TIM-
SCIENTIFIC WORKS

OF

C. WILLIAM SIEMENS, KT

F.K.S., D.C.L., LL.D.
CIVIL ENOINEER.

UNIFORM WITH THE PRESENT WORK.

With Portraits mid Illustrations, 8vo, 16s.

LIFE OF SIR WILLIAM SIEMENS, F.E.S.,

D.C.L., LL.D., Member of Council of the Institution of
Civil Engineers. By WILLIAM POLE, F.R.S., Honorary
Secretary of the Institution of Civil Engineers.

" For many years Sir W. Siemens has been a regular attendant at our
meetings, anil to few, indeed, have they been more indebted fur success.
Whatever the occasion, lie had always new and interesting ideas, jmt forth
in language which a child could understand. It is no exaggeration to say
that the life of such a man was spent in the public service." — LI>I:I>
RAM.EIGH at Uritish Association, 1SS4.

."Mr. Pole had a straightforwaid story to tell, and has told it in a way
likely to interest and instinct a wide circle of readers. Sir William's
biography shows him to have been a man of high talent, which under
effective discipline and the teachings of experience, was successfully and
profitably applied in devising and carrying out undertakings which form
marked features in the history of the period with which this biography
deals." — Tiiiirx.

"So inseparably connected is the career of this remarkable mini with
some of the most wonderful inventions of the past half-century, that the
narrative of his life is in fact a history of the progress, in this country, of
applied science in the two great departments of heat and electricity. As
Mr. Pole justly observes, lie was a civil engineer according to the most
comprehensive definition of that profession — the art of directing the great
powers in Nature for the use and convenience of Man." — .Scofoman.

" The most interesting book of the kind that we have read since
Xasmyth's delightful autobiography." — futui'ilni/ Jit-rim:

"His success in life was doubtless due in no small degree 1o his innate
genius, but this would have been unavailing without the energy and deter-
mination which accompanied it. The record of his life and work should
be vead by every student, and should be in the library of every Mechanics'
Institute in the country. " — ISulltlnr.

" It is of much more than merely technical or scientific interest

his biographer with admirable skill maintains the personal interest of the
narrative throughout." — Manclwster GtuinUan.

" A story of great interest, one sure to become familiar in English-speak
ing households, as well as in those of Germany."— Liverpool Mercury.

" The book will be found full of instruction and interest."— Nature.

THK

SCIENTIFIC WORKS

OF

C. WILLIAM SIEMENS, KT

F.R.S., D.C.L., LL.D.
CIVIL ENGINEER.

A COLLECTION OF

PAPERS AND DISCUSSIONS.

EDITED BY

E. F. BAMBER, C.E.

VOL. I.
HEAT AND METALLURGY.

WITH 47 PLATES.

LONDON :
JOHN MURRAY, ALBEMARLE STREET.

1889.

[All Rights reserved.]

LONDON :
BRADBURV, AONEW, <fe CO., PRINTERS, WHITEFRIARS.

PREFACE.

SIR WILLIAM SIEMENS in his will expressed the wish that
his scientific papers should be published in a collected form,
and his Executors entrusted the task of editing them to
Mr. E. F. Bamber, who had been the private secretary to
Sir William Siemens for the last ten years of his life.

With a view of showing more fully what a variety of
subjects iSir William had thoroughly grasped, extracts
from the discussions of learned Societies have been added,
which show in many cases the independent and individual
point of view from which Sir William Siemens regarded
even familiar subjects.

The papers have been divided, according to their subjects,
into three volumes, dealing with —

HEAT AND METALLURGY (Vol. I.),

ELECTRICITY AND MISCELLANEOUS SUBJECTS (Vol. II.),

ADDKKSSKS AND LECTURES (Vol. III.),

being arranged in each volume in chronological order, and
much care has been bestowed upon making the index really
useful.

VI PREFACE.

A perusal of these volumes will show that, like other
inventors, Sir William Siemens was not uniformly fortunate
enough to attain the results he anticipated from his inven-
tions ; the faithful record of his attempts to overcome diffi-
culties, and of his uniform practice to investigate thoroughly
the scientific basis of the various problems, will, it is hoped,
be of use to those who are following similar lines of research.

These volumes may be regarded as complementary to
" The Life of Sir William Siemens," by Dr. William Pole,
F.R.S. ; and the Executors take this opportunity of publicly
expressing their thanks for the admirable manner in which
Dr. Pole has accomplished his arduous task, and for the
kindness with which Messrs. Macmillan have permitted the
use, as frontispiece, of the portrait of Sir William Siemens
published in Nature shortly after Sir William Siemcns's
death.

The Executors desire to express their appreciation of the
careful manner in which these volumes have been edited by
Mr. Bamber ; and for the valuable assistance rendered in
their compilation, the special thanks of the Executors are
due to Mr. J. Head, who for many years held the responsible
and important position of manager of Sir William's metal-
lurgical business, and who, since Sir William Siemens's
death, has continued to act in the same capacity for Mr.
Frederick Siemens.

It has, moreover, been possible to present the papers in
their present complete form only through the kindness and

PREFACE. vil

courtesy of the Royal Society, Institution of Civil Engineers,
Institution of Mechanical Engineers, Society of Arts, British
Association, Society of Telegraph Engineers, Iron and Steel
Institute, Royal Institution of British Architects, Institute
of Naval Architects, Physical Society, Royal United Service
Institution, Royal Institution, Chemical Society, Society of
Chemical Industry, who most generously placed their trans-
actions and blocks at the disposal of the Executors, for which
they beg also to express their cordial thanks.

THE EXECUTORS.

CONTENTS OF VOLUME I.

PAPERS.
HEAT.

PACK

CONDENSERS FOB LAND AND OTHER STEAM ENGINES ... 1

ON A NEW REGENERATIVE CONDENSER FOR HIGH-PRESSURE AND
LOW-PRESSURE STEAM ENGINES 3

ON THE EXPANSION OF ISOLATED STEAM AND THE TOTAL HEAT

OF STEAM 17

ON THE CONVERSION OF HEAT INTO MECHANICAL EFFECT . . 29
ON A REGENERATIVE STEAM ENGINE 50

ON A NEW CONSTRUCTION OF FURNACE. PARTICULARLY APPLIC-
ABLE WHERE INTENSE HEAT IS REQUIRED 62

ON A REGENERATIVE GAS FURNACE AS APPLIED TO GLASSHOUSES,
PUDDLING, HEATING, ETC 81

ON UNIFORM ROTATION 107

See also ON AN IMPROVED GOVERNOR FOR STEAM ENGINES,

Mcch. Eng. Inst. Proc. 1853, pp. 75-83.
and DESCRIPTION OF AN IMPROVED CHRONOMETRIC GOVERNOR

FOR STEAM ENGINES, Much. Eng. Inst. Proc. 1866, pp. 19-31.

ON A STEAM JET FOR EXHAUSTING AIR, ETC., AND SOME RESULTS

OF ITS APPLICATION 141

See also ON THE STEAM BLAST, Brit. Assoc. Rep. 1871 (Sect.)
pp. 240-241.

ON THE USE OF COAL GAS AS A FUEL 177

See also ON GAS SUPPLY BOTH FOR HEATING AND ILLUMIN-
ATING PURPOSES, Brit. Aasoc. of Gas Managers, 1881.

METALLURGY.

ON THE REGENERATIVE GAS FURNACE AS APPLIED TO THE

MANUFACTURE OF CAST STEEL 209

ON PUDDLING IRON . ... 237

X CONTENTS OF VOLUME I.

I'AGE

ON SMELTING IKON AND STEEL 283

See also ON THE MANUFACTURE OF IRON AND STEEL BY A
DIRECT PROCESS, I. & S. lust. Jour. 1873, p. 253.

SOME FURTHER REMARKS REGARDING THE PRODUCTION OF IRON

AND STEEL BY DIRECT PROCESS 340

ON THE PRODUCTION OF STEEL AND ITS APPLICATION TO MILI-
TARY PURPOSES . 378

DISCUSSION OF PAPERS, ETC.

HEAT.

ON HEATED-AIR ENGINES 2(i, 1(51

ON HORIZONTAL PUMPING ENGINES . . . . . . . 68

ON A COMBINED VAPOUR ENGINE r.9

ON THE APPLICATION OF SUPERHEATED STEAM . . . . 72

ON COMBINED STEAM 76

ON GJFFARD'S INJECTOR 78. 79, 101

ON COKE MANUFACTURE yy

ON THE STEAM-STEERING ENGINE IN THE "GREAT EASTERN" . 124

ON LIQUID FUEL 125, 130

ON COAL GETTING 128

ON REGENERATIVE HOT BLAST STOVES . . . . . 132

ON THE AERO-STEAM ENGINE 13<>

ON STEAM ENGINE GOVERNORS . 139, 159

ON THE EJECTOR CONDENSER 157

ON MODERN LOCOMOTIVES 105

ON PEAT-FUEL MACHINERY 166

ON STEAM-ENGINE CONDENSERS 167

ON COMBUSTION OF REFUSE VEGETABLE SUBSTANCES . . . 168
ON EXPERIMENTS ON STEAM BOILERS ... 170

CONTENTS OF VOLUME I. XI

MM

NI,\V i;i;\ 1.1; .IN<, \ND Kxi'ANsivi; X'.Ai.vi: (\ KAR . . . 178

\S 1'KHDI . 157

ON STEAM-HEATING FOR TOWNS . . ... . . . !!»<>

ON An; Ki.ri;n;i;i:ATiN<; M. \cni\KKV 1'.'-'

i IN 'in i; <i AS K.N<;INK -(>3

METALLURGY.

<>.\ ULAST FURNACES 255, 2(54, 273, 276

ON THE STRENGTH OF IRON AND STEEL -'<;;>

ON TESTINI; KAILS .271

ON PUDDLING FURNACES 280, 353

ON TIIK MANUFACTURE OF IRON AND STEEL BY DIRECT PRO-
CESSES -'503, 328

TIIK PRESIDENT'S (MR. I. LOTHIAN BELL'S) ADDRESS . . . 309

ON Si'lEGELEISEN 311

ON IRON AND STEEL FOR SHIP-BUILDING 313

O.v STEEL MANUFACTURE 315

ON FLUID COMPRESSED STEEL AND GUNS 320

ON A RETORT FURNACE 324

ON ANTHRACITE COKE 325

ON RECENT METALLURGICAL PROCESSES 330

ON USING MOLTEN METAL DIRECT FROM THE BLAST FURNACE . 331

ON CASTING ARRANGEMENTS 335

ON THE PERMANENT WAY OF RAILWAYS 338

ON THE SEPARATION OF CARBON AND IMPURITIES FROM IRON

AND STEEL 353

ON THE PRESENTATION OF THE BESSEMER MEDAL TO PROFESSOR

TUNNER 3(50

ON THE SEPARATION OF PHOSPHORUS FROM PIG-IRON . . . 3<>2

ON RECENT IMPROVEMENTS IN THE MANUFACTURE OF IRON

SPONGE 305

ON STEEL FOR SHIPBUILDING .... :W7, 409, 426, 448
ON STEEL FOR MARINE BOILERS 361)

ON THE MECHANICAL AND OTHER PROPERTIES OF IRON AND MILD

STEEL, AND THEIR USE IN CONSTRUCTION . . 373. 409 419

Xll CONTENTS OF VOLUME I.

PAGE

Ox IKON AND STEEL FOR ARCHITECTURAL CONSTRUCTION . . 402
ON THE ELIMINATION OF PHOSPHORUS .... 413, 416, 436

ON IRON AND STEEL AT Low TEMPERATURES 422

ON STEEL COMPRESSION 429

ON HARDENING IRON AND STEEL : ITS CAUSES AND EFFECTS . 433

ON STEEL-MAKING MACHINERY 437

ON THE CORROSION OF IRON AND .STEEL . . . . " . . 440
ON STEEL PLATES .444

ON A MECHANICAL AGITATOR FOR STEEL MANUFACTURE, AND

THE DISTRIBUTION OF ELEMENTS IN STEEL INGOTS . . 450

ON THE METALLURGY AND MANUFACTURE OF MODERN BRITISH

ORDNANCE 452

ON STEEL FOR STRUCTURES 458

ON THE INFLUENCE OF THE BOARD OF TRADE RULES FOR BOILERS
UPON THE COMMERCIAL MARINE 4<ii>

ON THE USE OF STEEL CASTINGS IN LIEU OF IRON AND STEEL
FORCINGS . 4(U

LIST OF ILLUSTKATIONS.

