at its lower extremity like a spring, and
falling away through the air. Again, the magnet was thrown
out of action, and a glass with lime-water placed beneath the
lower end of the catch-tube ; no carbonic acid appeared there,
though the fluid in the glass was continually stirred; but the
instant the magnet was made, the carbonic acid appeared in
the catch-tube, fell into the glass and made the lime-water
turbid. This gas therefore is diamagnetic in air.

Carbonic oxide. — This gas was carefully freed from carbonic
acid before it was used. It was employed as a descending
stream, and was apparently very diamagnetic : but it is to be
remarked, that a substance which is so nearly the specific gra-
vity of atmospheric air is easily dispersed right and left in it,
and therefore that the facility of dispersion is not a correct
indication of the diamagnetic force. By introducing a little
ammonia into the mica chamber, it was, however, easily seen

of Flame and Gases. 411

that carbonic oxide was driven away equatorially with consi-
derable power; and I judge from the appearance, that it is
more diamagnetic than carbonic acid.

Nitrous oxide. — This gas was moderately, but clearly, dia-
magnetic in air. Much interest belongs to this and the other
compounds of nitrogen and oxygen, both because they contain
the same elements as air, and because of the relations of ni-
trogen and oxygen separately.

Nitric oxide. — I tried this gas both as an up and down
current, but could not determine its magnetic condition. What
with the action of the oxygen of the air, the change of the
nature of the substances, and the heat produced, there was so
much incidental disturbance and so little effect due to magnetic
influence, that 1 could not be sure of the result. On the whole
it was very slightly diamagnetic ; but so little, that the effect
might be due to the smoke particles which served to render it
visible.

Nitrous acid gas. — Difficult to observe, but I believe it is
slightly magnetic in relation to air.

Olefiant gas was diamagnetic, and well so. The little dif-
ference in specific gravity of this gas and air, even creates a
difficulty in following the course of the olefiant gas, unless it
be watched for on every side.

Coal-gas. — The coal-gas of London is lighter than air, being
only about two-thirds in weight of the latter. It is very well
diamagnetic, and gives exceedingly good and distinct re-
sults.

Sulphurous acid gas is diamagnetic in air. It was generated
in a small tube containing liquid sulphurous acid ; this being
connected, in place of the gas bottle, with the delivery-tube
and mouthpiece by the vulcanized rubber tube. The presence
or absence of the gas in the catch-tube was well shown by
ammonia, and still better by litmus paper.

Muriatic acid. — I'he retort in which it was generated was
connected, as just described, with the delivery-tube. The gas
was very decidedly diamagnetic in air.

Hydriodic acid was also diamagnetic in air. When there
was an abundant stream of gas, its entrance into and passage
through the side catch-tube, on rendering the magnet active,
was very striking. When there was less gas, the stream was
dispersed equatorially in all directions, and less entered the
tube.

Fluo-silicon. — Diamagnetic in air.

Ammonia, — This gas was evolved from materials in a retort,
and tested in the catch-tube above by muriatic acid in the
paper. It was well diamagnetic, corresponding in this respect

412 Dr. Faraday 07i the Diamagnetic conditions

with the character of its elements. It could also be very well
indicated by reddened litmus paper held over the tubes.

Chlorine was sent from the WoulPs bottle apparatus, and
proved to be decidedly diamagnetic in air. Either ammonia by
its fumes, or litmus paper by its becoming bleached, served
to indicate the entrance of the chlorine into the side catch-
tube every time the magnet was rendered active.

Iodine. — A piece of glass tube was so shaped at its lower
extremity as to form a chamber for the reception of iodine,
which chamber had a prolonged mouth directed downwards
so as to deliver the vapour formed within. On putting a little
iodine into the chamber, then heating it, and especially the
mouth part, by a spirit-lamp, and afterwards inclining the
apparatus, abundance of the vapour of iodine was generated
as the substance flowed on to the hotter parts, and passed in
a good stream from the mouth downwards. This purple
stream was diamagnetic in air, and could be seen flowing right
and left from the axial line, when not too dense. If very dense
and heavy, its gravity was such as to make it break through
the axial line, notwithstanding the action of the magnet; still
it was manifest that iodine is diamagnetic to air.

Bromine. — A little bromine was put into the horizontal part
of the delivery tube, and then air passed over it by the apparatus
already described. So much bromine rose into vapour as to
make the air of a yellow colour, and caused it to fall well in
a stream by the axial line. A little ammonia delivered near
the magnetic field showed that this stream was diamagnetic,
and hence it may fairly be presumed that the pure vapour of
bromine would be diamagnetic also.

Cyanogen. — Strongly diamagnetic in air.

Taking air as the standard of comparison, it is very striking
to observe, that much as gases appear to differ one from an-
other in the degree of their diamagnetic condition, there are
very few that are not more diamagnetic than it; and when the
investigation is carried forward into the relation of the two
chief constituents of air, oxygen and nitrogen, it is still more
striking to observe the very low condition of oxygen, which,
in fact, is the cause of the comparatively low condition of air.
Of all the vapours and gases yet tried, oxygen seems to be
that which has the least diamagnetic force. It is as yet a
question where it stands ; for it may be as low as a vacuum,
or may even pass to the magnetic side of it, and experiment
does not as yet give an answer to the question. I believe it to
be diamagnetic ; and this belief is strengthened by the action of
heat upon it, to be described hereafter; but it is exceedingly low
in the scale, and far below chlorine, iodine, and such like bodies.

of Flame and Gases. 413

All the compounds of oxygen and nitrogen seem to show
the influence of the presence of the oxygen. Nitrous acid
seems to be less diamagnetic than air. Nitric oxide mingled
with nitrous acid and warm, is about as air. Nitrous oxide is
clearly diamagnetic in air, though it contains more oxygen:
but it also contains more nitrogen than air, and is also denser
than it, so that there is more matter present; still I think
the results are in favour of the idea that oxygen is diamagnetic.
By referring to the relation of carbonic oxide to carbonic acid,
described further on, it will be seen that the addition of oxygen
seems to make a body less diamagnetic. But the truth may
be, not that oxygen is really magnetic, but that a compound
body })Ossesses a specific diamagnetic force, which is not the
sum of the forces of its particles.

It is very difficult to form more than a mere guess at the
relative degree of diamagnetic force possessed by different
gaseous bodies when they are examined only in air, because
of the many circumstances which tend to confuse the results.
First, there is the invisibility of the gas which deprives one of
the power of adjusting by sight so as to obtain the best effect :
then, there is the difference of gravity ; for if a gas ascend or
descend in a rapid stream, it may seem less deflected than
another flowing more slowly, though it be more diamagnetic;
and as to gases nearly of the specific gravity of air, whether
more or less diamagnetic, they are almost entirely dispersed in
different directions, so that little only enters the catch-tube.
Another modifying circumstance is the distance of the aperture
delivering gas from the axial line, which, to obtain the max-
imum effect, ought to vary with the gravity of the gases and
their diamagnetic force. Again, it is important that the mag-
netic field be not filled with the gas to be examined, and that
generally speaking only a moderate stream be employed ; which
however must depend again upon the specific gravity.

The only correct way therefore of comparing two gases to-
gether is to experiment with them one in the other. For the
experiments made with gases, in gases or in air are differential,
and similar in their nature with those made on a formeroccasion
with solutions (Experimental Researches, 2362, &c.) ; I there-
fore changed the surrounding medium in a few experiments,
substituting other gases for air ; and first chose carbonic acid
as a body easy to experiment with, and one that would, pro-
bably, be more powerfully than some other of the gases, dia-
magnetic (I speak as to the appearances or relative results only)
in air.

I constructed a kind of tray or box, by folding up a doubled
sheet of waxed paper; thus making a vessel thirteen inches

414 Dr. Faraday on the Diamagnetic conditions

long, five inches wide, and five inches high. This was placed
on the ends of the great magnet, and the terminal pieces of
iron before described, placed in it. The box was covered
over loosely by plates of mica, and formed a long square
chamber in which were contained the magnetic poles and field.
All the former arrangements in respect of the magnetic field,
the delivery- tube, the catch-tubes, &c., were then made; and,
lastly, the box was filled with carbonic acid by a tube, which
entered it at one corner; and was, from time to time, supplied
with a fresh portion of gas, as the previous contents became
diluted with gases or air. Everything answered perfectly,
and the following results were easily obtained.

Air passed axially, being less diamagnetic than carbonic
acid gas.

Oxygen passed to the magnetic axis, as was to be expected.

Nitrogen went equatorially, being therefore diamagnetic,
even in carbonic acid.

Hydrogen, coal-gas, olefiant gas, muriatic acid and ammo?iia
passed equatorially in carbonic acid, and were fairly diamag-
netic in relation to it.

Carbonic oxide was very fairly diamagnetic in carbonic acid
gas. Here the effect of oxygen seems to be very well illus-
trated. Equal volumes of carbonic oxide and carbonic acid
contain equal quantities of carbon ; but the former contains
only half as much oxygen as the latter. Yet it is more dia-
magnetic than the latter; so, that, though an additional volume
and quantity of oxygen, equal to that in the carbonic oxide, is
in the carbonic acid added and compresed into it, it does not
add to, but actually takes from, the diamagnetic force.

Nitrous oxide appears to be slightly diamagnetic in relation
to carbonic acid ; but nitric oxide gas was in the contrary re-
lation and passed towards the axial line.

Hence it seems that carbonic acid, though more diamag-
netic than air, is not far removed from it in that respect ; and
this position it probably holds because of the quantity of
oxygen in it. The apparent place of nitrous oxide close to it
appears, in a great measure, to depend on the same circum-
stance of oxygen entering largely into its composition. Still
it is manifest that the action is not directly as the oxygen, for
then common air would be more diamagnetic than either of
them. It seems rather that the forces are modified, as in the
case also of iron and oxygen, and that each compound body
has its peculiar but constant intensity of action.

In order to make similar experiments in light gases, the
two terminal pieces of the magnet were raised, so that they
might be covered by a French glass shade, which, with its

of Flame and Gases, 415

stand, made a very good chamber about them. The pipe to
supply and change the gaseous medium, and also that for
bringing the gas under trial as a stream into the magnetic
field, passed through holes made in the bottom of the stand.
The different gases to be compared with those employed as
media, were, except in the cases of ammonia and chlorine,
mingled with a trace of muriatic acid, as before described.
The gaseous media used were two, coal-gas and hydrogen.
Whilst using coal-gas, I observed the direction of the currents
of the other gases in it by bringing a little piece of paper, at the
end of a wire and dipped in ammonia solution, near the stream.
In the case of the hydrogen, I diffused a little ammonia through
the whole of the gas in the first instance.

Air passed towards the axial line in coal-gas, but was not
much affected.

Oxygen had the appearance of being strongly magnetic in
coal-gas, passing with great impetuosity to the magnetic axis,
and clinging about it; and if much muriate of ammonia fume
were purposely formed at the time, it was carried by the oxygen
to the magnetic field with such force as to hide the ends of the
magnetic poles. If then the magnetic action were suspended
for a moment, this cloud descended by its gravity ; but being
quite below the poles, if the magnet were again rendered active,
the oxygen cloud immediately started up and took its former
place. The attraction of iron filings to a magnetic pole is not
more striking than the appearance presented by the oxygen
under these circumstances.

Nitrogeji. — Clearly diamagnetic in coal-gas.

Olefianti carhojiic oxide, and carbonic acid gases were all"
slightly, but more or less diamagnetic in the coal-gas.

On substituting hydrogen as the surrounding medium in
place of coal-gas, more care was taken in the experiments.
Each gas experimented upon was tried in it twice at least ;
first in the hydrogen of a previous experiment, and then in a
new atmosphere of hydrogen.

Air. — Air passes axially in hydrogen when there is very
little smoke in it: when there is much smoke in the stream
the latter is either indifferent or tends to pass equatorially. I
believe that air and hydrogen cannot be far from each other.

Nitrogen is strikingly diamagnetic in hydrogen.

Oxygen is as strikingly magnetic in relation to hydrogen.
It presented the appearances already described as occurring
in coal-gas; but as the jet delivered the descending stream
of oxygen a little on one side of the axial line, its centrifugal
power, in relation to the axial line, was so balanced by the
centripetal power produced by the magnetic action, that the

416 Dr. Faraday on the Dlamagneiic conditions

stream at first revolved in a regular ring round the axial line,
and produced a cloud that continued to spin round it as long as
the magnetic force was continued, but fell down to the bottom
of the chamber when that force was removed.

Nitrous oxide. — This gas was clearly diamagnetic in the
hydrogen, and gave rise to a very beautiful result in conse-
quence of its following the oxygen ; for at the beginning of
the experiment, the little oxygen contained in the conducting
tube passed axially; but the instant that was expelled, and the
nitrous oxide issued forth, the stream changed its direction,
and passed off diamagnetically in the most striking manner.

Nitric oxide. — This gas passed equally in hydrogen, and
therefore is magnetic in relation to it.

Ammonia. — Diamagnetic in hydrogen.

Carbonic oxide, carbonic acid, and olejiant gases were dia-
magnetic in hydrogen ; the last most so, and the carbonic acid
apparently the least.

Chlorine was slightly diamagnetic in hydrogen. It was
clearly so; but the cloudy particles might conduce much to
the small effect produced.

Muriatic acid gas. — I think it was a little diamagnetic in
the hydrogen.

Notwithstanding the many disturbing causes which interfere
with first and hasty experiments of this kind, and produce
results which occasionally cross and contradict each other, still
there are some very striking considerations which arise in
comparing the gases with each other at the same temperature.
Foremost amongst these is the place of oxygen ; for of all the
gaseous bodies yet tried it is the least diamagnetic, and seems
in this respect to stand far apart from the rest of them. The
condition of nitrogen, as being highly diamagnetic, is also im-
portant. The place of hydrogen, as being less diamagnetic than
nitrogen, and of chlorine, which, instead of approaching to
oxygen, is above hydrogen, and also of iodine, which is pro-
bably far above chlorine, are marked circumstances.

Air of course owes its place to the proportion and the indi-
vidual diamagnetic character of the oxygen and nitrogen in it.
The great difference existing between these two bodies in re-
spect of magnetic relation, and the striking effect presented by
oxygen in coal-gas and hydrogen, bodies not far removed from
nitrogen in diamagnetic force, made me think it might not be
impossible to separate air into its two chief constituents by
magnetic force alone. I made an experiment for this purpose
but did not succeed ; but I am not convinced that it cannot
be done. For since we can actually distinguish certain gases,
and especially these by their magnetic properties, it does not

of Flame and Gases. 417

seem impossible that sufficient power might cause their sepa-
ration from a state of mixture.

In the course of these experiments I subjected several of
the gases to heat, to ascertain whether they generally under-
went the same exaltation of their diamagnetic power which oc-
curred with common air. For this purpose a helix of platina
wire was placed in the mouth of the delivering tube, which
itself was placed below the magnetic axis between the poles.
The helix could be raised to any temperature by a little vol-
taic battery, and any gas could be sent through it and upwards
across the magnetic field by means of the Woulf's bottle ap-
paratus already described. It was easy to ascertain whether
the gas went directly up between the poles, or, on making the
magnet, left that direction and formed two equatorial side-
streams, either by the sensation on the finger, or by a spiral
thermoscope formed of a compound lamina of platinum and
silver placed in a tube above. In every case the hot gas was
diamagnetic in the air, and I think far more so than if the gas
had been at common temperatures. The gases tried were as
follows : oxygen, nitrogen, hydrogen, nitrous oxide, carbonic
acid, muriatic acid, ammonia, coal-gas, olefiant gas.

But as in these experiments the surrounding air would, of
necessity, mingle with the gas first heated, and so form, in
fact, a part of the heated stream, I arranged the platinum
helix so that I could heat it in a given gas, and thus compare
the same gas at different temperatures with itself.

A stream of hot oxygen in cold oxygen was powerfully
diamagnetic. The effect and its degree may be judged of by
the following circumstances. When the platinum helix below
the axial line was ignited, the effect of heat on the indicating
compound spiral, placed in a tube over the axial line, was
such as to cause its lower extremity to pass through one and
a half revolutions, or 54-0°: when the magnetic force was
rendered active, the spiral returned through all these degrees
to its first position, as if the ignited helix below had been
lowered to the common temperature or taken away; and, yet
in respect of it, nothing had been changed. On rendering
the magnet inactive, the current of hot oxygen instantly re-
sumed its perpendicular course and affected the thermoscope
as before.

On experimenting with carbonic acid, it was found that
hot carbonic acid was diamagnetic to cold carbonic acid ; and
the effects were apparently as great in amount as in oxygen.

On making the same arrangement in hydrogen, I failed to
obtain any result regarding the relation of the hot and cold
gas, for this reason : — that 1 could not, in any case, either

F/iil. Mag, S. 3. Vol. 31. No. 210. Dec, 1847. 2 E

418 Dr. Faraday on the Diamagnetlc conditions

with or without the magnetic action, obtain any signs of heat
on the thermoscopic spiral above, even when the platinum
helix, not more than an inch below it, was nearly white hot.
This effect is, I think, greatly dependent upon the rapidity
with which hydrogen is heated and cooled in comparison with
other gases, and also upon the vicinity of the cold masses of
iron forming the magnetic poles, between which the hot gas
has to pass in its way upwards : and it is most probably con-
nected with the fact observed by Mr. Grove of the difficulty
of igniting a platinum wire in hydrogen.

When the igniting helix was placed in coal-gas, it was
found that the hot gas was diamagnetic to that which was
cold; as in all the other cases. Here, again, an effect like
that which was observed in hydrogen occurred ; for when
there was no magnetic action, the ascending stream of hot
coal-gas could cause the thermoscopic spiral to revolve through
only 280° or 300°, in place of above 540° ; through which it
could pass when the surrounding gas was oxygen, air, or
carbonic acid ; and that even when the helix was at a higher
temperature in the coal-gas than in any of these gases.

The proof is clear then that oxygen, carbonic acid, and
coal-gas, are more diamagnetic hot than cold. The same is
the case with air ; and as air consists of four-fifths nitrogen
and only one-fifth oxygen, and yet shows an effect of this
kind as strongly as oxygen, it is manifest that nitrogen also
has the same relation when hot and cold.

Of the other gases also I have no doubt ; though to be quite
certain, they ought to be tried in atmospheres of their own
substance, or else in gases more diamagnetic at common tem-
peratures than they. The olefiant and coal-gases in air easily
bore the elevation of the helix to a full red heat, without in-
flaming when out of the exit-tube: the hydrogen required that
the helix should be at a lower temperature. Muriatic acid
and ammonia showed the division of the one stream into two,
very beautifully, on holding blue and red litmus paper above.

There is another mode of observing the diamagnetic con-
dition of flame, and experimenting with the various gases,
which is sometimes useful, and should always be understood,
lest it inadvertently might lead to confusion. I have a pair of
terminal magnetic poles which are pierced in a horizontal
direction, that a ray of light may pass through them. The
opposed faces of these vertical poles are not, as in the former
case, the rounded ends of cones ; but, though rounded at the
edges, may be considered as flat over an extent of surface an
inch in diameter. The pierced passages are in the form of
cones, the truncation of which in this flat surface is rather

of Flame and Gases, 419

more than half an inch in diameter. When these poles were
in their place, and from 0"3 to 0*4 of an inch apart, a taper
flame, burning freely between them, was for a few moments
unaffected by throwing the magnet into action; but then it
suddenly changed its form, and extending itself axially, threw
off' two horizontal tongues, which entered the passages in the
poles; and thus it continued as long as the magnetism continued,
and no part of it passed equatorially.

On using a large flame made with the cotton ball and aether,
two arms could be thrown off" from the flame by the force of
the magnetism, which passed in an equatorial direction, as
before; and other two parts entered the passages in the mag-
netic poles, and actually issued out occasionally at their further
extremities.

When the poles were about 0*25 of an inch apart, and the
smoking taper was placed in the middle between them level
with the centres of the passages, the effect was very good ; for
the smoke passed axially and issued out at the further ends of
the pole passages.

Coal-gas delivered in the same place also passed axially,
/. e. into the pole passages and parallel to the line joining them.

A little consideration easily leads to the true cause of these
effects, and shows that they are not inconsistent with the
former results. The law of all these actions is, that if a par-
ticle, placed amongst other particles, be morediamagnetic (or
less magnetic) than them, and free to move, it will go from
strong to weaker places of magnetic action ; also, that particles
less diamagnetic will go from weaker to stronger places of
action. Now with the poles just described, the line or lines
of maximum force, are not coincident with the axis of the
holes pierced in the poles, but lie in a circle having a diameter,
probably, a little larger than the diameter of the holes; and
the lines within that circle will be of lesser power, diminishing
in force towards the centre. A hot particle therefore within
that circle will be driven inwards, and, being urged by succes-
sive portions of matter driven also inwards, will find its way
out at the other ends of the passages, and therefore seem to go
in an axial direction ; whilst a hot particle outside of that
circle of lines of maximum force will be drivenoutwards,and so,
with others, will form the two tongues of flame which pass off
in the equatorial direction. By bringing the glowing taper to
different parts, the circle of lines of maximum magnetic inten-
sity can be very beautifully traced ; and by placing the taper
inside or outside of that circle, the smoke could be made to pass
axially or equatorially at pleasure.

I arranged an apparatus on this principle for trying the

2 E2

420 On the Diamagnetic conditions of Flame and Gases.

gases, but did not find it better than, or so good as, the one
1 have described.

Such are the results I have obtained in verifying and ex-
tending the discovery made by P. Bancalari. I would have
pursued them much further, but my present state of health
will not permit it: I therefore send them to you with, probably,
many imperfections. It is now almost proved that many
gaseous bodies are diamagnetic in their relations, and probably
all will be found to be so. I say almost proved ; for it is not,
as yet, proved in fact. That many, and most, gaseous bodies
are subject to magnetic force is proved ; but the zero is not
yet distinguished. Now, until it is distinguished, we cannot
tell which gaseous bodies will rank as diamagnetic and which
as magnetic ; and, also, whether there may not be some
standing at zero. There is evidently no natural impossibility
to some gases or vapours being magnetic, or that some should
be neither magnetic nor diamagnetic. It is the province of
experiment to decide such points ; and the affirmative or ne-
gative may not be asserted before such proof is given, though
it may, very philosophically, be believed.

For myself I have always believed that the zero was re-
presented by a vacuum, and that no body really stood with
it. But though I have only guarded myself from asserting
more than I knew, Zantedeschi (and 1 think also De la Rive),
with some others, seem to think that I have asserted the gases
are not subject to magnetic action ; whereas I only wished to
say that I could not find that they were, and perhaps were
not: I will therefore quote a few of my words from the Ex-
perimental Researches. Speaking of the preparation of a
liquid medium at zero, I say, "Thus a. Jluid medium was ob-
tained, which practically, as far as I could perceive, had every
magnetic character and effect of a gas, and even of a vacuum^
&c." — Experimental Researches, 2423. Again, at (2433) I
say, "At one time I looked to air and gases as the bodies
which allowing attenuation of their substance without addition,
would permit of the observation of corresponding variations
in their magnetic properties, but now all such power by rare-
faction a/?/>^a;'5 to be taken away." And further down at(2435),
" Whether the negative results obtained by the use of gases
and vapours depend upon the smaller quantity of matter in a
given volume, or whether they are the direct consequences of
the altered physical condition of the substance, is a point of
very great importance to the theory of magnetism. I have
imagined in elucidation of the subject an experiment, &c., but
expect to find great difficulty in carrying it into execution, &c."
Happily P. Bancalari's discovery has now settled this matter

On the hifluence of Eleclro-magnetism upon 'Flame. 421

for us in a most satisfactory manner. But where the trne
zero is, or that every body is more or less removed from it on
one side or the other, is not, as yet, experimentally shown or
proved.

I cannot conclude this letter without expressing a hope that
since gases are shown to be magnetically affected, they will
also shortly be found, when under magnetic influence, to have
the power of aff'ecting light (Experimental Researches, 2186,
2212). Neither can I refrain from signalizing the very re-
markable and direct relation between the forces of heat and
magnetism which is presented in the experiments on flame,
and heated air and gases. I did not find on a former occasion
(Experimental Researches, 2397) that solid diamagnetic
bodies were sensibly affected by heat, but shall repeat the ex-
periments and make more extensive ones, if the Italian philo-
sophers have not already done so. In reference to the effect
upon the diamagnetic gases, it may be observed that, speaking
generally, it is in the same direction as that of heat upon iron,
nickel and cobalt ; /. e. heat tends in the two sets of cases,
either to the diminution of magnetic force, or the increase of
diamagnetic force; but the results are too few to allow of any
general conclusion as yet.

As air at different temperatures has different diamagnetic
relations, and as the atmosphere is at different temperatures
in the upper and lower strata, such conditions may have some
general influence and effect upon its final motion and action, sub-
ject as it is continually to the magnetic influence of the earth.
I have for the sake of brevity frequently spoken in this
letter of bodies as being magnetic or diamagnetic in relation
one to another, but I trust that in all the cases no mistake of
my meaning could arise from such use of the terms, or any
vague notion arise respecting the clear distinction between
the two classes, especially as my view of the true zero has
been given only a page or two back.

I am, my dear Sir,

Yours, &c.,
Richard Taylor, Esq., M. Faraday.

Ed. Phil. Mag., Src. Sfc.

LXV. On the Motions -presented by Flame when wider the
Electro-Magnetic Influence. By Prof. Zantedeschi.

THE most eminent philosophers have at all times maintained
the universality of the magnetism of bodies* ; and in
our days Faraday is the only one who has placed the expansi-

* Raccolta Fisico-Chimica Italiana, t. iii. Dei corpi magnetici e dia-
magnetici.

422 Prof. Zantedeschi on the Motions presented hy Flame

ble fluids at the zero of the scale of action among magnetic
and diamagnetic bodies. On the 21st of September 1847, at
the Physical Section of the Ninth Italian Scientific Congress
in Venice, Padre Bancalari, Professor of Physics in the Royal
University of Genoa, read a memoir on the universality of
magnetism ; and the argument was considered by philosophers
to be of such importance, that a desire arose to verify chiefly
the action of magnetism on expansible fluids. It was an-
nounced by the Reporter Belli at the sitting of the 27th of
September, that it had been proved in the presence of various
philosophers that, on the interposition of a flame between the
two poles of an electro-magnet, it was repulsed at the instant
the electric current was closed, to return to the first position
the instant it was broken. This discovery received well-
merited applause in the sitting of the 28th of September,
from the General Secretary and the Secretary of the Section of
Physics. A wish was expressed by some to witness the experi-
ment of Bancalari; and a Daniell's apparatus having been got
ready, often elements eighteen centimetres each in dimension, I
endeavoured to repeat the experiment in the Cabinet of Physics
of the Royal Imperial Lyceum of Venice; but I did not chance
to see the asserted phaenomenon. My temporary magnet
had the power of sustaining above 48 kilogrms. weight; but
as my principle is, that a negative argument never destroys a
positive one, I for my further information requested the
machinist Cobres to give me the particulars of the apparatus;
Belli not having treated of these in his report, and they having
escaped Prof. Zambra, the Secretary of the Section. I knew
that the two pieces of soft iron, which constituted the inter-
rupted anchor, were perforated in the axial direction. I
suspected that the repulsion of the flame was not the immediate
effect of the magnetism, but of two currents of air issuing from
the apertures of the perforated keeper generated by a vorticose
movement produced by the magnetism, as the celebrated
Faraday had observed in liquids* ; and I was confirmed in this
suspicion by the negative experiment which I had instituted in
Venice with solid pieces. On arriving in Turin, I communi-
cated my doubts to the well-known mechanicians Jest, father
and son, who to their professional abilities unite a rare courtesy.
They soon furnished me in their laboratory with a Bunsen's ap-
paratus, and constructed terminal pieces of soft iron forming the
interrupted anchor, both solid and pierced, of a parallelepipe-
don and cylindric form, as I pointed out to them ; and I have
repeated the experiments in their company : the temporary

* RaccoHa, cited above, t. ii. Relazione dell' influenza delle forze elet-
triche e magnetiche sulla luce ed il calorico.

ijohen und£r the Electro-Magnetic Influence. 423

magnet, made in the shape of a horseshoe, was formed of a
cyhnder of soft iron of the length of 0"^*335 and the diameter
of O'^'OIS; and its electro-magnetic spiral was formed of a
copper wire 33"^ long, and of a diameter of a millimetre and
a third ; the internal distance of the poles was 0'"'027; the
two solid parallelepipedon contacts, forming the interrupted
anchor, were 0'»-04 long; and of the sides 0«i'01 1 and 0°^-006 :
and the hollow terminal pieces were 0™*035 long ; and of the
side 0™*009. They were placed at a distance from one another
of four to five millimetres, the magnet being kept in a vertical
position with the poles turned upwards. In front of the in-
terval of the separation of the contact pieces was placed the
flame of a small candle, or of a little oil or alcohol lamp, so that
it surmounted with its top by nearly a fourth the thickness of
the contacts. The electric circuit was closed by copper wires,
and the metallic unions were maintained both at the magnetic
poles and at those of the pile by clamps : one of the wires
therefore was divided into two equal parts, and the ends
being dipped into a tumbler of mercury, allowed the closing
and opening of the circuit at pleasure.

/ have constantly observed repulsion ifi the act of closing the
circle, which lasted the xohole time that the magnetism was kept
up ; and, when in the act of opening the circle, I saw thejlame .
return to its primitive position. Well-satisfied with having
in this manner confirmed this important fact which reflects
honour on its discoverer, I applied myself to the study of the
phaenomenon, and I found —

I. That this happens with contacts of both solid and hollow
soft iron; whereupon I abandoned my suspicion that the
movement of the flame was attributable to currents of air; I
convinced myself that it was an immediate action of the mag-
netism upon the flame, — a fact of the greatest importance to
science.

II. That the repulsion, when it is quite distinct and the
flame quite pure, and terminated in a well-shaped top, is ac-
companied by depression : repulsion and depression are simul-
taneously observed at the closing of the circle ; the return of
the flame and rising of the same, at the opening of the circle.

III. That, ceteris paribus, the greatest effect takes place when
the flame is touching the convex of the magfietic curves indicated
b\j iron filings.

IV. That the action is tiull, or almost nullf when the flame
is placed in the centre of the interval which separates the two
contacts.

V. That in the manifestation of the effects stated above, it is
not necessary for the contacts to be entirdy separated: they may

424 On the Lifluence of Electro -magnetism upon Flame.

be placed at an angle and touch at two corners; the flame
placed within the base of this triangle, generally manifests the
two phsenomena indicated.

VI. That thei'e is a certain mass of the contacts {or keeper
pieces) "vohich is the most efficacious : bej/ond a limits nsohich can
he shown by experiment^ increase of the mass causes a diminution
of the effect: from this I found the cause of my negative
results, which I obtained in Venice in the first experiments
that I made.

VII. That the movements of the fame increase isoith the
number of the pairs {of battery plates). With one pair the
effect was not perceptible to me* : with two pairs the movements
began to show themselves ; with three pairs they became distinct,
and increased with the increase of the number of pairs up to ten,
which was the greatest that I employed in this experiment. The
pairs were of the known ordinary size.

On the repetition of the phaenomena as above stated, the
precaution was taken to cover the apparatus with a bell, which
was open above and supported by two discs below, which left
a free access to the air, by which to support the combustion :
in this manner all agitation and danger of disturbance under
the circumstances were avoided.

I must not forget, in concluding this article, to state that
the celebrated Prof. Gazzaniga, starting from his numerous
experiments, which demonstrate the influence of magnetism
upon the same aeriform fluids, in a manner therefore different
from that of Bancalari, was induced to consider the sun and
all the other celestial bodies as so many enormous magnets;
by which he established that attraction is merely an effect of
the magnetism of the great celestial masses placed at an
enormous distance, — an idea which reappeared in IH^Q in
Prussia, and in ISi? in France, as we see fi'om the Comptes
Mendus of the Royal Academy of Sciences at Paris, The
mystery that attraction operates at a distance without inter-
media would be removed in this case, and the phaenomena of
attraction would enter again into the class of those of common
dynamics.

Dalla Gazz. Piem,, Oct. 12, 1847, No. 242.

• Messrs. Jest prepared for me last evening an electro-magnet of a
circular form interrupted by a prismatic section having an interval of two
millimetres ; and I had, without need of contact pieces, the phaenomena
distinct with a single element. The most conspicuous movements here
appeared in the greater proximity of the flame to the section.

The complete apparatus, of a circular form, furnished with a glass bell
with its accessories is sold in Turin by Messrs. Jest, at the price of thirty
francs, not including the electro-motor.

[ 425 ]

LXVI. On Asymptotic Straight Lities, Plmiesy Cones and Q/-
linders to Algebraical Surfaces. By Thomas Weddle*.

IN the Cambridge Mathematical Journal, first series, vol. iv,
pp. 42-4' 7, the late D. F. Gregory gave a very excellent
method of determining the asymptotes to algebraical curves. I
here purpose considering the corresponding subject relative
to algebraical surfaces ; and as this seems to have as yet en-
gaged but little attention (if any), I trust the discussion will
not be unacceptable to the mathematical readers of this
Journal.

Definitions.

1. A straight line which passes through a point at a finite
distance and touches a surface at an infinite distance, is called
an asymptotic straight line, or simply an asymptote to the
surface.

2. 1( eve7-y straight line drawn in a plane be an asymptote
to a surface, the plane is styled a conical asymptotic plane to
the surface.

3. If all straight lines drawn in a plane parallel to a straight
line in that plane be asymptotes to a surface, the plane is de-
nominated a CYLINDRICAL asymptotic plane to the surface.

4. An asymptotic cone or cylinder to a surface is a cone or
cylinder having its generators asymptotes to the surface.

If <Pq{xyz) denote a homogeneous function of a;,y, z of the g'th
degree, it is plain that a surface of the p\h degree may be
denoted thus:

Let

x—u V—B z—y , , ^

I m n ^ '

be the equations of an asymptote to (1.) passing through the
point (a/3y) : hence

x—lr-\-a, j/=wr + /3, and ;^ = 7M- + 'y ;

substitute these values oi x,y and z in (1.) and develope each
term, the result is,

* Communicated by the Author.

t The axes may be either rectangular or oblique ; only in the former
case we shall have

but in tlie latter,

/,g, h denoting the cosines of the angles which the axes make with each
other.

426 Mr. T. Weddle on Asymptotic Straight Lines, Planes,

kh-s'''

fp-k-

2.3. ..(s-l)

+Df;,_,+, + fp-s)rP-'> =0*,

where D denotes the operation

^:^D— ^^_,+ !► (3.)

d ^ d d

dl

dm

dn

This equation will determine the values of r at the points
in which the straight line (2.) cuts the surface (1.); now for
all lines parallel to an asymptote, one of these points is evi-
dently at an infinite distance ; hence a root of (3.) being infi-
nite, we must have

f;> = 0; (4.)

and this equation determines the directions of the asymptotes.
The equation (3.) hence becomes

+

2. ..(5-1)

= 0;

(5.)

in which values of/, m, n satisfying (4.) must be substituted.
Now an asymptote being a tangent at an infinite distance, it
follows that the asymptote will be distinguished from all lines
having the same direction by a root of (5.) being infinite ; we
must therefore have

that is,

Dfj, + fp-i = 0;

^a+$^/3 +

d'Pn

dl- dm'-^-d^'^-^^^'-'^^-

(6.)

The equation (4.) shows that every asymptote is parallel to
some generator or other of the cone

^/.(•^i/2r) = 0; (7.)

* In this paper I restrict 6,^,^,x (either with or without a letter or figure
subscribed) to denote homogeneous functions only ; and when these sym-
bols stand alone, they are to be understood as functions of /, m, n ; in otiier
cases the symbols of quantity must be written ; thu? X(i(^I/^) ^^ homogeneous
function of x,i/,z of the qth degree) means the same function oix,y,z that x„
does of /, m, n.

Cones and Cylinders to Algebraical Surfaces, 427

and since («|3y) may be any point in each asymptote, (6.) de-
notes the locus (a, /3, y being the variable coordinates) of the
asymptotes parallel to the same generator of (7.) ; this locus is
therefore a cylindrical asymptotic plane, and it is parallel to
that tangent plane of the cone (7.) which touches along the
generator. Hence, to find the equation of a cylindrical asymp-
totic plane, we have only to take such values of /, m, n as
satisfy (4.) and substitute them in (6.). It thus appears that
when the cone (7.) is not imaginary, there is an indefinite
number of cylindrical asymptotic planes ; one indeed parallel
to every tangent plane of the cone (7.)> with a few excep-
tions, which I shall consider presently.

Should (4.), or, which is the same thing, (7.) be resolvable
into factors, then (7.) will in reality denote as many conical
surfaces ; and if any of these factors be of the first degree,
the corresponding conical surface will degenerate into a plane.

Let ^q be any factor of ip^, and put

hence (6.) becomes

when dj = 0, this reduces to

and this equation, together with fl,^=0, will supply the place
of (4.) and (6.) for those cylindrical asymptotic planes that
are parallel to the tangent planes of the cone $q{x}/z)=:0. Also
similar equations may be found for every factor of <pp.
If the equations

^^-^' ll -^' dm -^' dn -^*

can be satisfied by simultaneous values (/j m^ n^) of/, m, u, (6.)
cannot be satisfied unless (pp-i also =0; if <P;,-i should not
= 0, there will be no cylindrical asymptotic plane correspond-
ing to these values of I, m, n ; but if <p^j_i=0, so that we have

^^-^' if-^' f^' ■^-^' '^p-'-^^r V9.)

then (6.) will be satisfied independently of a, /3, y. We have

only to recur however to (5.), and equate to zero the coefficient

* Since ^p 'n a homogeneous function of /,j?j,n of the/>th degree, we have

dl dm d7i

hence the equations (9.) amount only to four independent equations — the
last four.

4-28 Mr. T. Weddle on Asi/mptotic StraigJit Lines, Planes,

of the first power of r that does not vanish independently of
any relation among a, /3, 7. If this coefficient be that of r^-^,
we have

(10.)

This equation denotes a surface which is evidently the locus
of the asymptotes which are parallel to that generator of (7.)

whose equations are — = ^ = — . Hence (10.) must denote

1 ^^ 1 ^^ I

a cylindrical surface; and as its generators are all asymptotes,
it is an asymptotic cylinder of the second degree (which may
in certain cases degenerate into one or two cylindrical asymp-
totic planes). Should the values of /,?«,» satisfying (9.) also
cause a, /3, y to vanish from (10.), there will be no correspond-
ing asymptotic cylinder, unless <Pp_2 = 0; and in this case we
must equate the coefficient of rP~^ in (5.) to zero, and we shall
have an asymptotic cone of the third degree ; and so on.

Hence, to determine the equations of the asymptotic cylin-
ders to the surface (1.), we must find such values (if any) of
l,m,7i as satisfy (9.), and substitute them in (10.); if all the
terms of (10.) also vanish, we must recur to the coefficient of
yrp-s in (5.) ; and so on. There will be as many asymptotic
cylinders as there are sets of values of /, m, ?i satisfying (9.),
unless, after substituting any set in (10.), &c., the only term
that does not vanish is that independent of a, /3, 7, in which
case there will be no asymptotic cylinder for this set of values.

If (p contain a factor of the form {5^}% the first four equa-
tions of (9.) will be satisfied by Qq = 0; and this, combined with
c j = 0, will give determinate values for the ratios l-i-m-^n,
and the corresponding asymptotic cylinders will be determined
in the way just mentioned. It may happen however that 9^ is
also a factor of (p^-i; and if so, all the equations (9.) will be
satisfied by fl^ = 0, and (10.) now admits of simplification as
follows. Let

9p={^y-^P-2g, and <Pp-, = Q^.^'p_^_,,
then it may easily be shown that when 9^=0,

(11.)

Cones and Cylinders to Algebraical Surfaces. 4-29
Hence (10.) becomes

that is,

which evidently denotes two parallel cylindrical asymptotic
planes ; also since /,?w, n are here only connected by the equa-
tion Q,j = Oy it appears that there are in general two cylindrical
asymptotic planes parallel to every tangent plane of the cone

fl,(^7/2) = 0.

Generally, let{0^}% {^J^'S {d^}'-^- 5, be factors of

<Ppy fp_i, <Pp-2 <Pp-s+i, and put

(here the subscribed letters relative to ^^ ^', &c. are omitted
for simplicity), then it may easily be shown that when 5^ = 0,
we have

Df^=0, B\=0 .... D-'fp=0, D*<p^=2.3...s.4/.{D9J*,&c.

Moreover, the equation to the asymptotic cylinder parallel to
a generator of the cone fl^(a:j/2;)=0, will, by equating to zero
the first coefficient of (5.) that does not vanish independently
of a, j3, y, be found to be

■2jZs^''^'-^2j:^{^=T)^'^^^^^
and this, by what precedes, reduces to

it is evident that the asymptotic cylinder degenerates into s
cylindrical asymptotic planes, all parallel to a tangent plane
of the cone (fg{xj/z) = 0; and there is in general the same num-
ber parallel to every tangent plane of this cone.