HEAT.

PLATES

REGENERATIVE CONDENSER 1_2

EXPANSION OF STEAM 3—4

CONVERSION OF HEAT INTO DYNAMICAL EFFECT ... 5

REGENERATIVE FURNACK 6

REGENERATIVE GAS FURNACK 7—18

frtiK Producer 7 — 8

Plate Glass Melt in tj Funnier 9 — 13

Ititiuid Flint (iltlx*. Furiinn- 14 — 15

Puddling Furnace 16 — 17

Steel Mi'Uhiij Fin-nun- 18

UNIFORM ROTATION 19 — 22

Oyrometric (ton rnor 20 — 21

Electric Clock 22

STEAM JET 23—30

As Exhauster or liloiccr 23 — 24

Applied to Pneumatic Despatch Tube 25 — 27

Applied to Raise Water 28

For Sugar Evaporating Pan 29

Applied to 6 as Producer 30

HEATED AIR ENGINE 31 — 32

COAL GAS AS FUEL . . . 33—35

Circular Gas Producer 33

Diagram of Gas Production 34

Regenerative Gas Fire and Gas Lamp 35

REFRIGERATING APPARATUS 36

GAS ENGINE 37

XIV LIST OF ILLUSTRATIONS.

METALLURGY.

PLATES

REGENERATIVE GAS FURNACE APPLIED TO MANUFACTURE OP

CAST STEEL 38—40

Gas Producer 38

Steel Melting Furnace 39 — 40

ON PUDDLING 41 — 42

Gas Producer 41

Puddling Furnace 42

ON SMELTING IRON AND STEEL 43—45

Steel Melt ing Furnace 43 — 44

Jllaxt Furnace I) tar/ mm ........ 45

ON THE PRODUCTION OP STEEL AND ITS APPLICATION TO

MILITARY PURPOSES \\\ — 47

Section of Steel Melting Furnace 4fi

Section of Rotator 47

HEAT.

SIK WM. SIEMENS, F.E.S.

HEAT.

CONDENSER FOR LAND AND OTHER STEAM-ENGINES.
BY C. W. SIEMENS.*

MR. SIEMENS submitted and explained his condenser for land
and other steam engines.

The action of the condenser is as follows : as the jet of steam
passes over the top of the condenser, water is forced through a
series of plates, and passes out at the top in a spray at the rate of
3 feet per second. When the water level falls below the bottom
of the plates, a jet of water is then thrown across the end of the
plates into the water well.

The water level falls to its lowest point during about |th part of
the stroke of the piston ; during the last |th stroke the water is
driven upwards.

The lower end of the plates of the condensers are about 150° of
heat, the upper portion are about 212°. The ascending water
rushes out at a temperature nearly equal to the temperature of the
plates themselves.

The gases are then carried into the hot well by the succeeding
jet of steam.

* Copy Extract from Minutes of Committee of Mechanics ot the Society of
Arts, May, 1850.

VOL. I. B

2 THE SCIENTIFIC PAPERS OF

The ejection pipe at the base of the condenser is furnished
with an air vessel for the purpose of obtaining a greater force
during the time the vacuum is greatest.

The whole condenser bears but a small proportion to the size of
the cylinder, and there would frequently be room for it within the
cylinder.

By Mr. Siemens's condenser, about 6 times the quantity of
water to condense the steam, about 20 are required by the ordinary
process.

The principal application of this condenser is looked for in
high-pressure engines. One of his condensers at work on 16
horse-power engine made by Cappel.

In low-pressure engines there would not be power to force it
into the atmosphere. In this engine three condensations take
place : one at 180°, this condenses the steam 200°, and the
remaining \ in. is steam first cooled by jet ; two, by contact sur-
faces ; and third, condensed at its maximum pressure.

Mr. Grampian sees nothing opposed to the condenser but the
cleaning, and that may easily be managed by simply having a
duplicate set of plates.

Mr. Glynn quite convinced that the principle involved is one
likely to be highly useful.

Resolved: — It is the opinion of this Committee that Mr.
Siemens's Regenerative Condenser is new, highly ingenious, and
appears to be highly valuable in its application to steam engines,
and is therefore well entitled to the Society's reward of their Gold
Medal.

Signed, C. VARLEY.

T. E,. CRAMPTON.

S/A WILLIAM SIEMENS, F.R.S.

ON A NEW REGENERATIVE CONDENSER FOR HIGH-
PRESSURE AND LOW-PRESSURE STEAM-ENGINES.
BY MR. C. WILLIAM SIEMENS.*

THE condenser of a steam engine has for its object the complete
discharge of steam from within the working cylinder after it has
served to propel the piston. This is effected by conducting the
expended steam into a closed chamber, containing an extended'
surface of comparatively cool substance, which absorbs the latent
heat of the steam and thereby reduces it to its liquid state. Cold
water is generally employed for this purpose, which is either
brought into immediate contact with the steam, as is the case in
Watt's injection condenser ; or through the medium of metallic
walls, as in the surface condenser by Hornblower, improved upon
by Hall and others. The more or less perfect condensation of the
steam depends, first, on the absence of air from the condenser, and
secondly, on the temperature at which condensation takes place.

The accompanying table shows the elastic force of steam in

TABLE OP THE PRESSURE OP THE VAPOUR OP WATER, PROM THE
FREEZING TO THK BOILING POINT.

Temperature.
Fahr.

Pressure.
Inches of Mercury.

Temperature.
Fahr.

Pressure.
Inches of Mercury.

32

0-20

130

4-34

40

0-26

140

5-74

50

0-37

150

7-42

60

0-52

160

9-46

70

0-72

170

12-13

80

1-00

180

15-15

90

1-36

190

19-00

100

1-86 ,

200

23-64

110

2-53

210

28-34

120

3-33

212

30-00

vapour, at various temperatures. It will be observed that, in
order to produce a perfect vacuum, the water should leave the
condenser at about 32° Fahr., or be introduced in the form of ice.

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1851, pp. 18—34.

B 2

4 THE SCIENTIFIC PAPERS OP

Condensing water, however, is generally obtained at the tempera-
ture of about 60° Fahr., and it leaves the condenser at about
110° Fahr., which latter temperature implies a remaining atmos-
phere of vapour equal to 2'5 inches of mercury, or in other words
a vacuum of 27'5 inches below the atmospheric pressure at 30'0
inches. If a smaller quantity of condensing water be used, it will
be raised to a proportionately higher temperature, and a less
perfect condensation will be effected. At 212° Fahr. the pressure
of the uncondensed vapour would be equal to that of the atmos-
phere, and the object of the condenser would be entirely frus-
trated.

In all cases where an abundant supply of condensing water
cannot be obtained, or where the heafc of the steam employed by
the engine is reclaimed for other purposes, steam engines are
worked without a condensing apparatus, or at high pressure, at
the sacrifice of an effective pressure nearly equal to that of the
atmosphere upon the working piston. The regenerative con-
denser, forming the subject of the present paper, redeems the
engine from this waste of heat in the one case and loss of
mechanical effect in the other case, being possessed of the peculiar
property of returning the condensing and condensed water at the
initial temperature of the steam previous to its discharge from the
working cylinder, commonly speaking at 212° Fahr. ; producing
nevertheless an efficient vacuum.

Fig. 1, Plate 1, shows a sectional elevation of the regenerative
condenser, as applied to a 10-horse-power high -pressure engine.
It consists of an upright rectangular trunk of cast iron A, the
lower end B of which is cylindrical, and contains a working piston
C. The trunk A is filled with metallic plates D, which are placed
upright and parallel to each other, with intervening spaces of not
less than 1-1 6th inch in breadth. The upper extremity of the
condenser communicates on one side E to the exhaust port of the
engine ; and on the other to the hot well F through a valve G.
A stop H prevents the opening of the valve beyond a certain
distance, in order that it may shut again more instantaneously.
The metallic plates D are fastened together by five or more thin
bolts, with small washers between the adjacent plates, which keep
them the required distance apart. They can easily be removed
from the condenser for the purpose of cleaning, by taking off the

.S7A' \\'ll.l. I.].M SIEMENS, F.R.S. 5

cover I and drawing out the whole of the plates. An injection
K enters the condenser immediately below the plates ; it is
with a small air vessel L, and a regulating cock. Various
modes have been provided to give motion to the displacing piston
C, among which a knee motion worked directly from the beam or
crosshead of the engine is generally found the most convenient, as
shown at M M in Fig. 1.

The action of the condenser is as follows. Motion is given to
the displacing piston C by the engine, causing it to accomplish two
strokes for every one of the engine. At the moment when the
exhaust port of the engine opens, the plates D are completely
immersed in water, a small portion of which has entered the
passage above the plates at A, and is, together with the air present,
carried off, by the rush of steam, through the valve G into the hot
well F, where the water remains, while the excess of steam escapes
at J into the atmosphere. An instant after the partial discharge
of the steam cylinder has commenced, the water recedes between
the plates D, and exposes them gradually to the steam, which
condenses on them in the manner following. The upper edges of
the plates, emerging first from the receding water, are enveloped
in steam. of atmospheric pressure, and in condensing a portion
thereof they become rapidly heated to nearly the temperature of
the steam, or about 210° Fahr. The partial condensation diminishes
the density and temperature of the remaining steam, which re-
quires additional and cooler surfaces for its further condensation.
This is provided for by the continual emerging of additional
portions of the metallic surfaces from the receding water. By
the time the water level leaves the bottom of the plates, the far
greater portion of the steam is condensed. The condensation of
the remaining portion of steam could not so readily be accom-
plished by means of metallic surfaces ; but the piston C, con-
tinuing to descend, causes it to come into immediate contact with
the jet of cold water from the pipe K, which completes the
vacuum in the manner of a common injection condenser. The
air vessel L, connected with the injection pipe, has the effect of
accumulating the injection water at the time when the water has
ascended between the plates, and of forcing it into the condenser
with increased intensity at the time when it is required to com-
plete the vacuum.

6 THE SCIENTIFIC PAPERS OF

Although the action of this condenser is strictly consecutive,
yet it does not check the continuous flow of steam from the
cylinder ; and it completes the vacuum when the working piston
of the engine has accomplished only one-tenth part of its stroke.
Both the engine crank and the crank driving the condenser are on
the top centre at the same moment, but the latter completes its
revolution in the time of half a revolution of the engine crank ;
consequently when the engine piston has passed through only one-
tenth of the whole stroke, the condenser crank will have travelled
through nearly half its stroke, by which time the whole process of
condensation will have been completed.

The principal part of the latent beat of the steam is stored up
in the plates D, the upper extremities of which are heated to
210° Fahr., and the lower to about 150° Fahr. The water, in
ascending again between the plates during the last tenth part of
the engine's stroke, absorbs heat from them in a similar successive
manner, passing first the coolest and by degrees the hottest portions
of their surfaces ; and it issues finally into the upper steam passage A
at a temperature approaching the boiling point, at which moment
a fresh discharge of steam from the engine takes place, which
carries it off into the hot well, as above described, and raises its
temperature fully to the boiling point.

Fig. 3, Plate 2, represents an actual indicator diagram, show-
ing the time occupied in completing the vacuum ; but it will be
observed that the loss of time and power may be diminished by
increasing the capacity of the displacing cylinder. As it is how-
ever, this loss does not amount to one-seventh part of a uniform
vacuum, an equivalent for which is obtained in the saving of the
power hitherto absorbed by the air pump ; for it will be observed
that the displacing piston works between two vacuums, and there-
fore meets with no resisting load.