The asymptotes to the surface (1.) passing through a given
point (a/3y) will be found by determining the ratios l-^m-r-n
by (4-.) and (6.), and substituting, in succession, each set of

'*"•••} . (12.)

4-30 Mr. T. Weddle on Asymptotic Straight Lines, Planes,

simultaneous values in (2.) ; the resulting equations will be
those of the asymptotes to the surface that pass through the
point («i3y).

Since (4.) is of thepth degree and (6.) of the (/?— l)th, the
equation resulting from the elimination of/ (suppose) from (4.)
and (6.) cannot exceed thep (jo— l)th degree, and consequently
there cannot be more than p (p—l) values of the ratio m-^n.
From this we learn, that through any point in space there
cannot be drawn more than /^(p— 1) asymptotes to a surface
of the pih degree.

This theorem suffers an exception, however, which I pro-
ceed to consider.

It may happen that the point (a/3y) through which the
asymptotes are to be drawn may be so taken as to cause (4.)
and (6.) to have a common factor ^^ (which I shall suppose
to be their greatest common measure). In this case (4.) and
(6.) will be satisfied if ;i^, = 0; and eliminating l,'m,n from
this equation by means of (2.), we have

%,(•«•-«> J/-/3, ^-7) = 0

for the equation to the asymptotic cone, which is the locus of
the innumerable asymptotes that pass through the point (a/3y).
(The factor ;)^,^ may sometimes be resolvable into other factors,
and then the preceding asymptotic cone of the qth degree will
in fact consist of several cones of inferior degrees.)

The division of (4.) and (6.) by X'l will give two equations,
^'p_qZ=0, andx"p-q-i = 0) which admit of no common measure.
Now (4.) and (6.) will be satisfied by these two equations; but the
equations ^'p_q = 0, x"p-q-i =^j will determine not more than
iP~^){P~9~^) ^^^^ of values of the ratios l-^-m-^n, hence
(excluding the generators of the cone corresponding to x<j)
not more than (/> — §')(/'— 2'—!) asymptotes can pass through
the point (a, /3, y).

In order to find those points (if any) which are the vertices
of asymptotic cones, eliminate one of the quantities /, »?, n from
(4.) and (6.), and find thosevaluesof a,/3,y thatwill renderallthe
coefficients of the resulting equation equal to zero. If no such
values be possible, the surface (1.) does not admit of an asymp-
totic cone ; but if values a^, /3,, y^ of a,/3,y can be found, then
the point (aj /Sj y,) will be the vertex of an asymptotic cone.
To find the equation of this cone, we must substitute «i, /3j, y^
for 01, ^,y in (6.), and ascertain Qg the common measure of (4.)
and (6.) thus modified ; then wilH,^(A'— «j, y— jSj, ^ — yj)=:0
be the equation to the asymptotic cone, having its vertex at
the point (aj /Sj yj). If the equation resulting from the elimi-
nation of/, m, or n from (4.) and (6.) can be rendered identically

Cones and Cylinders to Algebraical Surfaces. 431

zero by other simultaneous valuesof a,/3,y, there will be as many
asymptotic cones as there are sets of values. When the eli-
mination referred to above is effected by the process for the
common measure, the factor fl,^ will be the last of the remain-
<lers that do not vanish when «„|3i,yjare substituted fora,/3,y.
It will sometimes be found, however, that (4.) and (6.) have
a common measure independently of «,/3,y, arising from{6^^}^
and 6 being factors of <p^ and (p/,_i ; and in this case we must
proceed with this common measure in the way to be noticed
presently.

When we know that (4.) cannot be resolved into factors,
the determination of the asymptotic cone is very easy ; for
since (4.) admits of no measure but itself, and (6.) is of an in-
ferior degree^ it is evident that if there be an asymptotic cone,
(6.) must be identically zero ; hence if such values «i, /Sp yj can
be given to oL,^,y as to cause the coefficients of (6.) to vanish,
there will be an asymptotic cone of the pth degree, namely,

but if the coefficients cannot be rendered zero simultaneously,
there will be no asymptotic cone. Since «,i3,y enter (6.) in the
first degree only, there will evidently be at most only one set
of values of a,^,y that will render (6.) identically zero; and
hence a surface of the pih. degree may have one asymptotic
cone of the pih degree, but not more, and it is plain that there
cannot be an asymptotic cone of a higher degree.

If (4.) admits of being resolved into factors, and these fac-
tors can be found, the asymptotic cones may be determined
as follows. Let 9^ be one of the factors of ip^, and let fl^ itself
be irresolvable into factors. Arrange (6.), or rather (8.), and
6,/ according to the powers of either /, m or n {I suppose), and
divide the former by the latter until the remainder is of lower
dimensions in / than d, ; then since fl^ is irresolvable into fac-
tors, it is clear that this remainder must be identically zero :
find therefore a.^, /3j, y^ the values of a,|3,y, that make the coeffi-
cients of the remainder vanish, then ^^{x—Uy, y—^^ z-—'/^) = 0
will be the asymptotic cone. As ci,^,y enter (8.) in the first
degree and do not enter 9,^, there cannot be more than one
set of values of «,/3,y, it indeed there be any. The same pro-
cess being repeated with each of the other prime factors into
which (4.) is resolvable, we shall have all the asymptotic cones
which the surface admits of.

The preceding process requires modification when the
second or any higher power of 6,^ is a factor of (j&p. As an ex-
ample, suppose that {d }"* enters as a factor into <p , and put
^^=^|;.{fl^^)'* (d^ not being a factor of ^), When d^=0, Df^

432 Mr. T. Weddle on Asymptotic Straight Lines, Planes,

+ (Pp_i = 0, reduces to (Pp_i = 0, and consequently there will be
no asymptotic cone unless fl^ be a factor of fp_i ; if so, let <Pp-i
=4''.^^, then

becomes 4/'.D5^ + (pj5_2 = 0, which is of the first degree in a.,^,y,
and this (instead of (8.)) being combined with 9,y = 0, may give
an asymptotic cone. If {6 }2 however be a factor of <py,_i, then

becomes (pp-2 = 0, and there will be no asymptotic cone unless
$^ be a factor of (Pj,_2. If this be the case, assume f^-i =4/'. {5, }%
and <pp_2=\J/".fl^, then

reduces to

and this equation, which replaces (8.)} combined with fi,=0,
may give one or two asymptotic cones (but not more, as will
be shown below), unless 6^ should enter both fJ,_■^ and ip^.a in
a higher power than has been supposed ; we shall then have
(Pp_3 = 0; and hence fl,^ must (if there be an asymptotic cone)
be a factor of f^-s. Suppose therefore

<p^_j = vl/'.{aj3, ^^_^=4,//.|5j2^ and ^p_3 = vl;"'.0„ •

then 1 _. ,

becomes

and this equation (which cannot be satisfied independently of
a, /3, y, for 9^ is not a factor of rj/), combined with 5^ = 0, may
give four asymptotic cones.

Similarly, if {fl^}* be the highest power of 9^ that is a factor
of <Pp, it may be shown that «, /3, y enter the equation to be com-
bined with 3j = 0, only through D9^, and that this equation
may rise to any degree in D^^ (except the (5— l)th) not ex-
ceeding 5.

Hence when a power {s) of 5^ is a factor of <Pp, we must
ascertain the highest powers of fl^ that are factors of (p^,_„ <Py;_2,
...<Pp_<+i, also (pp^t the first term of (1.) that has not 9,^ for
a factor ; we must then equate to zero the coefficient (reduced
as above) of the highest power of r in (5.) that does not vanish
independently of a,^,y. If a, /3, y disappear from this equation

Cones and Cylinders to Algebraical Surfaces. 433

so that it becomes (Py,_<=0, there will be no asymptotic cone;
but if this be not the case, then the reduced equation must be
combined with fl,y = 0, in the same way as directed for (8.) and
d =0, and we may get asymptotic cones though not more than
s of them. I proceed to establish the last assertion.

It has been shown above that if {5^}* be the highest power
of ^q that is a factor of <P;;, then tlie equation to be combined
with 6^ = 0 will be of the form

rI/.{D5j' + vI;'.{D5j'-i + =0, . . (13.)

where 4/, \p' ... do not involve «, /3, y, and t may be equal to,
but cannot be greater than s. Now if there be a correspond-
ing asymptotic cone, let (ai/3,yi) denote its vertex j then if

(which I shall denote by Difl^) be substituted for DS, in (13.),
the resulting equation will be satisfied by aid (if necessary) of
d, = 0; hence (13.) must be divisible by D^, — D^^q, so that it
may be written

(D9^-D,y(vI/.{DfiJ'-» + ....) = 0. . . (U.)

Also,if a2j/32»72 beanothersetofvaluesofa,|3y, satisfying (1 3.),
they must reduce the second factor of (l-t.) to zero, for the first
is of a lower degree than 5 . Hence vJ/.jDfl^^}^"^ + . . . . must
be divisible by 'D^q — 'DcP)\ and so on. In this way we shall,
after a certain number {v) of divisions, get an equation,

4/.{D9j '-«+.... =0,

which either does not contain Dflg (and hence «,/3, y) at all, or

which cannot be satisfied by any values of a,/3,7. Rejecting

this factor then as affording no solution, (13.) is equivalent to

(D9^-D,g(D9^-D,y (D9^-D„y = 0,

and each of these factors will give but one set of values of«,|3,y;
hence there will be but v asymptotic cones,

^.(■^-"i. y-^v 2-yi)=0 .... e,(a:-«„, y-/3„, ^r-yj = 0;

and since v cannot exceed t, nor t exceed s, it follows that
there cannot be more than 5 asymptotic cones resulting from
a factor of (p^ of the form {9 }*•

When ^q is of the first degree, it is clear that instead of an
asymptotic cone we shall have a plane ; and since any point
in it may be regarded as the vertex, every straight line drawn
in it will be an asymptote ; hence the asymptotic cone will in
this case become a conical asymptotic plane : also since fl, is
here of the form A/+Bw + Cw,

Dfi^ = A« + B/34-Cy,

Phil. Mag. S. 3. Vol. 31. No. 210. Dec. 184-7. 2 F

434' Mr. T. Weddle on Asymptotic Straight Lines, Sfc.

which does not involve /, m or w. Hence to determine the
conical asymptotic planes (if any) to the surface (1.), we must
take those factors of spp that are of the first degree, and proceed
as directed above for asymptotic cones ; with this modification,
however, that Dd, not containing /, m or n must be regarded
as a single constant, and consequently the process will be much
simplified. If Vj, Vg^.V^ {t not -Ji s) be the values of T>Qq
corresponding to the factor

{S^Y={Al+Bm + Cn}%
we shall have

A.r+By + C^ = Vi, Ax+Bi/+Cz=y^... Ax+By+Cz=Yt
as the equations to the conical asymptotic planes relative to
this factor.

It appears from the preceding reasoning, that if the equation
(4'.), or, which is the same thing, the highest homogeneous
function in the equation to the surface (1.) can be resolved
into a factors of the first degree, b factors of the second de-
gree, c factors of the third degree, &c. (here a factor of the
form {fl }* is to be accounted s factors), then the surface may
admit of, but cannot have more than a asymptotic cones of the
first degree, that is, a conical asymptotic planes, b asymptotic
cones of the second degree, c asymptotic cones of the third
degree, &c. Some of these cones may have the same vertex ;
and since a + 2b + 3c . , . . =p, the degree of the aggregate of
all the asymptotic cones to a surface can never exceed that of
the surface itself.

It will be seen that unless equal factors enter the highest
homogeneous function, the asymptotic cones to a surface de-
pend only on the two highest homogeneous functions in its
equation ; and hence (the above case excepted) all surfaces
having the two highest homogeneous functions in their equa-
tions identical, will have the same asymptotic cones. Also
conversely, it is plain that those surfaces that have the same
asymptotic cones must have the two highest homogeneous
functions in their equations identical, providing the degree of
the equations to the surfaces be exactly equal to that of the
aggregate of the cones. Now this aggregate may be consi-
dered one of these surfaces ; hence if

Wl = 0, Uq=:0, .... Ui = 0

be the equations to cones, the aggregate of which is of the
j9th degree, the equation to all the surfaces of the j5th degree
having these for asymptotic cones may be denoted by

UiU^...Uf-i-Xp-2{^^z) +%p-3(a?2/2) +Xi('^^2;)+xo=0- (15.)

Wimbledon, Surrey, Nov. 10, 1847.

[ 435 ]

LXVII. Oti the Chemical Composition of the Substances em-
ployed in Pottery. By Mr. R. A. Couper*.

A LL kinds of earthenware are composed of two parts, viz.
-^^ the body and the glaze.

The body is the principal part of the vessel, being the base
or foundation, as indicated by the term itself. The glaze is
a thin transparent layer of glass which covers the body and
fills up its pores, giving it a smooth surface with a polished
and a finished appearance.

I. The substances principally employed to form the body
of earthenware are, clays of different kinds, flint and Cornish
stone.

Clay which constitutes the base of the body of earthenware
is distinguished from siliceous earth by becoming plastic when
mixed with water, and being very soft and not gritty to the
feel ; also when burned, it keeps its form, and becomes firm
and solid; whereas siliceous earth crumbles into a powder
when burned. Clay when intensely heated, as in porcelain
manufactories, does not regain its plasticity, which it loses in
the burning, although pounded very fine, in which state it is
technically termed potsherd.

Clay is obtained naturally from Cornwall, Dorset, and
Devonshire, and is the finer particles of decomposed felspar
deprived of its alkali.

1. The finest clay (termed China clay) used in Britain is
obtained artificially from Cornwall, by running a stream of
water over decomposed granite, which carries with it the finer
particles of felspar, and is then received into catchpools or
ponds where it is allowed to subside. The water is then run
off, leaving a fine sediment, which is removed and exposed to
the atmosphere for four or five months, when it is ready for
export. By analysis of this clay previously dried at 212°, I
found it to consist of —

I. ir.

Silica 46-32 46*29

Alumina 39*74 40*09

Protoxide of iron . . '27 '27

Lime '36 '50

Magnesia .... •44<

Water and some alkali 12*67 12*67

99*80 99*82

For the second analysis I am indebted to Mr. John Brown.
The more common clays, which are found naturally depo-

* R ad before the Philosophical Society of Glasgow, April 28, 1847,
and communicated by Dr. R. D. Thomson.

2 F2

436 Mr. R. A. Couper on the Chemical Composition of

sited, are supposed to have been produced in a similar manner
to the china clay ; the rains having washed from the hills the
decomposed rock into a lake or estuary, where it has subsided
and gradually displaced the water, and become in the course
of time perfectly firm and solid, forming fields of clay. The
clay is found in layers or strata lying over each other, each
layer possessing some distinctive property from the othei*,
which renders each clay fitted for a peculiar purpose.

2. Sandy clay (stiff or ball) is the upper layer of clay, and
is used by itself for making salt glazed ware ; it is well adapted
for this kind of ware, in consequence of the considerable quan-
tity of silica or sand which it contains. By analysis of this
clay, I found it to be composed of —

Silica 66-68

Alumina 26*08

Protoxide of iron .... 1 "26

Lime 'S*

Magnesia trace

Water 5'1*

100-00
being previously dried at 212°, specific gravity = 2-558.

3. Pipe clay is the second layer, which is used in making
tobacco pipes. This clay is not employed in manufacturing
earthenware, owing to its possessing the property of contract-
ing more than sandy clay. It was analysed by Mr. John
Brown, who obtained —

Silica 53*66

Alumina 32-00

Protoxide of iron .... 1*35

Lime '40

Magnesia trace

Water 12-08

99-49

4. Blue clay is of a grayish colour, and is considered the
best layer of clay in the whole series, owing to its burning
perfectly white, and approaching in character nearest to the
china clay. As analysed by Mr. John Higginbotham, it was
found to consist of —

Silica 46-38

Alumina 38*04

Protoxide of iron .... 1*04

Lime 1-20

Magnesia trace

Water 13-57

100-23

the Substances emjjloyed in Pottery. 437

also previously dried at 212°. There is a variety of other
clays obtained from these fields, which are of less value, and
need not be enumerated here, as they are similar in appear-
ance to those already noticed.

5. Red or brown clay, which is very abundant in the neigh-
bourhood of Glasgow, is a surface clay, and contains a large
quantity of peroxide of iron, which gives it a deep brown
colour. It is of this clay that common black ware, flower-
pots, and red bricks are made, which do not require a
very high temperature, else they would fuse. The analysis
gave —

Silica 49-44.

Alumina 34-26

Protoxide of iron .... 7*74

Lime 1-48

Magnesia ...... 1*94

Water 5-14

100-00

6. Yellow clay is obtained from various parts of the country,
and is so called from possessing a yellow colour both before
and after being burned, owing to the presence of iron.

By mixing sandy clay and red clay together, we gain an
artificial yellow clay, which is often employed.

Yellow clay, as analysed by Mr. John Brown, was found to
contain —

Silica 58-07

Alumina 27-38

Protoxide of iron .... 3-30

Lime -50

Water 10-30

Magnesia ....... trace

99-55

7. Fire-clay is also very abundant in this country, and oc-
curs both on the surface and several fathoms under ground.
It is termed marl, and is used principally in potteries for ma-
king saggars or vessels for placing the ware previous to burn-
ing to protect them from the flame; and owing to its coarse
particles, which cause the body to be very porous, is well
adapted for strong heats: crucibles, or large pots for glass
works, in which the glass is fused, are also made from fire-
clay, as well as bricks known under the name of fire-brick.
This clay was analysed by Mr. John Brown, who obtained —

438 Mr. R. A. Couper on the Chemical Composition of

Silica 66-16

Alumina 22*54

Protoxide of iron . . . . 5*31

Lime 1*42

Magnesia trace

Water 3'14

98-57

8. Flint as used in potteries is first calcined, then water-
ground, in which state it is used for mixing with clays, and is
called slop flint; but for glazes it is evaporated to dryness,
and used in the dry state with other articles which constitute
the glaze.

9. Cornish stone or granite is water-ground, then evapo-
rated to dryness for mixing in glazes, and is used in the slop
state for mixing with clays.

10. Plaster of Paris or gypsum, which is employed in form-
ing the moulds in which certain kinds of pottery are cast, is a
native sulphate of lime. It is a very important article to the
manufacturer of earthenware, owing to its singular property
of parting easily with the clay by the application of a slight
heat. Plaster of Paris requires to be dried at a high tempe-
rature before using it; but if it is over-dried, it will not again
set for making moulds ; the drier the stucco the harder are
the moulds that are made of it, and they will stand more
readily a greater degree of wear. Plaster of Paris casts, as
commonly prepared, cannot again be used for the same purpose.

11. The colours used for printing and painting on ware are
similar to one another, excepting that the colours for painting
may not be so expensive as for printing ; both however form
an important and extensive part of the materials of a pottery.
The manufacturers of earthenware are much occupied with
the improvement of the variety and beauty of the colours, as
well as of the patterns or styles that are produced, and hence
a great emulation exists among those employed in the trade.

1. The blue colour in printing is produced from cobalt,
which is used with flint, ground glass, pearlash, white lead,
barytes, china clay, and oxide of tin in reducing its strength.

2. The brown colour by ochre, manganese, and cobalt.

3. The black colour by chromate of iron, nickel, ironstone,
and cobalt.

4. The green colour by chrome, oxide of copper, lead, flint,
and ground glass.

5. The pink colour by chrome, oxide of tin, whiting, flint
ground glass, and china clay, which are mixed in various pro-
portions, fused together at a high temperature, then pounded
and mixed with oil, when it is ready for the printer's use.

the Substances employed in Pottery. 489

For the following analysis of a blue cobalt calx, I am in-
debted to Mr. John Adam : —

Silica 17-84.

Peroxide of cobalt 19'42

Peroxide of iron 25*50

Water 8'41

Carbonate of lime and magnesia . 28'45

99-62

The oil that is used for mixing with the colours, is made
by boiling the following substances together; viz. linseed oil,
rape oil, sweet oil, rosin, common tar, and balsam copaiba in
various proportions.

III. It is but recently since a new method has been applied
to cause the colours to flow or spread over the surface of the
ware. This object is effected by washing the saggars in which
the ware is placed previous to its being fired in the glost kiln,
mth a mixture of —

1. Lime, common salt, and clay slip. Dry flows are also
used, which answer equally well, the mixture being sprinkled
on the bottom of the saggar. The following are some of those
flows: —

2. Lime, sal-ammoniac and red lead.
S. Lime, common salt, and soda.

4. Whiting, lead, salt and nitre.

5. But there is a wash made of lime, clay slip, nitre, salt,
lead, in general use for washing all the saggars employed in
the glost kiln, which fuses on the inner surface of the saggar,
making it perfectly close and not porous, otherwise the gloss
required on the surface of the ware would not be obtained.

IV. The colours used in producing the dipt or sponged
ware are of a very cheap kind, as it is only for common pur-
poses that they are employed. The colours when used for
dipt ware are put on the ware before it is burned ; and when
used for sponged ware, are put on the ware in the biscuit
state. The following are some of those colours : —

L A black dip is made from manganese, ironstone and
clay slip.

2. A drab dip by nickel and slip.

3. A sage or a greenish-blue dip by green chrome and slip.

4. A blue dip by cobalt and clay slip.

5. A yellow dip by yellow clay alone, or a compound of
white and red clay, which produces the same results.

6. A red dip is produced from the red or brown clay ; but
it is not every quality of this clay that will answer, as it re-
quires to burn red.

The first four of these dips are prepared by mixing a little

440 Mr. II. A. Couper on the Chemical Composition of

of the colouring agent with a quantity of clay slip ; whilst the
two last-mentioned dips are mixed with water to produce the
slip state, in which state they are employed.

V. There are several kinds of bodies manufactured ; but
they may be all classed under two heads, viz. porcelain and
earthenware.

1. Porcelain or china is a rich, very smooth and transpa-
rent ware, and is the finest quality that has yet been manu-
factured. It is a fused body, and owes its transparency to
this circumstance ; it also requires a very high temperature to
burn it, and is manufactured in this country from flint, Cor-
nish stone (granite), china clay, and bone-earth ; the lime
employed acting as a flux, partly fusing it. By analysis of
two pieces of china from different manufactories in Stafford-
shire, I found them to be differently composed. The last of
these pieces was also analysed by Mr. Crichton, the three
analyses being as follows: —

No. 1, by R. A. C. No. 2, byR. A. C. No. 2, by W.C.

Silica 39-88

Alumina .... 21*48

Lime 10'06

Protoxide of iron ~\ nc.AA,

Phosphate of lime J

Magnesia

Alkali or difference 2* 14

40-60

39-685

24-15

24-650

14-22

14-176

15-32

15-386

•43

•311

5-28

5-792

100-00 100-00 100000

2. Foreign manufacturers do not employ bone-earth ; but
instead of it they use felspar, the alkali of which supplies the
place of the phosphate of lime. The Germans make the best
porcelain for chemical purposes, as that body is more vitrified
and less liable to be acted upon by acids, as well as being
capable of standing a very strong heat; and hence it is exten-
sively used by chemists. By the analysis of some specimens
of foreign porcelain, I obtained the following results: —

Berlin.

Silica 72-96

Alumina and protoxide of iron 24-78

Lime 1-04

Alkali 1-22

100-00

Specific gravity 2-419

VI. Earthenware is a very porous and less compact body
than china or porcelain, owing to its containing little or no
alkali, which is the great difference between these bodies. I had
a piece of ware manufactured, resembling in appearance porce-
lain, as regards the absence of porosity and its compactness,

Chinese

Porcelain,

superior.

inferior.

71-04

68-96

22-46

29-24

3-82

1-60

2-68

100-00

99-80

2-314

2-314

tJie Substances employed in Pottery, 44- 1

slightly transparent, and capable of standing a very strong and
sudden heat; it was produced by mixing soda to the extent ol' 3^
per cent, in a 1 ittle clay prepared for the common white body, and
was then fired in the biscuit kiln. The clay employed having
been previously well dried, so as to weigh it without water, the
proportional quantity of soda requisite was then calculated and
weighed out ; the clay was again mixed with water along with
the soda ; it was then formed into capsules, which after being
fired and then broken, presented the appearance of a vitrified
or fused body.

1. The common white ware or earthenware is made from
flint, Cornish stone, china clay, and blue clay, and does not
require such a high temperature in burning as the porcelain
does. By analysis of a piece of white ware manufactured in
this city, it was found to contain —

Silica 68-55

Alumina and protoxide of iron . 29'13
Lime 1-24

98-92
Specific gravity 2*36

Coloured ware is also manufactured from the same sub-
stances, but mixed with a colouring agent which stains the
body.

2. The toqua or blue-coloured ware is coloured by cobalt.

3. The sage or greenish- blue coloured ware, by nickel and
cobalt.

4. The drab or buff-coloured ware by chromate of iron.

5. The body for the cane or yellow-coloured ware is pro-
duced by a mixture of sandy clay and common red clay, the
same as used for red bricks, but is generally produced from
the natural yellow clay found in particular localities.

6. The last-mentioned body is also employed for making
Rockingham ware, which only varies from the cane ware by
possessing a different glaze.

7. The common black ware body is made from the red clay
alone.

8. The Egyptian ware body is made from ironstone, ball
and red clay.

These four last-mentioned bodies are not nearly so expen-
sive as the white ware, and do not require nearly such a high
temperature to burn them ; therefore they are, comparatively
speaking, soft bodies.

9. Salt glazed wai'e is made from sandy clay and a little
sand, to keep the body open, or make it less compact; but
for large salt glazed ware, potsherd, which is ware that has

4«42 Mr. R. A. Couper on the Chemical Composition of

been fired and then ground, is employed to render the body
still more open or porous, and also to give it a greater capa-
bility of standing sudden heats or colds. This ware is much
used in public works for chemical purposes : it is exposed to
the action of the flame during burning, whereas other kinds
of ware are protected by saggars from the flames.

VII. The glaze vitrifies the surface of the body, rendering
it generally capable of withstanding acids. It is a very im-
portant point with the manufacturer to obtain a glaze which
will adhere to the body without crazing or peeling off, as he
may discover a good body, but not find a glaze to answer it,
since every glaze will not adhere to the same body, and hence
every manufacturer has a glaze of his own composition.

1, The substances used in the preparation of the glaze for
white ware, are borax, china clay, flint, Cornish stone, Paris
white, and white lead.

In preparing the glaze, a substance technically termed frett
is first made, consisting of borax, china clay, flint, Cornish
stone, and Paris white, which are fused together in a kiln, and
when ready allowed to flow into water, which shortens it,
owing to the water being mechanically lodged in it, and keeps
it from adhering to the bottom of the vessel, rendering it much
easier to pound. Frett is a beautiful glass, coloured by a little
iron, and is pounded and water-ground along with Cornish
stone, flint, and white lead : this constitutes the glaze for white

ware.

Analysis of Analysis of
white glaze. frett.

Silica 43-66 55-98

Lime ;........ -52 2-52

Alumina and protoxide of iron 9-56 10-38

Borax 20-08 3M2

Carbonate of lime .... 10*88

Carbonate of lead .... 15-19

99-89 100-00

Specific gravity 2*345

A piece of earthenware was brought from America, having
been discovered several feet under ground, the glaze of which
was tested, and found to be composed of silica, iron, alumina,
lime, sulphate of lime and antimony, which was a beautiful
rich white glaze concealing a common red clay body.

2. The glaze of Rockingham ware possesses a beautiful
brownish metallic lustre, and is made from Cornish stone, flint,
manganese, red lead and clay slip, the latter substance being
a little clay mixed with water until it becomes of the consist-
ency of milk,

3. The glaze for compapo blftck ware is made from the same

the Substances employed in Pottery.

443

materials in different proportions, and has a brilliant black
appearance.

4. The glaze used for cane or yellow-coloured ware is made
from flint, red lead, and Cornish stone.

5. The Egyptian ware owes its value to the beautiful and
rich tinted black glaze, made from flint, Cornish stone, red
lead, and manganese, with which it is covered.

These four last-mentioned glazes are made by stirring the
substances together with a certain quantity of water, and pass-
ing it through a very fine sieve or search. Glazes do not re-
quire such a high temperature to fuse them on the surface of
the ware, as the body does to be burned.

6. The glaze for salt glazed ware is common salt, which is
thrown in at the top of the kiln through a number of small
apertures in the crown of it, and diffuses itself through all
parts of the kiln, giving the ware the required glaze. The
action that is supposed to take place, when the salt is thrown
into the kiln, is owing to its decomposition. The chlorine of
the salt combines with the hydrogen of the water, which is
mechanically lodged in the salt, forms muriatic acid gas, which
passes off, while the sodium combining with the oxygen of the
water then unites with the silica in the ware, forming a sili-
cate of soda which fuses on its surface. The salt is not thrown
in until the kiln has been raised to its greatest necessary tem-
perature.

Table of the Composition of Clays and Porcelain when free
from Water.

Cornish china clay

Cornish china clay

Sandy clay

Pipe clay

Blue clay

Red clay

Fireclay

Yellow clay

English china ware, No. 1 ...

No. 2...

No. 2..,

Berlin ware

Superior Chinese ware

Inferior Chinese ware

Common English white ware

53-16
5312
70-29
61-39
53-52
52-11
69-33
65-06
39-88
40-60
39-68
72-96
71-04
68-96
68-55

45-61

46-00

27-47

36-61

43-89

36-19

23-62

30-68

21-48

2415

24-65
24-78
22-46
29-24
29-13

-31
•31
1-33
1-54
1-20
8-17
5-56
3-70

-41

•57

-90

•46

1-39

1-56

1-49

-56

10-06

14-22

14-18

104

3-82

1-60

1-24

51
51

Trace
Trace
Trace
2-04
Trace
Trace
Trace
43
31
Trace
Trace
Trace
Trace

26-44
15-32
15-39

2-14

5-28
5-79
1-22
2-68

2-558

2-419
2-314
2-314
2-360

";^ [ 444 ]

LXVIII. On the Polarization of the Atmosphere. By Sir
David Brewster, K,H., D.C.L., F.Ii.S.y and V.F.R.S.

Edin.^

WHEN the light of the sun or of any self-himinous body
has been transmitted through certain crystallized sub-
stances, or has been reflected from, or refracted by, bodies
not metallic, it suffers a physical change, to which the name
o^ plajie polarization has been given. This physical change
consists in decomposing common light into two equal portions
of polarized light, one of which is polarized in a plane at right
angles to that in which the other is polarized. In doubly
refracting crystals, the two pencils are polarized in opposite
or rectangular planes ; and when common light is reflected
from any body not metallic, whether it is solid, or fluid, or
gaseous, a portion of the incident light enters the body ; and
of the portions thus reflected and refracted, precisely the same
quantity is polarized, — the light polarized by refraction being
polarized in a plane at right angles to that which is polarized
by reflexion.

If the earth had no atmosphere the sky would appear ab-
solutely black ; and when the sun sets we should be left in
utter darkness. The existence of twilight, however, the blue
colour of the sky, and the refraction of the rays which emanate
from the stars and planets, place it beyond a doubt that the
pure air in which we live and breathe is capable of acting
upon light like all other bodies, and consequently of producing
that physical change which constitutes polarization. The
polarization of the blue sk}', or of the atmosphere, was there-
fore observed and studied by different philosophers, both in
France and England ; and it was speedily ascertained, in con-
formity with the laws of polarization, that the polarization was
a minimum in the vicinity of the sun, where his light is reflected
at angles approaching to 90°, or where the incident and re-
flected rays form an angle approaching to 1 80° ; that it was
also a minimum in the region opposite the sun, where his light
is reflected at an angle approaching to 0°, or at a perpendi-
cular incidence; and that it was a maximum in those interme-
diate parts of the sky, which are distant about 90° from the
sun, and where his light is reflected at an angle of about 45°,
the polarizing angle for air.

Such was the first view which was naturally taken of the

* This paper is reprinted, with the permission of Dr. Berghaus and Mr.
A. K. Johnston, from the Seventh Part of their vakiablePh3sical Atlas now
in the course of publication. A map representing the four neutral points,
and the system of lines of equal polarization, will be found in that work.

On the Polarization of the Atmosphere. 4''t5

polarization of the atmosphere, and a considerable time elapsed
before its leading elements were determined, and its more
important phaenomena observed and measured. It is to M.
Arago, to whom this branch of science owes such deep obli-
gations, that we are indebted for the discovery of the first and
leading fact on which the law of atmospheric polarization
depends. In examining the region of the sky opposite to the
sun, he discovered a neutral point, or a point in which there
is no polarization whatever. This neutral point he found to
be 25° or 30° above the point diametrically opposite to the
sun, or what we may call the antisolar point ; and we shall
distinguish this pole of no-polarization by the name of M.
Arago's neutral point, or the antisolar neutral point. It is
best seen after sunset.

In the year 1840, M. Babinet discovered a second neutral
point, situated about the same distance above the sun as the
neutral point of M. Arago is situated above the antisolar point.
This point is most distinctly seen immediately after sunset,
but is generally much fainter than the other, owing to the
discoloration of the blue sky by the yellow light of the set-
ting sun.

Our readers are no doubt aware, that when light is reflected
from the surfaces of transparent bodies, a certain portion of
it, and at a particular angle the whole of it, is polarized in the
plane of reflexion, or positively'^', while precisely the same
quantity of the transmitted light is polarized in a plane at
right angles to the plane of reflexion or refraction, or nega-
tively. Now, in the part of the sky between the neutral point
of M. Arago and that of M. Babinet, the light is polarized
positively, while in the parts of the sky between the first of
these neutral points and the antisolar point, or between the
second and the sun, it is polarized negatively. Hence it
became obvious that the two neutral points must be produced
by a compensation, in which light polarized negatively neu-
tralized light polarized positively, and that the negative light
was either produced by reflexion in a plane at right angles to
that passing through the sun, the neutral point, and the ob-
server, or by refraction in a plane passing through these three
points, or by both these causes combined. But in whatever
way the negative polarization was produced, it was manifest
that the same cause ought to produce a neutral point beneath
the sun. After many fruitless attempts to discover this neutral
point — owing chiefly to the predominance of the sun's light

♦ These terms are used for the purpose of abbreviation. An account of
the laws of the polarization of light by reflexion and refraction, will be
found in my papers in the Phil. Trans., 1815, p. 129, and 1830, pp. 69, 133.

446 Sir David Brewster on the Polarization

at the part of the sky where it should be found — I at last ob-
served, under a very favourable state of the atmosphere, that
the polarization of the sky was negative in the space between
the risen sun and the horizon. This observation placed it
beyond a doubt that there must be a neutral point below the
sun, where that «^orflr^/t;e polarization passed mio positive pola-
rization; and by concealing the sun from view, and admitting
no light to the eye but what came from the probable place of the
neutral point, 1 succeeded in discovering it. After commu-
nicating this discovery to M. Babinet*, early in 1845, he made
several ineffectual attempts to confirm it; and it was not till
the 23rd of July IS^G, when the state of (he sky was peculiarly
favourable for the observation, that he succeeded in obtaining
a distinct view of itf.

Before proceeding to explain the map of the lines of equal
polarization in the pure blue sky, I shall give a brief account
of my observations on the three neutral points to which 1
have referred : —

I. On M. Arago's Neutral Point.

In the normal state of the lines of equal polarization,
namely, when the sun is in the horizon, this neutral point
is about 18^° above the horizon or above the antisolar point;
but when the sun is about 11° or 12° above the horizon,
and the antisolar point of course as much below it, the neutral
point is in the horizon, and consequently only 11° or 12°
above the antisolar point. As the sun descends to the horizon,
and the antisolar point rises, the distance of the neutral point
from the latter gradually increases ; and when the sun reaches
the horizon, the neutral point is 16^° above it, and therefore
18|° distant from the antisolar point. After the sun has
set, the distance of the neutral point from the antisolar
point increases ; that is, it rises faster than the sun descends,
and its maximum distance when the twilight is very faint, is
about 25°.

In the latitude of St. Andrews, M. Arago's neutral point is
above the horizon all the day between the middle of November
and the end of January. In the other months of the year it

* Comptes Rendus des Seances de t'Acad. des Scieiices, torn. xxii. p. 801-
803,1845,17th Mars.

•f- Comptes Rendus, &c,,i\n\\tt'^, 1846, torn, xxiii. p. 195; andAoiitS,
1846, torn, xxiii. p. 233. " M. Brewster," says M. Babinet, "a sans doute ^te
guide dans sa recherche par des vues theoriques ; autrement il me parait
pen probable qu'il eut fait, par observation seul de la polarization atmo-
spherique, la decouverte remarquable de ce point neutre si difficile a recon-
naitre, et que, depuis luij' avals plusieur fois tente inutilement de retrouver."
Ibid. p. 235.

of the Atmosphere, 447

never rises till the sun is within 11° or 12° of the horizon, and
never sets till the sun is 11° or 12° above the horizon.

II. On a secondary Neutral Point accompanying M. Arago's
Neutral Point.

I observed the first traces of this remarkable phsenomenon
on the 8th of June 184-1, at 5^ 50', when the positive polari-
zation was strongest close to the horizon, whether land or sea,
and to about 1^ above it. Hence, when M. Arago's neutral
point rose, it did not appear Jirst in the horizon^ but about l^-^
above it, the compensation being effected where the positive
polarization was weaker than in the horizon. When this took
place, we had the singular phaenomenon of a neutral point
xmth positive polarization on each side of it. When this phae-
nomenon was more fully developed, under a favourable state
of the horizon, the positive polarization was overcome by the
advancing negative polarization. The negative polarization
was then immediately below the ascending neutral point; but
at a certain distance, a few degrees below the neutral point,
the negative polarization was compensated by the excess of
positive polarization close to the horizon, and the beautiful
phaenomenon was seen of two neutral points separated by bands
of negative polarization ! This phaenomenon was best seen on
the sea horizon, which was marked by an obscure band a ^Q.yi
degrees high, that indicated the existence of a distant haze.
On the 21st of April 1842, I observed the secondary neutral
point under favourable circumstances. At 6^ 22' p.m., when the
primary neutral point was 15° high, the secondary one was
2° 50' high. At 7^ positive bands were still seen above the
sea line, and were strongest upon the obscure band above the
visible sea line.

III. On M. Babinet's Neutral Point.

This neutral point is situated about 18° 30' above the sun,
when he is rising or setting in a very clear sky. It is not so
easily seen as that of M. Arago, and was therefore longer in
being discovered. It is above the horizon during the greater
part of the year in great latitudes, and being above the sun,
it is of course always visible when the sun is above the horizon
in a clear sky. When the sun is in the zenith, this neutral
point coincides with the sun's centre. As the sun's altitude
diminishes, it separates from the sun's centre, its distance gra-
dually increasing till it becomes 18° 30', when the sun's alti-
tude is nothing, or at sunrise and sunset.

The neutral point of M. Babinet must^ like that of M.
Arago, be accompanied, in certain states of the horizontal

448 Sir David Brewster on the Polarization

sky, with a secondary neutral point; but I have never had an
opportunity of observing M. Babinet's neutral point when it
either rose above or set beneath the horizon, which, though
not essential, is the most favourable for observing a secondary
neutral point.

IV. On the Neutral Point below the Sutz.

This neutral point is, as we have previously noticed, much
more difficult to be seen than that of M. Babinet. In No-
vember, December and January, it cannot be seen in our lati-
tudes, unless when, early in November and late in January, a
higher degree of polarization in the sky brings it above the
horizon at noon.

As theory indicated the existence of this neutral point, I
long sought for it in vain ; but when I was assured of its ex-
istence by the discovery of negative polarization, which often
extended from the sun to thehorizon even when the sun's altitude
was 30°, I took such precautions for excluding all unnecessary
light from the eye that I at last observed it near the horizon,
with a small portion of positively polarized light beneath it.
I afterwards observed it repeatedly when the sun had higher
altitudes, and was able to measure its varying distance from
that luminary. On the 18th of February 1842, at noon, when
the sun's altitude was about 22°, I observed this neutral point
in the most distinct manner, the polarized bands being nega-
tive below the sun, and positive near the horizon. Its distance
from the sun, therefore, was about 15° or 16°. I afterwards
obtained the following measures of its distance from the sun : —

Distance of neutral point

h ,

1842, February 21,

12 39

... April 3,

11 45

6,

11 6

8,

2 7

from the sun.

15
13
12
16

0
0
0
0 estimated.

On the 20th of April, in a very fine day, the wind being
west and the barometer 30*02, I obtained the following mea-
sures:—

April 20,

The maximum polarization of the sky at the time of these
observations was equal to a rotation of 25^^°, about 4^° below
the greatest maximum.

12 10

Di

stance from sun,
11 20

12 37

10 40

2 21

12 0

3 45

12 35

of the Atmosphere. 449

On the 26th of April 184'2, when the barometer was at
SO'OO, and not a cloud in the sky from morning till night*, I
obtained the following measures : —

April 26,

11 1

Distance from sun.

0 /

12 15

11 46

12 30

3 30

14 35

3 35

15 5

4 10

17 45

At 10^ 53', the maximum polarization of the sky, or the
rotation, was 28f °, and at 1 1'^ 46', and 3*^ 42', it was 28|°.

On the 27th of April I observed a remarkable series of
phaenomena relative to this neutral point. The sky was sin-
gularly fine — the barometer at 30-04, and at 10^ 41' the maxi-
mum polarization of the sky 29|^°, the greatest that I have
observed. At lO*" 45', the distance of the neutral point from
the sun was 12° 3', and consequently about SS^° above the
horizon. At 12^ 12', a fog came rapidly from the sea. The
neutral point below the sun was driven beneath the horizon,
and Babinet's neutral point rose almost to the zenith. At 1^ 20'
the fog diminished. The neutral point below the sun reap-
peared near the horizon, oscillating up and down, through a
space of 5° or 6°, as the fog became alternately denser or
rarer !

When the sky is clear, the neutral point below the sun
approaches to the sun as his altitude increases, and finally
coincides with the sun's centre when he is in the zenith. Hence
it follows, that when the sun is in the zenith, the two neutral
points in his vicinity meet in the sun, and the system of pola-
rization lines becomes uniaxal.

Were the sky sufficiently clear, we should doubtless find a
secondary neutral point accompanying the primary one below
the sun ; but in our climate there is little chance of this phae-
nomenon being distinctly observed.

In his observations on the antisolar neutral pointy M. Arago
observed that it sometimes deviated from the plane passing
through the antisolar point and the eye of the observer; and
he justly ascribed this deviation to the influence of luminous
clouds situated out of this plane. The same phaenomenon
takes place in reference to the other neutral points, though
the deviation is in these cases less distinctly seen, from the
interference of the sun's light. But it is not merely the posi-
tion of the neutral point that is influenced by the intrusion of

* The lines in the spectrum were ilKdefined, from unequal refraction in
the air.

Phil. Mas. S. 3. Vol. 31. No. 210. Dec. 1847. 2 G

460 Sir David Brewster on the Polarization

light different from that of the sky ; the degree of polarization
is always affected whenever we measure it in parts of the sky
which have luminous clouds or illuminated terrestrial objects
in their vicinity, or any luminosity in the field of view of the
polarimeter. If the neutral point happens to be above or
below any such object, its distaiice from the antisolar point or
from the sun is increased or diminished*.

V. On the Maximum Polarization of the Sky.

After having ascertained the position of the neutral points,
or yoles of no-polarization as we may call them, the next most
important element to be determined is the maximum •polariza-
tion of the atmosphere.

When a ray of common light is reflected from any trans-
parent body, at an angle whose tangent is equal to the index
of refraction, it is completely polarized ; or when a ray of light,
completely polarized in a plane inclined 45° to the plane of
reflexion, is reflected from any such body, its plane of polari-
zation is brought into the plane of reflexion ; that is, its plane
is turned routid 45°. Hence complete polarization is measured
by a rotation of 45°. When the polarized ray is reflected at
angles above or below the angle of maximum polarization, its
plane is less turned round, and its rotation is more or less
than 45°, according as the angle of reflexion is more or less
distant from the angle of maximum or complete polarization f.

Different degrees of rotation below 45° may also be pro-
duced by the refraction of the polarized ray at one or more
surfaces of glass if, the rotation increasing with the angle of
incidence. Hence we may measure the degree of polarization
wherever it exists, by observing at what angle of incidence it
is compensated or neutralized, by reflexion from a transparent
surface, or by refraction at one or more such surfaces. I have
found the last method the most convenient, and have therefore
constructed a polarimeter which measures the polarization of
the sky, by observing with it either the varying angle at which

* On the 16th of May 1842, barometer 30*3, the sun was faintly seen
through a thick haze. At S*" 49' a.m. the polarization was positive all the way
from the sun to the horizon, so that the neutral point below the sun was
below the horizon. Immediately afterwards the sun was quite hid — a great
glare supervened, and a quaquaversiis polarization was observed, in which
the polariscope gave no coloured bands.

On the 17th of May, at 6" 30', the sun's disc was quite white through a
thick haze, and there was no neutral point either above or opposite the sun,
the polarization being everywhere positive. When the haze is thicker on
one side of the plane passing through the sun's spectrum, the neutral point
deviates from that plane.

f See Phil. Trans., 1830, p. 69. - + Ibid. p. 133.

of the Atmosphere. 451

it is compensated or neutralized by a fixed number of thin
glass plates, or the varying number of refracting surfaces, by
which the same effect may be produced at a fixed angle, ca-
pable also of being changed*.

With apolarimeter thus constructed, I have determined that
the maximum polarization of a clear blue sky is equivalent to
a rotation in the plane of a polarized ray of 30^ ; and that this
maximum takes place at a distance of from 88° to 92° from
the sun, and in the plane passing through the sun and the
zenith. This maximum is of course dependent on the state of
the atmosphere, both with respect to its magnitude and posi-
tion; but we shall assume 30° as its amount, and 90° from the
sun as its position in a normal state of the atmosphere, and
when the sun is in the horizon.

VI. On the Form of the Lines of equal Polarization in the
Atmosphere.

It is obvious, from the phsenomena already described, that
the polarization of the atmosphere, produced by the reflexion
of the sun's light from the matter which composes the atmo-
sphere, in planes passing through the sun, the point of re-
flexion, and the eye of the observer, would have been equal in
circles of which the sun and the antisolar point are the centre,
had there been no disturbing causes, or had the atmosphere
been a perfectly transparent medium. In this case the pola-
rization would have been complete, or 45° ; and this maximum
would have occurred at a distance from the sun, the half of
which was the polarizing angle of the medium. There is ob-
viously, however, a cause depending on the zenith distance of
the polarizing point of the sky, which acts in opposition to the
polarization produced by reflexion, and compensates it at the
neutral point already described. When the sun, therefore, is
in the horizon, these two actions are rectangular, as in biaxal
crystals ; and we must therefore determine the form of the
lines of equal polarization when the sun is in the horizon, and
when the atmosphere is perfectly pure. When viewed, con-
sequently, in their general aspect, the phasnomena of atmo-
spherical polarization may be represented by the formula

R=30°(sinDsinD'),;

where R = rotation, or degree of polarization, and

D and D' = the distances of the point whose polarization is

required from the tvoo neutral points.

This formula would make the lines of equal polarization

* See the Transactions of the Royal Irish Academy, vol. xix. part S.

2 G2

452 Sir David Brewster on the Polarization

Lemniscates, as in biaxal crystals, and consequently the po-
larization in the horizon greater than in die zenith, which is
contrary to observation. I have therefore added a correction,
depending on the zenith distance and azimuth, which makes
the formula coincide better with observation, namely,

R = 33|-°(sin D sin D')— 6° 34' (sin Z sin A) ;

Z being the zenith distance, and A the angle of azimuth.

Assuming, therefore, that the distance of the neutral points
from the sun and from the antisolar point is 18° 30', when the
sun is in the horizon, and that the atmosphere is perfectly pure
and uniformly transparent, the lines of equal polarization will
have the forms and the degrees of polarization represented by
the formula. The direction of the polarization follows the
same law as in biaxal crystals, the lines without bands or colour
corresponding with the black hyperbolic branches in the pola-
rized rings produced by these crystals, being distinctly seen
with the polariscope.

VII. On the Construction of the Map of the Lines of equal
Polarization.

Had the map been on a greater or a less scale than it is, it
might have been desirable to appropriate a single curve to
every single degree, or to every two degrees of rotation or
polarization. On the present scale, the curves would have
been too numerous and close had there been one to each
degree; and with only one to each two degrees, they would
have been too distant, in so far as that the form of the curves
round the neutral points would not have been sufficiently seen.
I therefore adopted such a number of curves, viz. 18|, as
enabled me to get the curves. No. 2, continuous round each
neutral point. Hence the formula became

N = 20-5 (sin D sin D')— 3-9 sin Z sin A,

or in the plane passing through the sun and the zenith, in
which Z and A become zero,

N=25-5 (sin D sin D').

In the zenith itself we have N = 18*45, and at P, P' we
have N = 0.

The curves thus obtained do not represent values of N in
degrees of rotation, but in numbers, each of which is equal to
l°-626. Hence R = N 1°'626, and the distance between each
curve is 1°*626. The following table contains the rotations or
degrees of polarization, indicated by each of the curves num-
bered from ^ to 18*45 in the map : —

of the Atmosphere. 453

Corresponding degrees of rotation
or polarization, or values of R.

6-813

1-626

2-439

3-252

4-065

4-878

6-504

8-130

9-756

11-382

13-008

14-634

16-260

17-886

19-Sil

21-137

22-764

24-396

26-016

27-642

29-268

30-000

Hence the maximum polarization of the atmosphere, as
measured by a rotation of 30°, is equal to that produced by
reflexion from a plate of glass at an angle of 65^°, and with a
refractive index of r4826, or to that produced by a surface
of diamond at an angle of 75^°. The number of refractions
at a given angle, or the angle, with a given number of plates
of glass, at which a rotation of 30° is produced, will be found
from the formula in my paper on the Compensations of Po-
larized Light*.

As the sun rises above the horizon, the lines of equal pola-
rization change their form, and the degree of polarization
varies at points of the sky whose distance from the sun is in-
variable. The neutral points above and below the sun approach
•his disc till he reaches the meridian, when the distance of each
from the sun is a minimum ; they then separate again, and
attain their maximum distance, when he reaches the horizon.
In countries where the sun passes across the zenith, these two
neutral points coincide with the sun, when he reaches the
zenith, and again separate.

• Transactions of the Royal Irish Academy, vol. xix. part 2. p. 13.

Values of N.

1
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

18-45

454 Mr. Smith on the Hydrates of Nitric Acid.

As the sun descends beneath the horizon, the neutral point
of M. Arago separates from the antisolar point, and when this
point is first seen in the morning before sunrise, its distance
from the antisolar point is a maximum ; it gradually ap-
proaches that point till the sun rises, and also till the neutral
point itself reaches the horizon, when its distance from the
antisolar point is a minimum.

When the altitude of the sun is 45°, the distance x of the
neutral point above the sun is about 13° 5', and the distance
a?' of the neutral point below the sun 6° 42'; at other altitudes
we have

X = A cos A,
and j/=AcosA,

tan Z,
A being 18|°, A the sun's altitude, and Z the zenith distance
of P', the neutral point below the sun.

An interesting paper, entitled Delle Leggi della Polarizza-
zione della Luce Solare ?iella Atmosphera Seretia, communicato
con lettera al David Brewster, LL.D., F.R.S., Lond. et
Edin., membro delle Principali Academie di Europa, del
Prof. A. B. Francesco Zantedeschi, will be found in the Rac-
colta Fisico-chimica Italiana, torn. i. fascic. 10. 1846. The
details in this paper are chiefly historical. The results ob-
tained by M. Zantedeschi himself, which are of a general
nature, differ in several respects from mine ; but whether this
difference arises from a difference in the methods of observa-
tion, or from the different states of the atmosphere under which
the observations were made, I am not able to determine.

In a Memoir on the Polarization of the Atmosphere, which,
I trust, will soon be published in the Transactions of the
Royal Irish Academy, I shall give a full account of my obser-
vations, and enter more deeply into the subject than would
have been proper in the preceding popular explanation of a
Map of the Lines of Equal Polarization.

LXIX. On the Hydrates of Nitric Acid. By Mr. Arthur
Smith, Assistant in the Laboratory of University College,
London^.

SOME doubt still hanging over the composition of the hy-
drates of nitric acid, especially of the first hydrate, I was
induced to try some experiments with a view of diminishing
this uncertainty. For this purpose a quantity of the red
fuming acid was procured, which I examined before com-
* Communicated by the Chemical Society; having been readjune 7.
1847.

Mr. Smith on the Hydrates of Nitric Acid, 455

mencing my experiments very carefully for chlorine, and found
to be perfectly free from that impurity, and to have a specific
gravity of 1"500.

Fourteen ounces of acid of the above-named strength were
mixed with 7 ounces of commercial oil of vitriol, and distilled
in a sand-bath over a gas flame ; the first 2 ounces that came
over were rejected, and the receiver changed directly the red
fumes of nitrous acid were observed to fill the interior of the
retort. The acid collected was almost as dark in colour as
the acid before distillation. Its specific gravity was 1*522,
and it turned out to be perfectly free from the smallest trace
of sulphuric acid.

I also examined the first two ounces of acid that came over
very carefully for chlorine, and found it to contain scarcely a
trace, nitrate of silver producing only a slight opalescence,
and that which came over afterwards, being the portion that
I selected for my experiments, contained none at all. This last
acid, when diluted with water, gave off nitric oxide gas with
a burst of effervescence, which was the principal reason why
it could not be employed to ascertain the exact amount of
real acid by saturation in its present dark-coloured condition.

The apparatus employed in decolorizing the nitric acid
consisted of a capacious retort, capable of holding about a
pint, to the beak of which Avas attached a large tubulated
receiver, which was kept surrounded with water, to condense
any little acid that might come over during the process ; to
the tubulure of this receiver was adapted a glass tube, bent
at right angles, fitting tightly with a cork, the other extremity
being in connexion with a large gas-holder, which was kept
constantly filled with water, to be used as an aspirator. To
the tubulure of the retort was also fitted a long glass tube
bent at right angles, the one end of which terminated within
an inch of its bottom, whilst the other was in connexion with
a couple of tubes, each 2 feet 11 inches long, arranged side by
side, and connected by means of a tube of a smaller diameter
bent like the letter U.

These long tubes, through which the air was to be aspired,
were filled, the one with dried chloride of calcium, and the
other with pumice-stone moistened with oil of vitriol, and by
these means the absence of all moisture from the air was en-
sured.

In decolorizing the acid a quantity amounting to 6 or 7
ounces was introduced into the retort, and after having ascer-
tained that the whole apparatus M'as perfectly tight, heat was
applied to the bottom of a small sand-bath in which the
retort was immersed, and the temperature kept up carefully
to 1 70° F. Then, by removing the plug at the bottom of the

456 Mr. Smith on the Hydrates of Nitric Acid.

gas-holder, and turning the stop-cock at the top, which was
in connexion with the apparatus, a constant flow of perfectly
dry air was caused to bubble through the nitric acid in the
retort, the level of which was kept 2 or 3 inches above the
orifice of the tube in the interior, the only passage for the air
being through the long desiccating tubes. Aspiration kept
up for two or three hours was found to be generally sufficient
to decolorize completely 6 or 7 ounces of nitric acid.

The acid before decolorization had a specific gravity of
1*522, and after the process fell to 1*503. Fifty grs. of the
colourless nitric acid were accurately weighed out in a stoppered
specific gravity bottle, to which was cautiously added, whilst
in the bottle, with a view to prevent any loss from splashing,
a known weight of perfectly pure carbonate of soda, recently
ignited in a porcelain crucible, until the solution was perfectly
neutral to test-paper. The absence of any sulphate or chlo-
ride in the carbonate had been previously ascertained.

I. Carbonate of soda required 40*23 grs.

II. Carbonate of soda required 40*23 grs.

The quantity of carbonate of soda that 50 grs. of acid re-
quired for saturation, then, was 40*23 grs., which corresponds
to 40*78 grs. of nitric acid, or 81*56 per cent.

An acid containing 1^ equiv. of water would contain in
100 parts —

Real nitric acid .... 80
Water 20

Too

A portion of the prepared acid, amounting to about 5
ounces, was introduced into a small retort, through the tubu-
lure of which was fitted tightly, by means of a stopping of
moist clay, a delicate thermometer, which was kept immersed
in the liquid. The acid began to boil at 190", and before the
distillation had come to an end it had risen to 250°. The
acid coming over between 190° and 200° was collected apart
to be examined by saturation.

50 grs. of the acid which remained in the retort boiling at
250° were then examined, and found to require 31*20 grs. of
carbonate of soda in the first experiment, and 31*07 in the
second, for saturation ; the mean of the two experiments would
correspond to 63*11 per cent, of nitric acid.

50 grs. of the most volatile portion, namely, that which
came over between the temperatures of 190° and 200°, were
then weighed out exactly ; this quantity was found to require
no less than 41*92 grs. in the first experiment, and 41*91 in
the second, corresponding to 84*96 per cent, nitric acid ; but
then it must be remembered that this acid had a very dark
red colour.

Mr. Smith on the Hydrates of Nitric Acid, 457

A quantity of this red acid was introduced into the decolo-
rizing apparatus, and a constant rapid stream of dry air made
to bubble through it for two hours ; at the expiration of that
time it was found to be perfectly limpid, and colourless as
water, and to have a specific gravity of 1*516 at 60°.

50 grs. of the last acid were weighed out and neutralized
with pure carbonate of soda as before. The numbers below
will show the amount required for saturation : —
Exp. Caib. of soda. Mean.

2* * * 41*69 l^^*^^^ corresponding to 42*27, or
«' * * ^1 /.J I 84*54 per cent, nitric acid.
3. . . 41*64 J ^

This acid began to boil at about 184°, the greater part di-
stilling over between the temperatures of 184° and 188°; it
afterwards rose when near the end to 200°.

The first portion that came over was collected apart, intro-
duced into the decolorizing apparatus, and dry air again drawn
through it until it was quite colourless. This was found to
be necessary after each distillation, on account of the decom-
position that it suffered upon boiling, which rendered it as
dark in colour as the original acid. 50 grs. of the colourless
acid, of the specific gravity of 1*517 at 60°, were weighed out,
and carbonate of soda very carefully added until neutral to
test-paper. The increase in the specific gravity this time only
amounted to '001.

Exp. Carb. of soda. Acid. Mean.

1. . . 41-79 = 42*361 .

2. . . 41-81 = 42*38/^"^ '^'*

Hence in 100 parts —

According to theory with 1 eq. water.

Real acid . . 84*74 Real acid . . 85*71
Water . . . 15-26 Water . . . 14*28
100*00 99*99

This would give, when compared with the theoretical compo-
sition of nitric acid with 1 equiv. of water, a deficiency of "97
in the acid, and an excess of '98 in the water.

This hydrate, when pure, was a perfectly limpid and colour-
less liquid, like so much water; it boiled at 184°, and had a
specific gravity of 1*517 at 60°. It was found not to have
the slightest action on tin or iron even when boiled. A por-
tion of this acid placed in a freezing mixture composed of ice
and salt suffered no change.

These experiments leave little doubt concerning the com-
position of the first hydrate of nitric acid, namely, that it is
the true mono-hydrate, consisting of 1 equiv. of nitric acid
and 1 of water, HO, NO5.

Deuto-Hydrate. — In preparing this hydrate, I set out by
obtaining a quantity of colourless strong nitric acid, the exact

458 Mr. Smith on the Hydrates of Nitric Acid,

amount of real acid in which was ascertained by saturation to
be 79*79 per cent. To reduce this acid to the proper strength,
so as to contain exactly 2 equivs. of water, it was found by
calculation that it would require 63*86 grs. of w'ater to every
1000 grs. of acid.

The proper proportions of acid and water were weighed out
carefully in a stoppered specific gravity bottle, and the two
mixed. This mixture was cooled down to 60° and found to
have the sp. gr. 1*486.

50 grs. of this hydrate were weighed out and saturated in
the usual way with recently-ignited carbonate of soda. The
quantities of carbonate of soda required were as follows : —
Exp. Carb. of soda. Mean.

2* ' ' 37*53 y^^*^^' °^ ^^'^^ P^^ *^^°^* ^^^^•

An acid containing 2 equivs. of water will contain 75 per
cent, real acid.

A portion of this acid was introduced into a small retort
and distilled. It began to boil, as nearly as could be judged,
at about 200°, it being difficult to come at the exact tempera-
ture on account of the very rapid rise of the thermometer,
which continued to take place until it had gained the tempe-
rature of 218°; it afterwards rose when near the end to 250°.

It appears, then, from these experiments, that no such thing
as a deuto-hydrate exists, but that when a mixture is made
in the proportions to form such a hydrate and subjected to di-
stillation, it divides spontaneously into the first and another,
at the same time suffering considerable decomposition ; and
the acid which is found remaining in the retort has the exact
boiling-point of the tetra-hydrate, namely, 250° ; and more-
over, the first portion that came over had the exact density ,
of the first.

A portion of this acid placed in a freezing mixture of ice
and salt, suffered not the least solidification.

Tetra-Hydrate. — This hydrate was prepared in the same
way as the first, namely, by preparing a quantity of colourless
acid, ascertaining its saturating power, and mixing it with the
proper quantity of distilled water, ascertained by calculation.
It was then tried afterwards by saturation to see if it was
correct ; the numbers below will show the difference : —
Exp. Cavb. ofsoda. Acid. Mean.

\'. \ '. 29*87 = 30*27 }^^'^2' ^"^ ^^'^"^ P^"" ''^''^- ""^^^ ^'''^-

According to theory with 4 eqs. water.
Real acid . . 60*64 Real acid ... 60

Water , , . 39*36 Water .... 40

100*00 100

On the Decomposition ofCuminate of Ammonia hy Heat. 459

The acid had a density of 1*424 at 60°; it began to boil at
250°, and distilled over perfectly colourless and unchanged ;
towards the end, when slight decomposition commenced, the
temperature rose to 260°.

Five or six ounces of very weak acid, of the density of 1*180,
were introduced into a retort and kept heated just below its
boiling-point for two or three hours ; the heat was increased
from time to time so as to make it boil briskly, and a ther-
mometer introduced through the tubulure ; when that which
remained in the retort boiled uniformly at 250°, the heat was
withdrawn and it was allowed to cool.

When the specific gravity of this acid was taken, it was
found to be close upon that of the tetra-hydrate, but not
exactly ; probably if I had operated upon a large quantity,
and carried it on for a longer time, it would have been more
so; as found, its density was 1*412 instead of 1*424, which
would make a difference of rather less than 1^ per cent, de-
ficiency in the acid.

This is, I have no doubt, the proper hydrate of nitric acid,
HO, NO5 + 3HO, as it is generally considered; and as Dr.
Dalton correctly observed, acids which are either stronger or
weaker than this acid, are brought to this strength by conti-
nued ebullition, the former losing acid and the latter water.

LXX. On the Products of the Decomposition of Cuminate of
Ammonia by Heat. By Mr. Frederick Field*.

THE peculiar mode of decomposition which the ammonia
salts of inorganic acids exhibit when exposed to the ac-
tion of heat, occurs likewise in the ammonia compounds of
organic acids, although the results in the latter instances are
usually of a more complicated nature. In most of these cases
a formation of water takes place, the hydrogen of which is
derived from the volatile alkali, while the acid furnishes the
oxygen, the residue of which combines in a more intimate
manner with the nitrogen of the ammonia. In decompositions,
however, of inorganic compounds this reduction seems to be
carried at once as far as it can go, the Avhole of the hydrogen
contained in the ammonia being converted into water ; while
in organic salts this hydrogen is eliminated only by degrees,
an intermediate body being produced between the original
ammonia salt and the final product of the decomposition.
Thus we find that nitrite and nitrate of ammonia, when ex-
posed to heat, are at once converted into water, and respect-
ively into nitrogen and nitrous oxide. Oxalate of ammonia,

* Coramunicated by the Chemical Society; having been read June 7.

1847.

460 My. Field on the Products of the

on the other hand, if submitted to a gentle heat, loses only
two equivalents of water, the residue of both base and acid
combining to form oxamide, and only by a strong and brisk
application of heat Doebereiner converted it into cyanogen,
the rest of the hydrogen being eliminated in the form of water.

The dry distillation of oxalate of ammonia thus affords the
prototypes of two series of compounds, which may arise from
ammoniacal salts by the elimination of two or four equivalents
of water respectively. There are few cases, however, in which
the decomposition of ammoniacal salts have been carefully
studied, and the instances in which we are acquainted with
the representative of the two types are exceedingly scarce.
We are indeed intimate with a very great number of amidogen
compounds analogous to oxamide (fumaramide, salicylamide,
succinamide, anisylamide, &c.), but only few of these have
been obtained from ammoniacal salts by the action of heat.
The greatest number of these bodies were produced by the
change most compound aethers suffer under the influence of
ammonia, a beautiful mode of decomposition pointed out
first by Professor Liebig in the transformation of oxalate of
ethyl into oxamide, or by the action of gaseous ammonia on
other substances related in some manner with the acid : thus
was chloride of benzoyle converted into benzamide by Wohler
and Liebig, and lately lactide into lactimide by Pelouze.

As yet, however, the members of the second class, those
compounds standing to other acids in the same relation as
cyanogen to oxalic acid, are very rare. From a beautiful ex-
periment of Pelouze, we know that the vapour of formiate of
ammonia, when passed through a red-hot tube, is converted
into water and hydrocyanic acid. In their investigation on
the radical of benzoic acid, Wohler and Liebig obtained a
peculiar oil by the action of heat on benzamide, which at that
time they did not study more closely. The same body was
at a later period obtained in the dry distillation of benzoate
of ammonia, and fully examined by Fehling, who found that
this interesting substance, to which he gave the name henzo-
nitrile, has the composition C14 H5 N, and is produced from
benzoate of oxide of ammonium, exactly in the same manner
as cyanogen and prussic acid are formed respectively from
oxalate or formiate of ammonia. These facts did not long
remain isolated. Schlieper, in an excellent examination he
has lately published on the products of oxidation of gelatine
by chromic acid, discovered that in these reactions, among
other products, the body C,o Hg N is formed, valeronitryle or
valerianate of ammonia — 4 equivs. of water.

The members of this class acquire every day a greater
degree of importance. A remarkable paper, read before the

Decomposition of Cuminate of Ammonia by Heat. 461

Chemical Society a short time since by Dr. Kolbe and Mr.
Frankland, has indeed opened a most interesting connexion
between these bodies and another class of substances, which
hitherto have been obtained by very different processes. The
conversion of cyanide of ethyl into metacetonic acid by means
of alkalies and acids, seems to indicate that cyanide of ethyl
is nothing else than metacetonitryle. This experiment is
likely to be of great importance, for it is exceedingly proba-
ble that the whole class of substances alluded to must be con-
sidered as a class of cyanogen compounds. It is evident that
similar considerations may be applied to cyanide of methyl
and cyanide of amyl, lately described by Balard ; and the con-
version of these cyanides respectively into acetic and caproic
acids, which w^e have a right to anticipate on treating them
W'ith alkalies or acids, will prove that these compounds are
the nitriles of acetic and caproic acids — acetonitryle and
capronitryle — which as yet have not been obtained by the
action of heat on the ammoniacal salts of these acids.

The following experiments on the action of heat on cumi-
nate of ammonia have been made with the hope of contri-
buting to the history of the nitryles, or organic cyanides, as
they perhaps should be more correctly designated.

The cuminic acid employed in my experiments was prepared
by the action of solid hydrate of potash on oil of cumin, and
the product perfectly freed from the least traces of cymol
which it might possibly contain by precipitating the potash
salt by hydrochloric acid, dissolving the precipitated cuminic
acid in ammonia, reprecipitating by hydrochloric acid, and
crystallizing from water. The acid was then dissolved in
strong ammonia, and the solution subjected to heat. The
first portions which passed over, although consisting chiefly
of water and ammonia, together with cuminate of ammonia,
which is always carried over with the steam, presented more
or less an opalescent appearance, indicative of traces of the
oil. On evaporating the solution in the retort to dryness, a
portion of the salt is decomposed, ammonia is evolved, and
cuminic acid condenses in beautiful plates upon the sides and
neck of the retort, separation going on even on raising the
temperature ; but simultaneously another decomposition takes
place, water is eliminated, in consequence of Avhich there are
produced a peculiar white crystalline body, difficultly soluble
in water, and subsequently a colourless oil of a most fragrant
odour ; although the operation may seem very simple, expe-
rience alone teaches the proper regulation of temperature ne-
cessary to obtain these two bodies.

Cuminamide. — Observing in ray first experiments evolution
of ammonia and sublimation of cuminic acid on heating cu-

462 Mr. Field on the Products of the

minate of ammonia, I thought that by heating it under pres-
sure, the ammonia then not being able to escape, the desired
change might be effected. Accordingly a portion of the salt
was placed in a strong glass tube, and after sealing the other
end, gradually heated in an oil-bath to nearly the boiling-
point of the oil, and allowed to cool. On cooling the mass
appeared to have been completely fused, but perfectly solid
and of a highly-crystallized texture. On examination it was
found to be insoluble in cold water and ammonia, but very
soluble in hot water, from which it solidified into a crystalline
mass as the temperature cooled ; this alone sufficiently indi-
cated that a complete change had been effected, the cuminate
of ammonia being readily soluble in cold water. In order to
ascertain the nature of the change it was dissolved in hot
water, and weak ammonia added to dissolve any cuminic acid
that might be mixed with it, and crystallized; the crystals
were separated by filtration, and once more dissolved in a hot
weak solution of ammonia, from which they separated on
cooling in brilliant white crystalline plates, similar in appear-
ance to benzamide. These were dried at 212° in a water-bath,
and analysed in the usual manner.

I. 0-174 grm. of substance burnt with oxide of copper
yielded 0*470 of carbonic acid and 0*128 of water.

II. 0*248 grm. yielded 0*670 of carbonic acid and 0*181 of
water.

III. To estimate the nitrogen, 0*287 grm. ignited with soda-
lime yielded 0*390 of ammonio-chloride of platinum*.

From these analytic results the following per-centages are
obtained : —

I. 11. III.

Carbon . . 73*66 73*67

Hydrogen'. 8*17 8*10

Nitrogen . 8*50

leading to the formula Cc^ Hjy NOg, as may be seen from the
following comparison of the theoretical and experimental
numbers : —

20 equivs. of Carbon ,
13 ... Hydrogen

1 ... Nitrogen

2 ... Oxygen .

This body therefore is cuminamide, NH2 Cgo Hn Og, having

* In this operation a large quantity of an oily body is produced, which
floats on the surface of the hydrochloric acid. It is evidently cumol.

Theory.

Mean of exp.

120

73*68

73*66

13

7*99

8*13

14

8*52

8*50

16

9*81

9*71

163

100*00

100*00

Decomposition of Cuminaie of Ammonia by Heat. 463

precisely the same relation to cuminate of ammonia as oxa-
mide to oxalate of ammonia.

In preparing large quantities of this substance the employ-
ment of close tubes would be very inconvenient, and I soon
found that it could be obtained in a retort by the continued
application of a heat sufficient to keep the salt in a state of
semi-fluidity. The analyses II. and III. vi^ere made with the
product obtained in this manner.

Cuminamide crystallizes like benzamide, in two forms, ac-
cording to the state of the solution ; if crystalUzed imme-
diately, or from a strong solution, it separates in the form of
crystalline tables of great brilliancy, but if the solution be di-
lute, it crystallizes after the lapse of some hours in long
opake needles, both forms having exactly the same compo-
sition. It is soluble in hot and cold alcohol in any propor-
tion, as also in aether. This new amide differs from most
others that have been described in remaining intact on the
addition of strong solution of potash, or mineral acids ; from
the former it crystallizes in large plates after some days.
Long boiling with alkalies or acids is scarcely sufficient to
produce the characteristic conversion of amides either into
ammoniacal salts or combinations of the base with the acid
and evolution of ammonia.

Cumonitrile. — On heating cuminate of ammonia until it is
perfectly fused, and keeping the fused mass in a state of brisk
ebullition, large globules of a light yellowish oil pass over
with water, evidently derived from the decomposition of the
salt ; when the globules began to diminish the process was
stopped, the oil was separated from the water in the receiver
by means of a pipette, the remaining distillate added to the
mass in the retort, and the process again repeated as before ;
in this manner, after some half-dozen distillations, nearly an
ounce of oil was obtained ; it was well- washed with ammonia
to remove cuminic acid, which seemed to be soluble in the
oil, then treated with hydrochloric acid to remove ammonia,
thoroughly washed with water, and digested with chloride of
calcium; after standing some days to separate chloride of
calcitim, it was distilled and carefully rectified, the first por-
tions being rejected, as possibly containing traces of water ;
the middle portion was reserved and placed in a retort with a
coil of platinum ; the liquid entered into ebullition at 239° C,
at which point it remained stationary while at least a quarter
of an ounce was passing over. This portion was employed
in the following analyses : —

I. 0*212 grm. burnt with oxide of copper yielded 0*644 of
carbonic acid and 0*145 of water.

464 Mr. Field on the Products of the

II. 0-225 grm. yielded 0*6835 of carbonic acid and 0-161
of water.

III. 0-244 grm. ignited with soda-lime yielded 0-364 of
ammonio-chloride of platinum*.

From these analytical results the following per-centages
are obtained : —

I. ir. III.

Carbon . . 82-82 82-84

Hydrogen . 7*59 7-96

Nitrogen . 9*34

leading to the formula Cgo Hji N, as may be seen from the
following comparison of the theoretical and experimental
numbers : —

Tlieory. Mean of exp.

20 Carbon . . 120 82-76 82-83

11 Hydrogen . 11 7-58 7-77 *

1 Nitrogen . 14^ 9-66 9-34

145
This body is therefore cumonitrile, Cgo H^l N, standing in the
same relation to cuminate of ammonia as cyanogen does to
oxalate of ammonia.

Cumonitrile is a perfectly clear and colourless liquid, pos-
sessing a high refractive power; it has a most powerful and
agreeable odour and a burning taste ; it is somewhat soluble
in water, causing turbidity in that liquid ; it is soluble in all
proportions of alcohol and aether ; it is lighter than water,
having a specific gravity 0*765 at 14° C. (57°Fahr.). The
boiling-point, when in contact with metal, is constant at
239° C. (462-2° Fahr.), at the barometric pressure 0-7585 m.
= (29-85 inches). The equivalent ofcuminic acid containing
3C2 Hg more than the equivalent of benzoic acid, it was in-
teresting to compare the boiling-points of benzonitrile and
cumonitrile. According to Fehling's experiments, the boil-
ing-point of benzonitrile is 191° C; on calculating from this
observation the boiling-point of cumonitrile according to the
rules first pointed out by Kopp, the boiling-point should be
191 -f- 3-1 9 = 248.

Dr. Fehling does not hoNvever mention that he had this
substance in contact with metal, and it is not improbable that
the true boiling-point of benzonitrile is somewhat lower ; the
vapour of cumonitrile is very inflammable and burns with a
bright flame, which deposits much carbon.

* Professor Fehling found it difficult to estimate the nitrogen in hetj-
zonitrile in the form of ammonia, drops of oil passing over into the hydro-
chloric acid. In the case of cumonitrile, this method gave very accurate
results ; oil drops also passed over, but they were evidently cumol.

Decomposition of Cuminaie of Ammonia by Heat. 465

The strongest nitric acid has but little action upon this
substance ; after boiling, however, and setting aside for some
days, crystals of cuminic acid are formed. On being heated
with potassium it darkened, and apparently another oil was
produced ; the mass on being washed and tested for cyano-
gen in the usual manner gave a copious precipitate of prus-
sian blue, which seems to be strongly in favour of the view
which Kolbe and Frankland have recently promulgated. A
strong alcoholic solution of potash has no immediate action
on cumonitrile, but after a day or two, on pouring the li-
quid into a watch-glass, it partially solidified into a yellow
crystalline mass, a mixture of the original substance with
white crystals. These crystals after purification had all the
appearance of cuminamide, and in order to be satisfied of
their composition —

I, 0*174 grm. burnt with oxide of copper yielded 0*472 of
carbonic acid and 0*124 of water; the calculated per-centage
of carbon and hydrogen from these numbers being —
Carbon . . 73*62
Hydrogen . 7*91
These numbers correspond to those of cuminamide, as may be
seen by a comparison with the former analyses.

It appears then that cumonitrile, on the addition of potash,
is not, as might have been expected, converted into cuminate
of ammonia, but into cuminamide, taking 2 instead of 4 atoms
of water— -C20 H,i N + 2HO = Cgo HigNO^, the latter body
being, as before remarked, in such a remarkable degree un-
affected by alkalies or acids.

Having obtained one amide with comparative ease, many
other ammoniacal salts were heated for the purpose of obtain-
ing analogous amidogen compounds. Benzoate of ammonia
was tried unsuccessfully, and it appears from the account
published by Fehling of his investigation of benzonitrile, that
the residue in the retort consisted entirely of benzoate of am-
monia, that salt appearing to have lost directly 4 equivs. of
water without undergoing an intermediate conversion into an
amide by the loss of 2 equivs. Nitrobenzoic acid was dis-
solved in ammonia, evaporated, and cautiously fused for a
considerable time ; when cold it was found to be insoluble in
water and ammonia at the ordinary temperature, but dissolved
by hot water, from which it crystallized in beautiful yellow
needles. On analysis, the following results were obtained : —

I. 0*222 grm. of substance burnt with oxide of copper
yielded 0*410 of carbonic acid and 0*080 of water.

II. 0*255 grm. yielded 0*472 of carbonic acid and 0*087 of
water.

P/iil. Mag. S. 3. Vol. 31 . No. 210. Dec. 1847. 2 H

466 On the Decomposition of Cuminate of Amnwnia by Heat.

From these results the following per-centages were ob-
tained : —

I. II.

Carbon . . . 50*36 50-43

Hydrogen . . 4*00 3-78

corresponding to the formula 0^4 Hg Ng Og, as may be seen
from the comparison of the theoretical and experimental
numbers : —

14 equivs. of Carbon . .

84

Theory.
50-60

Mean of exp,
50-39

6 ... Hydrogen .
2 ... Nitrogen .

6

28 .

3-62
16-87

3-98

6 ... Oxygen . .

48

28-91

166 100-00

This body is therefore nitrobenzamidej having the same rela-
tion to nitrobenzoate of ammonia as cuminamide has to cumi-
nate of ammonia.

This beautiful substance can only be obtained with diffi-
culty, as the nitrobenzoate of ammonia explodes violently
unless very great caution is employed.

A specimen of chlorobenzoic acid, made In the laboratory
for some other investigation, was dissolved in ammonia and
heated ; it fused readily, became perfectly insoluble in cold
water and ammonia, but soluble in hot water, crystallizing as
the solution cooled in long needles of great beauty. The
specimen of acid afforded me, being all that could be spared,
was insufficient for the manufacture of an amide ; I prepared
a portion of chlorobenzoic acid by acting upon benzoic acid
for some days with hydrochloric acid and chlorate of potash ;
after purification it was burnt with chromate of lead and gave
the following results : —

I. 0-394 grm. = 0-769 of carbonic acid and 0-114 of water.

From this result the following per-centage was obtained : —

periment.

Theory.

53-22

53-61

3-22

3-25

Carbon .
Hydrogen

leading to the formula HO, C,4 4 ^j* V Og, 1 equiv. of the

hydrogen of benzoic acid replaced by an equivalent of chlo-
rine.

This acid, however, on being subjected to the usual treat-
ment by solution in ammonia and subsequent heat, did not
fuse but blackened, charcoal being separated. Unfortunately
the specimen of ammoniacal salt from which I had made the
former compound was not analysed, probably it would have

On the equation x^ + j^ + As^zs'Man/z, Src, 467

proved to be C,4 {q T O3J HO, or C14 ^Qf^ O3, HO, a

dichlorobenzoic or a trichlorobenzoic acid, such existing.

These experiments were conducted in the laboratories of
the Royal College of Chemistry under the direction of Dr.
Hofmann, to whom I beg to offer my best thanks for his ad-
vice and assistance during their progress.

LXXI. On the General Solution {in certaiti cases) of the
equation a:^ + i^-{-Az^=Mxyz, i^-c. By J, J. Sylvester,
A.M., F.R.S., late Professor of Natural Philosophy in Uni-
versity College, London*.

T SHALL restrict the enunciation of the proposition I am
-*- about to advance to much narrower limits than I believe
are necessary to the truth, with a view to avoid making any
statement which I may hereafter have occasion to modify.
Let us then suppose in the equation

a^ -\-y^ -f- Az^ = Msyz
that A is a. prime number, and that 27 A — M^ is positive, but
exempt from positive prime factors of the form 6/+ 1. Then
I say, and have succeeded in demonstrating, that all the pos-
sible solutions in integer numbers of" the given equation may
be obtained by explicit processes from one particular solution
or system of values of x, y, z, which may be called the Primi-
tive system.

This system of roots or of values of x, y, z is that system in
which the value of the greatest of the three terms x, y, A^.z
(which may be called the Dominant) is the least possible of all
such dominants. I believe that in general the system of the
least Dominant is identical with the system of the least Content,
meaning by the latter term the product of the three terms out
of which the Dominatit is elected. I proceed to show the law
of derivation.

To express this simply, I must premise that I shall have to
employ such an expression as S'=f (S) to indicate, not that
a certain quantify, S', is a function of S, but that a certain
system of quantities disconnected from one another, denoted
by S', are severally functions of a certain other system of
quantities denoted by S; and, as usual, I shall denote <p<pS by
f^S, ffS by <p^S, and so forth.

Let now P be the Primitive system of solution of the equa-
tion

^ +y + A^^ = Mxyz,

* Communicated by the Author.
2 H2

468 Mr. J. J. Sylvester on the General Solution of

P denoting a certain system of values of and written in the
order of the letters a?, y, 2;, which may always be found by a
limited number of trials (provided that the equation admits
of any solution). That this is the case is obvious, since we
have only to give the Dominant every possible value from the
integer next greatest to A* upwards, and combine the values
of x^,i/^, Az^ so that none shall ever exceed at each step the
cube of such dominant, and we must at last, if there e^/5/ ani/
solution^ arrive at the System of the Least Dominant.