The quantity of condensing water required with this condenser
to condense 1 Ib. of steam of atmospheric pressure — taking the
initial temperature of the condensing water at 60° Fahr., the final
temperature at 210° Fahr., and the latent heat of steam of 212°

Fahr. at 960 units — is ^,-7^— - = 6*4 Ibs. of water to condense
ZiO — bU

1 Ib. of steam. The common injection condenser, supposing the
condensing and condensed water to issue at 110° Fahr., requires

SSA WILLIAM SIEMENS, F.R.S. ~

960 + (212 — 110)

= 2T2 Ibs. of water to condense 1 Ib. of

ll(i — 00

steam, in place of the 6'4 Ibs. which the regenerative condenser
requires. In the case of a locomotive or other high-pressure
engine, where the steam is released from the cylinder at a pressure
of say 80 Ibs. above the atmosphere, two-thirds would be allowed
to escape uucondensed, and the vacuum would be obtained with
only G'4 -*• 8 = 2*13 Ibs. of condensing water for every 1 Ib. of
steam passed through the cylinders. The small quantity of con-
densing water required renders the regenerative condenser ap-
plicable to engines in nearly every locality ; and pains have been
taken to render the apparatus itself equally light and compact.

The advantages resulting from the application of the regenerative
condenser to high-pressure engines are as follow.

1. Additional effective power, gained on account of the vacuum.
The indicator diagram, Fig. 3, Plate 2, illustrates this gain,
which, supposing the average steam pressure to be 40 Ibs. above
the atmosphere and the vacuum within the cylinder to be 10 Ibs.,
amounts to 20 per cent, irrespective of expansion. If both the
steam pressure and the work on the engine remain unchanged
after the condenser is applied, it is evident that the steam may be
worked expansively to a large extent without diminishing the
absolute driving power of the engine.

2. Heat saved in generating the steam, by the use of boiling-hot
feed water ; and the remaining portion of the hot water may be
advantageously used for heating buildings, dyeing, &c. High-
pressure engines are frequently provided with heating apparatus
for the feed water, which heats it on the average to about the
temperature of the condensing water from low-pressure engines,
or 110° Fahr. The regenerative condenser heats it to 210° Fahr.,

which constitutes a saving of -- r - = about 10 per cent.

When such heating apparatus is not provided, the saving amounts to

210- GO

= about 15 per cent.

3. The steam which is not condensed may be used to cause a
draught in the chimney, or for other purposes.

4. The displacing cylinder, unlike the air pump of the injection
condenser, abstracts no motive power from the engine.

8 THE SCIENTIFIC PAPERS OF

5. The condenser may be started and stopped at any time, by
turning the supply of injection water either on or off. If turned
on, it at once forms the vacuum, without involving the necessity
of blowing through ; and if turned off, it allows the engine to
proceed in the same manner as though no condenser had been
applied.

6. The air contained in the condenser is at the commencement
of each stroke bodily expelled, which is of great advantage to the
formation of a good vacuum, instead of the ordinary air pump
removing only a portion of the air at each stroke, and consequently
leaving a portion always in the condenser.

7. The regenerative condenser is more compact and even less
expensive than the ordinary injection condenser, being less than
one quarter of the size and having only one valve instead of three.
Its proportionate dimensions are as follows : — Area of plate-
chamber, three times the area of exhaust pipe ; length of plates,
one-quarter to one-third of stroke of engine ; thickness of plates,
l-292nd part of their length ; spaces between plates, the same,
but never less than 1-1 6th inch, it having been found that the
alternate rush of water and condensing steam prevents the settle-
ment of grease and earthy matter between the plates, if they are
not less than 1-1 6th inch apart : capacity of displacing cylinder,
one and a half times the capacity of the plate -chamber. The
total capacity of the condenser is equal to only about one-tenth of
the capacity of the working cylinder.

In applying the regenerative condenser to existing high-pres-
sure engines, a saving. of fuel of from 30 to '65 per cent, has been
effected, or an increase of power to that amount with the same
expenditure of fuel as previously. This saving may however be
still augmented considerably, if advantage be taken of the in-
creased effective pressure to work the engine expansively. In
most cases this may be easily effected, by merely adding to the
lap of the slide valve and increasing the lead of the eccentric
proportionately, whereby the additional advantage of a more
early discharge of the steam is obtained.

The first regenerative condenser was attached to a 16-horse-
power high-pressure engine at Saltley "Works near Birmingham in
September, 1849, where it has been found to answer, although it
is not perfect in its proportions, and could not be kept constantly

SfX WILLIAM SIEMENS, F.R.S. 9

in operation in consequence of a deficiency of injection water.
The actual indicator diagram, shown in Fig. 3, Plate 2, was
taken from this engine ; since then several more of the regenera-
tive condensers have been erected, and the results above referred
to have been obtained. The dotted line in Fig. 8 shows the
indicator diagram taken from the engine before the condenser was
applied ; and the full line shows the diagram from the engine
working with the condenser, and exerting exactly the same power
as in the former case. The shaded portion of the diagram shows
the power gained or saved by the use of the condenser.

The advantages attending the application of the regenerative
condenser to stationary engines being practically proved, the author
is desirous to extend it also to that important class the locomotive
engine. In inviting the attention of railway engineers to this
enquiry, he is prepared for practical objections being raised on
account of the great rapidity of motion, the necessity for the
greatest possible simplicity and lightness, the deficiency of con-
densing water, &c. ; but he thinks that the condenser under
consideration is peculiarly well adapted to meet these objections.

The peculiarities of the regenerative condenser in this respect
are as follow. ' It may be accommodated to any speed of piston,
by reducing the length and increasing the breadth of the condens-
ing plates, thus reducing the velocity of the displacing piston
proportionately. Its dimensions are proportionate to the capacity
of cylinder only, and not, as in other condensers, to the horse-
power of the engine. The total weight of a pair of condensers,
as applied to a locomotive engine with cylinders of 13 inches
diameter and 20 inches stroke, is about 3£ cwts. The power of
the blast remains nearly undiminished. The condenser requires
no attention in working the engine, and in case it should fail to
act from any accidental cause the engine will continue to work
high pressure as usual. Moreover it does not interfere with the
working parts of the engine.

The advantages which would result from a vacuum in the cylin-
der of a locomotive engine have been ably set forth by Mr. Edward
"Woods in his " Observations on the consumption of fuel and
evaporation of water in locomotive and other steam engines."
The present paper may therefore be limited to the means proposed
for that purpose.

10 THE SCIENTIFIC PAPERS OF

The two condensers are cast in one piece, and placed immediately
in front of the cylinders of the engine. Each of them closely resembles
the condensers above described ; only the length of the condensing
plates and the stroke of the displacing pistons are much reduced
in proportion to the steam cylinder stroke, in order that the
velocity of the water between the plates may not exceed certain
limits. The two displacing pistons are connected to opposite ends
of a short vibrating beam, which receives its motion from the
engine.

In addition to the exhaust valves leading into the hot well,
these condensers are provided with a second set of discharge
valves, of a somewhat peculiar construction, which with very
limited motion combine the advantage of opening a perfectly
clear passage for the exhaust steam of the engine into the chimney,
where its remaining expansive force is required to produce the
draught. Each' of these valves consists of a longitudinal rectan-
gular slot, in the upper wall of the steam passage which leads from
the cylinder to the condenser. At the ends of the slot are trian-
gular pieces, supporting the sides of two longitudinal lips which
cover the aperture, except at times when a higher pressure from
within forces them open : the extent of their motion is limited by
dead stops. The escape of steam, together with the hot water,
into the hot well is regulated by a blow-off valve from the hot well
into the atmosphere ; by this means a pressure above the atmos-
phere is obtained in the hot well, which acts favourably in forcing
the boiling-hot condensing water into the feed pump of the
boiler.

It has been shown previously that the ordinary supply of feed
water in locomotive engines, where two-thirds of the steam is
allowed to escape uncondensed, is of itself not quite half-sufficient
to maintain a vacuum within the condenser, and an additional
supply of water must be provided for. But considering how
small the excess of condensing water will be, especially if the
diameters of the working cylinders are reduced in proportion
to the additional effective power gained, and considering that
boiling-hot water will readily part with the principal portion of
its heat, it is proposed to take the excess of condensing water back
to the tender through a simple refrigerator, in which advantage is
taken of the rapid motion of the engine through the air for

WILLIAM SIEMENS, F.K.S. II

cooling the water. The refrigerator may be placed conveniently
on the back of the tender.

The application of the regenerative condenser to low-pressure
engines, as shown in Fig. 2, Plate 1, requires but a short notice
after what has been said already ; the letters refer to the same
parts as in the former description of the high-pressure condenser
shown in Fig. 1. In the low-pressure condenser the steam at the
time when it is released from the cylinder has not sufficient force
to expel the air and heated water from- the condenser into the
atmosphere, and a partially vacuous space must be provided for
their reception. For this purpose the side B of the displacing
cylinder, which in the arrangement hitherto described is always
empty, is put in communication with the exhaust valve G of the
condenser, and receives the charge of water and air at the time
when the piston C is at the opposite end ; a second valve 0 is
provided, through which the water is expelled into the hot well F
during the return of the piston C. For the convenience of
arrangement, the displacing cylinder is reversed in this case.

The chief advantages obtained by the application of this con-
denser to the low-pressure engine are : —

1. The requisite amount of injection water is reduced in the
proportion of 3 to 1.

2. The feed water of the boiler is obtained nearly boiling hot,

which constitutes a saving in fuel of = about 10 per

960

cent.

3. The whole amount of heat generated under the boiler is
given off by the engine in the form of water at 210° Fahr., which
in most cases may be advantageously employed for heating build-
ings, washing, dyeing, and other purposes.

4. A large proportion of the power required for working the air
pump is saved.

The author proposes to conclude this paper with a short historic
sketch of the steam engine condenser, to illustrate the distinct
features of this regenerative system.

In Newcomen's engine the condensation of the steam was
effected by the alternate introduction of a jet of cold water into
the steam cylinder itself. The cold water naturally cooled the
walls of the cylinder, which in their turn condensed a large portion

12 THE SCIENTIFIC PAPERS OF

of the succeeding charge of steam before it had forced the piston
upwards.

James Watt, in seeking a remedy against this loss of heat,
conceived the possibility of condensing the steam in a separate
closed vessel ; and in carrying his idea into effect, he not only
realised his immediate object, but at the same time rendered the
steam engine susceptible of that degree of perfection and general
application which it now possesses. The injection condenser of
Watt is the most effectual of its kind, and has maintained its
exclusive dominion to the present day. It consists of a closed
vessel, which communicates periodically with the steam cylinder.
The injection water, together with the condensed steam and the
air, which last is partly evolved from the injection water and
partly leaks in through the joints of the cylinder and exhaust pipe,
are continually discharged from it by means of the air pump.

Shortly after the introduction of Watt's condenser, a surface
condenser was proposed by Hornblower, which consisted of a close
annular vessel of thin metal plate, on the inner surfaces of which
the waste steam of the engine was condensed ; its latent heat being
continually carried off by a stream of cold water which surrounded
the vessel. A comparatively small air pump was provided, which
served to discharge the condensed water (to be again forced into
the boiler) and any air that might leak in through the joints.
This condenser failed in practice, for want of sufficient extent of
cooling surface.

An effective surface condenser would possess considerable advan-
tages over the injection condenser, especially in the case of
marine engines. Allowing the condensed steam to be continually
returned into the boiler, it prevents incrustation of the boiler, and
moreover dispenses with the necessity of blowing off. Its air pump
absorbs a much smaller proportion of the power of the engine, and
its functions require less personal attention. Stimulated by these
considerations, several attempts were made to improve on Horn-
blower's invention ; but since all these improvements partake
very much of the same character, it is thought sufficient for the
present purpose to mention only Hall's condenser, which has
obtained the greatest amount of notoriety. It consists of two flat
chests or closed chambers, connected together by means of a large
number of brass tubes, through which the condensing steam

\\-II.I.lAM SIEMENS, F.R.S. 13

circulates. These tubes are surrounded by cold water, which fills
up the space between the flat chests. A small air pump removes
the condensed water and air from the lower chest. The great
weight and cost of this condenser, its liability to derangement, and
the impossibility of removing the calcareous deposit of the water
from the tubes without taking the whole fabric to pieces, are found
to be serious practical objections.