Now, every system of solution is of one or the other of two
characters. Either x and y must be odd and z even, or x
and 3/ must be one odd and the other even and z odd. That
all three should be odd is inconsistent with the given condi-
tions as to A being odd and M even ; and if all three were
even, by driving out the common factor we should revert to
one or the other of the foregoing cases.

The systems of solution where z is even may be termed Re-
ducible, those where z is odd Irreducible. Let <p denote a
certain symbol of transformation hereafter to be explained.

Then the Reducible systems of the first order may be ex-
pressed by

<pP, f^P, <p^P, ad injinitum ;

or in general by <p"^ P n^ being absolutely arbitrar}'. I will
anticipate by stating that the function <^ involves no variable
constants ; that is to say, <p (S) may be found explicitly from
S without any reference to the particular equation to which
S belongs. Let now \{/ denote another symbol of transforma-
tion, also hereafter to be defined, and differing from (p insofar
as it does involve as constants the three values of x, y, 2 con-
tained in P : then the general representations of Irreducible
systems of the first order will be denoted by 4/ (p"'. P.

It is proper to state here that the symbol -^ is ambiguous ;
and \|/ f^iP, when P and n^ are given, will have two values,
according to the way in which the terms represented by P
are compared with x,y,z in the given equation

a^ +y + Az^ = Mxyz ;
for it is obvious that if <r=flr, y = ^, zz=c satisfies the equation,
so likewise will

xz=.b, y=a, z =c.

Each however of these values of -^ <f^'P gives a solution of
the kind above designated.

Proceeding in like manner as before, the Reducible system
of the second order may be designated by 4>"2 . ■], <p«i . P, the
Irreducible by 4/ f "2 . vf; ^«i . P ; and in general every ^possible

the equation x^ -^y^ -\- kz^ — Mxyz, 8^c. 469

system of values of x,y, z satisfying the proposed equation,
in which z is even, is comprised under the form

and every possible system of such values, in which z is odd,
is comprised under the form

^^f\.4,<p\-i.yl, . . . vI/(p~i.P:

the quantities n^ Wg • • • ^h being of course all independent
of one another, and unlimited in number and value.

Thus then we may be said to have the general solution of
the given equation in the same sense as an arbitrary sum of
terms, each of a certain form, is in certain cases accepted as
the complete solution of a partial differential equation.

As regards the value of the symbols ^ and (p, <p indicates the
process by which a, b, c becomes transformed into «, /3, 7, the
relations between the two sets of elements being contained in
the following equations :

a' = fl3 V = l^ c' = Ac3

/3 = a' b''' + b' c'^ + c'a'^-SaW
y=abc{a'^ + b'^ + c'^-a'b'-a'c' - b'c'}.

Next, as to the effect of the Duplex symbol rj/. Let eg i
be the elements of the Primitive system P: » being the value
of z and e, g of x and j/ taken in either mode of combination,
each with each, which satisfy the proposed equation

^ + j/^ + Az^ = Mxyz,

Let /, ntj n represent any system S,
A, ft,, V represent any system 4/(S),
\I/S has two values, which we may denote by vJ/'S, '\{/S re-
spectively, and accentuating the elements X, fx,, v accordingly
to correspond, we shall have

>J=3ffm{gl—em) + S\m{U—en) — M{gd^ - eHm)
ljJ=3Am{i7n —gl) -r Sel{ein —gl) — M{eim^—gHm)
v' = 3el{en — d) + 3g7n(gn — nn) — M {egn^— iHm) ;

we have then

and in like manner

'\I/ S being derived from '^ S by the mere interchange of e
and g one with the other.

I have stated that every possible solution of the proposed
equation comes under one or the other of the orders, infinite

470 On the equation j:^+^' + A«^= M^n/^, S^c,

in number and infinite to the power of infinity in variety of de-
gree, above given: this is not strictly true, unless we understand
that all systems of solution are considered to beequivalent which
differ only in a multiplier common to all three terms of each;
that is to say, which may be rendered identical by the expul-
sion of a common factor. So that 7nu, m^, my as a system is
treated as identical with a, /3, 7, which of course substantially
it is ; and it should be remarked that there is nothing to pre-
vent the operations denoted by <p and vf/ introducing a com-
mon factor into the systems which they serve to generate, and
the latter in particular will have a strong tendency so to do.

I believe that this theorem may be extended with scarcely
anymodification to the case where A, instead of being a prime,
is any power of the same, and to suppositions still more gene-
ral. I believe also that, subject to certain very limited restric-
tions, the theorem may prove to apply to the case where the
determinant 27 A — M^ becomes negative.

The peculiarity of this case which distinguishes it from the
former, is that it admits of all the three variables x, 3/, z in the
equation

x^ + j/^ + A^ = Ma^z

having the same sign, which is impossible when the determi-
nant is positive; or in other words, the curve of the third

M

degree represented by the equation Y^+X^+l=-rT XY

(in which I call the coefficient of XY the characteristic),
which, as long as the quantity last named is less than 3, is
a single continuous curve extending on both sides to infinity,
as soon as the characteristic becomes equal to 3 assumes to
itself an isolated point, the germ of an oval or closed branch,
which continues to swell out (always lying apart from the
infinite branch) as the characteristic continues indefinitely to
increase.

I ought not to omit to call attention to the fact that the
theorem above detailed is always applicable to the case of the
equation

^+y+A23=o,

when A is awj/ power of a prime number 7iot of the form
6i+l; in other words, the above always belongs to the class
of equations having Monogenous solutions, which for the sake
of brevity may be termed themselves Monogenous Equations *.

* Thus the equation j:'^+j/^4-A2^=0 alluded to by Legende is Monoge-
nous, and the Primitive system of solution is x=l i/=2 z= — 1, from which
every other possible solution in Integers may be deduced.

Mr. De la Rue on Cochineal* 471

On the probable existence of such a class of equations I
hazarded a conjecture at the conclusion of my last communi-
cation to this Magazine. As 1 hope shortly to bring out a
paper on this subject in a more complete form, I shall con*-
tent myself at this time with merely stating a theorem of much
importance to the completion of the theory of insoluble and
of Monogenous equations of the third degree; to wit, that the
equation in integers

a{a,^ + j/3 _j. 2:3^ ^ ^ ^^.2^ ^^2^ _^ ^2^ ^ ^^2 4.^^2 ^. ^^2^ _(_ ^^^ __ q

may always be transformed so as to depend upon the equation

Ji^ +gv^ + hisfi = {Qa — e)uvw,

wherein fgh —ae^— {f + 3a2)^+ 9a^— Sac^— gc^.

By means of the above theorem, among other and more
remarkable consequences, we are enabled to give a theory of
the irresoluble and monogenous cases of the equation

when m is some powet of 2, or of certain other numbers.

26 Lincoln's Inn Fields, J J S

Nov. 17, 1847.

Erratum. — In the October Number, at page 295, a little below the
middle, for _y= 11 7949000 read the same with the number 1 added at
the end. At page 293, last line, omit the words after ** provided " as far
as " divisible by 9 " on the following page, and read in lieu, " provided in
the second case that ABC is of the form 9»z+l and that D is divisible by 9."

LXXII. On Cochineal (Coccus Cacti). First Memoir,
By Warren De la Rue, Esq.'^

T^HE beautiful theoretical results which have been lately
* obtained by a closer examination of indigo blue and its
products of decomposition, made it desirable to undertake
similar investigations with other colouring matters. I made
choice of the colouring principle of cochineal {Coccus Cacti),
hoping that a detailed research might not only prove of in-
terest in a scientific point of view, but also throw some light
on its practical applications, and the more so, as the recent
investigations of Preisser had seemed to point out a very close

* Communicated by the Chemical Society ; having been read June 21,
1847.

472 Mr. De la Rue on Cochineal

analogy in the chemical properties of a variety of colouring
matters with indigo blue.

Before entering into the detail of my experiments, I think
it desirable to give a brief outline of the results obtained by
the chemists who have hitherto worked on this subject.

Dr. Jean Frederic John, in a quarto volume translated
from the German and entitled Tableaux Chimiques du liegne
Animal, appears to have published the first analysis of cochi-
neal : he does not describe his method, but merely states that
it contains the following per-centage : —

Colouring principle (semi- solid, soluble in"\ cn.nn

water and alcohol) j

Gelatine 10-50

Waxy fat 10*00

Modified mucus 14-00

Membrane 14-00

Alkaline phosphates and chlorides, phosO

phate of lime, phosphate of iron, and > 1'50

phosphate of ammonia J

100-00

Pelletier* and Caventou, in a very long memoir read before
the Institut de France in 181S, have gone very elaborately
into the examination of cochineal and obtained many interest-
ing results. In analysing this substance they employed the
following process : — They removed the fatty bodies by boiling
aether, in which they found the colouring matter but slightly
soluble ; these fatty substances, recovered by distilling off the
aether, were considered to consist of stearine, oleine, and an
aromatic acid, from which latter substance it was difficult to
remove the adhering colouring matter.

The cochineal, exhausted M'ith aether, M'as treated with al-
cohol of 40° Beaume, which dissolved the colouring matter,
together with a small quantity of fatty and nitrogenous sub-
stances.

On cooling, and by spontaneous evaporation, they obtained
a granular red residue of a semi-crystalline appearance, and
which they considered to be the colouring matter contami-
nated still with nitrogenous matter {maiiere animalisee) and
some fatty bodies, the greater part of which remained undis-
solved in strong cold alcohol; by repeating the operation
once or twice they considered that the substance was ob-
tained almost in a state of purity. To remove the last traces

* /nnales de Chiniie ef. de Physique, ser. 2, tome viii. p. 250. Journal
de Pharmacie, ser. 2, tome iv. p. 193.

Mr. De la Rue on Cochineal. 473

of foreign matter it was dissolved in strong alcohol, and
an equal volume of aether added, which precipitated the
colouring matter and retained the fat, which was still ad-
hering to it. The colouring matter thus purified they named
carmine {carminium), and described as being very soluble in
water, from which it did not crystallize, more or less soluble
in alcohol, according to its strength, and quite insoluble in
aether and the fixed and volatile oils. Acids did not precipi-
tate it from its aqueous solution if free from animal matter.
They found hydrochloric and sulphuric acid to decompose
it; the latter with elimination of carbon. By the action of
nitric acid they obtained an acid in prismatic crystals resem-
bling oxalic acid, but differing in some of its properties.

On heating the " carmine" it intumesced and gave off carbu-
retted hydrogen, a considerable quantity of oily substances,
a little acid water, but no trace of ammonia. Chlorine and
iodine decomposed it; the alkalies in the commencement
produced merely a change in colour, turning it violet, but by
the assistance of time or heat they effected a complete altera-
tion. They found an aqueous solution of " carmine " to exhibit
the following comportment with reagents.

Of the alkaline earths, lime only produced a precipitate ;
hydrate of alumina showed a marked affinity, absorbing the
whole of the colouring matter from an aqueous as well as an
alcoholic solution ; the presence of alum prevented this reac-
tion : iron, copper, and silver salts were without reactions ;
terchloride of gold destroyed the colour ; neutral salts of lead
merely changed it to violet, except the neutral acetate, which
precipitated it, the free acetic acid retaining a little of the
compound in solution; the colouring matter could be re-
covered by decomposing the lead compound with hydrosul-
phuric acid. The nitrate of mercury gave a purple, and the
pernitrate a scarlet-red precipitate; the bichloride no pre-
cipitate ; chloride of tin gave a violet precipitate ; the bichlo-
ride changed the colour to scarlet without causing a precipi-
tate. Albumen and gelatine had no marked action, but if
precipitated by reagents the colouring matter was carried
down.

In a later communication (1832), Pelletier* gave the com-
position of the colouring matter as prepared by himself and
Caventou. In a previous qualitative examination they had
failed to exhibit the presence of nitrogen which M. Pelletier
now detected. The substance was dried in vacuo at a gentle
heat to remove every trace of alcohol and aether, and burnt
with oxide of copper it yielded —

* Annales de Chimie et de Physique, sir. 2, tome li. p. 194.

474 Mr. De la Rue on Cochineal.

Carbon 49'33

Hydrogen 6'66

Nitrogen 3*56

Oxygen 40*45

100-00

M. Pelletier stated, however, that he did not greatly rely
on the correctness of this analysis.

After alcohol had dissolved out all the colouring matter
removable by it, they extracted the last traces, by repeatedly
M'ashing the residue with boiling water, and along with it a
little fatty and some nitrogenous matter ; the residue was a
brownish transparent mass. The later decoctions, contain-
ing no red colouring matter, left likewise on evaporation a
brownish transparent mass, which they considered identical
with the organic residue of the insect. This animal matter
had, according to them, some analogy with gelatine, but dif-
fered in many of its properties, as it did also from albumen
and fibrine, they therefore considered it as peculiar to the
cochineal insect; the alkalies and ammonia dissolved it
readily ; chlorine precipitated it ; all acids and acid salts pre-
cipitated it, as also acetate of lead, salts of tin and copper, and
nitrate of silver ; and they considered the latter reagent as a
good test of the purity of the colouring matter, as it did not
precipitate the latter if free from nitrogenous substances. If
the colouring matter were contaminated with nitrogenous sub-
stances, all the salts which precipitated the latter carried down
likcAvise some of the colouring matter.

An examination of the ashes showed the presence of phos-
phate of lime, carbonate of lime, chloride of potassium, and
phosphate of potash, to the extent of 0'7 per cent.

In the second part of the memoir they went into the theory
of the technical applications of the colouring matter : this
having no reference to the present subject, I do not think it
necessary to reproduce it here.

M. Lassaigne, in 1819*, examined Kermes {Coccus ilicis),
an insect common in the South of Europe, and employed as
a red dye before the discovery of America, and obtained by
following the methods of Pelletier and Caventou, substances
agreeing in their properties with the analogous ones found in
cochineal.

M. F. Preissert, in an elaborate paper on the origin and
nature of colouring matters, has again drawn the attention of
chemists to the subject. This gentleman, from a study of a
variety of colouring substances, comes to the conclusion that
all resemble indigo in its behaviour with reducing agents.

* Journal de Pharmacie, ser, 2, tome v. p. 435. f Ibid, p. 191.

Mr. De la Rue on Cochineal. 475

He affirms that he obtained by the action of hydrosulphuric
acid on the lead compounds of a great number of organic
colouring matters, substances bearing the same relation to
the original colouring matters as white indigo does to blue
indigo. In order to obtain the colourless modification of the
colouring matter of cochineal, he adds what he terms "hydrate
of oxide of lead " to an aqueous decoction of cochineal, the
fats being previously removed by aether. The colouring mat-
ter is entirely removed by the so-called oxide of lead, which
is nothing but a basic nitrate of lead, 2(3PbO, NO5) +3HO.
The lead compound suspended in water (hot?) was decom-
posed by a stream of hydrosulphuric acid ; the nearly colour-
less filtrate deposited on cooling needles of a pale yellow
colour, which became perfectly white by washing with aether
and pressing between bibulous paper ; these crystals, which,
according to his statement, are soluble in water and alcohol,
but less so in aether, assume in contact with the atmosphere
the purple-red of the colouring matter of cochineal. He more-
over asserts that his colourless modification gives a white lead
salt on adding acetate of lead to its aqueous solution, and that
this assumes a purple colour in contact with the air.

He proposes to give the name carmine, hitherto applied to
the red colouring matter, to the white crystals, and to desig-
nate the red substance by the name " Carmeine."

The statements of Preisser, generalizing most beautifully
under one head the chemical character of all colouring mat-
ters, making indigo as it were the prototype of them all,
could but induce other chemists to M^ork out more in detail
the relations cursorily pointed out in the memoir of this che-
mist. Unfortunately a careful repetition of these experiments
has not confirmed the basis on which his theory reposes.

M. A. E. Arppe repeated Preisser's experiments on the
colouring matter of cochineal*. He found that by proceed-
ing in the manner described by Preisser that he could only
obtain a red solution, which on evaporation was converted
into white crystals of oxalic acid by the nitric acid derived
from the basic lead salt,

Arppe now prepared a pure oxide of lead by precipita-
ting acetate of lead with potash. He found that this would
not take down the colouring matter in the cold, but by boil-
ing it is carried down as a blue lake, which he decomposed
by hydrosulphuric acid ; the supernatant liquor was nearly
colourless, the colouring matter he found adhered with great
obstinacy to the sulphuret of lead, from which water, alcohol
and ammonia failed to separate it ; but sulphuret of ammo-
* Liebig's Annalen, vol. J.v. p. 101.

476 Mr. De la Rue on Cochineal.

nium and acids rendered it perceptible. He likewise tried
to obtain the colouring matter in a state of purity by pre-
cipitating with hydrated oxide of lead an aqueous decoction of
cochineal (previously purified from nitrogenous substances by
nitrate of silver). On treating the precipitate by hydrosul-
phuric acid, he obtained a red liquor of strongly acid reaction,
the acid of which Avas not derived from the lead salt : on eva-
poration it left a dark red mass, emitting the smell of burnt
sugar. Wishing to separate the acid, which he thought
contaminated the colouring matter, he prepared a strong
aqueous decoction of cochineal, and after separating the ni-
trogenous matter by means of nitrate of silver, filtering, and
then saturating by ammonia, and afterwards adding the
hydrated oxide of lead, he found that the supernatant am-
moniacal liquor, which was nearly colourless, yielded by
evaporation an acid liquid ; and on decomposing the lead lake
with hydrosulphuric acid, he obtained a liquid slightly
coloured (the colouring matter adhering to the sulphuret of
lead), which was likewise acid. From this he concluded that
the colouring matter had not up to that period been obtained
in a state of purity.

Microscopic Examination of Living Cochineal.

By the kindness of Sir James Clark, who furnished me with
specimens of the living insect, I have been enabled to examine
the physical characters of the colouring matter as it exists in
the organism of this little insect before it is subjected to the
process of drying for commerce. On examination by the mi-
croscope, the white dust which covers the insect and the ad-
jacent parts of the cactus leaves, on which it feeds, has all the
characters of an excrement ; it has a curled cylindrical form,
is of very uniform diameter and of a white colour. On re-
moving the powder with a little aether and piercing the side
of the little creature, a quantity of a purplish red fluid exudes,
which contains the colouring matter in minute granules as-
sembled round a colourless and larger nucleus, and these
groups float in a colourless fluid. It is evident from this,
that, whatever may be the function of the colouring matter, it
has a distinct and marked form, and does not pervade as a
mere tint the fluid portion of the insect.

Separation of the Colouring Matter.

It became evident from a few preliminary experiments that
the investigation would be greatly facilitated by the employ-
ment of a large quantity of material ; and as in the course of
the inquiry different methods were adopted for the prepara-

Mr. De la Rue on Cochineal. 477

tion of the colouring matter, capital letters will be used to
designate the various preparations.

A. About 3 lbs. of ground cochineal (technically known as
shelly black) was introduced into 15 gallons of boiling distilled
water, and the mixture maintained at that temperature for
twenty minutes ; the decoction, strained through a sieve, was
allowed to subside for a quarter of an hour and then decanted
off; whilst still hot the transparent liquid was mixed with
basic nitrate of lead, added with caution to avoid excess ; a
fine purple lake was thus obtained, the supernatant liquor
retaining only a pale buff tinge. After decantation of the
supernatant liquor, the lake was thrown on a cloth filter and
washed with distilled water until the filtrate gave but a slight
opalescence with chloride of mercury, which was found to be
a test for the presence of nitrogenous matter. The lead lake
was then suspended in distilled water and treated with a co-
pious stream of hj'^drosulphuric acid, when a precipitate of
sulphuret of lead and a deep red supernatant liquid was ob-
tained ; on stirring the liquid this colour almost disappeared,
the colouring matter being evidently absorbed by the sul-
phuret, agreeing perfectly with Arppe's observation. A fresh
stream of gas reproduced the colour, which was again absorbed
on stirring; after continuous treatment with hydrosulphuric
acid, the lead lake being completely decomposed, the filtered
liquid was evaporated in a water-bath to a syrupy consistence,
and the evaporation finished as far as possible at a tempera-
ture of 38° C. The semi-solid substance thus obtained was
of a deep purple colour, had a strongly acid reaction, and
evolved the smell of burnt sugar, as noticed by Arppe. The
weight of this substance, which I call crude carminic acid,
was 3^ ozs., and 1 oz. more was obtained from the residue by
similar treatment.

B. On repeating the same process the whole product was
lost. An excess of the basic nitrate having been employed,
the nitric acid set free by the hydrosulphuric acid caused a
violent decomposition, with evolution of nitrous fumes, as
soon as the carminic acid arrived at a pasty consistence ; this
agrees also with Arppe's experience.

C. In this operation a decoction of cochineal, made in the
described manner, was precipitated with a solution of acetate
of lead acidulated with acetic acid (six parts by weight of
crystallized acetate, and one part of strong acetic acid). The
resulting lead lake, being very bulky, was washed by decan-
tation with boiling distilled water, collected on a filter, dried
in a current of warm air, and finely powdered; 17 ozs. of
crude cai'minate of lead were thus obtained.

478 Mr. De la Rue on Cochineal.

D. Half a pound of cochineal was boiled with five pints of
alcohol, spec. grav. '830. The filtered tincture deposited on
cooling a granular precipitate, consisting chiefly of fatty
matter retaining a portion of colouring matter; on concen-
trating the tincture by distillation a further quantity was
deposited, which was filtered off; the filtrate was evaporated
to dryness in vacuo, when after eight weeks a gummy resi-
due was obtained. This mass dissolved with great difficulty
in a large quantity of absolute alcohol, a red flocculent
substance consisting chiefly of nitrogenous matter remaining
undissolved. The alcoholic solution filtered off" from this
deposit, concentrated by distillation and finally evaporated in
vacuo over sulphuric acid, dried to a tenacious semi-solid
mass, covered with a colourless oily fluid, and containing cry-
stalline particles of a solid fat. After removal of the fats by
means of sether, this mass was digested in water at 38° C,
which partly dissolved it with a fine red colour, leaving a
brown mass of resinous aspect behind, more of which de-
posited on the cooling of the coloured liquid ; the decoction
was now evaporated to the consistence of a syrup, and finally
dried in vacuo over sulphuric acid.

These are all the processes employed to extract the colour-
ing matter from the cochineal; 1 may here remark, before
entering on the details of its further purification, that I ob-
tained other substances on evaporating the mother-liquors
from which the colouring matter had been separated by lead
salts, which will be hereafter described.

Purification of the Carminic Acid. — In my first attempts
to purify the colouring matter I proceeded in the following
way : — An aqueous solution of the crude carminic acid (A)
was precipitated with acetate of lead, the precipitate of car-
minate of lead well-washed and decomposed by hydrosulphu-
ric acid ; the red supernatant liquid was first concentrated on
the water-bath and finally dried in vacuo; a highly hygro-
scopic purple residue was thus obtained.

I could not, by whatever means I adopted, effect the deco-
lorization of the colouring principle. In several attempts I
heated the solution for some hours to 100° C, keeping up a
continuous current of hydrosulphuric acid, and in other ex-
periments a stream was made to pass for several days through
the disengaged colouring matter, but without the slightest
change in its aspect. From these experiments, made with
the greatest care and at several periods, I am led to the same
conclusion as Arppe, that Preisser must have been mistaken
in his results, and I regret that I cannot throw any light on
the probable cause of his error.

Mr. De la Rue on Cochineal. 479

Several combustions of the carminic acid thus purified were
made, the resulting numbers however became useless by the
subsequent observation that this acid was by no means pure.
A sufficient quantity being incinerated left a residue of acid
reaction, which was suspected to contain phosphoric acid.
Carminic acid burning only with great difficulty, it was re-
converted into carminate of lead, the oxide of lead dissolved
out of the residue obtained after fuming by acetic acid, which
left a white residue of metaphosphate of lead, together with a
little lead. The white residue was soluble in dilute nitric
acid, and exhibited, when treated before the blowpipe, the
characters of metaphosphate of lead; other tests likewise
confirmed the presence of phosphoric acid. It will hereafter
be seen that the process of extracting the colouring matter by
alcohol (D) does not exclude the phosphoric acid, which in
all probability existed in the colouring matter analysed by
Pelletier. It is further evident that the presence of phos-
phoric acid explains most satisfactorily the facts observed by
Arppe.

a. In order to separate the phosphoric acid, another por-
tion of crude carminic acid (A) was precipitated with acetate
of lead. Three-fourths of the carminate of lead produced were
decomposed by hydrosulphuric acid and evaporated to dry-
ness in the way already mentioned. The dry mass being dis-
solved in cold absolute alcohol, and filtered from a slight floc-
culent brownish residue, was heated to ebullition in a water-
bath and mixed with the remaining fourth of the carminate
of lead, which had been previously reduced to a fine powder ;
the ebullition was continued for a few hours. In this method
the free phosphoric acid combined with the lead, liberating
an equivalent proportion of carminic acid, which was taken
up by the alcohol. The alcoholic solution was filtered whilst
hot, concentrated by distillation, and then evaporated in vacuo
in the presence of sulphuric acid; it dried into a granular
mass of a deep purple-brown colour, detaching itself sponta-
neously from the sides of the vessel, and on examination by
the microscope was found to be a beautiful transparent crim-
son substance, exhibiting only slight, if any, signs of crystal-
line structure ; by pulverization it became of a fine scarlet
colour ; it left a mere trace of ash, and was found to be per-
fectly free from phosphoric acid. It was highly hygrome-
tric*.

* In consequence of this it was found convenient to dry the carminic
acid intended for analysis in little stoppered tubes in tlie air-pump, as tlie
stopper could be rapidly inserted after desiccation, and access of air effec-
tually prevented.

480 Mr. De la Rue on Cochineal.

Burnt with chromate of lead, —

I. '4647 grm. gave '9096 grm. carbonic acid and '2175 grm.
water.

JI. •4630 grm. gave '9105 grm. carbonic acid and '2140
grm. water.

For the latter analysis I am indebted to my friend Mr. Ni-
cholson.

b. A second preparation of carminic acid was made by
operating on the crude carminate of lead (C) and treating the
resulting crude carminic acid in the manner just described for
the preparation a. It left on incineration 0*2 per cent, of ash
(•1609 grm. giving '0003 grm. ash), which was neglected in
the following analyses : —

III. '3710 grm. gave '7316 grm. carbonic acid and '1710
grm. water.

IV. '3685 grm. gave '7235 grm. carbonic acid and '1722
grm. water.

c. To effect the purification of the carminic acid (D) ob-
tained by digesting cochineal in alcohol, it was dissolved in
water and precipitated by acetate of lead; the filtrate was found
to contain nitrogenous matter,' and the carminate of lead to
be contaminated wath phosphate of lead; it was therefore
treated in the manner already detailed.

V. "3925 grm. of this substance gave "7658 grm. carbonic
acid and •1780 grm. water.

d. A fourth preparation of carminic acid was obtained by
substituting phosphoric acid for hydrosulphuric in the de-
composition of the crude carminate of lead (C), and evapo-
rating the carminic acid to dryness in contact with a fresh
portion of carminate of lead ; this did not however separate
entirely the phosphoric acid, it was therefore redissolved in
boiling absolute alcohol, and maintained some time at that
temperature with more carminate of lead. On analysis —

VI. ^3805 grm. gave "7530 grm. carbonic acid and •1848
grm. water.

Pelletier having obtained in his analysis of " carmine"
(carminic acid) as much as 3*56 per cent, of nitrogen, all the
before-cited preparations of carminic acid were carefully exa-
mined qualitatively for nitrogen by heating with soda-lime,
and without exception gave indications of its presence; in
most cases but a mere trace was found, but I thought it ne-
cessary notwithstanding to make a few quantitative determi-
nations, especially as M. Berzelius* had pointed out the im-
probability of it being an essential constituent.

* Traite de Chim. t. iii. p. 808. Brussels, edit. 1839.

Mr. De la Rue on Cochineal, 481

The last preparation {d) appearing to contain more than
any of the others, it was chosen and burnt with soda-lime.

It was indispensable in experiments of this nature to test
the purity of the soda-lime as regarded the absence of am-
monia. A tube having 9 inches of its length filled with soda-
lime was heated to redness, just as in a nitrogen determi-
nation ; the hydrochloric acid, being treated with bichloride
of platinum in the usual manner, gave 7'5 milligrammes of
ammonio-chloride of platinum ; and a repetition of the expe-
riment gave a similar result. This allowance has been made
on all the nitrogen determinations by soda-lime.

•5938 grm. carminic acid {d) gave '0717 grna. ammonio-
chloride of platinum = 0*76 per cent, of nitrogen.

This quantity of nitrogen could not be supposed to belong
to the composition of the colouring matter, but was evidently
due to some foreign substance, and not improbably to am-
monia. In order to purify the carminic acid still more, the
same specimen {d) was dissolved in a small quantity of boil-
ing absolute alcohol and the filtered solution mixed with three
times its bulk of anhydrous aether ; a splendid scarlet precipi-
tate was immediately produced, which absorbed water rapidly
from the atmosphere, and agglutinated into a dark purple
mass ; when dried it weighed 0*3 grm. (e). The filtrate,
which was of a pale orange-red colour, left on evaporation
0*5 grm. of carminic acid (/").

•2635 grm. (e) burnt with soda-lime gave '0637 grm. am-
monio-chloride of platinum = 1*52 per cent of nitrogen.

•4732 grm. (/) gave "0150 grm. ammonio-chloride of pla-
tinum = 0^2 per cent, of nitrogen.

We have therefore (e) 0*3 grm. found to contain by ana-
lysis 1^5 per cent, nitrogen, and (/) 0^5 grm. 0^2 per cent.,

(3 X 1-5) -I- (5 X 0^2) ^„ . u- 1,

^ = '69 average per cent., which agrees

o

closely with '76, found previous to treat]Tient with eether.

(/. Another preparation of carminic acid was obtained
by precipitating crude carminic acid with acidulated acetate
of copper, which salt was found to carry down the carminic
acid, and to leave in solution by far the greater portion of the
phosphoric acid. The carminate of copper, which occupied
a long time in washing, was collected and decomposed by
hydrosulphuric acid. The filtrate was evaporated to dryness,
dissolved in boiling absolute alcohol, filtered, concentrated
by distillation, and again evaporated to dryness in vacuo. It
still contained a trace of phosphoric acid. On evaporating
the mother-liquor and filtering, from an impure carminate of
copper which deposited as the acetic acid was driven ofi^, and

Phil. Mag. S. 3. No. 2 11 . Suppl. Vol. 3 1 . 2 1

482 Mr. De la Rue on Cochineal,

again concentrating to dryness, a brown mass was obtained,
which on incineration left a greenish- white very hygrometric
ash, in which phosphoric acid, soda and copper were found.
Burnt with chromate of lead —

VII. '4020 grm. carminic acid {g) gave '7842 grm. carbonic
acid and '1662 grm. water.

This acid however still retained some impurities : on inci-
neration it left 0*4 per cent, of ash ('5489 grm. giving "0022
grm. ash), and examined for nitrogen it gave the following
numbers : —

•4731 grm. burnt with soda-lime gave 'OlSOgrm. ammonio-
chloride of platinum = 0*2 per cent, of nitrogen.

h. In order to separate these impurities the greater portion
was dissolved in boiling absolute alcohol, and filtered from a
slight residue ; the remainder, about an eighth, was converted
into carminate of lead and digested with the boiling alcoholic
solution for some hours ; the alcoholic tincture filtered off
whilst hot and mixed with about six times its volume of an-
hydrous aether ; this threw down a bulky precipitate of a fine
red colour, which was separated by filtration and the filtrate
concentrated in a retort, and finally evaporated to dryness in
vacuo {h).

i. The precipitate retained on the filter was dried in vacuo,
then dissolved in as small a quantity of alcohol as possible,
and again mixed with a large quantity of aether ; this deter-
mined a precipitate which was no longer of a fine red but of
a brown colour, and on re-solution and similar treatment it di-
minished in quantity and became darker in colour, leaving
the colouring matter in solution. From the filtrates a quan-
tity of carminic acid {i) was obtained on evaporating to dry-
ness in vacuo. It therefore appears that the aether precipi-
tates a nitrogenous body which carries down with it variable
quantities of carminic acid, according as a larger or smaller
relative proportion is present. The carminic acid {h) was
found to be free from phosphoric acid as well as nitrogen.

•3003 grm. burnt with soda-lime gave "001 5 grm. ammonio-
chloride of platinum = 0*03 per cent, of nitrogen.

From this analysis I venture to assert that the colouring
principle of cochineal contains no nitrogen, thus fully con-
firming the anticipation of Berzelius. We can now under-
stand from the preparation of the specimen of carminic acid
(e), that the method employed by Pelletier for the prepara-
tion of the substance he analysed was calculated to accumu-
late all the nitrogenous matter contained originally in his
alcoholic decoction ; a fact which fully explains the large
amount of nitrogen he obtained in his analysis.

Mr. De la Rue o« Cochineal. 48J

An analysis of the carminic {h) acid by chromate of lead
gave from —

VIII. '3167 grm. '6203 grm. carbonic acid and •1402
water.

The following table exhibits the per-centage results de-
duced from the foregoing analyses : the specimens were all
dried over sulphuric acid in vacuo, with the exception of ana-
lysis VII., in which the carminic acid was dried at 100° C.

I.

II.

III.

IV.

V.

VI.

VII.

VIII.

Mean.

Carbon... 53-38

53-63

53-78

53-55

53-21

53-97

53-20

53-42

53-51

Hydrogen 5-20

5-14

5-12

5-19

5-04

5-39

4-59

4-92

5-07

By the analysis of a copper salt of carminic acid hereafter
to be mentioned, it became probable that carminic acid might
still retain, when only dried in vacuo, a portion of the solvents
employed ; a presumption which was supported by the ana-
lysis VII., in which the substance analysed had been dried
at 100*^ C, and which gave a smaller per-centage of hydrogen.
A portion of carminic acid {i), being first dried in vacuo, and
then heated to a temperature of 121° C, was found to yield
a small quantity of acetic acid, and was not altered in its
properties, which were not in fact changed even at a tempe-
rature of 136° C.

In the following analyses the carminic acid, previously dried
in vacuo * and then at a temperature of 120° C, gave, on
burning with chromate of lead, the following results : —

IX. '3347 grm. {h) gave '6648 grm. carbonic acid and
•1381 grm. water.

X. '3583 grm. (i) gave '7108 grm. carbonic acid and •1504
grm. water.

These analyses give the following per-centage quantities : —

IX. X.

Carbon . . . 54*17 54-10

Hydrogen . . 4-58 4*66

The analysis IX. being of the same preparation as had
served for analysis VI II., it is fair to presume that all the
other specimens of carminic acid would have given the same
per-centage quantities as the specimen (A) if dried at 120° C,
as this particular specimen, dried in vacuo, yielded numbers
in close accordance with the mean of the other analyses.

These numbers converted into the most simple expression
lead to the following formula, Cj4 H^ Og ; but an analysis of
a copper salt renders it probable that this formula has to be

* The carminic acid fuses if exposed to a temperature of 120° C, with-
out having been previously dried.

2X2

4S4 Mr. De la Rue on Cochineal.

doubled, and that the composition of carminic acid is ex-
pressed by the formula,

C28 Hi4 O16,
as may be seen from the following table containing the com-

})arison of the theoretical per-centages with the mean of ana-
yses IX. and X.

Theory. Experiment.

C29 . 168 54-19 5413

Hi4 . 14 4-52 4-62

Oje- . 128 41-29 41-25

310 100-00 100-00

From the foregoing experiments, it seems that the best
method of obtaining pure carminic acid is to precipitate
the aqueous decoction by acetate of lead ; to decompose the
washed carminate of lead by hydrosulphuric acid, and to
throw down the carminic acid once more by acetate of lead,
previously mixed with acetic acid ; to decompose the carmi-
nate of lead by hydrosulphuric acid ; to evaporate to dryness
and redissolve the carminic acid in absolute alcohol ; then to
digest the alcoholic tincture with carminate of lead ; and lastly,
to precipitate the trace of nitrogenous matter by aether, the
pure carminic acid is obtained from the filtrate.

As thus prepared, carminic acid has the following proper-
ties. It is a purple brown friable mass, transparent when
viewed by the microscope, and pulverizing to a fine red
powder ; soluble in water and alcohol in all proportions, very
slightly soluble in aither, which does not however precipitate
it from its alcohoHc solution if free from nitrogenous matter.
It is soluble without decomposition in concentrated hydro-
chloric and sulphuric acids. It is decomposed by chlorine,
iodine and bromine, which change its colour to yellow, and
the latter on warming or by standing gives a yellow precipi-
tate soluble in alcohol. Nitric acid decomposes it even if
highly diluted : I shall have occasion to refer to this decom-
position presently. It bears a temperature of 136° C. with-
out decomposition ; on gradually increasing the temperature
a quantity of an acid liquor is produced, and at a red heat it
intumesces and gives oft" a small quantity of red fumes, which
condense : it gives no trace of oily matter.

The aqueous solution has a feeble acid reaction ; it does not
absorb oxygen. A volume of this gas contained in a tube
with carminic acid over mercury did not change by absorption
after exposure for several months. The fixed alkalies and
ammonia give no precipitate in the aqueous solution, merely
changing its colour to purple j in the alcoholic tincture they

Mr. De la Rue on Cochineal. 485

produce purple precipitates ; all the alkaline earths produce
purple precipitates ; sulphate of alumina gives no precipitate,
but on addition of a drop of ammonia the carminic acid is
immediately taken down as a beautiful crimson lake ; acetates
of lead, copper, zinc and silver give purple precipitates ; the
latter is immediately decomposed, and silver deposited ; the
nitrates of lead, mercury and silver reddish precipitates ; pro-
tochloride and bichloride of tin no precipitates, but change
the colour to a deep crimson.

The acid character of carminic acid being so very little pro-
nounced, I met with considerable difficulties in determining
its atomic weight ; it is only with great reserve that I bring
forward the formula before cited. Several attempts were
made to produce soda, baryta, lead and copper compounds,
but it was only with the copper salt that I obtained results
agreeing in different preparations.

It seems that carminic acid attaches itself to salts, for it
was found that the precipitants could be removed from the
precipitates only with the greatest difficulty. I omit several
soda, baryta and lead determinations which have not led to
any satisfactory result, and confine myself to the statement of
the result of the analysis of the copper compound. It was
obtained by acidulating an aqueous solution of pure carminic
acid with acetic acid, and then precipitating by the cautious
addition of acetate of copper, so as to leave an excess of car-
minic acid in the liquid. The precipitate was well-washed by
decantation (by which a great loss was sustained) and dried.
It formed into masses of a bronze colour, very hard and dif-
ficult to powder. Two specimens were prepared at different
times (« and b).

I. -2800 grm. (a) dried at 100° C. left, after ignition and
treatment with nitric acid and re-ignition, '0330 grm. oxide
of copper.

II. -3782 grm. (b) dried at 100° C. gave -0426 grm. oxide
of copper.

III. '4702 grm. (b) dried at 100° C. gave on burning with
chromate of lead '8210 grm. carbonic acid and '1743 grm.
water.

These numbers lead to the following per-centage results : —

I. II. III.

Carbon ... 47*62

Hydrogen ... 4*12

Oxide of copper 11-78 11-27
agreeing closely with the formula, C^ H^4 Ojg, CuO, as will
be seen from a comparison of the theoretical and experimental
numbers.

486 Mr. De la Rue on Cochineal.

Calculated.

Found.

Carbon ... 28

168

48-05

47-62

Hydrogen . .14

14

4-01

4-12

Oxygen • . .16

128

36-61

36-74

Oxide of copper 1

39-6

11-33

11-52

349-6 100-00 100-00

Action of Nitric Acid on Carminic Acid.

Nitrococcusic Acid. — ^When acting with nitric acid on " car-
mine" (carminic acid), MM. Pelletier and Caventou obtained
white acid crystals resembling oxalic acid, but differing from
this acid in several of its properties. M. Arppe found that
the acid produced was oxalic acid. In my experiments I ob-
tained the following results : — One pound and a half of crude
carminic acid was gradually introduced into ten pounds of
nitric acid, spec. grav. 1-4, and digested at a moderate heat;
a violent evolution of nitrous fumes succeeded each addition
of the carminic acid ; after the whole quantity had been in-
troduced and the action had somewhat subsided, the mixture
was transferred into a smaller vessel and the action continued
at the boiling-point for about two hours ; by this time the
greater part of the nitric acid had evaporated, and on with-
drawing the vessel from the fire and allowing the mixture to
cool, a crystalline cake was obtained, which on examination M'as
found to consist partly of a new acid and partly of oxalic acid.
To separate the oxalic acid, it was dissolved in a large quan-
tity of boiling water and treated with nitrate of lead as long
as any precipitate formed ; this was collected and decomposed
by boiling with dilute sulphuric acid ; the filtrate from the
sulphate of lead yielded a large quantity of prismatic crystals
of oxalic acid, which were obtained perfectly white and pure
after two or three crystallizations with the aid of a little animal
charcoal.

The yellow liquid filtered from the oxalate of lead was con-
centrated and separated from a fresh portion of oxalate which
deposited on concentration, the evaporation was then continued
until a large quantity of crystals formed; the solution on
cooling deposited a very bulky mass of yellow rhombic prisms,
which were drained and dried, and re-dissolved in a sufficient
quantity of boiling water, which on cooling deposited the acid
(for which I propose the name of nitrococcusic acid) in
beautiful crystals free from any lead salt ; it was recrystal-
lized twice more, by which means it was obtained perfectly
pure.

Several preparations were made, sometimes using pure car-

Mr. De la Rue on Cochineal, 4A%

minic acid, at other times carminate of lead, with similar re-
sults.

The analyses of four different preparations dried at 100°
C. gave, on burning with chromate of lead (unless otherwise
stated), the following numbers : —

I. '3152 grm. {a) gave '3892 grm. carbonic acid and '0561
grm. water.

II. "2500 grm. (a) gave '3080 grm, carbonic acid and '0445
grm. water.

(For this analysis I am indebted to Mr. Nicholson.)

III. '3068 grm. (a) gave '3820 grm. carbonic acid and
•0502 grm. water.

IV. '4498 grm. {b) gave '5626 grm. carbonic acid and
•0757 grm. water.

V. '4461 grm. (c) gave '5515 grm. carbonic acid and '0777
grm. water.

VI. '4503 grm. (c?) gave, on being burnt with oxide of
copper, '5585 grm. carbonic acid and '0757 grm. water.

VII. '4796 grm. (c) gave, on being burnt with oxide of
copper, and a layer of copper twelve inches long used so as
to completely decompose the binoxide of nitrogen, '5882 grm.
carbonic acid and '0815 grm. water.

The foregoing analyses lead to the following per-centage
quantities : —

I. II. III. IV. V. VI. VII.

Carbon . 33-67 33*60 33-95 34-11 33-72 33-82 33-44
Hydrogen 1-98 1-98 1-82 1-87 1-93 1-87 1*89

In the following experiments the nitrogen of the nitrococ-
cusic acid was ascertained by burning with oxide of copper
in an atmosphere of carbonic acid.

VIII. -6808 grm. {h) dried at 100° C. gave 84 cub. cent, of
moist nitrogen at 6°-5 C. and 0*7585 m., barometer corrected.

IX. '7162 grm. (c) dried at 100° C. gave 91-5 cub. cent, of
moist nitrogen at 17°*5 C. and 0*7641 m., barometer corrected.

These numbers correspond to the following per-centage
quantities : —

VIII. IX. Mean.

Nitrogen . . 15-03 14-92 14-97

X. In this experiment the nitrogen was determined accord-
ing to Bunsen's* method, which consists inhuming the sub-
stance mixed with oxide of copper in the presence of copper
turnings in a hard glass tube. The tube being freed from air
by a stream of hydrogen, is then exhausted, sealed hermetically,
and placed in an iron mould filled with plaster of Paris ; it is
then heated to redness and allowed to cool. After the com-

* Liebig's ^nnalen, vol. xxxvii. p. 27.

488 Mr. De la Rue on Cochineal.

bustion, the gas is transferred into a graduated jar over mer-
cury and its volume noted ; the carbonic acid being absorbed
by a potash ball, the volume is again read off. This analysis
gave the following numbers : —

Vol. Temp. Diff. of level. Barom.
Carbonic acid -f nitrogen (moist) 123 20°'7 C. 0"'-0640 0'"-7543
Nitrogen 22-2 20°-0 O^-IGSO 0-"7529

The height of the column of mercury in the eudiometer
above the level in the trough and the barometric column are
corrected for temperature.

Carbonic acid -\- nitrogen corrected to 0° C. and barom. 1™ := 76-84

Nitrogen 0° ... 1°' = 12-16

Carbonic acid 0° ... l'" = 64-68

■ - ., - = 5-32, which is the ratio of carbon equivalents to one

I2"lb

nitrogen equivalent.

The preceding analyses of nitrococcusic acid agree with the
following formula, confirmed by the analyses of several of its

compounds, C.,H,N3 0,„
as will be seen on referring to the table.

Theory.

Experiment.

Carbon . 16
Hydrogen 5
Nitrogen . 3
Oxygen . 18

96 33-45
5 1-74

42 14-63
144 50-18
287 100-00

33-75

1-91

14-97

49-37

100-00

By analysis VII., in which the precaution was taken of
using a very long layer of copper turnings, there was ob-
tained, carbon 33-44, hydrogen 1-89 ; these numbers agree
as closely as possible with the theoretical quantities, as does
likewise the nitrogen determination (X.) by Bunsen's method ;
in this experiment the ratio of carbon equivalents to nitrogen
equivalents was found to be as 5*32 to 1, or as 16 equivs. of
carbon to 3-007 equivs. of nitrogen ; taking analysis VII. as
the basis of calculation, it gives 14'67 per cent, of nitrogen,
the theoretical number being 14-63.

The acid, as it separates from its aqueous solution, contains
water of crystallization, which it loses at 100° C. ; four expe-
riments gave the following results : —

•4800 grm. lost -0289 grm. = 6-02 per cent.

-6613 -0395 ... =5-97

•6586 -0385 ... =5-84

•4804 -0289 ... =6-01

Mean . . =5*96

Mr. De la Rue on Cochineal. 489

This mean corresponds perfectly with the formula
C,6H5N3 0i8 + 2Aq,
as may be seen by a comparison of the theoretical and expe-
rimental numbers.

Theory. Experiment.

f ^ 1 Mean.

1 equiv. dry acid ... 287 94*10 94*04

2 ... water .... 18 5*90 5*96
1 ... crystallized niO 3^^ ^^^,^^ ^^^^.^^

trococcusic acid . .J
Properties of Nitrococcusic Acid. — It is of a yellow colour,
crystallizing in rhombic plates, and presenting very different
aspects, according to the circumstances under which it is cry-
stallized. Its solution stains the skin yellow, it is soluble in
cold, but considerably more so in hot water; soluble in al-
cohol, and very soluble in aether. All its salts dissolve readily
in water, and most of them in alcohol ; it deflagrates violently
on being heated ; it dissolves iron and zinc, becoming dark-
coloured. It is decomposed by sulphuret of ammonium with
separation of sulphur and the formation of the ammonia salt
of a new acid, which I have not yet examined.

Compounds of Nitrococcusic Acid.

Nitrococcusate of Potash. — I have prepared this salt by
two different methods.

a. A solution of nitrococcusic acid in boiling water was
accurately saturated with carbonate of potash ; by evaporation
to a small bulk and cooling, the salt was obtained in small
yellow crystals ; it was purified by draining and recrystallizing.

b. An aetherial solution of the acid was precipitated by the
cautious addition of an alcoholic solution of potash ; the pale
yellow precipitate washed with aether and dried, then dissolved
in as small a quantity of cold water as possible, and the solu-
tion poured into about five times its bulk of absolute alcohol ;
after standing some time the salt crystallized in well-formed
crystals ; it was washed with aether and dried. The aetherial
washings being added to the mother-liquor, a further portion
was obtained and washed with aether. The latter process is
less troublesome than the process a.

I. *5469 grm. (a) dried at 100° C. were dissolved in a small
quantity of boiling water and decomposed by sulphuric acid ;
dried in a water-bath, the nitrococcusic acid, removed by aether
and the residue ignited, gave '2606 grm. sulphate of potash.

IT. *4383 grm. [b) dried at 132° C. gave -2103 grm. sul-
phate of potash.

III. '6251 grm, {b) dried at 100° C. and burnt with chro-

490 Mr. De la Rue on Cochineal.

mate of lead, gave -6064 grrn. carbonic acid and -0662 grm.
water.

These numbers give the following per-centage quantities : —

I. II. ■ III.

Carbon ... 26-46

Hydrogen . ... ... 1*18

Potash . . 25-74 25-92

corresponding with the formula

Ci6H3N3 0,6+2KO,

as may be seen by comparing the theoretical and experimental
numbers.

Theory.

Experiment.
vj.6cin»

Carbon .

16

96

26-45

26-46

Hydrogen
Nitrogen .

3
3

3

42

-83
11-57

1-18

Oxygen .
Potash

16

2

128
94

35-26
25-89

25-83

363 100-00

I was not successful in preparing a nitrococcusate of potash
with one equivalent of fixed base ; the method I adopted was
saturating a given weight of acid with carbonate of potash,
and then adding the same amount of acid to the bibasic pot-
ash salt ; on washing with tether the greater part of the ex-
cess of acid was removed, leaving the bibasic salt behind.

Nitrococcusate of Ammonia. — This salt was prepared by
passing an excess of dry ammoniacal gas through an aetherial
solution of the acid dried in the atmosphere ; the solution be-
came turbid, and by standing for a short time deposited the
salt in clusters of needles adhering firmly to the sides of the
vessel ; these were removed, washed with jether, and dried on
bibulous paper. It is volatile, and sublimes on being heated,
most probably with decomposition,

I. -6011 grm. of the salt dried in vacuo was dissolved in a
small quantity of boiling water and decomposed by strong
hydrochloric acid, which immediately separated the acid in
f rystals ; the mixture was dried in a water-bath, and the ni-
trococcusic acid removed by aether, a little bichloride of pla-
tinum and alcohol being added to the aetherial washings to
precipitate a trace of chloride of ammonium. The residue,
precipitated as ammonio- chloride of platinum, gave -8208
grm. of the double chloride.

II. -6126 grm. dried in vacuo and burnt with oxide of cop-
per, the mixture being made in the combustion-tube, gave
.'6525 grm. carbonic acid and -2191 grm. water.

Mr. De la Rue on Cochineal,

491

These numbers correspond with the following per-centage
quantities : —

I. II.

Carbon . » 29-05

Hydrogen 3*97

Oxide of ammonium . . 15"91

agreeing closely with the following formula,
C,6H3N3 0,6,2NH4 0 + HO,

as may be seen by a comparison of the theoretical and expe-
rimental numbers.

Theory.

Carbon . 16
Hydrogen 12
Nitrogen . 5
Oxygen . 19

96
12

70
152

330

29-09

3-64

21-21

46'06

100-00

Experiment.

29-05

' 3-97

Or

Theory.

Experiment.

Acid 1 269

Water 1 9

Oxide of ammonium 2 52 15*76

15-91

330

Nitrococcusate of Baryta was prepared by adding an excess
of a solution of baryta to an aqueous solution of nitrococcusic
acid, a stream of carbonic acid gas being passed through the
solution to separate the excess of baryta. The solution was
warmed, filtered and evaporated in a water-bath, and again
filtered from a small quantity of carbonate of baryta. The
evaporation being continued until a pellicle formed on the
surface, the solution on cooling deposited this salt in minute
yellow crystals. It is insoluble in alcohol, which precipitates
it in the form of a jelly from the aqueous solution.

I. -6750 grm. of substance dried at 100° C. and decom-
posed by sulphuric acid, gave '3602 grm. of sulphate of
baryta.

II. -6439 grm. of nitrococcusate of baryta dried at 100°
C. and burnt with chromate of lead, gave -5185 grm. of car-
bonic acid and -0800 grm. of water.

These numbers correspond to the following per-centage
quantities: —

I. II.

Carbon 21'96

Hydrogen 1*38

Baryta 35-06

492 Mr. De la Rue on Cochineal.

agreeing with the formula CjgHgNgOig + 2BaO + 2HOj as
may be seen from the following table : —

Theory. Experiment.

(

1

Carbon . 16

96-00

21-80

21-96

Hydrogen 5

5-00

1-14

1-38

Nitrogen . 3

42-00

9-54

Oxygen . 18

144-00

32-71

Baryta . . 2

153-28

34-81

35-06

440-28 100-00

Nitrococcusate of Silver. — I attempted to make this salt by
boiling oxide of silver with an aqueous solution of nitrococ-
cusic acid, but there was an evident decomposition of the
acid, a large quantity of carbonic acid being evolved ; after
warming the filtered liquor a brown deposit was formed. On
filtering off this brown deposit a silver salt was obtained by
evaporation, which yielded on analysis —

Carbon .... 23-64
Hydrogen . . . 1-26
Oxide of silver . . 38-10

per-centage numbers not reconcilable with those of nitro-
coccusate of silver.

On decomposing a hot solution of this salt with hydro-
chloric acid a new acid was obtained, perfectly distinct from
nitrococcusic acid ; it crystallized in long needles ; very in-
soluble in water, but soluble in aether and alcohol. 1 refrain
from giving any further account of this acid until the study
is completed.

In order to avoid decomposition the nitrococcusate of silver
was prepared without the aid of heat, by dissolving carbonate
of silver in a cold aqueous solution of nitrococcusic acid and
evaporating the filtered solution in vacuo over sulphuric acid.
The salt crystallized in long bulky needle-like crystals of a
yellow colour; when dried at 100° C. the powdered salt be-
comes deep orange.

It is soluble in alcohol and water, and is highly explosive
when heated ; in small quantities it may be decomposed by
a progressive heat without any violent action ; but on attempt-
ing to decompose a quantity amounting to -45 grm. in a
porcelain crucible, heated in an air-bath, the salt exploded
with great violence, shattering the copper air-bath and driving
fragments of the crucible through the copper ; the tempera-
ture was noted just before the explosion, the thermometer
standing at 200° C. ; the silver was therefore determined as
chloride.

Mr. De la Rue on Cochineal. 493

I. -4698 grm. of substance (a) dried at 100° C. and decom-
posed by nitric acid and the silver precipitated by the addi-
tion of hydrochloric acid, gave -2675 grra. chloride of silver.

II. -5085 grm. of substance {b) dried at 100° C. gave
•2892 grm. chloride of silver.

III. -8184 grm. of substance («) dried at 100° C. and
Tjurnt with oxide of copper, gave -5700 grm. carbonic acid
and -0554 grm. water.

Corresponding to the following per-centage quantities : —

I. II. III.

Carbon ... 18-99

Hydrogen ... 0*75

Oxide of silver . . . 46-03 45-97

and agreeing closely with the following formula,

C,6H3N3 0,s+2AgO,

as may be seen by the following table : —

Theory.

Experiment.

Carbon ... 16 96 19-162
Hydrogen . . 3 3 -599
Nitrogen . . 3 42 8-383
Oxygen ... 16 128 25-549
Oxide of silver . 2 232 46-307

501 100-000

18-99

•75

46-00

Nitrococcusate of Copper. — This was made by dissolving
carbonate of copper in nitrococcusic acid and deposited on
evaporation in pale apple-green needles. I made no analysis
of this salt.

The following is a synoptical table of the analyses of nitro-
coccusic acid and its compounds : —

Hydrate of nitrococcusic acid . C,gH3N30i(. + 2HO.

Hydrate of nitrococcusic acidn 4_oTTn ■ oA

as crystallized from water j "i 4-

Nitrococcusate of potash +2KO.

Nitrococcusate of ammonia +2NH40 + Aq.

Nitrococcusate of baryta +2BaO + 2Aq.

Nitrococcusate of silver +2AgO.

The properties of nitrococcusic acid and its salts exhibit a
great analogy with those of a number of acids obtained by
the action of nitric acid on organic bodies, more especially
nitropicric and styphnic acids, from which it differs by the
greater solubility of its salts.

If we assume with many chemists the nitrogen of these

494 ' Mr. De la Rue on Cochineal.

acids to exist in the form of hyponitric acid, the formula of
nitrococcusic acid will be represented by

C,«/!lu To., 2HO.

'16

This acid would consequently derive from a non-nitrogenous
acid, having the composition expressed by the formula •

Ci6H6 04,2HO.

When I first began this investigation I imagined that a
similar relation might exist between nitrococcusic acid and
carminic acid ; the analysis of these acids, however, as well
as the simultaneous production of a large quantity of oxalic
acid in its oxidation, showed that this view was erroneous,
and that nitrococcusic acid was derived from carminic acid in
a more complex manner. Some attempts were made to pro-
duce the non-nitrogenous acid, the coccusic acid, but unsuc-
cessfully.

The experiments of MM. Cahours and Laurent on the
oxidation of the oils of anise and of tarragon [Oleum dracun-
culi) have made us acquainted with anisic acid, the composi-
tion of which is Cjg Hg Og. ■ The formula agrees with the
composition of the hypothetical hydrated coccusic acid.

Anisic acid, however, as well as nitranisic acid, being mono-
basic, it was not probable that the further introduction of the
elements of hyponitric acid Avould convert it into a bibasic
one ; nevertheless it was my intention to have studied the
further action of nitric acid on the acids mentioned, in order
to obtain if possible trinitroanisic acid, and to compare this
substance with the acid obtained from carminic acid, when an
account of some new experiments of M. Cahours came under
my notice, of the action of a mixture of concentrated sulphuric
and nitric acids on anisic acid, by which he has succeeded in
preparing trinitroanisic acid. The experiments of M. Ca-
hours have not yet been published in detail, and from his
short account it was not possible to decide on the identity or
non-identity of nitrococcusic and trinitroanisic acids. A small
specimen of anisic acid at my disposal was treated in the
manner described by him ; after acting for some time water
threw down an acid, from the insolubility of which I conclude
that these acids are only isomeric.

Investigation of the Mother-liquor from wJiick the Carminic
Acid had been separated.

On evaporating the mother-liquors of carminic acid and
separating the lead held in solution by means of hydrosul-

Mr. De la Rue on Cochineal. 495

phuric acid, they all gave the following results : on acquiring
a syrupy consistence, a white chalky-hke matter subsided;
this was separated by filtration, and proved to be a new cry-
stalline body. The liquor filtered off from this substance
deposited a small quantity more on further concentration, and
could only be dried to a soft tenacious mass, partly soluble in
alcohol, the rest being soluble in water. From three pounds
of cochineal five ounces of this soft matter were obtained,
showing that the precipitation by a lead salt had effected the
separation of carminic acid from a large quantity of foreign
matters. This gelatinous matter appears to be of a complex
character, but I have not yet examined it fully.

To purify the chalky-like matter, it was well-washed with
cold water and crystallized twice by solution in boiling water
and evaporation ; it w as then dissolved by boiling it in a just
sufficient quantity of water ; animal charcoal was now added,
and the ebullition continued for a few minutes ; the solution
filtered whilst hot deposited on cooling a mass of silky cry-
stalUne tufts, completely filUng the liquid, and when collected
and dried they aggregated into paper- like masses of a silky
aspect. I obtained in three experiments 4 parts of the new
body from 1000 of cochineal.

I. '4918 grm. of substance, preparation {a), dried in vacuo
and burnt with oxide of copper, gave 1*0705 grm. carbonic
acid and 0*2838 grm. water.

II. '5680 grm. of substance {b) gave 1*2416 grm. carbonic
acid and *3160 grm. water.

III. '4700 grm. of substance {b) gave 1*0210 grm. car-
bonic acid and *2660 grm. water.

For the latter analysis I am indebted to the kindness of
Mr. Nicholson.

A qualitative examination having pointed out the presence
of nitrogen, it Mas determined by Varrentrapp and Will's
method.

IV. '5046 grm. of substance (a) dried in vacuo and burnt
with soda-lime, gave *6131 grm. ammonio-chloride of pla-
tinum.

V. '5076 grm. of substance [b) gave '6239 grm. ammonio-
chloride of platinum.

From these numbers the following per-centages are calcu-
lated : —

I. II. III. IV. V.

Carbon . 59*36 59*62 59*25
Hydrogen 6*41 6*18 6*29

Nitrogen . ... ... ... 7*62 7'7l

496 Mr. De la Rue on Cochineal.

These per-centages, translated into the most simple expres-
sion, lead to the formula, Cgi Hu NOg as may be seen from
the following table : —

. 18
. 11
. 1
. 6

Theory.

Experiment.

Mean.

59-41

6-29

7-66

Carbon . .
Hydrogen .
Nitrogen .
Oxygen . .

108 59-668
11 6-077
14 7-735
48 26-520

181 100-000

Careful and repeated examinations for sulphur proved the
absence of this element as a component of the new white sub-
stance. I have been unable to produce a compound to con-
trol the proposed formula, though several methods were
adopted ; amongst others, I attempted to form a lead com-
pound by adding acetate of lead to an ammoniacal solution of
the substance; I obtained merely a bulky precipitate, con-
sisting of little else than oxide of lead.

This substance is sparingly soluble in cold water, much
more so in boiUng water ; insoluble in alcohol and aether ;
soluble in hydrochloric acid, which appears to be driven off
by evaporation, leaving the substance in large crystals. In a
large quantity of nitric acid it dissolves with a slight evolu-
tion of gas ; the solution evaporated spontaneously furnishes
long crystals, which are in all probability a new acid ; if dis-
solved in a small quantity of nitric acid, the mixture becomes
spontaneously heated, violent action takes place, and the pro-
duct is lost ; frequently the substance becomes blackened into
charred masses. It is soluble in ammonia, from which it is
again recovered by the evaporation of the ammonia. It is
soluble in the fixed alkalies, and is precipitated from these
solutions by saturating with an acid.

In a short paper, entitled "Valerianic Acid and a new
body from Casein," Baron Liebig* describes a new substance
obtained by fusing casein with hydrate of potash until an
evolution of hydrogen takes place along with ammonia. On
saturating with acetic acid the aqueous solution of the fused
mass an aggregate of fine needles was produced, which were
purified by repeated solution in carbonate of potash and re-
precipitation by acetic acid. A preliminary analysis led to
the formula Cjg Hg NO^, differing from the result I obtained
in the analysis of the white substance from cochineal by two
carbon, two hydrogen, and one oxygen. The properties of
the two bodies being however so analogous, it is extremely
probable that they are identical, a presumption I am sup-
• Liebig's Annalen, vol. Ivii. p. 127.

On Crystals in the Cavities oj Minerals. 4-97

ported in by a comparison of a specimen kindly furnished me
by Dr. Hofmann * ; further investigations will clear up this
point : in the meantime I refrain from proposing a name, as
Liebig f has lately proposed the name Tyrosine for the sub-
stance prepared from casein. As the latter body arises evi-
dently from a process of oxidation, and as I had obtained the
first crop of crystals from a liquid from which the colouring
matter had been precipitated by the basic nitrate of lead, I
thought that this body might owe its formation to the action
of the nitric acid liberated by the sulphuretted hydrogen ; but
this supposition proved to be erroneous, for in later experi-
ments in which acetate of lead had been used, the same body,
and in exactly the same quantity, was obtained. From this
we may assume that this substance is contained ready-formed
in the cochineal insect.

My engagements for the present preventing me from con-
tinuing these researches, I must defer for a future period their
completion, but hope to be enabled to communicate to the
Society a second paper. In conclusion I may be allowed to
express my thanks to my friend Dr. Hofmann for his valuable
instruction in the methods of organic research, and his kind
advice during the progress of this investigation.

LXXIII. On the Existence of Crystals tvith different jn-imitive

forms and j)hysical properties i7i the Cavities of Minerals ;

with additional Observations on the Nexv Fluids in which

they occur. By Sir David Brewster, K.H., LL.D.,

F.'R.S., and V.P.R.S. Edin.X

[With a Plate.]

IN 1823 and 1826 I communicated to the Society two
papers on the nature and properties of two immiscible
fluids, which I discovered, in contact with each other, in the
cavities of topaz and other minerals §. Although the facts
contained in these papers were of so extraordinary a nature
as to be received with scepticism by some, and with ridicule
by others, yet I am not aware that, during the twenty years
which have elapsed since their publication, any person has
either repeated my observations, or advanced a single step in
the same path of inquiry. In showing to strangers some of
the leading phaenomena of the two new fluids, my attention
has been frequendy recalled to the subject ; but it was not till

* This specimen had been prepared by Baron Liebig himself. — A.W.H.

f Researches on the Chemistry of Food, p. 16.

X Read before the Royal Society of Edinburgh on the 17th of February
1845, and published in their Transactions, vol. xvi. part 1. p. U.

§ Edinburgh Transactions, vol. x. p. 1 and 407.
Phil. Mag. S. 3. No.211. Stippl. Vol. 31. 2 K

498 Sir David Brewster on the Existence of Crystals

last spring, when I discovered cavities in topaz filled with the
most beautiful crystals of various form, that 1 was induced to
undertake a new investigation of their nature and properties.
In this investigation I have examined, with various magnifying
powers, and both in common and polarized light, more than
900 specimens of topaz from Scotland, New Holland, and the
Brazils ; and I have had the good fortune to observe many
new phaenomena connected with mineralogy, chemistry, and
physics, which, in addition to the interest which they may
possess as scientific facts, promise to throw a strong light upon
the existing theories of crystallization, and to bring before us
some of those recondite operations which had been going on
in the primitive rocks of our globe, before the commencement
of vegetable or animal life.

1. On the Form and Position of the Strata ifi which the Cavities

lie.

The cavities which contain the two new fluids, and their
accompanying crystals, sometimes occur single, and in groups
more or Jess numerous ; but, in general, they exist in millions,
occupying extensive strata, which affect the transparency of
the mineral, and render it unfit for the use of the jeweller, or
even for the cabinet of the collector, who has not learned that
it is in the deviations from her ordinary laws that Nature often
discloses her deepest mysteries.

Although the strata of cavities sometimes occur, as in arti-
ficial salts, in planes parallel to the primary or secondary
forms of the crystal, yet they occupy every possible position m
reference to these planes ; and we therefore cannot account
for them by supposing that certain spaces have been left in
the crystal, without the primitive molecules which ought to
have been there deposited. The strata of cavities, too, have
every possible curvature. From a plane surface they pass
into a curved one, sometimes of variable curvature, and some-
times of contrary flexure, cutting and intersecting each other
in the most capricious manner.

In the shape of the strata the same irregularity presents
itself; their outline is sometimes rectilineal, sometimes curved,
and sometimes singularly irregular. In some specimens the
whole crystal is intersected with the strata; and it is extremely
probable, though it is impossible to determine the fact, that in
every specimen some edge or angle of the stratum touches the
surface.

The succession of the cavities in composing the stratum,
and their form in relation to the character of the stratum,
present interesting phasnomena. I have found specimens in

in the Cavities of Minerals. • 499

which the cavities lie in concentric arches, and have their sides
concentric, and, as it were, a portion of the same arches, as if
they had been formed under the influence of a rotatory force.
In other cases they occupy parallel lines, and are sometimes
so equidistant that they might be advantageously used as mi-
crometers for microscopes. In one remarkable specimen they
radiate from a centre, each radiation having a character of its
own. One radiation will -sometimes throw off a diverging
branch, while two or more radiations will converge and then
diverge again, subsequently uniting themselves into a single
radiation.

When different strata of cavities lie parallel to each other
in the specimen, which they sometimes do, to the number of
four ovjivey each stratum has generally a distinct character ;
flat and exceedingly thin cavities occupying one stratum, very
deep cavities occupying another, minute cavities which the
highest magnifying powers can scarcely resolve occupying a
third, while a fourth consists of the most irregular and inde-
scribable forms.

When the forms of individual cavities are related to that of
the stratum which contains them, they, of course, cut at all
angles the primary and secondary planes of crystallization ;
and the same is true of insulated cavities of great length, which
are sometimes turned, and twisted, and bent in the most ca-
pricious manner. It is impossible to read these details, and
still more so to study the phaenomena themselves, without
being driven to the conclusion, that the strata of cavities must
have been formed under the influence of forces propagated
through a soft and plastic mass, and carrying along with them
gases and vapours which came to a position of rest previous
to the regular crystallization of the topaz. This conclusion,
which I have been led to draw, in another paper, from a series
of entirely different facts, will be still further confirmed by
the phaenomena of imbedded crystals, to which I shall have
to refer in another section.

2. Additional Observations on the Nature and Properties oj
the two New Fluids.

In re-examining the phaenomena exhibited by the two nev^
fluids, I have found no occasion to modify or to correct any
of the results contained in my former papers. In the cavities
which appear to contain only one fluid, namely, the dense
fluid, I have sometimes found a very small quantity of the
volatile fluid, which, with a slight rise of temperature, passes
into vapour, and prevents the apparent vacuity from disap-
pearing by the application of a strong heat. When there is

2 K2

500 Sir David Brewster on the Existence of Crystals

no volatile fluid present in such cavities, the vacuity is a real
one, and disappears entirely by the application of such a heat.
It* the heat is not instantly withdrawn on the disappearance of
the vacuity, the crystal never fails to burst with great violence.

In some specimens of Brazil topaz I have found cavities
with two fluids, and without any vacuity in the volatile fluid
at the ordinary temperature of an apartment. In such cases
I have generally produced a vacuity by the application of ice.
Had heat been applied, the crystals would have burst, as there
were no empty .spaces into which the fluids could expand.

When the cavities are flat, and have their faces perpendi-
cular to the axis of the crystal, or parallel to the planes of
easy cleavage, the application of heat does not burst the cry-
stal, but produces a very remarkable phaenomenon. The
cavity opens at its weakest point, and the fluid passes by starts,
through a succession of resting places, to another part of the
crystal where it finds the readiest exit. The fluid penetrates,
as it were, the solid gem, and the laminae which it has forced
asunder in its passage, again close into optical if not into me-
chanical contact. If the heat is withdrawn when the first
minute drop has passed, the laminae unite, and we can
discharge the rest of the fluid whenever we please till the
cavity is exhausted. This phaenomenon is represented in
Plate III. fig. 1, where ABCD is a shallow cavity in a plate
of topaz MN, and EF another cavity, which has been emptied
of its fluid contents by reaching the surface at N, where it had
been broken through. Upon looking at the cavity ABCD
when slightly heated, I observed dark portions of fluid rushing
from its sharp termination at D through the cavity at a, and
then reappearing at b and c, and then passing into the empty
cavity EF. The small lakes, as we may call them, at «, b
and c, disappeared entirely when the discharged portions of
fluid had passed, and reappeared with a change of form and
size when the operation was repeated.

In a specimen of topaz possessed by Major Playfair, and
seen by many individuals, a white ball passed from one cavity
to the edge of the specimen, as if projected from a mortar ;
but by the application of too strong a heat it was shattered in
pieces.

In my first paper of 182:3*, I have described and figured
a phaenomenon of an analogous kind ; but as it appeared un-
expectedly, and was instantly followed by the explosion of the
crystal, I could neither observe it accurately, nor confirm what
I did observe, by a repetition of the experiment. I have,
therefore, some satisfaction in describing a similar phaeno-
* Edinburgh Transactions, vol. x. p. 11, plate 1. fig. 5, 6.

in the Cavities of Minerals. 501

menon, seen frequently, and under more favourable circum-
stances, not only from its intrinsic interest, but because a
distinguished philosopher had treated with an air of incredu-
lity an observation which I had made of a similar kind.
There can be no higher testimony to the novelty and import-
ance of a scientific fact, than when a competent judge raises it
to the supernatural.

I come now to describe a property of the dense fluid, so
new and remarkable that it cannot fail to excite the attention
of chemists. This fluid occupies the whole of a large cavity
ABCDE, fig. 2, with the exception of a bubble at A, which
must be either a vacuum, as it is in all cavities containing only
this fluid, or a bubble of the expansible fluid, or the vapour
of the dense fluid, or some gaseous body. It cannot be a
vacuum, because it expands with heat, in place of being filled
up by the expansion of the fluid. It cannot be the expansible
fluid, because cold would contract it, and produce a vacuity.
It cannot be the vapour of the expansible fluid, because there
is no expansible fluid to throw it oif, and it has not the optical
properties of its vapour. It cannot be the vapour of the fluid
in the cavity, for it does not disappear by the application of
cold, and does not become a vacuit}', which fills up by the
expansion of the fluid. It is therefore an independent gas,
which exhibits the following phaenomena.

When heat is applied, the bubble A expands, not by the
degradation of its circular margin passing into vapour, as in
the vapour cavities described in a former paper, but by the
rapid enlargement of its area. When it attains a certain size,
it throws off a secondary bubble B, which passes over a sort
of ridge or weir inno^ in the bottom of the cavity, and settles
at B. If the heat is continued, these two bubbles increase in
size ; but it was instantly withdrawn when B had begun to
swell. As the topaz began to cool, both the bubbles A and
B quickly contracted. The primary bubble A returned gra-
dually to its original condition, and B, when reduced to a
single speck, would have disappeared, had the cooling not
been stopped. This speck swelled again by the application of
heat, and so did the bubble A. When the speck at B was
allowed to vanish, which it did on the spot which the bubble
occupied, the fresh application of heat did not revive it at that
spot, but merely expanded the primary bubble A, which
again threw off" a secondary bubble B, which exhibited by
heat and cold the same phaenomena as before. These ex-
periments I repeated many times with the same result. It will
naturally be asked, what was the condition of the fluid itself
wliich has the properly of expanding by heat ; and what be-

502 Sir David Brewster on the Existence of Crystals

came of it while a part of the space which it occupied was
appropriated by the bubble B, and the addition to the bubble
A? An accidental circumstance enables me to answer this
question, which would have been otherwise a very perplexing
one. Having applied too strong a heat to the specimen, the
bubble A threw off beside B two or three smaller ones, which
moved along the upper edge AE. My attention having been
thus directed to this part of the specimen, I was surprised to
observe a great number of capillary lines or pipes PQ, rising
from the edge AE of the cavity, and into which the fluid was
forcing itself, oscillating in these minute tubes like the mercury
in a barometer, and sometimes splitting the laminae between
them. The force of cohesion, thus overcome by the expan-
sive efforts of the fluid, predominated over the capillary attrac-
tion of the tubes and surfaces, and pressed back all the fluid
into the cavity, when the body of fluid had contracted in
cooling.

If we now consider the body which occupies the vacuity A
as a gas, and, consequently, the other bubble B as the same,
it follows that the whole of the gas in B was absorbed by the
fluid while cooling, and again given out by an increase of tem-
perature. The gas, when in the act of being discharged, took
its course to the locality of the speck at B, and to the bubble
A ; but to the bubble A alone when the speck had disap-
peared.

Upon repeating these observations the cavity burst ; and I
have now before me its two halves, forming its upper and its
under surface. The portion of" the cavity at A has the same
depth as the portion below mno, all the restof the cavity being
much shallower. There was a fine doubly refracting crystal
at MN, which polarized the blue of the second order; and its
outline is still left on the cavity. There was a sort of crystal-
line powder disseminated round MN to a considerable di-
stance, and the roof of the bubble B, when the roof of the
cavity was entire, was always mottled with this powder.

In a former paper, I have distinguished vapour cavities
from common cavities, by the manner in which the vacuity in
the expansible fluid disappears. In the one case, the vacuity
gradually enlarges by the degradation, as it were, of its mar-
gin, as the fluid passes into vapour ; in the other, the vacuity
gradually diminishes till it disappears. I have since found
cavities of an intermediate character, in which the vacuity, on
the first application of heat, diminishes, and then, when it has
contracted to a certain size, it begins to expand ; and its
margin becoming thinner and thinner, it finally passes into
vapour.

in the Cavities of Minerals. 503

3. On the Form and Position of Crystals in the Cavities of

Topaz.

In a former paper I have described a moveable group of
crystals of carbonate of lime, which I discovered in a cavity
in quartz from Quebec, containing a fluid with the properties
of water. The crystals to which I am about to call attention
are of a very different kind, and possess a very different kind
of interest.

The crystals which occupy the fluid cavities of topaz are
either fixed or moveable. Some of the fixed crystals are often
beautifully crystallized. They have their axes of double re-
fraction coincident with those of the crystal, and, as I have
ascertained by the examination of exploded cavities, they ac-
tually form part of the solid topaz, though they exist in the
fluid cavity. One or two of these are shown in fig. 4, plate
19, of my paper of 1826*, and they may be distinguished
by their attachment to the sides of the cavity. In the same
figure, as well as in figs, 10, 13, 20, and 21 of my paper of
1823tj I have drawn others which I then believed to be fixed,
but which I have no doubt are moveable, and produced from
one or other of the new fluids.

In re-examining my specimens of topaz, I have been sur-
prised at the great number of cavities which contain crystals.
In some there are only one; in very many there are two,
three, and four; and in a great number of specimens the
cavity is so ci*ammed with them, like a purSe full of money,
that the circular vacuity has not room to take its natural shape,
and often can scarcely be recognised, in its broken-down con-
dition, among the jostling crystals.

The crystals of which I am treating are sometimes found in
the volatile, and sometimes in the dense fluid, but chiefly in
the latter. They are often found in an amorphous state in the
narrow necks and narrow extremities of cavities, positions in
which they remain fixed while they continue solid ; and some-
times regularly formed crystals remain fixed between the pris-
matic edges of cavities, in consequence of having either fallen
into that position, or of having been formed there.

The crystals in topaz cavities are, in genei'al, beautifully
crystallized, and have a great variety of forms. I have ob-
served the following : —

1. The tetrahedron.

2. The cube.

3. The cube, truncated on its edges and angles.

4. The rhombohedron.

• Edinburgh Transactions, vol. x. f Ibid, plates 1 and 2.

' 504 Sir David Brewster on the Existence of Crystals

5. The prism, with plain and pyramidal summits.

6. The flat octohedron, truncated on its edges and angles,

7. Rhomboidal plates.

8. Hexagonal plates.

9. Long rectangular plates.

Besides these, there are amorphous crystals and crystallized
masses of various characters.

4. On the Physical Properties of the Crystals in Topaz Cavities.

Although it vk^ould be desirable to submit these crystals, as
well as the fluids which contain them, to chemical analysis,
yet the task is loo difficult to be accomplished in the present
state of chemical science. I must therefore limit my obser-
vations to such of the physical properties of these crystals as
can be rendered visible to the eye.

When I first applied heat to the crystals under considera-
tion, I employed a very fine specimen, with large and nume-
rous crystallized cavities, ofaprismatical form, containing both
the new fluids. In this specimen there wei'e seven cavities
unlike all the rest, and each of them containing a single cry-
stal, and apparently but one fluid, namely, the dense one. The
cavities were exceedingly flat, and irregular in their shape,
and very unlike one another. Upon applying the heat of only
a lighted paper match beneath the plate of glass on which the
specimen lay, I was surprised to see the crystals gradually
lose their angles, and then slowly melt, till not a trace of them
was visible. In this state one of the cavities had the appear-
ance shown in fig. 3, where V was the vacuity, and v^ x/ other
two bubbles, one of which v soon joined the principal one V.
In all the other six cavities the crystals were speedily repro-
duced, always at the point where they disappeai'ed, provided
a small speck remained unmelted ; but otherwise in diflferent
parts of the cavity. In the cavity AB, however, fig. 3, the
crystal was very long in appearing. In the course of an hour,
however, a fasciculus of minute crystals appeared in the centre
of the vacuity, as in fig. 4, and to them the principal crystal
attached itself, as in fig. 5, which exhibits a perfect rhomboidal
plate, truncated on its obtuse angles. The elliptical vacuity
was pressed into the shape of a heart : and, by the application
of ice, I succeeded in precipitating the vapour of the expan-
sible fluid, which existed in a very minute quantity in all the
seven cavities. The expansible fluid is shown between the two
heart-shaped outlines in the figure, and I repeatedly threw it
into vapour, and reduced that vapour to a fluid state. The
phaenomenon now described, of the melting of the crystals,
and their subsequent recrystallization, I have shown to various

in the Cavities of Minerals. 505

persons ; and it is very remarkable that they generally reap-
pear in this specimen of the same form, though with consi-
derable modifications.

Upon applying heat to other cavities containing several
crystals, I obtained very different results. Some of them
melted easily, others with greater difficulty ; and some were
not in the slightest degree affected by the most powerful heat
I could apply. When the crystals melted easily, they were
as quickly reproduced; sometimes reappearing more perfectly
formed than before, but frequently running into amorphous
and granular crystallizations.

In some specimens of topaz all the crystals in the cavities
refuse to melt with heat, and seem not to suffer the slightest
change in their form. Hence we are entitled to conclude,
that the crystals possessing such different properties must be
different substances; and this conclusion is amply confirmed
by an examination of their optical properties.

In making this examination, I used a polarizing microscope,
so constructed that the plane, passing through the optical axis
of the topaz, could be readily placed either parallel or per-
pendicular to the plane of primitive polarization. In this case
the field of the microscope is wholly obscure, in so far as the
depolarizing action of the plate of topaz is concerned ; but if
there is any crystal in the topaz, either imbedded in its mass,
or included in its cavities, that crystal will exhibit its doubly
refracting structure, if it has any, by its depolarizing action.
It may, indeed, happen, — and it does happen, — that the plane
passing through their optical axes coincides, either accurately,
or so nearly, with that of the topaz, that its depolarizing action
is a minimum ; but an experienced observer will have no dif-
ficulty in distinguishing this want of depolarization by position,
from the v/ant of it by structure.

When the specimen of topaz is rich in cavities full of cry-
stals, the display of luminous and coloured crystalline forms in
the dark field of the microscope, indicating, too, the impri-
sonment of fluids, and the condensation of gases before vege-
table or animal life had visited our primaeval globe, was as
interesting to the imagination and the judgement as it was
beautiful to the eye. Having had the privilege of being the
first to see it, I felt the full influence of the sight ; and 1 have
again and again contemplated it with renewed wonder and
delight. When the cavities are so numerous as to mock cal-
culation, and so infinitely small as to yield no visible outline
to the highest powers, the bright twinkle of a crystalline atom
within them reveals to us their nature as well as their contents.