In the year 1847 the author had occasion to apply a surface
condenser in a situation where economy of space and material
were essential. In considering the most rational distribution of
surfaces, he happened to find an arrangement which, with less than
one half the amount of material used in Hall's condenser, produced
a very satisfactory result, and which paved the way to the more
important improvement chat forms the principal subject of this
paper.

The surface, condenser referred to, shown in Figs. 4 and 5,
Plate 2, consists of a number of copper plates A A, of ^Vnds inch
thickness and about 4£ inches broad by 2 feet long, which are
fixed together by two longitudinal flattened wires of the same
metal between the adjacent plates ; and the whole pile is screwed
up tight together between the sides of a rectangular cast iron
vessel, which constitutes the body of the condenser. The ends of
the plates project through the top and bottom of the condenser,
and are planed flush with its exterior surfaces. The joints at top
and bottom are secured by means of india-rubber rings, which are
screwed down under small cast iron frames, and yield to the
difference of expansion between the two metals. The flattened
wires are laid parallel, about 3 inches apart from each other, and
form with the plates a large number of narrow passages, through
which the cold condensing water flows in a vertical direction,
without entering the vacuous space of the condenser, into which
the edges of the plates outside the flattened wires project, forming
the condensing surfaces.

The rationale of this condenser is as follows. The transmission
of heat in a surface condenser is threefold : first, from the con-
densing steam to the internal metal surfaces ; secondly, from the
internal surfaces through the body of the metal to its external
surfaces ; and thirdly, from the external surfaces to the surround-
ing water by which it is carried off. The first operation (condensa-

14 THE SCIENTIFIC PAPERS OF

tion) would, it is presumed, proceed with undefined rapidity, if
it were not retarded by the second and third, or by the presence
of some permanent gases, which accumulate on the condensing
surfaces and prevent their immediate contact with the steam. The
second (conduction) varies in direct proportion with the conduct-
ing power of the metal, and with its thickness ; but the conducting
power of copper is so great that its thickness seems to exercise no
appreciable influence on the amount of heat transmitted in a given
time. This interesting fact is proved by Dr. lire's experiment
with two copper pans, of the same internal area, but very unequal
thickness of bottom, the thicknesses being in the proportion of I
to 12 ; which were both filled with water, and dipped into a hot
solution of muriate of lime. It was found that the water in the
thick pan evaporated the quickest, which may be accounted for by
its slightly increased external surface in contact with the heating
solution ; and this affords additional evidence that the limit of
transmission does not lie within the metal, but rather between the
metal surface and the liquid. That the absorption of the heat by
the water is a slow process may be inferred from the circumstance
that water, although possessing a large capacity for heat, is a very
bad conductor, and depends for its power to absorb heat on the
slow circulation over the heating surface caused by the lower
specific gravity of the heated particles of water. A strong artificial
current along the heating surfaces greatly accelerates the process.

The surface condenser above described was arranged in accord-
ance with these observations. It contains : — heat-absorbing surfaces
(by the water), 18 square feet per horse power ; condensing sur-
faces, 9 square feet per horse power ; computed mean thickness
of metal through which the heat is transmitted, 1| inch ; weight
of copper, 60 Ibs. per horse power ; space occupied by plates, 0'4
cubic foot per horse power, or about one-tenth of the space occupied
by the tubes in the tubular condenser.

The essential features of this condenser are its comparative
cheapness of construction, and the easy access which it affords to
the water channels between the plates. It also requires less con-
densing water than previous surface condensers, in consequence
of the repeated and close contact in which each particle is brought
with the heating surfaces, before it can reach the upper reservoir
or hot well. The author considers that the surface condenser just

SIR WILLIAM SIEMENS, F.R.S. 15

described may be advantageously applied to marine engines ; and
not being subject to a patent, he hopes it will receive a sufficient
trial.

I'.cing required to save the waste steam of a low-pressure engine
in the form of slightly heated water, for Mr. John Graham of
Manchester, the author in the spring of 1847 conceived the idea
of a regenerative condenser. Figs. 4 and 5, Plate 2, show his
first arrangement as already described, which may be termed a
regenerative surface condenser. It consists of a revolving valve
B, which admits the waste steam of the engine first to the atmos-
phere at C, and then successively into the separate compartments,
D, E, F, G, where it is condensed at various densities. The cold
water enters at H, and first passes between the plates within the
last compartment G, and by degrees through those within the first
compartment D, where the steam is of nearly atmospheric pressure,
and consequently heats the water to nearly 212° Fahr., when it
passes out at I.

The next step was a regenerative injection condenser on the
same principle, as represented by Fig. 6, Plate 2. The revolving
valve B admits the waste steam of the engine first to the atmos-
phere at C, and then successively into the separate compartments,
D, E, F, G, where it is condensed at various densities. The cold
water is injected at H, and is passed down through the steam in
each compartment in succession by means of the displacing
pistons K K, which work all on the same piston rod through each
of the divisions between the compartments ; and the heated water
passes out at the bottom at I. L L are overflowing distributing
trays, for the purpose of bringing the water more rapidly and
completely into contact with the steam. M is a small pump to
extract the air that is mixed with the steam and water.

The regenerative condenser in its present form partakes of the
nature of both the surface and the injection condensers.

Attempts have been made from time to time to condense the
steam of a high-pressure engine without the aid of an air pump, by
blowing the steam into a small injection condenser which is
provided with a large exhaust valve. It is clear that the steam of
high pressure will at first partially blow through the condenser, and
rid it of its air and condensing water ; and that by degrees the jet
of cold water will overpower the influx of steam, and consequently

1 6 THE SCIENTIFIC PAPERS OF

produce a vacuum. An arrangement of this description, although
simple, is at least very imperfect ; because it is a matter of
considerable difficulty so to proportion the injection of cold water
that the first rush of steam may not forthwith be condensed, but
may exert its expansive force in a cold vessel, and yet that an
instant afterwards a complete condensation of the remaining steam
may be effected. If too much water be used, the air and water
will not be expelled, and consequently no vacuum will be formed ;
if too little, no final condensation will take place. The quantity of
injection water must be very large, because the whole of the steam
has to be condensed ; and having to complete the condensation in
the same vessel, the water must leave it at a low temperature.

The principle of the regenerative condenser has been carried
still further in the regenerative engine, which has been executed
on a large scale by Messrs. Fox, Henderson & Co., under the
superintendence of the author. In this engine the steam, after it
has served to propel the working piston to the end of its stroke,
is received into a series of consecutive chambers, from which it
returns to the working cylinder an indefinite number of times.
On a future occasion the author will be glad to bring the particu-
lars of this engine before the Institution.

MR. SIEMENS, in answer to enquiries, replied that the experiment
at the Saltley Works, had been tried with one week's working with
the condenser, and then one week without it ; and the saving of fuel
with the condenser was at the rate of 18 per cent. The apparatus
with which the condenser worked was however too light, and had
not been made for the purpose ; also that condenser was the first
that had been made, and the proportions had been improved in the
subsequent ones.

There was a deficiency in the supply of condensing water, which
sometimes interfered with the regular working of the condenser, as
well as the defects arising from the gearing being too light for
working it, and these had caused irregularities in the working of
the engine ; there was also a difficulty in regulating the engine as
a condensing engine with the present governor. The steam pressure
was 30 Ibs. per inch ; but a smaller supply of condensing water
would be sufficient if a higher pressure of steam were employed.

As regarded its use in locomotive engines, Mr. Siemens replied

SIR WILLIAM SIEMENS, F.R.S. 1 7

that it would only be necessary for the condenser to work quick
enough to condense one cylinder-full of steam before the next
cylinder-full was discharged ; and this he thought would easily be
effected by widening the plates of the condenser to a proportionate
size, and shortening the stroke of the condenser piston, so as to
reduce its velocity as far as might be required. The condensed
steam would then return to the tender in pipes, between which air
was caused to circulate by means of the rapid motion of the
engine. He expected the condensing water would be cooled down
to about 100° before it was returned to the tender, by the process
of passing through the pipes of the refrigerator, from the rapid
motion of the engine through the air ; and the water was not re-
quired to be so cold as in the ordinary condenser, since only the
last portion of the steam was condensed by injection.

Mr. Siemens showed by a comparative indicator diagram that
with the application of the condenser to a locomotive engine the
steam might be cut off at about one-third of the stroke, instead of
at two-thirds as usual, and thereby a saving of one-half the steam
would be effected with the same power.

ON THE EXPANSION OF ISOLATED STEAM, AND
THE TOTAL HEAT OF STEAM.

BY MB. CHAKLES "W. SIEMENS, of LONDON.*

THE object of this paper is to lay before the members the
results of certain experiments on steam, purporting, in the first
place, to corroborate Regnault's disproval of Watt's law, "that
the sum of latent and sensible heat in steam of various pressures
is the same ; " in the second place, to prove the rate of expansion
by heat of isolated steam ; and, in the third place, to illustrate
the immediate practical results of those experiments in working
steam engines expansively.

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1852, pp. 131—141.

VOL. I. C

1 8 THE SCIENTIFIC PAPERS OF

The author pursued these experiments at long intervals since
the year 1847, with no other object in view than to extend his
own information ; and, consequently, without pretence to
generalisation or extreme accuracy. The question, however, is
one of great practical importance to engineers, and with the
advantage of valuable suggestions and the co-operation of his
friends, Mr. Edward A. Cowper and Mr. William P. Marshall,
the author has again taken up the experiments, which, having
been referred to at the previous meeting by Mr. Cowper, he feels
himself called upon to lay before this Institution in their present
state, though incomplete.

The amount of heat required to convert one pound of water
into steam of different pressures has occupied the attention of
natural philosophers from the earliest periods of the modern
steam engine.

Dr. Black observed, about a century ago, that a large quantity
of heat was absorbed by water in its conversion into steam (not
accompanied by an increase of temperature), which he termed
" the latent heat of steam." His apparatus consisted simply of
a metallic vessel containing water, which he exposed to a very
regular fire ; and from the comparative time which was occupied,
first in raising the temperature of the water to the boiling point,
and, secondly, in effecting the evaporation, he approximately
determined the amount of latent heat. Resuming the experi-
ment, in conjunction with Dr. Irvine, he employed a different
apparatus, consisting of a steam generator, and of a surface
condenser, or 'a serpentine tube, surrounded by a large body of
cold water.

The steam which condensed in the serpentine tube was care-
fully collected and weighed, and the rise of temperature of the
surrounding water was observed, which, multiplied by its known
quantity, represented the total quantity of heat which the steam
had yielded.

The quantity of heat requisite to raise the temperature of one
pound of water through 1° Fahr. being taken for the unit of heat,
Black and Irvine obtained for the total quantity of heat in

Steam of atmospheric pressure, the number . . 954
Southern . 1021

SSK WILLIAM SIEMENS, F.R.S. 19

Watt obtained the number 1140

Regnault 1145

Dr. Ure 1147

Desprez, 11 8G, but later li:»-2

Brix 1152

Gay Lussac and Clement 1170

Count Rumford 1206

All of these eminent experimentalists employed essentially the
same apparatus, and the differences between their results proves
its great liability to error. Brix, of Berlin, was the first to
investigate those errors, and to calculate approximately their
effect upon the results obtained.

"While such a large amount of labour and talent has been
expended to determine the latent heat in steam of atmospheric
pressure, a far more important question seems to have been
passed over with neglect, namely, What is the relative amount of
heat in steam of various densities ?

The celebrated Watt justly perceived the importance of this
question, but contented himself with one experiment upon which
he based his law, "Uiat the sum of latent and sensible heat in
steam is the name under all pressures."