In the examination of the individual crystals, many interest-

506 Sir David Brewster on the Existence of Crystals

ing facts present themselves to our notice. The crystals of
the tessiilar class, which are modifications of the cube, are
very numerous, and have no action upon polarized light.
Many of them melt easily, while others refuse to yield to the
action of heat ; and hence there must be two different sub-
stances in the cavities which assume the same shape. In like
manner, some of the doubly refracting crystals melt readily,
others with very great difficulty, and others not at all ; so that
there must be three different substances, which belong to the
classes of forms that give double refraction ; a conclusion
which is confirmed by the different secondary forms which I
have already enumerated.

I have seldom found any crystals in these cavities which
depolarize white light, or the highest order of colours. I have
found some that depolarize ybz/r orders of colours ; and when
the crystal which does this is a flat hexagonal plate, it is highly
interesting to see it pass through all the tints which these
orders include, while slowly melting, and again reproducing
them during its recrystallization.

In a cavity which was so placed as to be entirely black from
the total reflexion of the light which fell upon it, I observed
three iiohite openings, a, b, c, of a crystalline form (see fig. 6).
These appeared to be fixed crystals, or rather parts of the
topaz, surrounded by a cavity. I found, however, that the
hexagonal one C depolarized white light, while the rest had
no action upon polarized light. Upon applying heat, the
crystal c melted, and took up a position at e, fig. 1 5, in a nar-
rower part of the cavity, where it remains of an irregular
form, having been repeatedly melted and recrystallized. Upon
turning the cavity into a position where it became transparent,
I found that there was no fluid whatever in the cavity ; so that
we have here an example of a crystal melting and recrystal-
lizing without having been dissolved in one of the fluids. From
the irregular state of the laminae close to this cavity, there is
every appearance of the fluids having escaped from one of its
extremities.

In the course of these observations, I observed a phaeno-
menon, produced by heat, of the most novel and surprising
kind, and one which 1 feel myself utterly unable to explain.
It piesented itself when I was studying the very interesting
collection of crystals in the cavity AB, fig. 8. This cavity is
filled with the dense fluid, in which there is a vacuity V : the
fluid swells to sucli a degree with heat as to diminish very
perceptibly the size of this vacuity ; and as I can find no trace
of any portion of the volatile fluid, I have no doubt that this
vacuity would disappear by an increased degree of heat. The

in the Cavities of Minerals. 507

fear, however, of bursting so rare and interesting a cavity,
has prevented me from making this experiment. The cavity
contains a great number of crystals of different forms, not one
of which melts with heat, and almost all of which possess
double refraction. When I first submitted this cavity to the
microscope, there were Jive small crystals lying between D and
the vacuity V; one a flat prism, another a hexagonal plate, a
third amorphous, and a fourth and fifth two irregular halves
of a hexagon. Upon the first application of heat, one or two
of these crystals leapt from their resting place, and darted to
the opposite side of the cavity. In a iew seconds the others
quitted their places one after another, performing the most
rapid and extraordinary rotations. One crystal joined an-
other, and, at last, four of them thus united revolved with such
rapidity as completely to efface their respective shapes. They
then separated on the withdrawal of the heat, and took the
position which their gravity assigned them. On another oc-
casion, a long flat prism performed the same rotation round
its middle point; and I have repeated the experiment so often,
in showing it to others, that the small crystals have been driven
between the inclined edges of the cavity, from which I cannot
extricate them. I have succeeded, however, in conducting a
fine octohedral crystal, truncated on its edges and angles, into
the arena at D, where I have just seen it perform its rotation,
as indicated by the concentric circles on the right-hand of D.

In seeking for the cause of so extraordinary a phaenomenon,
we are reminded of the rotations of camphor and other vola-
tile substances ; but in this case no gas or matter of any kind
could be thrown off without becoming visible in the fluid. The
pyro-electricity of topaz next suggests itself as a moving power;
but though it might produce attractions and repulsions, we
cannot see how it could turn a crystal upon its axis. The
experiments of Libri and Fresnel, on the repulsions which
heated bodies exert upon each other at sensible distances,
afford us as little aid. They may enable us to account for
the mere displacement of the crystals by the application of
heal, or for their sudden start from their places of rest, but
they do not supply us with a force fitted to give and to sustain
a rapid rotatory movement.

I have already had occasion to state, that the cavities often
burst when too much heat is applied to the specimen. This
generally takes place by a separation of the laminae, which
fly off in splinters; but when the burst cavity is large and
insulated, a piece of the solid crystal is scooped out on its
weakest side. Sometimes a great number of cavities explode
at the same time, and when they are small, or exist in a part

508 Sir David Brewster 07i the Existence of Crystals

of the crystal where there are no large ones, the explosive
force is not strong enough to separate the laminae. The fluid
is merely driven between the laminae to a small distance around
the cavity, and shows itself as a dark brown powdery matter,
encircling the cavity as the burr of a comet does its nucleus.
When the cohesion of the laminae is great, it resists the ex-
plosive force over a large cavity, and the contents of the cavity
are thrown to a considerable distance around it, and remains
between the laminas, either as a sort of powder, or as a con-
geries of minute crystals, which are sometimes large enough
to show their depolarizing action. When the laminae sepa-
rate, we find this crystalline matter either fluid or indurated ;
exhibiting, when fluid, the extraordinary properties described
in my former papers. If we breathe upon the indurated
matter it becomes fluid, recrystallizes in new spicule and cry-
stals; and, on several occasions, I have found fine examples
of circular crystallization.

After the explosion of cavities containing only the dense
fluid, I have been surprised to find, and that in large cavities,
that no trace of matter was left upon the sides of the cavity or
around it. Whether this arose, as the fact seems to indicate,
from the dense fluid being a condensed gas, or from some
other cause, it will require new experiments to determine.

In a very remarkable specimen, in which the cleavage plane
passed through a great number of large flat cavities, the brown
matter has been lodged near to the edges of each cavity, and
marks them out even to the unassisted eye. These cavities
were filled almost solely with the volatile fluid ; and since the
faces of the cavities are corroded as if by the action of a sol-
vent, developing crystalline forms, there is reason to think that
the fluid has exercised this action, and that the phasnomenon
is analogous to that external action, on the faces of hundreds
of Brazil topazes in my possession, which I have described in
the Cambridge Transactions*, and the singular optical figure
formed by which, I have represented in a late volume of the
Transactions of this Society f.

The only chemical experiment on the contents of these
cavities, which I have had occasion recently to make, is per-
haps worth reporting. One angle of a cavity was blown off"
by its explosion, and though the fluids escaped, a pretty large
prismatic crystal remained within the cavity. I introduced
water and alcohol successively into the cavity, and raised them
to a considerable heat ; but they had no effect in dissolving
the crystal.

* Vol. ii. plate 1. fig. 16.

t Edinburgh Transactions, vol. xiv. plate 10. figs, 1, 2.

in the Cavities of Minerals. 509

5. On Solid Crystals and Crystalline Masses imbedded in
Topaz.

Among the new phaenomena which this section embraces,
there is at least one intimately connected with the subject of
the fluid cavities. How far the other phaenomena may have
any such connexion, it remains to be seen.

The imbedded crystals to which I refer, presented them-
selves to me while the specimens which contain them were
exposed to polarized light. Mineralogists have been long
familiar with the beautiful crystals of titanium, imbedded in
quartz, and I have found the same mineral imbedded under
still more interesting circumstances in the Brazilian amethysts.

In topaz, however, the imbedded crystals have never been
noticed, and I have fortunately obtained specimens, in which
they are displayed with singular beauty. Their axes of double
refraction are not coincident with those of the topaz; and
hence they are seen in the obscure field of the microscope
splendent with all the colours of polarized light. These cry-
stals are equally transparent with the topaz, with a ^evt slight
exceptions. They sometimes polarize five or six orders of
colours ; and, in general, they have very beautiful crystalline
forms, which can be seen by the microscope in common light.
In some cases they are mere crystalline masses, often of a
reniform shape, but still with regular axes of double refraction.

In some specimens of Brazil topaz, the crystals occur in
branches or groups of singular beauty, consisting of prisms and
hexagonal plates, connected apparently by filaments of some
opake matter.

I have occasionally met with another interesting variety
of them, which have no visible outline by common light, and
which could never have been detected but by the polarizing
microscope. In one of these cases, the crystalline mass, which
is nearly spherical, lies in a crowded group of small fluid ca-
vities, none of which enters its mass ; a complete proof that
the cavities were formed in the soft mass of topaz, when it en-
circled the indurated crystal.

Along with these interesting phaenomena, another occasion-
ally occurs, which may still require a further examination. I
have observed apparent doubly refracting crystals, which differ
in some essential points from those which have been described.
They depolarize a uniform, or nearly a uniform tint, notwith-
standing the different thicknesses through which the polarized
light passes ; and that tint is less brilliant than in the real im-
bedded crystals. I conceive, therefore, that they are crystal-
lized cavities, having their inner surfaces coated with a doubly

510 Observations on Chloric Acid and the Chlorates,

refracting crust. This is^ in itself, a very natural supposition,
seeing that the fluid may have discharged its gaseous portion,
and left behind it the matters which it held in solution. The
cavities however, of this kind, which I have described in a
former paper, have no depolarizing action ; and I find that
those now under consideration have regular axes of double
refraction. Hence the matter which covers them must be a
regular crystalline shell, with optical and crystallographic
axes — a phaenomenon which has no parallel in mineralogy.

St. Leonard's College, St. Andrew's,
February 15, 1845.

LXXIV. Observations on Chloric Acid and the Chlorates.
By Lewis Thompson.

To the Editors of the Philosophical Magazine and Journal.

Gentlemen,
A N easy and oeconomical mode of preparing chloric acid
-^^*- and some of the chlorates has not been described in any
chemical work that I am aware of: the following will be fonnd
to answer extremely well.

Dissolve in two separate portions of boiling water one atom
(122-81) of chlorate of potash, and one atom (168-34) of bi-
tartrate of ammonia; mix the two solutions together, and set
the whole aside in order that the bitartrate of potash may
crystallize; then mix the clear solution with an equal bulk of
alcohol, and filter or pour off the alcoholic solution of chlorate
of ammonia, which must now be boiled in a flask or other
narrow-necked vessel, with an excess of recently-precipitated
carbonate of baryta, until the ammonia is expelled, water being
occasionally added ; then filter the fluid, evaporate, and cry-
stallize. In dissolving the chlorate of potash and bitartrate of
ammonia, as little water must be used as possible.

The chlorates of strontla and lime may be prepared in a
similar manner; and the metallic chlorates are easily prepared
by decomposing the chlorate of baryta by means of asulphatef
of the base required.

Chloric acid is best obtained by dissolving a given weight
of chlorate of baryta, and adding no more sulphuric acid than
is sufficient to combine with the base ; several hours or eveif
days, however, appear necessary to effect this decomposition
in the cold ; after which the whole may be filtered and care-
fully evaporated at a low heat. When sulphuric acid is added
to a solution of the chlorate of baryta, as long as it gives ^
precipitate, I have always found an excess of it in the chloric
acid.

Sir W. Rowan Hamilton on Qtiaternions* 511

The bitartrate of ammonia may be readily made by dissol-
ving tartaric acid in water, saturating one-lialf of the solution
with ammonia or its carbonate, and adding to this the remain-
ing half of the liquid tartaric acid; the bitartrate of ammonia
immediately precipitates.

For pyrotechnical purposes, the chlorates of baryta, stron-
tia, lead, &c. may be made without alcohol. With combus-
tibles containing hydrogen, the chlorate of baryta produces a
green flame of surpassing brilliancy ; and the chlorate of
strontia, although somewhat deliquescent, is much superior as
a crimson to the nitrate of that earth.
I am, Gentlemen,

Your most obedient Servant,
Byker Bar, Newcastle-on-Tyne, Lewis Thompson.

October 14, 1847.

LXXV. On Qiiaternions ; or on a New Si/stem of Imag manes
in Algebra. By Sir William Rowan Hamilton, LL.D.^
V. P.R.I. A., F.R.A.S., Corresponding Member of the In sti'
tute of France^ S^c., Ajidrews' Professor of Astronomy in the
University of Dublin j a?id Royal Astronomer of Ireland.
■[Continued from p. 293.]
51. TT has been shown* that if the two symbols «, x denote
J- certain constant vectors, perpendicular to the two
cyclic planes of an ellipsoid, and if v, t denote two other and
variable vectors, of which the former is normal to the ellipsoid
at any proposed point upon its surface, while the latter is tan-
gential to a line of curvature at that poin^, then the directions
of these four vectors », x, v, t are so related to each other as to
satisfy the condition f

S . VT»Tx = 0 (49.), article 47 ;

S being the characteristic of the operation of taking the scalar
part of a quaternion. And because the two latter of these
four directions, namely the directions of the normal and tan-
gential vectors v and t, are always perpendicular to each other,
this additional equation has been seen to hold good :

S.vT=0 (36.), article 45.
Retaining the same significations of the symbols, and carrying
forward for convenience the recent numbering of the formulae,

* See the Philosopliical Magazine for October 1847; or Proceedings of
the Royal Irish Academy for July 1846.

t Inadvertently transcribed as S . vtikt^O, towards the end of the last
communication to this Magazine: but correctly printed in the formula
(49.) here referred to.

512 Sir W. Rowan Hamilton on Quaierm'ons.

it is now proposed to point out some of the modes of combining,
transforming, and interpreting the system of these two equa-
tions, consistently with the principles and rules of the Calculus
of Quaternions, from which the equations themselves have been
derived.

52. Whatever two vectors may be denoted by t and t, the
ternary product tit is always a vector forrn^ because (by
article 20) its scalar part is zero ; and on the other hand the
square t^ is a pure scalar : therefore we may always write

T*T = /AT% Ti = |XT, (52.)

where jw. is a new vector, expressible in terms of » and t as
follows :

H* = T<T-'; (53.)

so that it is, in general, by the principles of articles 40, 41,
42, 43, the reflexion of the vector « with respect to the vector
T, and that thus the direction of t is exactly intermediate be-
tween the directions of < and jw,. In the present question, this
new vector ]«,, defined by the equation (53.), may therefore
represent the reflexion of the first cyclic normal «, with re-
spect to any reflecting line which is parallel to the vector t,
which latter vector is tangential to one of the curves of cur-
vature on the ellipsoid. Substituting for m its value (52.), in
the lately cited equation (49.), and suppressing the scalar
factor T% we find this new equation :

S.v/*x=0; (54.)

which, in virtue of the general £'g'Mfl!//ow o/'co/'/awanVj/ assigned
in the 21st article (Phil. Mag. for July 1846), expresses that
the reflected vector jt*, the normal vector v, and the second
cyclic normal x, are parallel to one common plane. This result
gives already a characteristical geometric property of the lines
of curvature on an ellipsoid, from which the directions of those
curved lines, or of their tangents (t), can generally be assigned,
at any given point upon the surface, when the direction of the
normal (v) at that point, and those of the two cyclic normals
(« and x), are known. For it shows that if a straight linejw, be
found, in any plane parallel to the given lines v and x, such
that the bisector t of the angle between this line ju, and a line
parallel to the other given line » shall be perpendicular to the
given line v, then this bisecting line t will have the sought
direction of a tangent to a line of curvature. But it is pos-
sible to deduce a geometrical determination, or construction,
more simple and direct than this, by carrying the calculation
a little further.

Sir W. Rowan Hamilton o« Quaternions. 513

53. The equation (52.) gives

(|a + ,)T = T. + JT = V-10, .... {55.)

this last symbol V~'0 denoting generally any quaternion of
which the vector part vanishes; that is any pure scalar, or in
other words any real number, whether positive or negative or
null. Hence /«, + » and t denote, in the present question, two
coincident or parallel vectors, of which the directions are
either exactly similar or else exactly opposite to each other ;
since if they were inclined at any actual angle, whether acute
or right or obtuse, their product would be a quaternion, of
which the vector part would not be equal to zero. Accord-
ingly the expression (53.) gives this equation between tensors,
TjL. = T,; [56.)

so that the symbols jtx, and i denote here two equally long
straight lines ; and therefore one diagonal of the equilateral
parallelogram (or rhombus) which is constructed with those
lines for two adjacent sides bisects the angle between them.
But by the last article, this bisector has the direction of t (or
of — t) ; and by one of those fundamental principles of the
geometrical interpretation of symbols, which are common to
the calculus of quaternions and to several earlier and some
later systems, the symbol jm< + i denotes generally the interme-
diate diagonal of a parallelogram constructed with the lines
denoted by ju, and » for two adjacent sides : we might there-
fore in this way also have seen that the vector ]«, + « has, in
the present question, the direction of +t. This vector /* + j is
therefore perpendicular to v, and we have the equation

0 = S.v(/x, + i), or S.viu,= — S.v*. . . . (57.)

But by {56.), and by the general rule for the tensor of a pro-
duct (see art. 20), we have also

T.v/* = T.v«; (58.)

and in general (by art. 19), the square of the tensor of a qua-
ternion is equal to the square of the scalar part, minus the
square of the vector part of that quaternion ; or in symbols
(Phil. Mag., July 1846),

(TQ)2=(SQ)2-(VQ)2.

Hence the two quaternions vjw, and vt, since they have equal
tensors and opposite scalar parts, must have the squares of
their vector parts equal, and those vector parts themselves
must have their tensors equal to each other; that is, we may
write

(V.Vi*)2 = (V.v.)S TV. via = TV. VI : . . (59.)
Phil. Mag. S. 3. No. 2 1 1 . Suppl. Vol. 3 1 . 2 L

5H Sir W. Rowan Hamilton on Quaternions.

and may regard these two vector parts of these two quater-
nions, or of the products vjw- and v», as denoting two equally
long straight lines. Consequently the vector +vt, which has
the direction of the line represented by the pure vector pro-
duct v(|«. + »), or by the sum V.vjw, + V.v< of two equally long
vectors, has at the same time the direction of the sum of the
two corresponding versors of those vectors, or that of the sum
of their vector-uiiits ', so that we may write the equation

/vT=UV.vjct + UV.vj, (60.)

where U is (as in art. 19) the characteristic of the operation
of taking the versor of a quaternion, or of a vector; and t is
a scalar coefficient. Again, the equation 0=S.v/ax, (54.),
which expresses that the three vectors v, jix,, x are coplanar,
shows also that the two vectors V. vju, and V. vx are parallel to
each other, as being both perpendicular to that common plane
to which V, /x, and x are parallel ; hence we have the following
equation between two versors of vectors, or between two vec-
tor-units,

UV.Vi(;t=±UV.vx; ...._. (61.)
and therefore instead of the formula (60.) we may write

^T = v-iUV.v.±v-aiV.vx (62.)

In this expression for a vector touching a line of curvature, or
parallel to such a tangent, the two terms connected by the
sign + are easily seen to denote (on the principles of the
present calculus) two equally long vectors, in the directions
respectively of the projections of the two cyclic normals » and
X on a plane perpendicular to v; that is, on the tangent plane
to the ellipsoid at the proposed point, or on any plane parallel
thereto. If then we draw two straight lines through the point
of contact, bisecting the acute and obtuse angles which will in
general be formed at that point by the projections on the tan-
gent plane of two indefinite lines drawn through the same
point in the directions of the two cyclic normals, or in direc-
tions perpendicular to the two planes of circular section of the
surface, the two rectangular bisectors of angles, so obtained, "will
he the tangents to the two lines of curvature : which very simple
construction agrees perfectly with known geometrical results,
as will be more clearly seen, when it is slightly transformed as
follows.

54t. If we multiply either of the two tangential vectors t by
the normal vector v, the product of these two rectangular vec-
tors will be, by one of the fundamental and peculiar^ prlnci-

* See the author's Letter of October 17, 1843, to John T. Graves, Esq.,
printed in the Supplementary Number of the Philosophical Magazine for
December 1844: in which Letter, the three fundamental symbols i,J,i
were what it has been since proposed to name direction-units.

SirW. Rowan Hamilton on Quaternions, 515

pies of the calculus of quaternions, a third vector rectangular
to both ; we shall therefore only pass by this multiplication,
so far as directions are concerned, from one to the other of the
tangents of the two lines of curvature: consequently we may
omit the factor v~* in the second member of (62.), at least if
we change (for greater facility of comparison of the results
among themselves) the ambiguous sign + to its opposite.
We may also suppress the scalar coefficient i, if we only wish
to form an expression for a line t which shall have the required
direction of a tangent, without obliging the length of this line
T to take any previously chosen value. The formula for the
system of the two tangents to the two lines of curvature thus
takes the simplified form :

T=UV.v» + UV.vx; (63.)

in which the two terms connected by the sign + are two vec-
tor-units, in the respective directions of the traces of the two
cyclic planes upon the tangent plane. The tangents to the
two lines of curvature at any point of the surface of an ellipsoid
(and the same result holds good also for other surfaces of the
second order), are therefore parallel to the two rectangular
straight lines which bisect the angles between those traces ; or
they are themselves the bisectors of the angles made at the
point of contact by the traces of planes parallel to the two
cyclic planes. The discovery of this remarkable geometrical
theorem appears to be due to M. Chasles. It is only brought
forward here for the sake of the process by which it has been
above deduced (and by which the writer was in fact led to
perceive the theorem before he was aware that it was already
known), through an application of the method of quaternions,
and as a corollary from the geometrical construction of the
ellipsoid itself to which that method conducted him*. For
that new geometrical constructioii has been shown (in a recent
Number of this Magazine) to admit of being easily retranslated
into that quaternion form of the equation f of the ellipsoid,
namely

T(ip + px)=x2— ,2^ equation (9.), art. (38.),

as an interpretation of which equation it had been assigned by
the present writer; and then a general method for investiga-
ting by quaternions the directions of the lines of curvature on
any curved surface whatever, conducts, as has been shown (in

* See the Numbers of the Philosophical Magazine for June, September,
and October 1847; or the Proceedings of the Royal Irish Academy for
July 1846. ,

t Another very simple construction, derived from the same quaternion
equation, and serving to generate, by a moving sphere, a system of two
reciprocal ellipsoids, will be given in an early Number of this Magazine,

2 L2

516 Sir W. Rowan Hamilton on Quaternions.

articles46and47),tothe equation of those lines for the ellipsoid,

from which, when combined with the general equation S.vtisO,
the formula (63.) has been deduced, and geometrically inter-
preted as above.

55. Another mode of investigating generally the directions of
those tangential vectors t which satisfy the system of the two
conditions in art. 51, may be derived from observing that
those conditions fail to distinguish one such tangential vector
from another in each of the two cases where the variable nor-
mal V coincides in direction with either of the two fixed cyclic
normals, » and x ; that is, at the four umbilical points of the
ellipsoid, as might have been expected from the known pro-
perties of that surface. In fact if we suppose

v = mi, S.«T = 0, (64.)

where w is a scalar coefficient, that is if we attend to either of
those two opposite tmibilics at which v has the direction of »,
we find the value

vt»tx = w(»t)^x, (65.)

which is here a vector-form, because by (64.) the product »r
denotes in this case a jnire vector, so that its square {like that
of every other vector in this theory) 'will be a negative scalar, by
one of the fundamental and ■peculiar'^ principles of the present
calculus ; the scalar part of the product vtjtx therefore vanishes,
or the condition (49.) is satisfied by the suppositions (64.).
Again, if we suppose

v = m^)i, ....... (66')

m' being another scalar coefficient, that is if we consider either
of those two other opposite umbilics at which v has the direc-
tion of X, we are conducted to this other expression,

vr»Tx = »i'xT<Tx; (67.)

which also is a vector-form, by the principles of the 20th
article. In this manner we may be led to see that if in general
we decompose, by orthogonal projections, each of the two
cyclic normals, » and x, into two partial or component vectors,
i', j", and x', k", of which •' and x' shall be tangential to the
surface, or perpendicular to the variable normal v, but i" and
x" parallel to that normal, in such a manner as to satisfy the
two sets of equations,

, = ,' + ."; S.i'v = 0; V.,"v=0; \ ^ ^ ^gg^^
x = x' + h"; S.x'v=0; V.x"v = 0;J

* See the author's letter of October 17, 1843, already cited in a note to
article 64.

';.} (70.)

Sir W. Rowan Hamilton on Qiiaiernions. 517

then, on substituting these values for i and x in the condition
(4-9.), or in the equation 0=S.vt<tx, the terms involving i"
and x" will vanish of themselves, and the equation to be satis-
fied will become

0=S. vtj'tx'; (69.)

which is thus far a simplified form of the equation (49.), that
three of the four directions to be compared (namely tho!>e of
*', x', and t) are now parallel to one common plane, namely to
the plane which touches the ellipsoid at the proposed point,
and to which the fourth direction (that of v) is perpendicular.
Decomposing the two quaternion products, t*' and tx', into
their respective scalar and vector parts, by the general formulae,

Tl'=S.Tl' + V.T*';
TX'=S.TX' + V.TX'

and observing that the vectors V. n' and V. tx' both represent
lines parallel to v, because v is perpendicular to the common
plane of t, »', x'; so that the three following binary pi'oducts,
V.Tj'.V. tx', vV. t«', vV. tx', are in the present question scalars ;
we find that we may write

S.vt«'tx' = vS.t/.V.tx' + vV.t.'.S.tx'. . . (71.)

Hence the equation (69.) or (49.) reduces itself, after being
multiplied by v~', to the form

S.t/.V.tx' + V.t.'.S.tx'=0; .... (72.)

which gives, in general, by the rules of the present calculus,

V.t'T_V.Tx'

sT?7-"sr;v' ^^^'>

and by another transformation,

V.i't-1 _ V.x't-' , ^

[s:?F^-~s:^?^' (^^•)

which may perhaps be not inconveniently written also thus :
V i' _ V x'
S"'f"'"S*7' C^^')

in using which abridged notation, we must be careful to re-

V

member, respecting the characteristic ^, of which the effect

is to form or to denote the quotient of the vector 'part divided
by the scalar part of any quaternion expression to which it is
prefixed, that this Jiew characteristic of operation is not (like S
and V themselves) distributive relatively to the operand. The
vector denoted by the first member of (74.) or of (75.) is a line
perpendicular to the plane of *' and t, that is to the tangent

618 Sir W. Rowan Hamilton on Quaternions,

plane of the ellipsoid ; and its length is the trigonometric
tangent of the angle of rotation in that plane from the direction
of the line t to that of the line »'; while a similar interpretation
applies to the second member of either of the same two equa-
tions, the sign — in that second member signifying here that
the two equally long angular motions, or rotations, from t to
i', and from t to x', are performed in opposite directions.
Thus the vector t, which touches a line of curvature, coincides
in direction with the bisector of the angle in the tangent plane
between the projections, i' and x', of the cyclic normals there-
upon ; or with that other line, at right angles to this last bi-
sector, which bisects in like manner the other and supplemen-
tary angle in the same tangent plane, between the directions
of »' and — x' : since x' may be changed to — x', without alter-
ing essentially any one of the four last equations between t, *', x'.
Those two rectangular and known directions of the tangents
to the lines of curvature at any point of an ellipsoid, which
were obtained by the process of article 53^ are therefore ob-
tained also by the process of the present article ; which con-
ducts, by the help of the geometrical reasoning above indicated,
to the following expression for the system of those two tan-
gents T, as the symbolical solution (in the language of the pre-
sent calculus) of any one of the four last equations (72.),. .(75.):

t=j!'(U.'±Ux'); (76.)

where t' is a scalar coefficient.

The agreement of this symbolical result with that marked
(62.) may be made evident by observing that the equations
(68.) give

,' = v-iV.v»; x' = v-iV.vx; . .-. . (77.)

so that if we establish, as we may, the relation

^^'=(Tv)-S (78.)

between the arbitrary scalar coefficients t and t', which enter
into the formulae (62.) and (76.), those formulae will coincide
with each other. And to show, without introducing geome-
trical considerations^ that (for example) the form (73.) of the
recent condition relatively to t is symbolically satisfied by the
expression (76.), we may remark that this expression, when
operated upon according to the general rules of this calculus,
gives

Tx'.V./t=±/'V..'x'; Tx'.S.i'T=^'(-T.i'x'±S..'x');'l ,^^ ^
Ti'.V.Tx'=/'V.i'x'; Ti'.S.Tx'=^'(S..'x' + T.i'x'); J ''

and that therefore the two members of (73.) do in fact receive.

On the Chemical History of Gun-Cotton and Xyloidine. 519

in virtue of (76.), one common symbolical ralue, namely one
or other of the two which are included in the ambiguous form

v.»y

S.»'x' + T..'x'' ^^^''

respecting which form it may not be useless to remark that
the product of its two values is unity.
[To be continued.]

LXXVI. Contributions to the Chemical History ofGun-Cotton
and Xyloidine. By Mr. John Hall Gladstone, of Uni-
versity College, London^.

A T the commencement of the present year, having perceived
-^^ that considerable doubt rested on the ultimate composi-
tion of gun-cotton, I undertook a series of experiments with
a view to ascertain it, if possible ; and during my investiga-
tion my attention was drawn to various papers that appeared
on the subject, where I found contradictory accounts, not
only of the results of analysis, but also of the action of va-
rious reputed solvents. The experiments detailed below,
although they are far from exhausting the subject, may serve
to explain some of these anomalies, and to point out a few
facts, which, as far as I have been able to learn, have not
been hitherto noticed.

The cotton employed was that used by jewellers, well-
carded, perfectly white, and free from imperfections. An
analysis of the substance by combustion with oxide of copper
in a stream of oxygen yielded the following results : —

Cotton employed 3*16 grs.

Carbonic acid produced . , . 5*14 ...

Water produced 2*06 ...

These proportions are, —

Carbon . . . 44-37
Hydrogen . . 7*24
Oxygen . . . 48-39
100-00
Lignine calculated from the formula C24 H20 Ogo •' —
Carbon . . . 44-44
Hydrogen . . 6-17
Oxygen . . . 49-39
100-00
The excess of hydrogen doubtless arises from moisture
absorbed by the oxide of copper during the unavoidable delay
in mixing it with the cotton.

* Communicated by the Chemical Society; having been read June 7>
1847.

520 Mr. Gladstone on the

This cotton, which may be considered as pure lignine, was
steeped until thoroughly wetted in a mixture of nitric acid of
spec. grav. 1*502, and nearly an equal bulk of strong sul-
phuric acid, then well-washed with water, and dried at a
temperature not exceeding 212°. In one instance 38*38 grs.
of cotton became 66*84 grs,, being an increase'of 28*46 grs.,
or 74*15 per cent. In a second experiment 59*3 grs. of cot-
ton gave an increase of 43*7 gi^s., or 73*7 per cent. The gun-
cotton, or pyroxyline, thus produced resembled the original
cotton in physical properties very closely, and exploded at
about 370°, producing no smoke and leaving no residue.

The action of various solvents and reagents upon this sub-
stance was found to be as follows : — It is absolutely insoluble
in pure water, and nearly so in strong alcohol, aether, m hether
hydrated or anhydrous, and in a mixture of aether with y'^th
part of alcohol ; but acetic aether instantly destroys its fibre,
and dissolves it in large quantity. The solution yields on
spontaneous evaporation a white powder of the same weight
as the original pyroxyline, but I have found it very difficult
to drive off the last traces of the solvent. The action of sul-
phuric acid upon it differs from that exerted upon unaltered
cotton ; for, while the latter is instantly dissolved by the
strong acid, and charred upon a slight elevation of tempera-
ture, pyroxyline dissolves with difficulty unless the acid be
warmed, evolving at the same time nitric oxide and other
gases, and not being charred even upon boiling. With the
aid of heat it dissolves immediately in a solution of potash.
By means of these three last-mentioned tests I was able to
prove the absence of any unaltered cotton in the product
under examination. The action of other reagents upon gun-
cotton was not so decided ; it was dissolved, but not without
long boiling, by ammonia, the alkaline carbonates, hydro-
chloric acid, acetic acid, both glacial and dilute, and weak
sulphuric acid. These solutions, as well as the two preceding,
contained nitric acid; nothing could be precipitated from
them by dilution or neutralization; and when evaporated
they yielded only a dark brown amorphous matter. It is
evident that none of these reagents restore the lignine in its
original condition; and they do not afford any means of
ascertaining whether the compound contains the elements of
nitric or hyponitric acid.

As there exists a great discrepancy in the accounts given
of the increase of weight in making gun-cotton, I examined
whether the length of time it was immersed in the acid liquor,
or the proportions of the acids employed, were the cause. The
length of immersion 1 found to produce no alteration ; but

Chemical History of Gun-Cotton and Xyloidine. 521

upon employing two measures of sulphuric acid to one of
nitric acid, 1 obtained a product resembling in all respects
ordinary pyroxyline, yet 42*77 grs. gave an increase of only
24-31 grs., or 56*84 per cent. Upon a repetition of this ex-
periment I found the increase to be 59*93 per cent., and again
70*6 per cent. Suspecting from the disparity of these results
that something might be dissolved in the acid liquor, I im-
mersed 6*7 grs. of cotton in a large quantity of the mixed
acids, but it increased 4*9 grs., or 73*1 per cent. Perceiving
that I had obtained an opposite effect to that anticipated, I
treated 12*64 grs. of cotton with just sufficient of the mixture
to wet it thoroughly : the fibre was evidently somewhat de-
stroyed ; the increase in weight was only 6*54 grs., or 51 '74
per cent., and the acid liquor squeezed from the cotton, neu-
tralized with ammonia, evaporated to dryness, and heated,
gave abundant evidence of organic matter being present.
Lest however it might be supposed that the whole had
not been converted into pyroxyline, it was treated again with
the mixed acids, but that produced an increase of only 0*12
gr. The action of various solvents confirmed its identity
with ordinary pyroxyline, while its solubility in potash proved
that the transformation had been very nearly complete. A
repetition of the experiment gave similar results. It thus
appears that the small increase in weight in the preparation
of pyroxyline takes place when there is not sufficient nitric
acid present to prevent the peculiar action of the sulphuric
acid, namely, that of dissolving and altering it. When how-
ever the increase amounted to about 74 per cent., I was never
able to detect the presence of oxalic acid or other organic
matter in the acid liquor ; and as no gas is evolved during
the'preparation of pyroxyline, it may be concluded that there
is no secondary product containing carbon.

Subsequently, when Dr. Schcinbein had specified his me-
thod of making gun-cotton, I treated 18*78 grs. of cotton
with a mixture of three parts of sulphuric acid and one of
nitric acid, sp, gr. 1*5, following his directions. The result
was 32*92 grs. of a substance similar to that produced in
former experiments, being an increase of 75*20 per cent. On
another occasion SO* 95 grs. of cotton gave an increase of
61*10 grs., or 75*47 per cent. The action of solvents and re-
agents confirmed the identity of this pyroxyline with that
obtained in my previous experiments, and I was equally able
to establish the absence of any secondary product containing
carbon.

In determining the ultimate composition of pyroxyline
several precautions were found to be necessary. In the ana-

522 Mr. Gladstone on the

lyses recorded below it was cut into small pieces, and, after
the weight was taken, mixed carefully with oxide of copper.
To prevent its caking together the admixture of a little as-
bestos was found useful. This was introduced into a long
combustion-tube, then some fresh oxide of copper, and upon
it again some fused into lumps so as to fill the whole bore for
about 7 inches. Lastly, was added a mixture of copper turn-
ings and reduced copper for about 9 inches. The combus-
tion conducted cautiously in the usual manner gave the fol-
lowing results ; the pyroxyline burnt in the sixth experi-
ment having been prepared by Schbnbein's method.

I. II. III. IV. V. VI.

Pyroxyline employed 4-09 4-61 3*57 4-85 4-55 2-905

Carb. acid produced 4*20 4-52 3*42 4*88 ... 2-84

Water produced . . 1-19 1*36 1*34 0-87

Hence in 100 parts, —

I. II. III. IV. V. VI.

Carbon . 27*90 26*74 26-10 27*44 ... 26-65

Hydrogen 3*22 3*27 3*27 3*32

In order to determine the amount of nitrogen the differ-
ential mode was adopted, as the method of MM. Will and
Varrentrapp is inapplicable to substances containing this
element in so highly oxidized a state. The same precautions
were taken as in the estimation of carbon ; and the collected
gases gave the following results after due correction for baro-
metrical pressure : —

I. II. Another specimen.

Carbonic acid . 25-0 38*5 23-9

Nitrogen ... 5*5 8-5 5*1

These proportions are, —

Nitrogen.

1

Carbonic acid.
4-55

1

4-53

1

4-68

The volumes of the gases represent respectively equivalents
of carbon and nitrogen, and since no secondary product is
formed in the conversion of lignine into pyroxyline, the 24
equivalents of carbon in the former must be found in the
latter. This will give the following ratio in equivalents of
carbon and nitrogen according to the three experiments above
cited : —

I. II. III.

Carbon .... 24-0 24*0 24-0
Nitrogen . . . 5*28 5-3 5'12

Chemical History of Gun-Cotton and Xyloidine. 523L

or 24 : 5, which accords with the proportions assigned by
M. Pelouze *.
The formula which best agrees with these results is the

following : — C24 -I ctuq f^^i which reckoned to 100 parts^

gives —

Carbon 26*23

Hydrogen 2*73

Nitrogen 12-75

Oxygen 58*29

In order to compare pyroxyline with xyloidine, I treated
starch with fuming nitric acid until the whole was converted
into a gelatinous mass. The addition of water then threw
down a white powder, which was subsequently well-washed
and dried. The iodine test proved the absence of all unal-
tered starch. The xyloidine thus obtained explodes at about
360°, leaving a carbonaceous residue. It is slightly soluble
in aether, with which it is capable of forming a peculiar com-
pound not yet investigated ; more so in alcohol, but most of
all in aether mixed with a small proportion of alcohol, or in
acetic aether. It is dissolved by strong sulphuric acid with-
out the aid of heat, and by boiling solutions of potash, am-
monia, hydrochloric acid and dilute sulphuric acid. These
solutions contain nitric acid, and nothing is precipitated
from them by dilution or neutralization. Xyloidine is also
soluble in strong acetic acid, or in nitric acid, whether fuming
or of sp. gr. 1*25, but is reprecipitated from either by dilu-
tion.

It was also found that nitric acid of ordinary strength (sp.
gr. 1*45) answered equally well in the preparation of this
substance ; but when acid of sp. gr. 1*41 was employed no
such result was obtained. Starch treated with a mixture of
equal measures of nitric and sulphuric acids produced a sub-
stance of greater combustibility, and more closely resembling
pyroxyline, but differing from it in being soluble in glacial
acetic acid, and in a mixture of aether with one-tenth part of
alcohol, as also in the action that acetic aether exerts upon it,
Xyloidine also when subjected to the mixed acids gave a pro-
duct identical with the above, as far at least as the action of
solvents can prove.

Xyloidine burnt by means of oxide of copper, with the
usual precautions, gave the following results. The sub-
stance employed in the third experiment was made from
arrow-root.

* Compttt Rendus, Jan. 4.

II.

III.

24-0

24-0

3-10

3-57

524 Mr. Gladstone on the

I. II. III.

Xyloidine employed . . 4' 7 7 5*23 6*75

Carbonic acid produced . 5*30 5*91 7*87

Water produced . . . 1-84 1*96 2-80

Hence in 100 parts, —

I. II. III.

Carbon . . 30-30 30*82 31-79

Hydrogen . 4-28 4-16 4*60

In the determination of nitrogen by the differential method
the proportions of the gases obtained were, —

I. II. III.

Carbonic acid . 70-7 53-4 53-8
Nitrogen . . . 10*6 6*9 8-0

These are in the proportion of —

I.
Carbon . . 24-0
Nitrogen . . 3*59

These numbers suggest the simple substitution product

{HI
3 NO I ^20' ^"^ which the per-centage of carbon would

be 31-37, and of hydrogen 3*70 ; yet the amount of nitrogen
is somewhat too great, and there is far from being sufficient
evidence to prove the definiteness of the substance itself.
The Avide difference also in the results obtained by various
chemists can scarcely be accounted for, except upon the sup-
position that they have operated upon very different sub-
stances.

The solubility of xyloidine in nitric acid led me to examine
whether any alteration could be effected upon pyroxyline by
similar means. The most dilute acid which I found to have
any effect upon it in the cold was that of sp. gr. 1*414 ; but
the alteration took place by means of this only after long
standing, and but to a slight extent. Nitric acid of sp. gr.
1*45 how- ever is capable of dissolving pyroxyline, and alters
both its composition and properties, as will be presently de-
scribed ; whilst fuming nitric acid has not the slightest effect
upon it. The new product just mentioned is acted upon
somewhat differently by various solvents, according to whe-
ther it exists in a fibrous condition, or in powder as precipi-
tated from solution ; yet I have found by experiment that no
alteration in weight is effected by this change of condition.
When in fibre it is slightly soluble in strong alcohol, aether, a
mixture of aether with one-tenth part of alcohol, and acetic
aether ; but when in the pulverulent state it is very soluble
in these menstrua, and in glacial acetic acid. In either con-

Chemical History of Gun-Cotton and Xyloidine. 525

dition it leaves a carbonaceous residue on combustion, is
dissolved by nitric acid, whether of sp. gr. 1'25 or 1*5, and
reprecipitated upon dilution. Strong sulphuric acid also
dissolves it in the cold, and chars it at a temperature below
180°. These two last properties show that the original pyr-
oxyline was perfectly free from admixture with this new sub-
stance.