Southern repeated the experiment, and found that steam of
greater density contained absolutely more heat than steam of
lower pressure, which induced him to adopt the hypothesis that
" the latent heat of steam was the same at all pressures"

Subsequent experiments and general reasoning seemed to be in
favour of Watt's law, which enjoyed the general confidence until
it was attacked, only a few years since, by Regnault, of Paris,
who proved by a series of exceedingly elaborate and carefully
conducted experiments, that neither the law of Watt nor that of
Southern was correct, but that the truth lay between the two.
The apparatus employed by M. Regnault may be said to be a
refinement upon those previously employed, and with the
advantage of Brix's labours to determine the amount of errors,
he seems to have succeeded in measuring the absolute amount of
heat in steam of various pressures with surprising accuracy.

The costly and complicated nature of the apparatus employed
by M. Regnault has hitherto prevented other experimentalists

c 2

20 THE SCIENTIFIC PAPERS OP

from repeating the experiment, and in the meantime practical
engineers still continue to adhere to Watt's law.

Shortly after the publication of Regnault's experiments by the
Cavendish Society, in 1848, the idea occurred to the author of
the present paper that their results might be brought to a
positive test by a simple apparatus, which he places before the
Meeting in operation, shown in Fig. 1, Plate 3. It consists of
an upright cylindrical vessel of tin-plate A, which is surrounded
by an outer vessel filled with charcoal B B, or other non-conducting
material. A steam-pipe C, with a contracted glass vein D, enters
the inner vessel in a slanting position, in order that the water of
priming from the boiler, and of condensation within the pipe,
may return to the former, allowing only a small jet of pure steam
to enter the vessel, where it suddenly expands and communicates
its temperature to the bulb of a thermometer E, which is inserted
through a stuffing-box from above. The lower extremity of
the inner vessel A is connected on the one hand to a mercury
gauge Gr, and on the other to a condenser, by means of a stop-
cock to regulate the pressure. The pressure and temperature of
the steam within the boiler being known, and the temperature of
the expanded steam observed, it will be seen whether that tempera-
ture coincides with the temperature which is due to pressure
indicated by the mercury gauge. If it did, then Watt's law
would be confirmed, but since the temperature rises higher than
is due to the pressure, it follows that the high-pressure steam
contains an excess of heat, which serves to super-heat the
expanded steam. All losses of heat from the apparatus would
tend to reduce the temperature, and be in favour of Watt's law ;
but it will be shown that those losses may be entirely eliminated,
and a true quantitative result be obtained. For this purpose the
pressure in the boiler should first be raised to its highest point,
and the indicating apparatus be well penetrated by the heat ; the
fire under the boiler should thereupon be reduced, and observa-
tions made simultaneously, and at regular intervals, of the
declining pressure within the boiler, and temperature of the
expanded steam of constant pressure. The pressures being nearly
equal, the fire under the boiler is again increased, and the
observations continued until the maximum pressure is once more
obtained ; and the loss of heat by radiation, &c., may be cor-

WILLIAM SIEMENS, F.R.S. 2 1

rectly estimated, by a comparison of the two series of observa-
tions.

Tin; second portion of this paper relates to the rate of expansion
of isolated steam by heat, that is, steam isolated from the water
from which it is generated.

The author has not been able to meet with any direct experi-
ments on this subject, except some at a recent period by Mr. Frost,
cl1 America, which, however, do not seem entitled to much con-
fidence. The rate of expansion of air and other permanent gases
by heat was first determined by Dalton and Gay Lussac simul-
taneously, who determined that all gases expanded uniformly, and
at the same absolute rate, amounting to an increase of bulk equal
to 4-J-oth Part °f the total bulk at 32° Fahr. for every one degree
Fahr., or -o^th part of the total bulk at 212°. Dulong and Petit
confirmed the law of Dalton and Gay-Lussac, but it appears that
these philosophers confined their labours to the permanent gases
and atmospheric pressure, and merely supposed the general appli-
cability of their discovery.

Being interested in the application of " super-heated " steam,
the author tried some direct experiments on its rate of expansion,
in the year 1847, which confirmed his view, that vapours expand
more rapidly than permanent gases, or in other words, tliat the rate
of expansion of different gases and vapours is equal, not at the same
absolute temperature, but at points equally removed from their
points of generation,

The apparatus employed in these experiments has been placed
before the meeting, and its simplicity, when seen in operation, is
such that the result, it is hoped, can hardly be doubted.

It is shown in Fig. 2, Plate 3, and consists of a metallic trough
A A, containing oil, which is placed upon a furnace B B, heated by
gas flames. One end of the trough is provided with a stuffing-
box, through which a glass tube C, of about TVth inch diameter,
and sealed at one end, may be slipped, which will rest horizontally
upon a scale below the surface of the oil. The mouth of the
glass tube is connected to an open mercury syphon G, with
either the one or the other leg filled with mercury, to produce the
desired pressure within the horizontal glass tube. A small drop
of water and a piston of mercury P being introduced into the
bottom of the tube, it is placed in the oil bath, and connected

22 THE SCIENTIFIC PAPERS OF

to the syphon. The oil bath is then gradually heated, and the
temperature observed. As soon as the boiling point of water
under the pressure in question is reached, the mercury piston will
move rapidly forward, until all the water is converted into steam.
The temperature continuing to increase, the piston will continue
its course more slowly upon the scale, where its progress is noted
from time to time, together with the temperature.

The experiment is continued until the temperature reaches about
400", when the oil begins to boil. The gas flame is then with-
drawn, and the bath allowed to cool gradually. The observations
of the temperature and the position of the mercury piston are
continued until the steam contained behind it is recondensed. A
comparison between the two series of observations gives the
correct mean of the experiment, by which the effects of the
friction of the mercury piston, any possible slight leakage of
steam past it, and faults consequent on the slow transmission of
heat, are completely neutralized.

The curve A on Fig. 3, Plate 4, has been drawn, expressing
the rate of expansion of isolated steam according to these ex-
periments. The results of nine separate experiments very nearly
coincide (as shown by the dotted lines, which give the ex-
treme variation in the experiments), except at the starting-
point, where the rate of expansion is so very great that it is
difficult to obtain correct observations ; changes in the barometer,
moreover, affect the curve in the vicinity of the boiling point.
To obviate the effect of these inaccuracies, the unit of volume in
laying down the curves from each of the nine experiments was
taken, not at the absolute boiling point, but at 250°, where the
expansion had already assumed a definite course.

The diagram also shows a straight line B, expressing the rate of
expansion of common air, which at first diverges greatly from the
hyperbolic curve of expansion of steam, although the asymptote
of the latter seems to run parallel to the former. The author
considers it therefore highly probable, " that the rate of expansion
of all gases may te expressed by one hyperbola, which starts from
the condensing point of the ffas," and that the apparently uniform
rate of expansion of the permanent gases may be accounted for by
their great elevation, at the ordinary temperature, above their
supposed boiling point, in consequence whereof the true curve

.s/A1 \VII.UAM SIEMENS, l-.R.S. 2$

approaches so nearly to its asymptote that the difference cannot
be detected by experiments.

The general result obtained from the above experiments may be
stated as follows : that steaui generated at 212*, and maintained
at a constant pressure of one atmosphere, when heated out of
contact with water to

230° is expanded 5 times more than air would be.

240° „ 4

260° „ 3

370° „ 2

The author intends to extend the range of his experiments upon
gases and vapours under high pi'essure, and will be glad to
communicate the further results to the Institution.

The diagram contains another curve, C, showing the results of
Mr. Frost's experiments, alluded to before, which, from the very
sudden and irregular rise at the commencement, appears to be
affected by some serious source of error.

The two curves of pressure and density, P and D, show the rate
at which saturated steam increases in pressure and in density, with
the rise of temperature marked at the bottom of the diagram. It
will be observed that the pressure increases at a rather greater rate
than the density ; and it is a remarkable circumstance, that the
difference, or the rate at which the pressure increases faster than
the density, which is in effect the rate of expansion of saturated
steam with the increase of sensible temperature, exactly coincides
with the line B, representing the rate of expansion of atmo-
spheric air.

An extension of our knowledge on the properties of steaui is a
matter of such evident importance to engineers, that it would be
useless to dwell upon its practical importance. Suffice it to say,
that it has been theoretically demonstrated that a perfect Boulton
and Watt Condensing Engine (abstracting friction and all losses of
heat in the furnace and through radiation) would only yield about
seven per cent, of the mechanical force which would be equivalent
to the expended heac. It may be argued from this, that the
steam engine is destined to undergo another great modification in
principle, and in the author's humble opinion this crisis will be
accelerated by inquiries into those properties of gaseous fluids
which have hitherto excited but little attention, and especially
into the properties of dry steam, or isolated steam.

24 THE SCIENTIFIC PAPERS OF

The present paper will be confined to showing the effect of the
above experiments upon the rate of expansion of steam within the
steam cylinder of an engine. It was demonstrated by the first-
named experiments, that expanded steam is super-heated steam;
and, by the second, it is shown what is the expansion of bulk due
to an increase of temperature.

Supposing the results of the experiments to be correct, the
expansion curve as laid down by Tambour, and which is based
upon Watt's law, requires a modification due to the excess of
temperature in expanded steam ; and it will be observed that this
correction in the curve of expansion is in favour of working
engines expansively, as a greater average pressure is obtained
during expansion than would be the case if the expanded steam
were not thus super-heated. Its correctness is corroborated by
some actual observations by Mr. Edward A. Cowper in taking
diagrams of expansive engines, previous to his acquaintance with
the above experiments. It moreover appears that in Cornwall,
engineers have been practically acquainted with the fact, that
expanded steam is super-heated steam, and more economic in its
use than saturated steam ; for it is a practice with them to
generate the steam at very high pressure, and to expand it down to
the required pressure previous to its reaching the steam cylinder.

Another remarkable practical observation is, that a jet of high-
pressure steam does not scald the naked hand, while a jet of low-
pressure steam does, although the high-pressure steam is the hotter
substance. The cooling effect of a jet of high-pressure steam is so
powerful, that, as the' author has been informed, ice has been
actually produced in the heat of summer in America, by blowing a
powerful jet of steam of 400 Ibs. pressure per square inch against
a damp cloth. This phenomenon may be explained by the perfectly
dry and under-saturated state of expanded steam, which, with a
strong tendency to re-saturate itself, produces a powerful evapora-
tion on moist surfaces with which it comes in contact.

The rapid rate of expansion of steam by heat, when still near
its boiling point, proves the economy of heating the steam
cylinder either by a steam jacket, or by the application of fire.
It is, however, important to observe, that the specific heat of
steam seems to diminish, the more the temperature exceeds the
boiling point. The annexed table of observations gives the data
from which the curve A in Fig. 3 has been drawn.

WILLIAM SIEMENS, F.R.S.

TABLE OK EXPERIMENTS ON THK EXPANSION OP ISOLATED
ATMOSPHERIC 8TKAM.

Ti-ini>«-r:i-
ture.
BjMMM

1

Is*

<~

2

J-

|.S

3

T3

g ti

4

i«

4-s

5

•3
£ SO

6

•i
Sg>

El

7

i

g ti

8

•a

g to

Descend- ^ \
ing.

209

05-0

210

5-00

212
213

8-00

8-00

10-00

10-00

9-70

8-00

...

'.••D|

I'll
216
: 2174
220
2224

L'L':."
2274

230
-2324

8-80
9-00

no

,,.._,.,

932

:i;.vi

8-40
8-68
8-90
9-11

<J;:i4
»58

10-40
10-52
10-66
10-84
10-94
11-01
11-11
11-21
11-29

10-'l2
10-20
10-30
10-48
10-68
10-60
10-68

i o-or,

10-16
10-32
10-50
10-61
10-70

l6-88

...

9-50
9-90

10-50
10-70

11-00

9-30

9-50
9-60
9-68
9-75
9-81

!)-30

;t-4:,
9-67
9-65
9-74

9-91

240
L'|:,
250
266

2(10
861

9-68
9-80
9-94
10-10
10-21
10-31

9-70
9-85
9-96
10-05
10-15
10-25

11-34
11-46
11-58
11-70
11-80
11 "JO

10-84
10-94
11-04
11-18
11-30
11-40

11-00
11-12
11-23
11-35
11-47
11-59

11-90

11-16
11-34
11-49
ll-(iO
11-71
11-83

946

10-06
10-19
10-29
10-40
10-50

10-02
10-13
10-23
10-33
10-44
10-54

266

268

10-41

11T-35

12-00

11-51

11-70

12-10

11-94

10-60

10-64

270
276

278

10-51
10-60

10-44
10-53

...

11-61
11-73

11-80
11-91

12-18
12-30

12-08
12-16

10-70
10-80

10-75
10-85

280

284

10-70

10-62

11-85

12-02

12-35
12-45

12-28

10-90

10-96

285
290
294

10-80
10-90

10-72
10-81

...

n-ys

12-10

12-14
12-26

12-66

12-64

12-40
12-50

11-00
11-10

11-06
11-17

295

298

10-98

10-91

12-20

12-39

12-75

12-60

11-20

11-27

300
305
310
B16
320
836
330
335
340

11-08
11-18
11-26
11-36
11-46
11-56
Ll-68
11-73
11-83

11-01
11-11
11-21
11-31
11-42
11-52
11-64
11-75
11-85

...

12-30
12-40
12-51
12-62
12-73
12-85
12-98
13-10
13-21

12-50
12-58

1L'-6'.I

12-80
12-90
13-02
13-15
13-25
13-36

12-79

12-88
13-00
13-10

12-70
12-80
12-95
13-08

11-30
11-40
11-50
11-60
11-71
11-81
11-91
12-02

11-38
11-48
11-58
11-69
11-79
11-89
11-99
12-08

M6

11-93

11-95

13-33

13-41

860

12-02

12-05

18-48

18-60

355

12-11

12-18

360

12-20

12-26

365

12-30

12-40

370

12-40

12-50

375

12-50

12-55

880

12-60

12-60

26 THE SCIENTIFIC PAPERS OF

MR. SIEMENS explained the action of the two instruments,
and showed their process in operation.

Mr. Crampton inquired whether the charcoal in the casing of
the instrument would not get heated by the tube of high-pressure
steam passing through it during the experiment, and so super-
heat the steam in the internal cylinder ?

Mr. Siemens explained, that it was not possible for such an
effect to take place, as the end of the steam-pipe was exceedingly
small, and was protected by a thick non-conducting casing. He
had also observed several times during the experiments, that
whenever any priming took place in the boiler, and a drop of
water came out with the steam, and fell on the bulb of the
internal thermometer, the mercury fell immediately to 212°, or the
boiling point of water, and remained steadily there for four or
five minutes, until the whole of the priming water was converted
into steam, when the mercury again gradually rose to its former
temperature. This showed that the increased temperature above
212° in the internal cylinder was entirely due to the extra heat
in the expanded high-pressure steam, and not to any heat derived
from the charcoal casing. As a check on the accuracy of the
observations, he had tried them successively in an ascending and
descending series, when an error from the source alluded to would
have been apparent, and^ been doubled in effect, but he could not
detect more than one degree difference in the observations.

In the discussion of the Paper

"ON THE USE OF HEATED AIR AS A MOTIVE
POWER," by Mr. B. CHEVEKTON,

ME. C. W. SIEMENS,* after sketching a diagram explanatory of
the action of Ericsson's engine, stated that he had not seen the
machine, but he believed the description which had been given of

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers
Vol. XII. Session 1852-1853, p. 345 etscq.

.v//i' \\-Il.LIAM SIEMENS, F.R.S. 27

the arrangement was substantially correct. He had followed its
progress with considerable interest, having himself been engaged,
for a number of years, in maturing an engine, in which steam
was employed in a highly heated state.

The pistons being on their bottom stroke, the air from the
reservoirs was admitted below, urging the working piston upward.
Being heated, in its passage through the regenerator to 400° Fahr.,
the volume of the air was increased in the proportion of 2 to 3,
and hence the reservoirs were deprived of only two-thirds the
contents of a working cylinder of compressed air. An additional
means of economising the supply of compressed air was, by shutting
off the admission, before the upward stroke was completed, allow-
ing it to act expansively.

The pumping cylinder had, in the meantime, discharged its
contents of fresh atmospheric air into the reservoir, to make up
for the supply of the working cylinder, so that the pressure of
10 Ibs. per square inch was always maintained. The position of
the slide-valve being then reversed, the air from beneath the
working piston was free to escape into the atmosphere, but having
to pass through the regenerator, the free and sensible heat con-
tained in it was restored to the metallic wire gauze, in the inverse
order to that in which it was taken up, issuing finally at a little
above the temperature at which it entered from the reservoir.

The descending stroke was effected by the mere weight of the
pistons. When completed, the position of the slide-valve was
again altered, and the air from the reservoir entered and forced
the piston upward. In its passage through the regenerator it
absorbed the heat which had been deposited there, and with an
additional supply from the fire, its volume was again doubled, in
filling the working cylinder.

Supposing the action of the regenerator could be made perfect,
so that the air left the regenerator at exactly the same temperature
at which it had entered, it might seem at first sight, that the
engine would work without any expenditure of heat, beyond the
mere losses from radiation, &c.

This view had indeed been maintained by Captain Ericsson and
others ; but upon consideration it became apparent, that there
was a theoretical consumption of heat, which might be very accu-
rately calculated, from the fact that the air entered the regenerator

28 THE SCIENTIFIC PAPERS OF

in a compressed state and returned through it, after expansion to
atmospheric pressure had taken place. This expansion was accom-
panied by a diminution of temperature of some 70° or 80° Fahr.,
which became latent and had to be replaced by the fire.

The theoretical consumption of a perfect caloric engine,
amounted to only one-fourteenth part of the theoretical consump-
tion of a Boulton and Watt condensing engine. The practical
arrangement of Ericsson's engine, however, rendered the attain-
ment of such a result impossible, for the following reasons : — •

Fully two-thirds of the power of the engine must be expended
in working the air pump independent of the resisting pressure of
the atmosphere, which was equal to three-fifths of the total
working pressure. The consequence was, that to produce the
effective displacement of the piston, for one single volume of air at
its full pressure, from 7 to 8 volumes had to be cooled and heated
alternately.

The working piston of Ericsson's engine had, moreover, to
work air-tight in a heated cylinder, which Mr. Siemens had
practically found to be a matter of great difficulty. The lubri-
cating material would become rapidly carbonized, and would fill
up the meshes of the regenerator.

The extent of heating surface provided, also appeared to be too
small for the quantity of heat required to be transmitted. It was
understood, from good authority, that the present working cylinders
of 14 feet diameter, were now being replaced by others of 16 feet,
which was the greatest size the breadth of beam of the vessel would
permit.

Mr. Armstrong's views required correction, owing to the
evolution of heat in compressing the air in the pumps which
would produce expansion and increased resistance. A corre-
sponding cooling effect was, moreover, produced through its
expansion in the working cylinder, which would diminish the
power shown by him to be obtainable. These objections would,
he expected, mar the anticipated results of this interesting experi-
ment.

SfK WILLIAM SIEMENS, F.R.S. 29

ON THE CONVERSION OF HEAT INTO MECHANICAL

EFFECT.
BY CHARLES WILLIAM SIEMEXS,* Assoc. Inst. C.E.

THIS subject may be considered under three heads.
First, an inquiry into general qualitative and quantitative
relations between heat and mechanical effect.

Second, the theoretical and practical consideration of actual
engines, including those of Stirling and Ericsson.

Third, the definition of the characteristics of a perfect engine.
The first portion relates to a purely theoretical question, and
would, separately considered, fall beyond the usual limits of
discussion at this Institution ; but the author is obliged to ask
for an exception in his favour, finding it would be impossible to
establish the ultimate object in view, without having proved his
premises, which are based upon evidence of recent discoveries.

In discussing the succeeding heads he will have to rely, to a
considerable extent, on his individual judgment and experimental
researches.

First. On the relations between " Heat and Mechanical
Effect."

The power obtained from a steam-engine depends upon the
increase of volume given to the water in its transformation into
steam, by the action of the fire under the boiler. Dr. Black
observed, in 17G3, that the effect of the fire was, for the most part,
required to effect the conversion, after the water had been raised
to the temperature of the steam itself ; and, moreover, that it
made no difference whether the evaporation took place in the open
air, or in a closed vessel under pressure. Upon these facts he
grounded his theory, that steam was a compound of water and heat,
which heat, on entering into combination with the water, lost its
individual properties, or became latent.

This " material theory " of heat has been generally adopted, in
preference to the " theory of undulation," according to which heat

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers,
Vol. XII. 1852-1853, pp. 571-590.

30 THE SCIENTIFIC PAPERS OF

is regarded as the undulatoiy motion of a supposed eetherial
fluid pervading all nature.

The supporters of the material theory explain the different
phenomena of sensible, radiant, and latent heat, by the free or
combined state of their supposed fluids ; the " specific heat of
bodies by their different degrees of affinity for that fluid." The
affinity of materials for heat is supposed to be invariably increased
by increase of volume, and the evolution of heat, by friction be-
tween solids, is supposed to arise from permanent compression of
their particles.

The latter supposition has been disproved by Sir Humphry
Davy, who showed, that heat was evolved by friction between two
pieces of ice, which caused them to melt, and could not, therefore,
arise from permanent compression of the solid particles.

Dulong proved, moreover, that the specific heat of gases is the
same before and after compression, showing that the heat lost in
their expansion is not absorbed into the gas, and cannot be
accounted for according to the " material theory." Joule, of Man-
chester, produced heat by agitating water in a closed vessel, and
also by an electric current, which, in its turn, was produced by
power, in turning the handle of a magneto-electric machine.

The latter experiments are not only proofs against the suppo-
sition that heat is material ; but their greater value consists in
showing an intimate connection between heat and the mechanical
force by which it was produced, and they are the foundation of
the " dynamical theory of heat."

According to this theory, in its general form, heat, mechanical
force, electricity, chemical affinity, light, and sound, are but
different manifestations of one great and infinite cause —
" motion."

Eecent discoveries and experimental researches all accord with
this great principle, which seems destined to open a new era of
natural sciences.* Dulong and Gay-Lussac have proved, by their
experiments on sound, that the greater the specific heat of a gas
the more rapid are its atomic vibrations. Elevation of tempera-
ture does not alter the rapidity, but increases the length of those
vibrations, and in consequence produces " expansion " of the
body.

* Vide Grove "On the Co-relation of Physical Forces," 1842 and 1850.

.S7A' WILLIAM. SIEMENS, FJLS. 31

The specific heat and temperature of a body determine the
vihmtiim velocity of the material particles, the square of which
multiplied by the weight of the particles, gives their inherent
force, or "vis viva."

In solids the "vis viva" is least remarkable, in fluids it is
greater, and in gaseous fluids it predominates so strongly over
the gravitation, that the latter force becomes practically inap-
preciable.

Joule explains the elastic pressure of a gas by the rebound of
its particles against the side of the vessel containing it,* and proves
the correctness of his views by calculation. If one side of a vessel
gradually yields to the pressure, as is the case with a working
piston, then the rebound of the particles will be less than their
impact, and consequently the length of their vibrations must
diminish, in proportion to the outward motion produced.

It is thus shown, that vibratory motion, or heat, is converted
into its equivalent of onward motion, or dynamical effect.

To express this equivalent by number, it is necessary to agree, in
the first place, on an arbitrary unit of heat, which is usually the
heat required to raise the temperature of 1 Ib. of water through 1 °
Centigrade, or Fahrenheit, and also, on a unit of mechanical effect,
which is usually the " foot-pound," or the power required to lift
1 Ib. through 1 foot. The data for these calculations may be
taken from any materials, the specific heat and density of which
are well known.