There occurs a considerable decrease of weight through
this transformation. In the first experiment 32 grs. of sub-
stance operated upon gave 25*82 grs. of the new product; in
the second 43*64 grs. of the one yielded 34*68 of the other.
Now assuming the increase in the preparation of pyroxyline
to be 75 per cent., the weight of the new product above that of
the original cotton would be, as calculated from these figures,
41*1 and 39*05 per cent.

When this new product, whether in the fibrous or the pul-
verulent condition, was treated with a mixture of equal parts
of nitric and sulphuric acids, it increased considerably in
weight, and the resulting substance had all the properties of
pyroxyline as prepared in the usual manner, 11*16 grs. of
the one yielded 13*56 grs. of the other ; the quantity that
should theoretically have been obtained, calculating it from
the decrease in making the new product, is 13*84 or 14*04
grs. Again, 12*35 grs. of the substance as precipitated from
solution gave 15*75 grs., the theoretical amount would have
been 15*31 or 15*54 grs. This result proves the distinct-
ness of the new product from xyloidine, a fact that could not
have been ascertained from the action of the before-mentioned
solvents.

Whilst engaged in obtaining these results, I also examined
the action of nitric acid of various degrees of strength upon
pure cotton. By treating it with nitric acid of sp. gr. 1*5
I obtained a product evidently different from gun-cotton, but
as it did not appear to be homogeneous throughout, I passed
on to investigate the action of a weaker acid. That of sp. gr.
1*45 gave a substance which proved to be identical with the
product of the action of the same acid upon pyroxyline.
Upon a repetition of the experiment 68*54 grs. increased in
weight 14*61 grs., or 21*31 per cent. — a smaller increase
than might have been anticipated, but which may easily be
accounted for by the fact that the whole cotton had not been
transformed, as was proved by a considerable portion being
left undissolved by a boiling solution of potash. Nitric acid
ofsp.gr. 1*414 produced the same alteration, but only to a
small extent, and after long standing. 23*75 grs. of cotton
soaked in nitric acid of sp. gr. 1*516 became a hard mass.

526 Mr. Gladstone on the

and increased in weight 13-49 grs., or 56-8 per cent. ; the
action of various solvents upon the resulting substance indi-
cated that it was a mixture of pyroxyline and the new pro-
duct. On another occasion, when the transformation by
means of nitric acidsp. gr. 1-47 proved to be complete, 29-52
grs. of cotton increased 9-51 grs., or 32-89 per cent. But in
order to obtain a substance sufficiently pure for analysis 16-29
grs. of cotton were treated with enough nitric acid to dissolve
the whole ; the new product was precipitated by dilution,
and the increase in \veight was found to be 5-34 grs., or
32*78 per cent. In these instances there occurred a secondary
product containing carbon not precipitable by water.

When this was subjected to combustion with oxide of
copper, the following results were obtained : —

I. II. Another specimen.

Substance employed . 3-15 2*985 3*165

Carbonic acid produced 3-58 3-39 3'55

Water produced . . . 1-00 1-01 1*14

Hence in 100 parts, —

Carbon . 30*99 30*97 30*59

Hydrogen 3*52 3*75 4*00

I was unable to obtain any very accurate estimation of ni-
trogen by the differential method : the results most to be
depended upon were —

Carbonic acid

. 120*7

76-7

Nitrogen . . .

13*6

8*3

In the proportion of

Carbon . .

. 24*0

24*0

Nitrogen . .

. 2*7

2*6

These numbers lead me to think that there are 3 equiva-
lents of nitrogen in the compound, especially as I observed
during the combustion that the substance became charred
even 1 or 2 inches beyond the glowing charcoal, which will
account for the deficiency of nitrogen when compared with
the carbonic acid. Hence the composition of the new pro-
duct coincides very nearly with that calculated from the

formula C24 \^'^q \^2q^ namely.

Carbon . .

. . 31-37

Hydrogen

. . 3*70

Nitrogen . .

. . 9-15

Oxygen . .

. . 55-78

10000

Chemical History of Gun-Cotton and Xyloidine. 527

Under this supposition the increase in weight in the pre-
paration would be 4r66 per cent. ; very similar to that cal-
culated from the results obtained by the action of nitric acid,
sp. gr. 1'45, on pyroxyline, namely, 3905 and 4ri per cent.

In order to add an additional proof of the identity of the
two substances obtained by the action of nitric acid of sp. gr.
1*45 on cotton and on pyroxyline, and also of the fact that
pyroxyline is reproduced by the action of mixed sulphuric
and nitric acids upon the new product, the experiment was re-
peated with a portion of the substance made from pure cotton :
the result was pyroxyline. In the transformation 26*56 grs.
became 38'04 : now these 26*56 grs. were produced from
21*81 grs. of the original cotton ; hence the increase upon the
cotton itself would be 16*23 grs., or 74*4 per cent., coinciding
with the amount usually obtained in the preparation of pyr-
oxyline.

I. From these results it appears that in the treatment of
woody fibre by nitric acid raised to its highest degree of
strength by the addition of sulphuric acid, 5 equivalents of
the acid combine with 1 of lignine to produce pyroxyline,
displacing 5 equivalents of the elements of water, as indicated

by the formula C24S c-vf/^ fOao* The amount per cent, of

carbon and hydrogen hence deduced closely agrees also with
that assigned by Mr. Ransome * and M. Pettenkofer f.

Calculated. Ransome. Pettenkofer.
Carbon . . 26*23 26*28 26*26

Hydrogen . 2*73 3*16 2*75

In this case the synthetical experiment would give an in-
crease of 69*44 per cent. — nearly the amount obtained in the
best experiments. My own analyses however have yielded a
somewhat larger amount of carbon.

II. If lignine be treated with nitric acid combined with
more than 1 equivalent of water, another compound is pro-
duced, containing a smaller proportion of the elements of

nitric acid, most probably C34 < „ ^q > 0<^^ and very closely

resembling, but not identical with, pyroxyline.

C24H,oO^ + 3(NO„2HO) = C^|f^7^Jo,o + 9HO.

Also if pyroxyline itself be treated with nitric acid con-
taining 3 equivalents of water, the same compound results :

* Phil. Mag., January ] 847.

\ Pharmaceutisches Central BlattfTiec.ZOi^lS'lQ,

528 Royal Astronomical Society.

^2^{5N04}^=><'+2(N05, 3H0) = Cs^l J^jj'oj0jo+4(N05, HO).

And this transformation may be reversed.

Whilst completing my examination of this substance, my
attention M^as drawn to the communication of M. Payen in
the Comptes Rendus of Jan. 25th, where some properties of
" coton hypoazotique " are described. It is possibly the
same; yet, in order to express its distinctness from pyr-
oxyline, I would propose as the appellation of my substance
cotton-xyloidine. "

Before concluding I would acknowledge my obligations to
several chemists whose published investigations on the same
subject have suggested many of my experiments, and more
particularly to Professor Fownes for the valuable advice with
which from time to time he has favoured me.

LXXVII. Proceedings of Learned Societies.

ROYAL ASTRONOMICAL SOCIETY.

[Continued from p. 389.]

June 11, /^N the Opinion of Copernicus with respect to the

1847. ^^ Light of the Planets. By Professor De Morgan.

The common story is, that Copernicus, on being opposed by the
argument that Mercury and Venus did not show phases, answered
that the phases would be discovered some day. The first place in
which I find this story is in Keill's Lectures, It is also given by
Dr. Smith, in his well-known Treatise on Optics, by BaiUi, and by
others. But I cannot find it mentioned either by Melchior Adam or
Gassendil, in their biographies of Copernicus ; nor by Rheticus, in his
celebrated Narratio, descriptive of the system of Copernicus ; nor
by Kepler, nor by Riccioli, in their collections of arguments for and
against the heliocentric theory ; nor by Galileo, when announcing
^and commenting on the discovery of the phases ; and, what is most
to the purpose, Miiler, in his excellent edition of the great work of
Copernicus, when referring to the discovery of the phases of Venus,
as made since, and unknown to, Copernicus, does not say a word on
any prediction or opinion of the latter.

This story may then be rejected, as the gossip of a time posterior
to Copernicus. If we try to examine what the opinion of Copernicus
on this matter really was, a point of some little curiosity arises. It
depends on one word, whether he did or did not assert his belief in
one or other of these two opinions, — that the planets shine by their
own light, or that they are saturated by the solar light, which, as it
were, soaks through them. I support the affirmative : that is to
say, I hold it sufficiently certain that Copernicus did express him-
self to the effect that one or the other of these suppositions was the
truth.

If we take the first edition of the work Z)e Revolutionibus, which

Royal Astronomical Society. 529

was printed from the manuscript furnished by Copernicus himself,
there is little doubt about the matter. There are but two passages
which bear or can bear upon the question. The first is in the ad
lectorem, in which the writer (Osiander, though even Delambre make
him Copernicus) asks whether any one acquainted with geometry or
optics can receive the Ptolemaic epicycle then used to explain the
motion in longitude of Venus ? But the meaning of the allusion to
optics is explained in the next sentence, by a reference (and by no
means a fortunate one) to the changes of apparent diameter of Venus
derived from that epicycle ; changes which, as they made the peri-
gean diameter more than four times as great as the apogean, were
assured to be falsified by common experience. The second passage
is the one on which this discussion must turn. In book i. chap, x.,
after noting that some had theretofore believed Mercury and Venus to
come between the earth and sun, he mentions the difficulty arising
from the absence of the remarkable phase, which we now call the
transit over the sun's disc. He describes the opinion just mentioned
favourably, referring, not to his own view, but to that of those
others who had held it. This is not an uncommon idiom : persons
advocating an unpopular opinion are very apt to describe the main-
tainers of it in the third person, though themselves be of the number.
But when he comes to describe what he takes to be the necessary
consequence of the opinion, he lapses into the first person as fol-
lows : — " Non ergo fatemur in stellis opacitatem esse aliquam lunari
similem, sed vel proprio lumine, vel solari totis imbutas corporibus
fulgere, et idcirco solem non impediri "

These are the words of the first edition (Nuremberg, 1543)*
That Copernicus could have answered any objection, either by word
or writing, is impossible, since he drew his last breath within a few
hours of the time when, not able to open it from weakness, he saw
the first printed copy. The second edition (Basle, 1566) is usually
said to have been edited by Rheticus. The reason of this is that the
name of Rheticus appears in the title-page. But this appearance
only arises from the Narratio, &c. of Rheticus being added to the
edition ; and it is only the description of this edition which brings
Rheticus into the title-page. There is no mark whatever of his
having been the editor ; and as the work was printed at Basle, where
I cannot find that Rheticus ever sojourned, and as the latter was
deeply engaged at the time in his enormous trigonometrical calcu-
lation, some proof of his editorship must be given before it is ad-
mitted. As the point is of importance, I will notice, that unless
Rheticus had made some stay at Basle, it is very unlikely he should
have edited a work printed there. He did not edite the first edition,
only because it was found convenient to print it at Nuremberg in-
stead of at Wittemberg ; and it was accordingly entrusted to Osiander.
Now, if ever there were a connexion between two men, and between
one of them and the book of the other, which made it desirable and
even necessary that the first should edite the second, it was the case
of Rheticus and the first editio.n of the De Revolutionibus, &c. ; and
yet no arrangement could be made by which the sheets printed at

Phil. Mag, S. 3. No. 21 1 . Suppl. Vol. 3 1 . 2 M

630 Royal Astronomical Society.

Nuremberg could be revised at Wittemberg. It is very unlikely,
then, that Rheticus should have edited the second edition, when, as
far as we know, a similar impediment existed.

The third edition, by MUler (Amsterdam, 1617), has no authority
as to the text above that of the second.

Now both the second and third editions change the wox6^ fatemur
into fatentur, thus causing Copernicus to throw the opinion in
question upon his predecessors, instead of directly making it his
own. Not that it would be conclusive, even if the emendation were
adopted : for, as I have said, Copernicus is evidently speaking with
approbation of the opinions which he describes ; and it would be
difficult to say why comperiunt or putant in one sentence should
imply approbation, and fatentur, in the next, should be at least dis-
avowal, if not disapprobation. If Rheticus, who knew the mind of
Copernicus better than any one, had been the editor, I can conceive
that stress ought to be laid upon the change of the first into the
third person as an emendation ; that is, I should be somewhat stag-
gered by Rheticus having thought it necessary to make such an
alteration.

But, Rheticus not being in the question, as I think, for the rea-
sons given above, the next best authority on an opinion of Coper-
nicus is Galileo. Now the latter, in speaking of the phases of Venus,
expressly attributes to Copernicus the maintenance of one of the
two alternatives, — that the planet is either self-luminous or perfo-
rated by the solar rays. Of these alternatives, he says, in his letter
to Velser (Works, vol. ii. pp. 88, 89), " Al Copernico medesimo
convien amettere come possibile, anzi pur come necessaria una delle
dette posizioni." And that such was the opinion of Copernicus is
also assumed by the writer of the note on the Sydereus Nuncius in
the volume just mentioned, and by others, even down to our own
time ; as by Mr. Drinkwater Bethune, in his life of Galileo. In fact,
with the exception of the unsupported story mentioned at the be-
ginning of this paper, there is nowhere, that I can find, anything
against my conclusion. And it is to be remembered, that Copernicus
nowhere shows any of that acumen in matters of phj^sics, apart from
mathematics, which has often enabled the cultivators of the former
to make steps more than proportionate to their knowledge of the
latter. Ptolemy, the great promoter of the old theory, and Coper-
nicus, its destroyer, were both mathematicians in a peculiar sense ;
Ptolemy being far the more sagacious in questions of pure experi-
ment. Their grounds of confidence are mathematical ; and Coper*
nicus, in particular, dares to face his own physics (for there is no
reason to suppose he was beyond his age in mechanical philosophy)
with reasons drawn entirely from probabilities afforded by mathe-
matics.

There is much reason to regret the practice of associating with
the names of those who have led the way in great discovery the
glory which is due to their followers. The disadvantage is twofold.
In the first place, it introduces into the history of science an index
error of from one to two centuries ; secondly, those who come to in-

Royal Astronomical Society. 531

quire are disappointed to find that they must lower their opinion of
great men, and are perhaps led to do it to a greater extent than jus-
tice requires. Our usual popular treatises speak of Copernicus as if,
besides himself, he had in him no inconsiderable fraction of Kepler,
Galileo, Newton and Halley. What is a person to think who comes
from those histories to actual investigation, when he finds in Coper-
nicus himself the immovable centrum mundi (only reading sun for
earth) of the Ptolemaists, their epicycles, and a suspicion, at least,
of the solid orbs ?

On the Formation and Application of Fine Metallic "Wires to
Optical Instruments. By Mr. Ulrich.

Dr. WoUaston, in the Philosophical Transactions for 1813, pro-
posed a method of forming wires of gold or platinum of any degree
of tenuity. The discovery does not appear to have been much used,
owing, as Mr. Ulrich supposes, to the difficulty of application.

Mr. Ulrich forms the fine wire by inserting a gold or platinum
wire in the centre of a silver cylinder of much larger dimensions,
which is afterwards drawn out by the usual process. When the
silver wire has been sufficiently extended, Mr. Ulrich cuts it into
short lengths and attaches platinS. rings to each end. The rings are
hooked upon a hooked fork, and the whole is plunged into heated
nitric acid, when the silver coating is dissolved.

The artist may now wire his cell according to his fancy. Mr.
Ulrich's plan seems to be, to hold one end by an overplate ; then to
allow the wire to be stretched by its platina ring, and to fix the other
oveqjlate. He recommends using a cell of the same material as the
wire, as, otherwise, a difference of expansion might break or slacken
the w^ires.

On the properties of Rock as a foundation of the Piers of Meridian
Instruments, with an Account of the Detection of a hitherto un-
suspected Cause of Error in the Edinburgh Transit. By Professor
C. P. Smyth.

Some years ago doubts were expressed of the fitness of a rock
foundation for an observatory. It does not appear that any experi-
ments were made, or that any reason was adduced beyond this, that as
tremor was unfavourable to the performance of large telescopes, and
as rock was more capable of transmitting tremors than less compact
material, therefore rock was to be avoided when choosing a site for
an observatory. The author or authors of this opinion were pro-
bably but ill- acquainted with the mode of working an observatory,
or the requisites for obtaining accuracy in meridian observations ;
yet it is certain that an undue importance was attached in some cases
to these very idle surmises. At the present time it is not likely that
any intelligent person would be misled by such authorities, and it
is therefore unnecessary to mention here the mischief they have
caused *. It is to be wished that the founders of future observatories,

* The effect of tremor on a telescope is probably familiar to every read-
er of tfiis notice. It causes a sort of burr round the object, and destroys
the sharpness of outline and definition. This is probably more injurious in
reflecting than in refracting telescopes ; but we may fairly doubt whether it

2 M2

532 Royal Astronomical Society,

•who can command a rock foundation, should make use of their good
fortune ; and that those who cannot, would look carefully to the
possible effects of moisture, which are probably more extensive, and
vary more rapidly, than those of temperature.

The observatory of Edinburgh is placed on the Calton Hill. This
is chiefly of a porphyritic formation. The apex w^as blasted away
to obtain a level area, on which the observatory was erected. The
site of each pier was cut away until a sound part of the rock was
arrived at (it was not necessary to go deeper for this purpose than
six or nine inches), when the exact size of the foundation was at
once marked out and the space carefully levelled. The foundation
stone was also carefully smoothed, and then laid in its place with
milk of lime. As the foundation and stone were both rather hollow,
except for three inches at the outer edge, which was polished, the
fitting was very perfect. There are no vertical joints, and each stone
was laid in the same manner as the foundation stone. As one of
the principal thoroughfares of Edinburgh runs about 100 feet below,
and only 300 feet distant from, the observatory, tremors were con-
fidently predicted by the alarmists. Professor Henderson, however,
found none, nor any interruption to his observations in mercury.
Professor Smyth adds that he finds no annoyance from the railroad
about 300 feet below, and at a horizontal distance of 500 feet.

So far the observatory founded on a rock came out victoriously
from its ordeal, but Professor Henderson, in the course of his work,
found a well-marked annual variation of the level of the transit,
which he attributed to the expansion of the rock. This variation
seemed so intimately connected with temperature that he latterly
took his factor for level correction from the thermometer, having
found a constant agreement between this and the indications of the
spirit level. The maximum of this change amounted to between
0^-2 and 0**3 in the value of the level factor, and the variations were
tolerably regular.

On computing the azimuthal factors for 1841, Professor Smyth

is more felt on solid than on loose foundations. In a standard observatory,
where observations are made principally in the meridian, tremor scarcely
affects the accuracy of observation at all, unless it is so excessive as to
change the position of the microscopes, piers, &c. Now this is obviously
the least likely to happen when the foundation is on rock ; the tremors are
propagated through the substance, without in any respect altering its form.
Sudden and discontinuous changes, which obey no law, are those only which
are to be feared in a well-directed observatory. Tremor is chief!}' objec-
tionable as disturbing the mercurial horizon, which, however, is now mostly
used as a verification, not as the ordinary mode of observing; and when
this inconvenience only occurs occasionally, it can generally be avoided or
palliated by a little contrivance or foresight. Unless the adjustments are
kept in a fluctuating and uncertain state by occasional small oscillations
(and we believe no careful experiments have been directed to this point),
they are minor evils. The experience of the Oxford and of the Edinburgh
Observatory is, so far as it goes, conclusive against any danger from mo-
derate exposiu'e to tremors in a well-founded and well-managed observa
tory.— S.

Royal Astronomical Society. 533

was very much disturbed on finding variations, which sometimes al-
tered the factor as much as 0*"3 in a day, and more than P*0 in the
course of the year. These changes in azimuth had been remarked by
Professor Henderson, and were attributed to the irregular action of
the counterpoises, which were consequently removed. On a com-
parison of these errors with the indications of thermometers plunged
in the rock there were apparent marks of correspondence.

There are several thermometers inserted at different depths in the
rock near the observatory, which had been carefully observed in the
year 1841*. The indications of these thermometers were projected
on paper, and the curves thus formed compared with a curve traced
according to the course of the azimuthal deviation. It was thus
made evident, that the curve of azimuthal deviation, though having,
like the other curves, an annual maximum, did not otherwise resem-
ble the curves belonging to the deep-seated thermometers at all ; and,
in fact, it came nearest the curve traced out by the thermometer at-
tached to the barometer and by the free thermometer exposed to the
outer air. Hence the cause of the deviation was not to be looked
for in the effect of temperature on the foundations or on the massive
transit piers, but on smaller parts more readily affected, such as the
metallic mounting. These were accordingly examined. In the
azimuthal Y, the construction was found to be much as usual, but
the artist has adopted an adjustment for the vertical Y, which seems
liable to suspicion. There are two vertical screws applied from be-
low ; one, pushing, on the north side of the middle, and the other,
pulling, at the south side. The Y is prevented from turning in a
vertical plane by jamming horizontal screws, which press a plate
against the north face of the Y so as to bring the whole tightly
against a stopping-piece, which blocks the south face. Professor
Smyth's present opinion is, that the effect of expansion on the two
screws, which are in contrary states of constraint, is to alter the ad-
justment ; certainly the arrangement looks unmechanical. In the
ordinary mode of construction, in this country at least, the elevating
Y is either raised by one central screw, or by two screws, one on
each side of the centre ; in which case a drawing-screw may be placed
at the centre. There is thus no tendency to twist, and the side-
plates which confine the Y laterally have to exert little restraining
force. Professor Smyth has communicated with MM. Repsold, the
makers of this magnificent instrument, and is awaiting their reply
before adopting any remedy t-

* Some years ago, Professor J. D. Forbes had four thermometers sunk in
the rock with their bulbs at the depths of 24, 12, 6, 3 French feet and a
fifth on the surface merely covered with sand.

t Sudden and lawless changes in azimuth forbid independent determina-
tions of the azimuthal deviation (which are also the best;, viz. from the
consecutive semidiurnal transits of circumpolar stars. The possessor of an
imperfectly mounted instrument must content himself with assuming the
fundamental places of his close circumpolar stars, and determine his azi-
muthal error from each of them. This will, with proper caution, be found
quite sufficient for objects not too near the pole, especially when the clock-
error stars are pretty numerous, and situated above and below the object
to be determined.

[ 534 ]

LXXVIII. Intelligence and Miscellaneoics Articles.

ON OSMIAMIC ACID. BY MM. J. FRITZSCHE AND H. STRUVE.

WHEN caustic ammonia is added to a solution of osmic acid in
excess of potash, the deep orange colour of the liquid becomes
rapidly a bright yellow, and a new salt is produced and separated,
either immediately or by evaporating the liquid at a gentle heat,
which is a yellow crystalline powder.

The formation of this new compound does not necessarily depend
on the presence of potash or any other oxide, but uniformly upon
that of ammonia ; the ammoniacal salt is, however, subject to altera-
tion, and decomposes during evaporation. It is better therefore to
cause a basic oxide to intervene.

M. Gerhardt remarks that the formula of the osmiamates which the
authors have given requires correction ; they agree, he states, with
the formula OS, 03 N (M).

The properties of the osmiamates are as follows : they decompose
by heat with explosion ; and several of them undergo the same de-
composition when struck. Among the products of this decomposi-
tion are metallic osmium, an osmiate, or a less oxygenated osmic
compound. Protosraiamate of mercury volatilizes without explosion,
when heated quickly ; and it diffuses a strong smell of osmic acid.

Osmiamic acid can be obtained only in solution in water. In
order to prepare it, osmiamate of barytes is to be cautiously decom-
posed by sulphuric acid, or recently prepared and moist osmiamate of
silver is to be decomposed by dilute hydrochloric acid. After fil-
tration a bright yellow-coloured solution is obtained, which may be
preserved for several days, if it be sufficiently dilute ; on the other
hand, if too concentrated, it becomes brownish and decomposes with
the disengagement of gas, osmic acid is set free, and a black non-
explosive substance is deposited which contains osmium.

The same metamorphosis occurs when the weak acid is evaporated
over sulphuric acid.

Osmiamic acid not only expels carbonic acid from carbonates,
but also decomposes chloride of potassium. In fact crystals of os-
miamate of potash are obtained, if a crystal of chloride of potassium
with a drop of solution of osmiamic acid be exposed to evaporation
on a strip of glass.

Zinc dissolves in solution of osmiamic acid, with the evolution of
a little gas ; part of the acid is decomposed, and the zinc is covered
with a very adherent black deposit, and flocculi appear in the liquid
which possess the odour of osmic acid. When all the undecomposed
acid is saturated with zinc, the metamorjihosis ceases.

In the cold, acids do not decompose osmic acid or the osmiamates :
sulphuric, nitric or hydrochloric acid may be added to their solutions
without inconvenience ; but decomposition readily occurs when heat
is applied, and it is rendered apparent by the disengagement of osmic
acid and by the brown colour of the liquor ; the products vary ac-
cording to the nature of the acid employed.

Intelligence and Miscellaneom Articles. 635

Osmiamates are obtained, either directly by the action of osmic
acid on a solution of bases in ammonia, as the salts of potash, zinc,
and silver, or by precipitating the potash salt by metallic salts, or
by decomposing the silver salt by chlorides.

MM. Fritzsche and Struve have stated that these salts yield no
hydrogen by analysis. In two experiments the potash salt gave by
combustion vsrith oxide of copper only 0*072 and 0033 of hydrogen ;
whereas 0'34 are required for one equivalent of hydrogen.

The osmiamates undergo an interesting decomposition by the
action of hydrochloric acid. The products vary according to the
concentration of the acid. If the potash salt be sprinkled with con-
centrated acid, energetic action immediately ensues, accompanied
with the disengagement of chlorine and probably of its oxide ; the
hydrochloric acid assumes a fine purple tint, and the crystals of
osmiamate of potash are covered with a crust of small red crystals of
two different kinds ; if the salt employed be powdered, and the action
of the hydrochloric acid be long enough continued, all the osmiamate
undergoes this change ; the nature of which the authors have not
hitherto succeeded in explaining.

If dilute hydrochloric acid be added to a solution of osmiamate of
potash saturated cold, no decomposition occurs at common tempera-
tures, the metamorphosis taking place only at a higher temperature.
It is then more complicated, the liquor temporarily assumes a red
and brown tint, and soon emits a smell of osmic acid, which is
abundantly disengaged as soon as the liquor is heated to ebullition.
If the solution be evaporated to the crystallizing point, as soon as it
ceases to emit osmic acid, a mixture of salts is obtained, among
which, as shown by the microscope, are hexagonal green tables, green
needles, and another red salt, &c. These salts appear to be decom-
posed by water, for they were not obtainable by solution and re-
crystallization. — Journ. de Ph. et de Ch., Octobre 1847.

ON THE PREPARATION AND PROPERTIES OF SOME OSMIA-
MATES. BY MM. FRITZSCHE AND STRUVE.

Osmiamate of Potash. — This salt is best prepared by dissolving
golid osmic acid in a concentrated solution of caustic potash, with
the addition of ammonia during the agitation of the mixture. The
solution becomes of a bright yellow tint, and the osmiamate of potash
is deposited in the state of a yellow granular powder. The product
of the distillation of osmic liquors may also be directly passed into a
solution of potash, containing ammonia and properly cooled; the
simultaneous distillation of nitrous vapours must be carefully avoided,
as they would decompose the osmiamate of potash.

In both cases, the mother-water which has deposited osmiamate
of potash is to be evaporated with a gentle heat ; carbonate of pot-
ash may be used instead of caustic, but not so advantageously ; the
osmiamate of potash is to be dissolved in a very small quantity of
boiling water ; on cooling the solution yields small crystals of the
salt of a lemon-yellow colour ; these crystals are of considerable size

536 Intelligence and Miscellaneous Articles.

when prepared from a cold saturated solution by spontaneous evapo-
ration, their form being an acute octahedron with a square base.

Osmiamate of potash is much less soluble in alcohol than in water ;
it dissolves without alteration, and decomposes but very little when
the solution is evaporated ; it contains no water of crystallization ;
it may be heated to 356" F, without decomposing, but it becomes
brownish and is rapidly decomposed at a higher temperature, with
violent projections.

This salt yielded by analysis —

Osmium 67*900

Nitrogen 4-126 4*820

Potash 16*126

M. Gerhardt gives as an amended formula 0S03N(K).

Osmiamate of Soda is best obtained from the silver salt and chlo-
ride of sodium ; the crystals are prismatic, contain water of ci*ystal-
lization, and are very soluble in water.

Osmiamate of Ammonia is prepared in the same manner. It forms
large anhydrous crystals, which appear to be isomorphous with the
salt of potash ; at 258° F. it decomposes with explosion. This salt
is very soluble in water and in alcohol.

Osmiamate ofBarytes readily crystallizes in yellow brilliant needles
of several lines in length. 1'his salt is readily soluble in water, and
explodes at about 300° F. It yielded by analysis —

Barytes 23*88

Osmium „ 61*07

Nitrogen 4*269

the formula according to M. Gerhardt being 0S03N(Ba).

Osmiamate of Ammonia and Zinc is obtained either by dissolving
osmic acid in a solution of a salt of zinc in caustic ammonia, or by
mixing a solution of osmic acid in ammonia with the solution of a
salt of zinc. A yellow, bright crystalline powder is soon deposited,
which is deprived of the mother-water by washings with ammonia.

This compound is very permanent ; it may be dried in the air, and
remains without losing ammonia. It is nearly insoluble in ammo-
nia, water decomposes it even when cold ; when boiled in water it
is completely decomposed with the deposition of oxide of zinc, the
disengagement of half of its ammonia, and yielding osmiamate of
ammonia. Formula according to M. Gerhardt 0S03N(Zn), 2NH3.

Osmiamate of Lead. — A solution of nitrate of lead is not precipi-
tated by a concentrated solution of osmiamate of potash ; after some
time some crystals are however formed, which are not sufficiently
stable for examination. A solution of acetate of lead gives with the
solution of the osmiamates a non-crystalline precipitate, which is at
first of a dirty yellow colour, but it soon becomes of a purple tint
with the extrication of osmic acid.

If a solution of chloride of lead, or a solution of nitrate of lead with
the addition of hydrochloric acid, be added to a solution of osmia-
mate of potash, a yellow crystalline precipitate is soon obtained, which
the authors consider to be a compound of equal equivalents of chloride
eind osmiamate of lead.

Intelligence and Miscellaneous Articles. 537

Protosmiamate of Mercury forms a bright yellow precipitate ; it is
not crystalline, and is insoluble in water ; the perosmiamate of mer-
cury forms prismatic crystals.

Osmiamate of Silver is obtained directly by dissolving osmic acid
in an ammoniacal solution of a salt of silver, and afterwards super-
saturating with nitric acid. It may also be obtained by adding to a
solution of osmic acid in ammonia nitric acid in excess at first, and
then a salt of silver. It may also be prepared by double decompo-
sition with the salts of silver and soluble osmiamates,

Osmiamate of silver is a crystalline powder of a lemon-yellow
colour ; it is very slightly soluble in water and in cold nitric acid,
more soluble in ammonia, and may be combined with it. It may be
dried in the dark without blackening, in vacuo, over sulphuric acid ;
eventually, however, it suffers decomposition, and then gives out
osmic acid ; at 176° F. it decomposes suddenly and with violent de-
tonation ; it is also decomposed by percussion, and likewise when
sulphuretted hydrogen is passed over the dried salt ; nitric acid de-
composes it readily when heated ; the liquor first acquires a brown
tint, and gradually becomes colourless with the disengagement of
osmic acid.

This salt yields by analysis —

Oxide of silver .. 32-08 32*060 32-13

Osmium 55-011

M. Gerhardt gives as its formula 0S03N(Ag). — Journ. de Ph. et de
Ch., Octobre 1847.

ON SULPHATO-CHLORIDE OF COPPER, — A NEW MINERAL.
BY ARTHUR CONNELL, ESQ.

This mineral occurs in small but very beautiful fibrous crystals, of
a fine blue colour, which is pale when the fibres are delicate, but
much deeper when they become somewhat thicker. Their form,
according to Mr. Brooke, is a hexagonal prism with the edges re-
placed, thus belonging to the rhombohedral system. They possess
considerable translucency, and have a vitreous lustre. On account
of the small quantity which he possessed, Mr. Connell was unable
to state the specific gravity, hardness, or fracture. Their locality is
Cornwall. Mr. Brooke is aware of the existence of only a very few
specimens of the mineral : one is in the British Museum.

Like atacamite, this mineral colours the blowpipe flame as well
as the simple flame of a candle, of a fine greenish-blue, indicating the
presence of chloride of copper. Reduced to powder, and mixed in
sufficient quantity with charcoal powder, and then heated in a close
tube, it gives decided, although not strongly marked, indications of
the presence of sulphuric [sulphurous ?] acid by the smell, and
partial bleaching of Brazil wood paper, the remainder of the paper
being reddened, doubtless by muriatic acid vapours. Alone, in the
close tube, it yields a little water, and other appearances resembling
those afforded by atacamite. Heated alone on charcoal before the
blowpipe, it decrepitates strongly ; but when previously deprived

5S8 Intelligence and Miscellaneous Articles.

of the greater part of its water by gentle heat, and then powdered
and moistened, and heated on charcoal, it gives no traces of arsenic,
although arseniate of copper is associated with it in the specimens.
The residue is a dark reddish slag or globule.

The crystals are njjt soluble in boiUng water, but dissolve entirely
and pretty readily in nitric or muriatic acid, especially by the aid of
a gentle heat. The solutions have the colour belonging to copper
solutions ; and in the act of dissolving a very few bubbles of gas
may be observed to arise, indicating probably the presence of a mi^
nute quantity of carbonate. The solutions yield, with barytic salts,
a white precipitate insoluble in acids ; and the nitric solution gives,
with nitrate of silver, a white and curdy precipitate insoluble in acida
or water, but soluble in ammonia. Ammonia in excess, added to
the original solution, gives the fine deep blue of copper.

These appearances, in conjunction with the blowpipe reactions,
are sufficient to show that the constituents of the mineral are sul-
phuric acid, chlorine, copper, and a little water ; but Mr. Connell
had not sufficient of the mineral to determine the proportions of its
constituents. The chloride is apparently more abundant than the
sulphate. — Jameson's Journal, October 1847.

ON THE FORMATION OF VALERIANIC ACID. BY M. THERAULT.

The author remarks that it has been long known that the oil of
potatoes yields valerianic acid under the influence of the caustic alka-*
lies; and it has also been stated that the oil of valerian gave analogous
results. M. Therault thought it would be an interesting subject of
inquiry to determine in what manner this transformation occurs, and
whether it is complete or only partial ; and in the latter case to ex-
amine into the nature of the non-acidifiable product ; whether the
alkalies directly produced a true chemical reaction on the elements
of the oil ; and lastly, whether the intervention of other agents is not
requisite to effect the transformation.

In order to resolve these questions, the following experiments were
performed, care being taken to operate with oil perfectly deprived of
any trace of acid.

1 . A portion of oil was mixed with distilled water, and divided
into two parts, one of which was exposed to the contact of the air,
and the other put into a bottle to prevent its action.

2. Another portion of the oil was mixed with caustic potash, per-
fectly dry and reduced to powder, and divided as in the preceding
experiment.

3. A mixture was prepared of six parts of oil and three parts of
potash, previously dissolved in one part of water, and the mixture
was divided as in the foregoing experiments.

The following observations were made on these mixtures. After
the contact of a month, that portion of the mixture of the first ex-
periment which had been submitted to the action of the air, had
become sensibly acid ; in the second portion no change had occurred.
In the second experiment no sensible trace of valerianic acid was

Intelligence and Miscellaneous Articles, 589

produced ; the potash and the oil had nearly retained their original
properties.

Circumstances were quite different in the third experiment. The
mixture had hardly been made when it became of the consistence of
honey, and of a red colour of considerable intensity : perfect saponi-
fication might be suspected. M. Bonastre had previously remarked
this action of the caustic alkalies on some essential oils, and had
proposed it as a means of distinguishing mixtures of them ; and he
noticed the partial combination of the oil of valerian with soda. This
fact might induce the belief that this oil was a substance of a com-
plex nature ; M. Therault is, however, of opinion that this is not
the case, but that the observation of M. Bonastre was derived from
the circumstance of the oil which he employed containing valerianic
acid, which would explain in this case the partial combination with
soda. The author attentively examined the nature of this mixture :
it was perfectly homogeneous, and comparable to crotonic soap.
Treated with water and suffered to remain undisturbed, the oil soon
collected on the surface ; it was separated, and the filtered liquor was
saturated with acetic acid. No sensible trace of oil was reproduced,
nor w«s the formation of valerianic acid detected ; it was therefore
certain that no chemical action had occurred ; and the name of com-
bination given to this mixture appears to the author to be improper
under these circumstances.

M. Therault relies upon this last fact as corroborating the result
of the third experiment. The portion of the mixture kept from the
contact of the air, underwent no change of properties after one
month ; no combination had occurred between the oil and the pot-
ash ; no valerianic acid was formed, or at any rate no appreciable
quantity.

On the contrary, that portion which had been exposed to the action
of the air contained valerianic acid, in minute quantity certainly,
but it was very appreciable : the matter was slightly decolorized.
The action of the air was allowed to continue, and after six months'
exposure nearly the whole of the oil had disappeared, and the mix-
ture was almost entirely decolorized, a slight amber tint only re-
maining.

The decoloration occurred nearly in direct proportion to the quan-
tity of acid formed : the action of the air was however continued, to
ascertain whether the whole of the oil could be acidified ; the opera-
tion required nearly six months, but it was then complete ; the
mixture had assumed the aspect of white, slightly grumoua honey ;
well-defined crystals of valerianate of potash had formed, and were
dispersed throughout the mass. It was covered with a solution of
this salt and of potash, without any trace of oil.

It may be concluded from the experiments detailed, that valerianic
acid does not pre-exist in valerian root ; that it is entirely the pro-
duct of the oxidizement of its oil ; that this oxidizement is due to
the oxygen of the air ; and that water and the caustic alkalies greatly
facilitate this oxidizement. The author also concludes that the
caustic alkalies exert no direct action on the elements of the oil ;
that they act only by the property which they po§ses^ pf forming an

540 Intelligence and Miscellaneous Articles.

intimate mixture with the oil, and exposing it in a state of perfect
division to the oxygen of the air. Lastly, M. Theraultisof opinion
that oil of valerian is not a substance of a complex nature, and that
it may be entirely converted into valerianic acid.

This conversion is readily explained. According to Ettling, the
formula of valerianic acid isC*" H'^ O^ + H^ O ; if that of oil of va-
lerian be C'^o H-o O^ and if tw^o atoms of oxygen be made to inter-
vene, one of vvrhich combines with two atoms of hydrogen to form
water, and the other be added, we shall have the following equation :

C20HW 02 + 20=020 HIS 03 + H'^O.

M. Tberault finishes his memoir with observing, that the process
for extracting valerianic acid, proposed by Messrs. Smith of Edin-
burgh in the Journal de Phannacie for January last, appears to be
a good one ; excepting that he would propose to use a caustic
instead of a carbonated alkali, and after having boiled the mixture, to
expose it for a month to the air, stirring it several times a day ; not
to subject the residue to pressure, and to distil with the roots, which
greatly facilitate the operation. When afterwards the distilled liquor
has been saturated by means of an alkali, and the valerianate of pot-
ash or soda has been concentrated, it is essential not to employ an
excess of sulphuric acid to separate the valerianic acid ; it would be
better to leave a small portion of the valerianate undecomposed, for
towards the end of the distillation, the organic matter mixed with
the salt is carbonized and sulphurous acid is formed, which appears
to react upon the valerianic acid. — Journ. de Pharm. et de Ch., Sep-
tembre 1847.

NOTE ON THE MEASUREMENT OF THE DOUBLE SULPHATES OF
ZINC AND SODAj AND OF MAGNESIA AND SODA. BY PROF.
W. H. MILLER.

The crystals were not good enough for me to obtain a very satis-
factory result from a few observations ; they are however sufficiently
good to show that they are isomorphous.

The crystals belong to the oblique prismatic system.
The angles between normals to the faces are — for the oxide of
zinc salt,

nn' 113° 4
M u' 74° 12

the angle between a normal to c and the
intersection of

M, m'=10°22'
u c 83° 46'
r c 49° 54'
The symbols of the simple forms, in the notation which I use, are —

c 001, rlOl, « 110,

u 120, e Oil, 5 121.
The faces of the magnesian salt are more irregular than those of
the former, so that I cannot pretend to determine the diflference be-
tween the angles of these crystals.

Intelligejice and Miscellaneous Articles. 541

The angles given above must be considered as rough approxima-
tions only. In a little time perhaps I may be able to obtain more
accurate values of them. — From the Proceedings of the Chemical
Society, vol. iii. p. 391.

NATIVE CARBONATE OF NICKEL.