The nature of the material cannot affect the result, for if one
should be more favourable to the production of mechanical force
by heat than another, and the second be more favourable to the
reverse process, it would follow that by a judicious selection of
materials, a machine might be devised, which would reproduce
more than its own cause of motion. This would be " to ascribe
creative power to matter," in contradiction to the laws of nature.

The limits of this paper do not permit of the train of reasoning,
by which the numerical equivalent of power for heat has been
ascertained, in different ways, by English, French, and German
natural philosophers, within the last ten years ; the author must,
therefore, content himself with merely mentioning their names and

* Vide Transactions of Philosophical Society of Manchester, 1848.

32 THE SCIENTIFIC PAPERS OF

publications. Carnot and Clapeyron produced the first general
formula?, which contained, however, an uncertain function, and
were still based upon the supposition that heat was material.*
Holtzman, of Manheim, in pursuing the views of Carnot and
Clapeyron, obtained a complete mathematical solution in 1845.f
Joule, of Manchester, solved the problem experimentally about
the same time.J

The " dynamical theory " was more fully developed by Helm-
holtz in 1847,§ and Mayer.|| Mr. W. J. M. Eankine, C.E., and
Professor Thomson, of Edinburgh, have much extended the
dynamical theory of heat, and applied the same to calculate the
power of steam and air engines. IF

M. Regnault, of Paris, has, by careful experimental researches
on the total heat of steam, &c.,** provided some important data for
the development of the dynamical theory of heat.

The following are the results obtained in units of power, or foot
Ibs., for one unit of heat, by different authors : —

Centigrade
Thermometer.

Fahrenheit's
Thermometer.

By Holtzman's formula ....
By Joule's experiments . . . .
By Rankine's formula ....
By Thomson's formula . . . .
For the best Cornish engine, by M. de \

1227 foot Ibs.
1386 „
1252 „
1390 .,

148 „

682 foot Ibs.
770 „
695 „
772 .,

82 .,

For a perfect low-pressure condensing \
engine /
For an actual Boulton and Watt's engine

90-8 ,.
46 „ .

50-4 „
25-5 „

The comparatively small effect produced by the steam-engine of
the present day would seem to indicate that there is still much
room for improvement.

Practical engineers will probably receive with incredulity, and

* Vide Annales de Chimie et Physique, XXIII.

t Vide his pamphlet ' ' Ueber die Waerme und Elasticitaet der Gase und
Daempf e. "

J Vide Transactions of the Philosophical Society of Manchester, Vol. XVIII.

§ Vide Ueber die Erhaltung der Kraft.

|| Vide Mechanische Aequivalent der Waerme, 1851.

\ Vide Transactions of Royal Society of Edinburgh, 1849-1850 and 1850-1851.

** Vide Comptes Rendus.

S7X WILLIAM SIEMENS, F.R.S. 33

certainly derive but little advantage from, the preceding numerical
statement, the result of abstract calculation, unless it can be proved
by simple demonstration, and in such a manner that the essential
difference between the actually and the theoretically perfect engine
is clearly pointed out.

The author proposes to accomplish this by means of a diagram
(Fi.u. 1, Plate t>), which is, in effect, the expansion curve of
saturated steam indefinitely prolonged.

The vertical lines and figures at the bottom signify the pres-
sures of saturated steam in Ibs. per square inch ; the horizontal
lines and figures on the sides denote the volume of steam compared
with the volume of water from which it is produced ; and the
horizontal lines, with figures on the curve, express the tempera-
tures of the steam corresponding to the pressure and volume of the
same.

The outer curve, a a «, is that usually employed in calcu-
lating the power of expansion engines, being the expression of
Watt's law " that the total heat in steam is the same at all
pressures."

The inner curve, b 1 7>, is the corrected one, in accordance with
the recent discoveries of Regnault, " on the total heat of steam," *
and may be termed " the curve of equal heat."

The fields between the horizontal dotted lines represent the
power given out by the steam, in losing equal decrements of 10°
of temperature in its expansion, and it is important to observe,
that the areas of these fields are nearly alike, between the limits to
which the pressure and temperature of steam are experimentally
known, increasing only slightly, and in a uniform ratio inversely
as the temperatures.

This gradual increase may be ascribed to the fact, that the
curve in question is one of equal heat, whereas it has been
shown, that in expanding behind a working piston, the steam
must lose heat in the dynamical proportion to the power given
out.

Messrs. Rankine and Clausius have first drawn attention to
this circumstance, and proved that the expansion of steam, behind
a piston, must be attended by partial condensation.

* Vide the Publications of the Cavendish Society, Vol. I.
VOL. I. D

34

On the other hand, experiments made by the author* prove
that steam, when expanded spontaneously, is superheated steam,
being a verification of Eegnault's discovery, that the total heat of
steam increases with its pressure.

When steam is expanded behind a working piston the excess of
free heat is first absorbed, or changed into dynamical effect, and
if that does not suffice, partial condensation must take place. It
appears, however, to the author, that Messrs. Rankine and
Clausius undervalue .the amount of free heat, and, therefore, over-
estimate the amount of condensation during expansion, by taking
the specific heat of steam at 0*305 (Regnault's co-efficient of
increasing heat), which there seem good reasons to believe is far
too low, because : —

1st. The specific heat of an elastic fluid must be proportionate
to its rate of expansion by heat. It has been shown, however, in
experiments instituted by the author above referred to, that the
rate of expansion of steam near its point of saturation is about
three times greater than that of air at the same temperature,
which would make its specific heat 3 x -267 = ' 801, diminishing
however rapidly with the increase of temperature.

2nd. The actual forms of diagrams, taken from the best ex-
pansive steam engines, do not show the effect of condensation :
the ordinates of the lower portion of the curve are indeed higher
than those given by Watt's law, starting from the same point.
This is shown by Fig. 2, Plate 5, which is a diagram taken from
the Old Ford Engine, by Mr. W. Pole,t in which a a a is the
actual curve, b b J-the curve representing Watt's law, and c c c
the curve of equal heat. The rise of the actual curve towards the
end may in part be owing to the generation of steam from the
heated sides of the cylinders ; but it could not be supposed that
the effect of such generation would equal that of spontaneous
condensation throughout the body of the steam. Moreover the
actual curve proves to be almost the perfect dynamical curve, as is
proved by the equal areas, or fields of power, obtained by drawing-
lines from points of the curve of equal progression of temperature.
If the limits of the sheet of diagrams had admitted of a con-

* Vide Proceedings of the Institution of Mechanical Engineers, for June 1,
1852, p. 24, ante.

•f Vide Treatise on the Cornish Engine, by W. Pole, Part III.

.svA- nv/././. M/ .SY/-;.)//-:.V.V, /--.as. 35

Initiation of the curve horizontally (Fig. 1, Plate 5), it would
have exhibited continually decreasing volumes, and increasing
temperatures, until finally a point would have been reached, where
the volume of the steam was equal only to the water producing it.
It may be assumed, that the temperature would, at that point, be
640° Centigrade (11HO° Fahr.), or in other words, that the entire
heat of the steam would be sensible. Supposing this steam (which
would have at least 2,o()0 Ibs. pressure per square inch) could be
introduced below a piston, and in giving out its power be ex-
panded, until its temperature was reduced to zero, then the entire
640 degrees of heat would have been converted into their
equivalent of power, of which the field of the diagram would
represent the integral. The theoretical equivalent of mechanical
force for heat is thus represented to the eye ; and in computing
the area of the entire figure, it is found to coincide nearly with,
but somewhat to exceed, the results of abstract calculation and of
'Mr. Joule's experiments. That portion of the diagram (Fig. 1)
which is shaded darker than the rest, represents the power of a
perfect low-pressure condensing engine ; it covers only about l-14th
part of the entire area.

The diagram shows that the expansive steam-engine would
be theoretically a perfect engine, if the water was heated in a
close boiler to 1180° Fahr., and being then introduced below
the working piston, under a pressure of at least 2,000 Ibs., would
resolve itself entirely into steam, and was allowed to expand 2000
times before it was discharged into a vacuous space, which, in
this case, would not necessarily be a condenser. The impracti-
cable nature of such an engine is manifest, and it becomes
necessary to seek for other means of obtaining from heat its full
value of power.

It may not be considered out of place to mention here, the well-
known experiments on steam-guns by Mr. Perkins, which went far
to prove the actual possibility of charging water with sufficient
heat, in close vessels, that, upon liberation, it would resolve itself
entirely into steam.

Before leaving this part of the paper, it will be necessary to
show the effect produced by the expansion of air, or any other
permanently elastic fluid, under a working piston. In the
diagram (Fig. 3, Plate r>), the figures on the base and vertical

D 2

36 THE SCIENTIFIC PAPERS OF

lines denote the pressure in Ibs. per square inch, and the figures
on the side denote volumes of air, as compared with the volume
of the same weight of water. The curve a a a represents
Marriotte's law, and is a curve of equal heat. The curve b b 1) is
the dynamical curve, representing the real rate of expansion of
air, behind a working piston. The difference between the two
curves arises from the loss of sensible heat, which is converted
into effect. The figures upon the curve show the rate of pro-
gression of temperature, in compressing air of atmospheric
pressure at 60° Fahr.

In constructing this curve, the specific heat of air at constant
volume has been taken at '267 as determined by Delaroche and
Bernard. It furnishes itself at least an approximate proof of the
correctness of this number, because the curve agrees with the ob-
served fact, that in compressing air, to double its original
pressure, its temperature is raised 70° Fahr. The dotted hori-
zontal lines limit the uniform fields of power obtained for every
10° decrease of temperature. This curve is directly applicable to
air, which when reduced to atmospheric pressure has a temperature
of 60° Fahr. It can, however, easily be corrected, for any degree
of temperature, by adding the difference of temperature at a
corresponding pressure throughout, and by adding to each volume
the same difference of temperature, divided by 508 (the ratio of
expansion of air at 60° Fahr.) *

Secondly. — " On the performance of actual engines, including
the air-engines of Stirling and Ericsson."

* The dynamical theory of heat must not be considered the creation of the last
few years, but, like all abstract truths in nature, it seems to have presented itself
to the minds of the greatest philosophers in all ages, to be, after them, again
superseded by theories moulded, as it were to order, to explain some isolated
phenomenon, such as the radiation of the sun, the heating flame of a fire, or the
latent heat in steam ; until at length the means of observation were sufficiently
perfected to cover with absolute proof, what could before be reached only by the
imagination. It may here be mentioned, as instances, a quotation by Baron
Humboldt from Aristotle, " who considered the first principle in nature to be a
unity in all its manifestations, and the manifestations themselves he reduced
always to motion as their foundation." Again, in Lord Bacon's Aphorisms, the
chapter on "The first Vintages of the Force of Heat," occurs the following
remarkable passage : — "From the instances taken collectively, as well as singly,
the nature whose limit is heat, appears to be motion." And further on, — " But
that the very essence of heat, or the substantial self of heat, is motion, and
nothing else limited,'' <fcc., &c. Bacon fails, however, in his attempts to prove
his philosophy, in confounding the visible motion of heating water, or of fire,
with the intrinsic motion of the particles that manifests itself as heat.

.SYA" WILLIAM SIEMENS, F.R.S. 37

I ii accordance with the principles put forth, the power of an
miue is expressed by a simple formula:

in which c is the velocity of the piston in feet per minute ; a the
area of the same in square inches ; t the temperature of the
steam, or air, on entering the cylinder ; t' its temperature on
leaving the same ; v expresses the volumes of the steam, or air,
on entering the cylinder, as compared to one volume of water ;
r the ratio of expansion, or fraction of stroke at which the supply
is shut off; A a constant, denoting the mechanical equivalent,
per unit of heat, being (as shown by the diagrams) for steam =
400, and for air = 0'267 x 400 = IOC ; p the pressure of the
fluid on entering the cylinder (pressure in boiler) in Ibs. per square
inch ; and p' the pressure against the working piston (back
pressure) in Ibs. per square inch.