This new mineral was exhibited last year at the Philosophical
Society's exhibition in Glasgow, and was examined at the time, at
the request of Dr. R. D. Thomson, by his pupil Mr. John Brown,
in the College Laboratory. It occurs in the form of thin green cry-
stalline layers, on the surface of chrome iron ore from America. It
dissolves with effervescence in dilute hydrochloric acid. The solu-
tion is precipitated black by sulphohydret of ammonia ; is precipi-
tated and dissolved in excess by caustic ammonia, yielding a cha-
racteristic coloured solution. Caustic soda precipitates the solution
green, without resolution. It is accompanied, apparently in union,
by carbonates of lime and magnesia — isomorphous bodies. The fact
of its occurring on the surface of chrome iron, and having been mis-
taken for sesquioxide of chrome, renders it probable that oxide of
nickel may exist in that mineral occasionally. — R. D. T.

AN EXAMINATION AND ANALYSIS OF THE " NADELERZ," OR
NEEDLE ORE OF BISMUTH. BY. E. J. CHAPMAN, ESQ.

The "needle ore" occurs in thin prismatic crystals, generally forming
more or less radiated groups imbedded in quartz, at Ekatherinen-
burg in Siberia, the only known locality in which it has been hitherto
found. The crystals are too imperfect to admit of measurement ;
but they appear to belong to the Trimetric or Prismatic system, and
to have for the primary form a right rectangular prism, or perhaps
more correctly a right rhombic one, in which the angle MM closely
approaches a right angle.

The colour of this mineral is dark steel-gray on the fractured sur-
face, but externally the true colour is usually masked by a yellow
tarnish. The powder or "streak" is black; the degree of hardness
2*0 to 2*5, or between that of rock-salt and calc-spar ; and the spe-
cific gravity about 6*1.

Before the blowpipe it fuses instantly and may be almost entirely
volatilized, forming a yellow incrustation of the mingled oxides of
lead and bismulii on the support. The presence of bismuth and
copper may be ascertained by fusion with " microcosmic salt" and
a little tin on charcoal in the reducing flame, when the lead, which
is clear whilst hot, becomes on cooling of a grayish-black colour
with red patches. With carbonate of soda on charcoal in the same
flame, it forms an alkaline sulphuret. The lead is best detected by
boiling a fragment in nitric acid, filtering, dissolving the residue
(sulphate of lead) in caustic potash, diluting the solution, and re-
precipitating by sulphuric acid.

This ore was first described by Karsten and analysed by John;
and although a considerable period has elapsed since the date of this
analysis, yet, probably from the rarity of the mineral, its composition

J43 Intelligence and Miscellaneous Articles.

has been examined by but one other chemist, Frick, in PoggendorfF's
* Annalen,' xxxi. p. 529.

These analyses have given —

1. By John :—

Per cent. Atomic relations.

Sulphur. 11-58 0-057 3

Bismuth 43-20 0-032 2

Lead 24-32 0-018 1

Copper 12-10 0-031 2

Nickel 1-58

Tellurium 1-32

Gold 0-79

94-89

2. By Frick :—

Sulphur 16-61 0-0826 6

Bismuth 36-45 0-0410 3

Lead 36-05 0-0270 2

Copper 10-59 0-0267 2

99-70
The first analysis does not admit of any rational formula; but if
we consider the loss, more than 5 per cent., to be sulphur, we may
obtain by a little latitude,

Cu2 S + PbS + Bi2 S», or 2 1 ^^^ \ + Bi^ S^,

a formula analogous to that of the kobellite from Ilvena in Sweden,
analysed and named by Setterberg, and in which the electro-negative
atoms in the base are to the electro-negative atoms in the acid as 2
to 3. The following is its formula : —

The second analysis yields also but an inexpressive and unsatis-
factory result. The formula, if such it can be termed, obtained
from it is —

Cu« S + 2PbS + 3BiS, or otherwise Cu^ S, BiS + 2(PbS, BiS).

I have now to enter into the details of a third analysis, executed
lately by myself, on a specimen kindly given to me by Colonel
Jackson, F.R.S., who brought it with him from Russia. The ore
was accompanied in the quartz by minute tufts of malachite, which,
together with the matrix, were carefully removed, by the aid of a
microscope, from the substance analysed.

8-38 grs. of the mineral in powder were boiled in strong nitric
acid.

A residue of 4-92 grs. of sulphate of lead remained, and 0-26 grs.
of sulphur. The 4-92 grs. of sulphate of lead (obtained, it should
be stated, after solution of the residue in potash and subsequent re-
conversion) = 3-36 grs. of lead and 0*52 of sulphur.

Carbonate of ammonia in excess was then added to the clear so-
lution ; and after remaining for three hours at a gentle heat, it was
filtered from the precipitate, which (after being well-washed with

Intelh'ge?ice and Miscellaneous Articles. 643

the same reagent, and the "washings" added to the original solu-
tion) was dissolved in acetic acid ; and a slip of pure lead being im-
mersed in the solution, the whole was covered up immediately and
suffered to stand for four hours. The slip of lead weighed 22*63 grs.

The four hours having elapsed, the lead was taken from the solu-
tion, and, after separation of the precipitated bismuth, dried and
weighed. It weighed 19*21 grs.; loss, 3*42 grs. On the addition
of sulphuric acid, 5*03 grs. of sulphate of lead were obtained, which
are equal to 3*435 of lead ; and this amount corresponding so nearly
with the loss in the metallic precipitant, the whole of the lead pre-
sent in the mineral may be considered to have been converted into
PbO, SO^ by the first operation.

The bismuth precipitate wad washed with cold distilled water
(which had been boiled), dissolved in nitric acid, and again thrown
down by carbonate of ammonia. The oxide of bismuth weighed
2"60 grs., equivalent to 2*33 grs. of bismuth.

To tiie original solution (containing carbonate of ammonia) a few
drops of ammonia were added, and it was then gently evaporated
until the ammoniacal odour was entirely destroyed. Solution of
potash was then added, and the whole boiled. The black oxide of
copper, well-washed with hot water, ignited, and weighed in a
covered crucible, came to 1*31 grs., an amount equal to 1*05 of
copper.

Finally, chloride of barium was added to the potash solution,
which produced a precipitate of sulphate of baryta weighing 5*72
grs., an amount corresponding to 0*79 of sulphur. The whole of
the sulphur present in the mineral was therefore

1*57 grs. (0*52 + 026 + 0*79).

The following table exhibits the above analysis and its atomic de-
ductions :— -

Per cent. Atomic relations.

Sulphur 1-57 18*78 0*0935 3 or 18

Bismuth 2*33 27*93 0*0315 1 or 6

Lead 3*36 40*10 0*0309 lor 6

Copper 1*05 12*53 0-0317 1 or 6

8*31 99*34
3Cu2 S, Bi^ S9 + 2(3PbS, Bi"- S^.

This formula is identical with that of the bournonite (from the
analyses of H. Rose, Smithson, &c.), substituting only Bi^ S' for
Sb^ S^ as below : —

Bournonite = 3Cu2 S, Sb^ S^ + 2(3PbS, Sb^ S^)-

in each ore, the electro-negative atoms in the basic compounds
are, to the electro-negative atoms in the acid compounds, as 1 to 1,
as expressed in the accompanying general formula : —

„ / Cu2 s \ , / Sb2 sn

As the bournonite crystallizes also in the same system as the
needle ore, and indeed affects probably the same primary form within
close measurements, the isomorphous relationship of these minerals
is sufficiently apparent.

544f Intelligence a7id Miscellaneous Ar fides.

Most English mineralogists give a right rectangular prism for the
primary form of the bournonite ; but Dufrenoy, in his recent Treatise,
vol. iii. p. 18, after an examination of numerous crystals, considers
a right rhombic prism in which the angle MM = 93" 40' to be the
correct primary form. The modified rectangular prism in which
the bournonite usually occurs is in this light a secondary form, de«
rived from the primary by the replacement of its lateral edges by
the planes g' h' in the notation of Hauy.

The specimen of the needle ore which furnished the above ana-
lysis, exhibited here and there in the quartz transverse rhombic
sections, in which an accustomed eye might easily perceive that the
obtuse angle was included between 90° and 100°.

I could not detect in this specimen the presence of tellurium,
found by John in the needle ore ; it is however perfectly conceivable
that, under certain circumstances, a portion of the PbS may be re-
placed by PbTe. — From the Chemical Gazette for September 1, 1847.

ACTION OF ANHYDROUS PHOSPHORIC ACID ON AMMONIACAL
SALTS. BY M. DUMAS.

The author finds that when anhydrous phosphoric acid is made to
react upon crystallized acetate of ammonia, there distils a liquid the
fixed boiling-point of which is 170° F., and which is miscible with
water in all proportions. When purified by digestion with a saturated
solution of chloride of calcium, and then distilled from solid chloride
of calcium and from magnesia, it still possesses the boiling-point
above mentioned.

Analysis gave the following numbers : —

Experiment. Calculation.

Carbon 57*4 58-5

Hydrogen 7"4 7'3

Nitrogen 34*4 34-2

99-2 100-0

The density of the vapour gave the number 1*45. The above
results lead to the very simple formula C"* H^ N, which differs from
acetate of ammonia by four equivalents less of water. Its composi-
tion is similar to that of nitroguret of acetyle.

But a point of view, which the reactions will warrant, would give
to this substance the following rational formula. C^ NH, C- H^, which
would make hydrocyanate of methylene of it, or an isomeric of it.

The reactions w^hich have been examined gave rise to some curious
phaenomena. Thus solution of potash at a boiling heat disengages
ammonia and regenerates acetic acid ; chromic acid has no action ;
nitric acid is not decomposed by this liquid even when heated to
ebullition. Potassium acts vividly in the cold, and with the disen-
gagement of heat ; cyanide of potassium is formed, and an inflam-
mable mixture of free and carburetted hydrogen gases is evolved.

It is well known that M. Fehling obtained a substance of analo-
gous composition to that now described by distilling benzoate of
ammonia with a naked fire : he did not however attach to the dis-
covery the views which have been now developed, nor did he study

Meteorological Observations. B^s

its reactions. M. Dumas proposes to examine, under the new point
of view described, the action of anhydrous phosphoric acid on the
ammoniacal salts formed by the volatile organic acids.

M. Dumas remarks that if the product which he has obtained
should constitute a compound identical with hydrocyanate of methy-
lene, all these ammoniacal salts, treated in the same manner, should
yield aethers corresponding to certain alcohols, according to the ge-
neral formula —

C»H'»0^ NH3=C~H«-« N=C«-2H'»-2, C2NH.

In decomposing the latter by potash, there might be produced
alcohol C«-2H''-2, 2H0, and prepared by this method, all the
alcohols from the fatty acids. — Comptes Rendus, Septembre 13, 1847.

METEOROLOGICAL OBSERVATIONS FOR OCT. 1847.
Chiaunck. — October 1. Hazy : cloudy. 2. Cloudy. 3. Light clouds and fine :
overcast. 4. Foggy : fine. 5. Fine : light clouds : clear at night. 6. Dense
fog : very fine : lightning and rain at night. 7. Fine : rain : lightning at night :
clear. 8. Very fine. 9, 10. Rain. 11. Rain in forenoon: clear at night.
12. Slight fog: very fine. 13. Foggy : hazy : cloudy at night. 14. Hazy and
drizzly: cloudy. 15. Hazy and cold : slight rain, 16. Foggy : very fine. 17.
Foggy, with slight drizzle : very fine. 18. Slight fog : rain. 19. Exceedingly
fine : rain. 20. Very fine : rain at night. 21. Rain : clear at night. 22. Fine.
23. Densely clouded and boisterous : rain. 24. Slight showers. 25. Very clear :
fine: clear and frosty. 26. Frosty: uniformly overcast. 27. Fine: rain. 28.
Hazy and mild. 29. Exceedingly fine. 30, Overcast and mild. 31. Cloudy
and mild.

Mean temperature of the month ., 52^*14

Mean temperature of Oct. 1846 50 '37

Mean temperature of Oct. for the last twenty years 50 '42

Average amount of rain in Oct 2*60 inches.

Boston. — Oct. \.—5. Cloudy. 6. Rain. 7. Fine: rain p.m. 8. Fine. 9. Fog:
eclipse of the sun invisible until three-quarters over : fog. 10. Rain : rain a.m.
11—13. Fine. 14,15. Cloudy. 16. Fine. 17, 18. Fog. 19,20. Fine.
21. Cloudy: rain a.m. 22. Fine. 23. Cloudy: rain p.m. 24 — 26. Fine.
27. Rain: rain a.m. and p.m. 28. Fog. 29. Rain: rain a.m. 30. Fine:
rain A.M. 31. Cloudy.

Sandwick Manse, Orkney. — Oct. 1. Clear: cloudy. 2. Cloudy: clear. 3.
Cloudy. 4. Cloudy : drops. 5. Bright : showers. 6. Showers. 7. Drizzle.
8. Drizzle : clear : aurora. 9. Clear : cloudy. 10. Cloudy : drizzle. II. Clear :
fog. 12. Fog. 13. Cloudy : clear : aurora. 14. Cloudy : clear. 15, 16. Clear:
cloudy. "17. Showers : drizzle. 18. Rain. 19. Damp : rain. 20,21. Showers:
clear. 22. Showers: rain. 23. Showers: sleet- showers. 24. Sleet -showers.
25. Clear. 26. Drops: showers. 27. Bright: drops. 28. Cloudy. 29. Cloudy:
shower : lightning. 30. Showers : rain. 31. Bright : cloudy.

Applegarth Manse, Dumfries-shire, — Oct. 1, 2. Chill and droughty. 3, 4.
Dull, but fair. 5. Fair a. m. : showery p.m. 6. Heavy rain a.m. 7. Heavy
rain a.m. : flood. 8. Frequent showers. 9. Fine a.m. : rain p.m. 10. Heavy
rain. 11. Fair: rain in the night preceding. 12. Fair and fine. 13. Fair,
but raw and cloudy. 14, 15. Fair, though chilly. 16. Very fine clear day.
17. Dull and cloudy. 18. Dull and cloudy : rain p.m. 19. Heavy rain. 20,
21. Occasional showers. 22. Rain a.m. : very heavy p.m. 23. Rain early
a.m. : fine day. 24. Heavy showers. 25. Fair : fine : clear. 26. Rain nearly
all day. 27. Heavy rain and flood. 28. Fog : cleared p.m. 29. Fair and fine.
30. Fair a.m. : heavy rain p.m. 31. Rain early a.m. : cleared.

Mean temperature of the month 49°'5

Mean temperature of Oct. 1846 49*5

Mean temperature of Oct. for twenty-five years 49 "6

Average rain in Oct. for twenty years 3*56 inches.

Rain in Oct. 1847 509 „

Fhil. Mag, S. 3. No. 2 11 . Suppl. Vol. 31 . 2 N

*1^

^ w

-S^.

■5 g

w c

WO

^ >
Hoi

5j S

•ajiqs
-saijjmnQ

•JiaijASiqo

?iiu,^"o^ s si ii^^ i s « ^- «! s is ^: » ^ i « i p^- 1' « s i| ^^ i «

•ajiqs

•wd I

& C

S P e c B «^ •«* » «■ S " c' c" c' ^ «' s'" ^' 6= St' I ^ ^ ^ fe ^' &* > ^" «• g

I q B g c « u s" i%iili « ol-g " I s s^ ^' i &• tt "^11 ^ ^

. S B

iii'^iiiii'^iji'i'^'^o'ii'iiiiiiiji «• ^; i « g'

O0irt<0-^I>->0-<'00(NC^OOIOCTNOOI<NOMOO'--^'^'^00-<^010'^

3 «

Q

•U!M

•XBK

•Ul'B ?8

•uoisog

OOO^OMOO^XOOt•0•^— 0^^Ot^OC^«c^l0^00^mt0l0^0CJOO^t^lC

I :i

(-!« —tei pHiei F^lei Mjei «Ht« r-bi -^ki mI« fftei .nie* -^Ici

OMOOi coo coco<N lo C5s-< t^-^tci inio inio(o rj<— -^roo m r^iocoioco-*

<N(^^Qoa^co(^^t^co<Moo>0(NO^ooo<Sloo^e^o^^^■rflGOLo^o^t>-^ooooo--

moo^o<oco-«*co<Nvoioc<-^'>*io<N>-<Nr^t^QoaM?rocotor-r^<MoO'^

«s

>

u

0)

1-

»S

«o

S

-ha

:a^

<S '*■

fc. tn
o o

PP

O rt

-wg

CO in •^ t^QO <o-^<oc^coio(0— ocooO"— co<^^^o■^'Oco'toooo^■^co
■>»cocO'-o^o^^*oo^t;^c»o^— ooO'ocOT'coip-rt'— .-r^or^-M(»(»o

O O O O CJ^0^O^0^(3^0^<j^C^^O O O O CMj^0^|^0^0^0^0^0^O (j\0 0^0^0
C0C0C0C00l<M<M«N0t(MCN0IC0C0C0C00l(NC«<N01C^<M0l<NC00<OC»<MC0

lnco<MOcooo>^cooco^Cl^^cooooocooolr^ooI^-<*oo>oo-^o■<*oo^
4> O o 0^o^0^O^o^c^l0^o^O^0^c^lO^0^C^C^lC^l0^O^G^^0^0^O o o o o 0^o^

COCOCOOI<Nr>tC»<N<NCNCNC4C»0<C^C<OIO><M(M(NClCl«NCOCOCOCOC'5«NO<

^ r-- r-> <o v^^ r^ -^t^^^ t^o\<x <x> <3\<x> -^ r^ -^is^^ cr,c^\^^ a\'^ <^<^ op
O O O O O^O^O^O^O^O^O^O^O^O^^O^O^O^O^O^^O^O^CT^O^O C^O 0^0 O^
COC0COC0C<<NOJ(N<MOICH0<O)CNOI(MC<OI<MCN«NOI(NOIC<C0(Nr0OIC0(N

•ta-B?8
•uojson

JO sjJbq

a\co
0^0

o —
6 6

CM 00 rj- -^
-« op lO ip
O On CTi 0^
CO CM C< C»

■^ 0^ O t^
lO CO >o <"0

O CO o -^

lO *0 00 o

•jr r^GO op

CT\ On 0> Oi
C» C4 C< (N

in int^t—

VO 00 OC 00

oi'S'
ro-^
00 00

■<*oo

00 00

COOMO

On CO ON

On -^ uo

« lO «
ON 00 00

»* CO 00
■^ ■^ On
OnOO tO

o ys 00
ooo —
ONO^ t^

C^ 00 00
00 o -^
ror^l>

0» On On
CM CM CM

VO VO ^~
t^ lO CO

Ttoo 9
6n6n6

CM CM CO

CM Ol o

CO in o

ON-^C^

Oico
CO -rf

ON CO

— cor«.

CO t^iO
COCM —

'^i

lo in

O 00

— o

6 6

coco

in ONOO
in OvCM

-« lO t^

covo

00 VO
CM CO

— in r>.

■t — c

CO CO CM

6 6 6

CO CO CO

CM Ol
00 CO
CM CM

6 6

CO CO

-^c<co-^u^v6t^odc^6'He4co'^»n<ot^odo^6'-eMco-^inidi>.o6oi6--'

-li-lF^^F-^r-i-lrHrtCMCMOICMeMOtC^CMCMCMCOCO

« 0 P\ O ^

547

INDEX TO VOL. XXXI.

ACETAL, on the preparation and com-
position of, 77.

Acidii: — hippuric, 127; prussic, 146; ar-
senious, 151; cinnamic, 153; nitrocin-
namic, 154; margaric, 167; oxalic,
233; metacetonic, 266; valerianic, 348,
538; pectic and metapectic, 390; nitric,
454 ; cuminic, 459 ; carminic, 478 ; ni-
trococcusic, 486; chloric, 510; osmi-
amic, 534.

Adams (Mr.) on an important error in
Bouvard's tables of Saturn, 143; on
the elements of Neptune, 380.

Adie (R.) on some experiments with gal-
vanic couples immersed in pure and in
oxygenated water, 350.

Airy (Mr.) on inequalities in the motion
of the moon, 384.

Albumen, action of induced electric cur-
rents on, 249.

Alcohol, action of chlorine on, 77.

Aldebaran, projection of, on the moon,233.

Algebraic equation of the fifth degree, on
the, 341.

Algebraical surfaces, on asymptotic
straight lines, planes, cones and cylin-
ders to, 425.

Alizarine, 46.

Anderson (Dr. T.) on certain products of
decomposition of the fixed oils in con-
tact with sulphur, 161.

Annular eclipse of Oct. 9, 1847, obser-
vation on the, 228.

Antimony, on the salts of, 230.

Apparatus, chemical, on some improved
forms of, 156, 393.

Arseniates, on a new test for, 258.

Arsenious acid, on two varieties of, 151.

Astringent substances, on the means of
testing the comparative value of, 150.

Astronomy, on a new notation for express-
ing various conditions and equations
in, 134.

Atmosphere, on the polarization of the,
444.

Aurora borealis of Oct. 24, 1847, obser-
vations on the, 369.

Balsam of Tolu, on some products de-
rived from the, 153.

Bancalari (M.) on the magnetism of
flame, 421.

Barrcswil (M.) on the dehydration of
moaohydrated sulphuric acid, 314.

2

Bath, analysis of the water of the thermal
spring of, 56.

Bile of the sheep, on the composition of
the, 366.

Binney (E. W.) on fossil calamites found
standing in an erect position in the
carboniferous strata near Wigan, Lan-
cashire, 259.

Birt (W. R.) on a new kite-apparatus for
meteorological observations or other
purposes, 131.

Bismuth, analysis of the needle ore of, 541.

Blowpipe, improvements in the construc-
tion of the hydro-oxygen, 356.

Books, new, notices respecting, 67, 219.

Braconnot (M.) on the urine of the calf
and the sheep, 49.

Brewster (Sir D.) on the modification of
the doubly refracting and physical
structure of topaz, by elastic forces
emanating from minute cavities, 101 ;
on the polarization of the atmosphere,
444 ; on the existence of crystals in
the cavities of minerals, 497.

Bronwin (Rev. B.) on the inverse calculus
of definite integrals, 12; on the alge-
braic equation of the fifth degree, 341.

Brown (J.) on the molybdate of lead, 253.

Buchner (M.) on the presence of arsenic,
copper and tin, in the mineral waters
of Bavaria, 392.

Bussy (M.) on two varieties of arsenious
acid, 151.

Caffeine and its compounds, on the com-
position of, 115.

Calamites, fossil, observations on, 259.

Callan (Rev. N. J.) on a new voltaic bat-
tery, and on a cheap substitute for the
nitric acid of Grove's platina battery, 81.

Cambridge Philosophical Society, pro-
ceedings of the, 130, 301, 376.

Carminic acid, researches on, 478.

Catalysis, observations on, 96, 192.

Chapman (E. J.) on the constitution of
the needle ore of bismuth, 541.

Chloric acid and the chlorates, observa-
tions on, 510.

Clays, on the composition of, employed in
pottery, 435.

Coathupe (M.) on the preparation of gun-
cotton, 152.

Cochineal, researches on, 471, 481.

Colouring matters, on the action of a
N2

548

INDEX.

mixture of red prnssiate of potash and
caustic alkali upon, 126.

Comet of 1264 and 1556, on the expected
reappearance of the, 50.

Commutators, remarks on, 241.

Connell (A.) on the precipitate produced
in spring and river waters by acetate
of lead, 122 ; on the sulphato-chloride
of copper, 537.

Continuity, on the principle of, 137.

Copernicus, on the opinion of, with re-
spect to the light of the planets, 528.

Copper, on the sulphato-chloride of, 536.

Cotton, detection of, in linen, 157.

Couper (R. A.) on the chemical compo-
sition of the substances employed in
pottery, 435.

Creatine, observations on, 236.

Cuminate of ammonia, on the products
of the decomposition of, 459.

Daubeny (Prof.) on active and extinct
volcanoes, 399.

De la Rive (A.) on the voltaic arc, 321.

De la Rue (W.) on a modification of the
apparatus of Varrentrapp and Will for
the estimation of nitrogen, 156 ; on
cochineal {Coccus cadi), 471.

De Morgan (Prof.) on the structure of
the syllogism, and on the application
of the theory of probabilities to ques-
tions of argument and authority, 130;
on the opinion of Copernicus with re-
spect to the hght of the planets, 528.

Domeyko (M.) on vanadiate of lead and
copper, 319.

Doveri (M.) on some properties of silica,
315.

Drach (S. M.) on eliminating the signs
in star-reductions, 251.

Dumas (M.) on the action of anhydrous
phosphoric acid on ammoniacal salts,
544,

Durocher (M.) on the extraction of silver,
317.

Earth, on the amount of radiation of heat
from the surface of the, at night, 69 ;
on the determination of the mean den-
sity of the, 73.

Ebelmen (M.) on the artificial production
of minerals, and especially of precious
stones, 311; analysis of kupfernickel,
314; analysis of gray copper from Al-
geria, 313.

Elastic medium, on the symbolical equa-
tion of vibratory motion of an, 376.
Electric telegraph, on the determination
of differences of longitude by means of
the, 338.
Electro-magnetic influence on flame and

gases, 401, 421,
Equations, on the solution of linear dif-

ferential, 372; monogenous, observa-
tions on, 467.
Euler's theorem, notice in reference to

the extension of, 123.
Faraday (Prof.) on the diamagnetic con-

ditions of flame and gases, 401.
Field (F.) on the products of the decom-
position of cuminate of ammonia by
heat, 459.

Figuier (M.) on the preparation and com-
position of lignine, 397.

Flame, on the diamagnetic conditions of,
401, 421.

Flax, on the chemical composition of the
ashes of, 36, 105.

Fluid motion, on some cases of, 136.

Fluxions, on the invention of, 35.

Forster's (T. J, M.) memoir on meteors
of various sorts, notice of, 219.

Frankland (E.) on the chemical constitu-
tion of metacetonic acid, and some
other bodies related to it, 266.

Fremy (M.) on the gelatinous substances
of vegetables, 389.

Fritzsche (J.) on the preparation and pro-
perties of osmiamic acid and some os-
miamates, 534.

Galloway (R.) on the water of the ther-
mal spring of Bath (King's bath),
56.

Galloway (T.) on the proper motion of
the solar system, 74.

Galvanic couples, account of some experi-
ments with, 350.

Gases, on the re-absorption of mixed, in
a voltameter, 72 ; on the diamagnetic
conditions of, 401, 421.

Geometry, on a new notation for express-
ing various conditions and equations
in, 134; contributions towards a system
of symbolical, 139.

Gladstone (J. H.) on the chemical history
of gun-cotton and xyloidine, 519.

Glaisher (J.) on the amount of the radia-
tion of heat, at night, from the earth,
and from various bodies placed on or
near the surface of the earth, 69 ; on
the Aurora Borealis, as it was seen on
Sunday evening, Oct. 24, 1847, 369.

Gregory (Dr. W.) on the preparation of
hippuric acid, 127.

Griffith (Dr. J. W.) on the composition of
the bile of the sheep, 366.

Grove (W. R.) on certain phsenomena of
voltaic ignition and the decomposition
of water into its constituent gases by
heat, 20, 91, 96 ; correlation of physical
forces, noticed, 67.

Gruner (M.) on bisilicate of iron or fer-
ruginous pyroxene, 78.

Gun-cotton, history of the discovery of,

INDEX.

7 ; on the preparation and composition
of, 152, 519.
Hall (Dr. M.) on the effects of certain
physical and chemical agents on the
nervous system, 72.
Hamilton (Sir W. R.) on quaternions ; or
on a new system of imaginaries in al-
gebra, 214, 278, 511.
Hansen (M.) on inequalities in the motion

of the moon, 382.
Hare (Dr. R.) on the fusion of iridium
and rhodium, 147 ; on certain improve-
ments in the construction and supply
of the hydro-oxygen blowpipe, 356.
Hargreave (C. J.) on the solution of linear

differential equations, 372.
Hearn (G. W.) on the cause of the discre-
pancies observed by Mr. Baily with
the Cavendish apparatus for determi-
ning the mean density of the earth, 73.
Heat, on the amount of radiation of, from
the earth's surface, 69 ; on the mecha-
nical equivalent of, 173.
Hebe, notice respecting the planet, 158.
Heintz (M.) on creatine, 236.
Higginbottom (J.) on the number of spe-
cies and the mode of development of
the British Triton, 74.
Hind (J. R.) on the expected reappear-
ance of the celebrated comet of 1264
and 1556, 50 ; observations of Hind's
second comet in full sunshine, 145 ;
on the planet Hebe, 158; on the new
planet Iris, 237.
Hippuric acid, on the preparation of, 127.
How (H.) on the analysis of the ashes of

the orange-tree, 271.
Hutchinson (J.) on the function of the
intercostal muscles, and on the respi-
ratory movements, with some remarks
on muscular power, in man, 222.
Induction, memoir on, 241.
Ink, invisible, on a new, 176.
Integrals, on the inverse calculus of de-
finite, 12.
Iridium, on the fusion of, 147, 365.
Iris, notice respecting the new planet, 237.
Jacobi (Prof. M. H.) on the reabsorption
of the mixed gases in the voltameter, 72.
Jones (C. H.) on the structxxre and de-
velopment of the liver, 224.
Joule (J. P.) on the theoretical velocity
of sound, 114; on the mechanical equi-
valent of heat, as determined by the
heatevolvedby thefriction of fluids, 1 73.
Kane (Sir R.) on the composition and cha-
racters of certain soils and waters be-
longing to the flax districts of Belgium,
36, 105.
Kindt (G. C.) on the detection of cotton
in linen, 157.

Koenig (F.), inventor of the printing-ma-
chine, 297.
Kolbe (Dr. H.) on the chemical constitu-
tion of metacetonicacid, and some other
bodies related to it, 266 ; on the decom-
position of valerianic acid, by means of
the voltaic current 348.
Kopp (M. E.) on balsam of Tolu, and some
products derived from it, 153 ; on the
action of hydrochloric acid in the for-
mation of oxalic acid, 233.
Ledoyen's disinfecting fluid, remarks on,

233.
Lefroy (Capt. J. H.) on a great magnetic
disturbance on the 24th of September
1847, 346.
Liebig (Prof.) on a new test for prussic
acid, and on a simple method of pre-
paring the sulphocyanide of ammonium,
146.
Lignine, on the preparation and composi-
tion of, 397.
Linen, on the detection of cotton in, 157.
Liver, on the structure and development

of the, 224.
Longitude, on the determination of differ-
ences of, by the electric telegraph,
338.
Loomis (Prof.) on the determination of
differences of longitude by means of the
electric telegraph, 338.
Lubbock (Sir J.) on the perturbations of
planets moving in eccentric and inclined
orbits, 1, 86 ; on the heat of vapours, 90 ;
on the development of the disturbing
function R, 144.
Madder, on the colouring matters of, 46.
Magnetic declination at St. Helena, on the

diurnal variation of the, 70.
Magnetic disturbance, on a great, 346.
Magnetism, influence of, on the voltaic

arc, 328.
Malaguti (M.) on the extraction of silver,

317.
Mannite, action of nitric acid on, 316.
Margaric acid, observations on, 167.
Mechanics, on a new notation for ex-
pressing various conditions and equa-
tions in, 134 ; contributions towards a
system of symbolical, 139.
Mercer (J.) on the action of a mixture of
red prussiate of potash and caustic al-
kali upon colouring matters, 1 26.
Merck (G.) on the water of the thermal

spring of Bath, 56.
Meridian instruments, on the properties
of rock as a foundation of the piers of,
531.
Metacetonic acid, on the constitution of,

266.
Metapectic acid, 389.

550

INDEX.

Meteor of September 25, 1846, notice re-
specting the, 368.

Meteors, observations on, 219.

Meteorological observations, 79, 159, 239,
319, 399, 545 ; on a new kite-apparatus
for, 191.

Meteorology, suggestions for promoting
the science of, 238.

Methylene, on the hydrocyanate of, 544.

Miller (Prof. W. H.) on the measurement
of the double sulphates of zinc and
soda, and of magnesia and soda, 540,

Mineral waters, analyses of, 56, 124 ; on
the presence of arsenic, copper and tin
in some, 392.

Minerals : —ferruginous pyroxene, 78 ; mo-
lybdate of lead, 253 ; gray copper from
Algeria, 313 ; kupfernickel, 314 ; vana-
diate of lead and copper, 319 ; sulphato .
chloride of copper, 537; native car-
bonate of nickel, 541 ; needle ore of
bismuth, ib.

Minerals, on the artificial production of,
311 ; on the existence of crystals in the
cavities of, 497.

Molybdate of lead, analysis of, 253.

Moon, on inequalities in the motion of
the, 382.

Muscles, on the function of the intercostal,
222.

Neptune, on the elements of, 380.

Nervous system, on the effects of certain
physical and chemical reagents on the,
72.

Nicholson (E. C.) on the composition of
caffeine and some of its compounds, 115.

Nickel, on the native carbonate of, 541.

Nitric acid, theoretical views on the na-
ture of, 7 ; on the hydrates of, 454.

Nitrococcusic acid, on the preparation and
composition of, 486,

Nitrogen, on some modifications of the
apparatus for determining, 156, 393.

Numbers, on certain properties of prime,
70; account of a discovery in the theory
of, 189 ; on an equation in, 293 ; on the
partitions of, 301.

O'Brien (Rev. M.) on a new notation for
expressing various conditions and equa-
tions in geometry, mechanics and astro-
nomy, 134 ; on a system of symbolical
geometry and mechanics, 139 ; on the
symbolical equation of vibratory motion
of an elastic medium, whether crytal-
lized or uncrystallized, 376.

Odmyle, on the sulphuret of, 170.

Oils, fixed, on certain products of the de-
composition of, in contact with sulphur,
161,

Optical instruments, on the formation and
application of fine metallic wires to, 53 1 .

Orange-tree, analysis of the ashes of the,
271. ^

Osraiamic acid and osmiamates, on the
preparation and properties of, 534.

Oxalic acid, formation of, 233.

Ozone, on a new test for, 176.

Papyrine, on the preparation and compo-
sition of, 398.

Pectic acid, 389.

Peligot (M. E.) on the preparation and
composition of the salts of antimony,
230.

Phosphoric acid, anhydrous, action of, on
ammoniacal salts, 544.

Pierre (M. I.) on chlorosulphuret of sili-
cum, 78 ; on the equivalent of titanium,
155; on the solubility of chloride of
silver in hydrochloric acid, 398,

Planets, on the perturbations of, 1,86;
on the opinion of Copernicus with re-
spect to the light of the, 528.

Platinum, on the fusion of large masses
of, 356.

Playfair (L.) on transformations produced
by catalytic bodies, 193.

Pollock (Sir F.) on certain properties of
prime numbers, 70.

Pottery, on the chemical composition of
the substances employed in, 435.

Poumarede (M.) on the preparation and
properties of lignine, 397.

Printing-machine, invention and first in-
troduction of, by Kceuig, 297,

Prus«ic acid, on a new test for, 146.

Pyroxene, ferruginous, analysis of, 78.

Pyroxyline, contributions to the chemical
history of, 7, 152, 519.

Quaternions, on, 214, 278, 511.

Reviews : — Grove's Correlation of Physi-
cal Forces, 67 ; Forster on Meteors, 219.

Rhodium, on the fusion of, 147, 365.

Richardson (T.) on the ashes of rough
brown sugar and molasses, 336.

Ronalds (Mr.) on a new kite-apparatus
for meteorological observations or other
purposes, 191.

Roth (M.) on the preparation of the prot-
oxide of tin, 392.

Rowney (T. H.) on the ashes of the orange-
tree, 271.

Royal Society, proceedings of the, 69, 222,
372,

Royal Astronomical Society, proceedings
of the, 143, 380, 528,

Sabine (Lieut.-Col. E.) on the diurnal
variation of the magnetic declination
of St. Helena, 70.

Salt, culinary, on the solubiUty of, in
alcohol, 393.

Salts, ammoniacal, action of anhydrous
phosphoric acid on, 544.

INDEX.

551

Saturn, on an important error in Bou-
vard's tables of, 143.

Schcenbein (Prof.) on the discovery of
gun-cotton, 7 ; on a new test for ozone,
176.

Schunck (Dr.) on the colouring matters of
madder, 46.

Silica, observations on, 315.

Silicium, on the chlorosulphuret of, 78.

Silver, on the extraction of, 31 7 ; solu-
bility of the chloride of, in muriatic
acid, 398.

Slatter (Rev. J.) on the meteor of Sep-
tember 25, 1846, 368.

Smith (Mr. A.) on the hydrates of nitric
acid, 454.

Smyth (Prof. C. P.) on the properties of
rock as a foundation of the piers of
meridian instruments, and on the de-
tection of a cause of error in the Edin-
burgh transit, 531.

Sobrero (M.) on nitric mannite, 316.

Soils and waters of the flax districts of
Belgium, on the composition and cha-
racters of, 36, 105,

Solar system, on the proper motion of the,
74.

Sound, on the theoretical velocity of, 114.

Spinelle, on the artificial production of,
312.

Star-reductions, on ehminating the signs
in, 251.

Stas (M.) on the action of chlorine on
alcohol — formation of acetal, 77.

Stokes (G. G.) on some cases of fluid mo-
tion, 136; on the theory of oscillatory
waves, 138.

Storms, observations relating to the laws
of, 338.

Struve (H.) on the preparation and pro-
perties of osmiamic acid and some os-
miamates, 534.

Sugar, analyses of the ashes of rough
brown, 336.

Sulphocyanide of ammonium, simple me-
thod of preparing, 146.

Sulphuric acid, on the dehydration of,
314.

Sulphates, on the measurement of some
double, 540.

Syllogism, on the structure of the, 130.

Sylvester (J. J.) on a discovery in the
theory of numbers relativatj^o the equa-
tion kx^+By^ + Cz^^Bayz, 189,
293 ; on the general solution (in certain
cases) of tbe equation or^+J/'+A^^
= M xyz, &c., 467.

Taylor (R.) on the invention and first in-
troduction of Mr. Koenig's printing-
machine, 297.

Taylor (T.) on some improved forms of

chemical apparatus, 393.

Therault (M.) on the formation of valeri-
anic acid, 538.

Thompson (L.) on chloric acid and the
chlorates, 510.

Thomson (Dr. R. D.) on a test for arse-
niates, 258.

Tin, on the preparation of the protoxide
of, 392.

Titanium, on the equivalent of, 155.

Tolene, composition of, 153.

Topaz, on the modification of the doubly
refracting and physical stmcture of,
101 ; on the crystals in the cavities of
the, 504.

Triton, on the number of British species
and mode of development of, 74.

Tyrosine, on the properties and composi-
tion of, 496.

Ulrich, Mr,, on the formation and appli-
cation of fine metallic wires to optical
instruments, 531.

Urine of the calf and the sheep, compa-
rative analysis of the, 49.

Valerianic acid, on the decomposition of,
by the voltaic current, 348 ; on the for-
mation of, 538.

Vapours, on the heat of, 90.

Veall (S.) on a means for promoting the
science of meteorology, 238.

Vegetables, on the gelatinous substances
of, 389.

Voltaic arc, researches on the, 321.

battery, description of a new, 81.

current, on the decomposition of

valerianic acid by the, 348.

ignition, on certain phaenomena of,

20, 91.

Voltameter, on the reabsorption of mixed
gases in a, 72.

Wagner (M.) on the solubility of common
salt in alcohol, 393.

Walter (Mr. John), false statements in
the Times newspaper and Mechanics*
Magazine concerning him as regards
Koenig's printing-machine, 297.

Warburton (H.) on the partitions of num-
bers, on combinations, and on permu-
tations, 301.

Warington (R.) on the means of testing
the comparative value of astringent
substances for the purposes of tanning,
150.

Wartmann (Prof.. E.) on induction, 241.

Water, decomposition of, by heat, 20, 91 ;
on the decomposition of, by platinum,
177.

Waves, oscillatory, on the theory of, 138.

Weddle (T.) on asymptotic straight Unes,

68i

INDEX.

planes, cones and cylinders to algebraical

surfaces, 425.
Weld (C. R.) on the invention of fluxions,

35.
"Wilson (Dr. G.) on the decomposition of

water by platinum and the black oxide

of iron at a white heat, 177.
Young (Prof. J. R.) on the extension of

Euler's theorem, 123 ; on the principle
of continuity in reference to certain re-
sults of analysis, 137.

Xyloidine, contributions to the chemical
history of, 519.

Zantedeschi (Prof.) on the motions pre-
sented by flame when under the electro-
magnetic influence, 421. .

END OF THE THIRTY-FIRST VOLUME.

PRINTED BY RICHARD AND JOHN E. TAYLOR,

RED LION COURT, FLEET STREET.

FLAMMAM.

^K-

J^.5.

A

Tig. 2.

^>

S/^

/P

A -^M

/I

Fvg.5.

S

y

rtaiia.nn del.

f

%

I%a. Miff. S.SYolSKZMFl.l

Fi^.3.

Fi^.4.

Fi^.6.

Fi^.6.

Fig. 7.

J'.Sasire, So.

^-:. ^.

J^*-

.J?

5;5*«ew*