The power required to work air, or feed pumps, has to be
deducted from the result of this formula.

Take, for example, an expansive and condensing steam-engine
of 16 inches diameter and 200 feet velocity of piston ; the total
pressure of steam in the boiler GO Ibs., cut off at one-fourth part
of the stroke ; the vacuum in the cylinder averaging 11 Ibs.
(having 4 Ibs. resisting pressure) ; the initial temperature of the
steam = 295° ; and the final temperature If = 207° Fahr. ; the
volume at 60 Ibs. pressure would be = 460 (see diagram, or any
table on the pressure, temperature, and volume of steam).

The indicated horse-power of this engine will be

lOZ)) - 36-7,

The evaporation in the boiler is 2 ^ - = 9'1 cubic feet of
water per hour.

The result agrees with that obtained by the usual method of
computing the contents of the expansion curve, and is certainly
more accurate and more expeditiously arrived at.

r A (t — t') has no value,
For non-expansive engines, the factor -

because / = i\ and r = 1.

38 THE SCIENTIFIC PAPERS OF

In applying this formula, to ascertain the power of an air-
engine, the value of the constant A is = 106, as shown by the
diagram, Fig. 3, Plate 5.

If the object is to ascertain merely the relative .economy of
an engine, as compared with a perfect engine, it suffices to
determine —

1st. The total units of heat which are imparted to the working
fluid, and, 2nd, the units of heat which disappear in producing
useful effect ; and inasmuch as the former exceed the latter, so
the engine falls short of producing a full equivalent of mechanical
effect for the heat expanded.

Take the example of an air-engine consisting of a working
cylinder, A, an air-pump B, and a reservoir D between them
(Fig. 4, Plate 5). To obtain the greatest effect, the admission of
air into the working cylinder should, under all circumstances, be
so regulated, that it may expand down to atmospheric pressure
before it is discharged.

Supposing that nothing was known of the proportion between
the cylinders, of the working pressure, or of the rate of expan-
sion, but that the temperature of the air was known to be, —

60° Fahr. On entering the pump at m.

130° „ On entering the vessel D, which would be the case
if compressed to half its original volume.

710° „ On entering the working cylinder, which would be
the heat required to double its volume at con-
stant pressure, and

570° „ On being discharged, having lost 140° in its expan-
sion down to the atmospheric pressure.

Then the heat supplied by the fire would be = 710— 130 =
580° Fahr. and the difference of temperature of the air, on
entering and on leaving the engine, would be = 570 — 60 =
510° Fahr. It follows, that 580 — 510 = 70° of heat have
been converted into their equivalent of mechanical effect, and
the duty performed by the engine, for every one unit (Fahr.) of

70
heat employed, is = -^ 770 = 91 '2 Ibs. lifted 1 foot high.

The expansive air-engine is, therefore, theoretically superior
to Boulton and "Watt's condensing engine, but inferior to a good

WILLIAM SIEMENS, F.R.S. 39

» xjiansive engine. Practically considered it is certainly inferior
to both, because one-half of the gross power of the working piston
is absorbed by the pump, and the losses by friction and leakage
are trebled in consequence. Moreover, it has been found by its
earliest promoter, Mr. Stirling, of Dundee, that the working of &
tight piston, in a highly -heated cylinder, is attended with almost
insurmountable difficulty.

The most essential difference between the steam-engine and the
air-engine is, that in the former, the unproductive heat is expended
in the boiler, where it becomes latent, in effecting increase of
volume without displacement of the piston, whereas, in the latter,
it presents itself as free heat at the exhaust port.

Mr. Stirling, and after him Captain Ericsson, in taking advan-
tage of this circumstance in favour of the air-engine, employed
the free lost heat to warm the fresh air on entering the reservoir
from the pump.

Ericsson constructed an engine on this plan, in 1833, which,
according to received accounts, worked with considerable economy
of fuel, but failed, in consequence of a defective heating apparatus,
and the continual derangement of the working piston in a heated
cylinder.

The apparatus he employed for recovering the lost heat re-
sembled a locomotive boiler, through the tubes of which the cold
air passed, while the heated and expanded air circulated in the
opposite direction, through the intervening spaces. He termed
this apparatus the " regenerator," because he supposed, that by
its means the same heat might be used perpetually over and over
again, to produce motive power.

Stirling, of Dundee, patented an air-engine in 181 G, and
improvements in 1827 and in 1840. He constructed one of
these machines, an account of which he brought before the
Institution in 1845.* In it he had combined several important
advantages over former attempts, namely —

1st. The hot working piston was dispensed with.

2nd. The working pressure was increased, by using the same
body of highly-compressed air over and over again.

* Vide Minutes of Proceedings of the Institution of Civil Engineers for 1845,
Vol. IV. p. 348.

40 THE SCIENTIFIC PAPERS OF

3rd. The reabsorption of the waste heat was carried out to
greater perfection, by means of a series of thin iron plates, pre-
senting a very large aggregate surface, which were held a small
distance apart by fillets, to allow of the passage of air between
them.

The lower extremities of these plates were heated by the fire to
about 650° Fahr., while the upper extremities were maintained at
the lowest possible temperature, by coils of cold-water pipes.
The air was made to pass to and fro between the same surfaces,
for every stroke of the engine, by means of a large displacing
plunger, which was not required to work tight in its cylinder.
In descending, the air absorbed heat, in the gradual proportion of
the increasing temperature of the plates, and in consequence its
elastic pressure was increased during ihe ascending stroke. By
the reverse process, a fall to the former temperature and a
decrease of pressure, inversely proportionate to the temperature
and the space occupied, was effected, which space was, in the
meantime, increased by the amount of the capacity of the working
cylinder.

The opposite ends of the working cylinder communicated with
two distinct heating apparatuses, the displacing plungers of which
were attached to the opposite ends of a beam and made stroke,
while the working piston was on its dead centres.

The excessive pressure of the heated air, beneath the one, over
the cold air above the other displacing plunger, constituted the
working pressure ; and the capacities of the displacing cylinders
had to be so proportioned to the working cylinder, that the work-
ing pressure was not exceeded by the resisting pressures at the end
of each stroke ; or supposing the volume of the air was doubled
by the heat, it follows, that the net capacity of each displacing
cylinder had to be equal to at least twice the capacity of the
working cylinder.

Fig. 5, Plate 5, is a theoretical diagram of Stirling's engine ;
the curve a a a represents the entire expansion of the air, from
the time when it is all confined in the heated space, below the
displacing plunger, to the moment when it occupies the cold
extended space, above the displacing plunger, and the working
cylinder. The power due to this expansion is measured by the
field a a a x y z, and is equivalent to 123 units of heat, as shown

WILLIAM SIEMENS, F.K.S. 41

by Fig. 8, Plate 5 ; whereas the power actually imparted to the
working piston is represented by the portion of the diagram
marked b b b, the corners of which are rounded off, in the ratio
produced by the two cranks working over right angles, and is
equivalent to 27'5 units, being about -Jth parts of the entire area.
The field b b b y z, immediately below the sectioned portion,
represents the back pressure on the working piston, or the power
required to force the air from the working cylinders back into the
displacing vessels.

The theoretical effect produced by a Stirling's engine, is to
that of a perfect engine, as the units of heat expended in the
entire expansion to the units producing useful effect, or as 123
to 27-5.

There remains to be considered the necessary mechanical loss
of heat, owing to the difference of temperature between the air
on entering the cool ends of the metallic plates (the regenerators),
and on returning from the same, which loss has- been proved, by
experiments hereafter to be explained, to amount to about 25°
Fahr., or to f x 25 = 19° Fahr. if distributed upon the entire
quantity of air employed, because only three-quarters of the same
are actually heated each time.

The real effect of a Stirling's engine, as compared with a perfect
engine, is therefore as 123 + 19 : 27'5 = 5'1G : 1, or it yields

770

the power of — - r == 149 Ibs. lifted 1 foot high, for every water
O'lb

unit of heat expended.

It follows from the above, that the Stirling's engine con-
verts into dynamical effect about one-fifth of the heat supplied,
which is about equal to the performance of the best Cornish
engines.

For calculating the actual power of a Stirling's engine of given
proportions, the formula assumes the following more simple
form : —

Ind- H-p- = x iQ which A = 10° and

t - 11 the decrease of temperature, by the expansion of the air
from the greatest to the lowest pressure.

The cause of the failure of Mr. Stirling's engine in practice
may apparently be traced, chiefly to insufficiency of heating

42 THE SCIENTIFIC PAPERS OF

surface, occasioned apparently from misapprehension of the
principle involved, it having been thought that the same heat
would serve over and over again to produce power, and that the
necessary expenditure of heat consisted only in the mechanical
loss by imperfect action of the respirative 'plates, which were
approached to each other to the utmost limits, consistent with an
unobstructed passage of the air.

By the aid of the dynamical theory of heat it has been shown,
that there is another and far more important expenditure of heat
which should have been provided for.

Another great practical defect in Stirling's engine, arises from
the necessity of providing a reservoir of highly-compressed air to
start with, and from the difficulty of preventing the escape of that
stock of air, through the stuffing-boxes, &c.

Ericsson patented in 1851 another form of engine, which has
lately been executed on a gigantic scale and continues to excite
public attention.

This engine, of which Fig. G, Plate 5, represents the theoretical
diagram, differs from the expansive air-engine (Fig. 4, Plate 5),
only in the application to it of Stirling's respirator, or regenerator,
and in the proportion between the capacities of the working and
pumping cylinder, which in his engine are as three to two. A is
the working cylinder, the bottom of which is made of wrought
iron, and exposed to the fire ; B is the pumping cylinder, which
draws in atmospheric air through the valve u and delivers it into
the air reservoir C through the valve v ; D is a slide valve, regulating
the admission of air to and from the working cylinders ; E is the
respirator, or regenerator (a box filled with wire gauze), which is
heated at the bottom by the fire, but is maintained cool towards
the upper end, by the alternate rush of cold air downwards. The
top of the working cylinder and the bottom of the pumping
cylinder are left open to the atmosphere, and the two pistons
are attached to the same rod by which motion is imparted to the
crank.

The pressure in the air reservoir 0 is said to be maintained at
10 Ibs. per superficial inch (— 25 Ibs. total pressure). In the
diagram, the figure a b c d e represents the entire pressure below
the 'working piston amounting to 28 Ibs. pressure per inch, up to
the point c, where the admission is supposed to be cut off, in order

S/X WILLIAM SIEMENS, F.R.S. 43

to expand the air down to atmospheric pressure, before it is dis-
charged, whereby the maximum effect will be obtained. From
this gross effect has to be deducted, first the resisting pressure of
the atmosphere against the working piston, which is represented
by the field a bfe, and secondly the power absorbed by the
pumping cylinder. 7?, as represented by the field b g h f. By
laying this field of resistance upon the field of power b c df of
the working cylinder, there remains the field b c d h g representing
the entire effective pressure upon the working piston. This
pressure amounts, on the average, to nearly 3 Ibs. per square inch
on the working piston.

In order to estimate the comparative economy of Ericsson's
engine, it is necessary to consider the total quantity of heat
absorbed for one revolution, the proportion of it which is trans-
ferred into useful effect, and the difference between the two, which
of necessity escapes in the form of sensible heat.

If the atmospheric air enters the pumping cylinder at a tem-
perature of 60° Fahr., it will be raised, by compression, to 10 Ibs.
additional pressure and to a temperature of 111° Fahr., as is
shown in the dynamical diagram (Fig. 3, Plate 5).

This air has to be heated on its passage to the working cylinder,
so that its volume is increased in the proportion of two to three,
in order that the air delivered by the pumping cylinder may each
time suffice to fill the working cylinder. To effect this it must
be heated to 391° Fahr.

The expansion that takes place in the working cylinder will
reduce that temperature to 314°, which is the temperature at
which the pumping cylinder full of air at 60° will fill the working

508

cylinder of one-half greater capacity (for — + GO = 314).

z

The temperature lost, during expansion in the working cylinder,
is 391 — 314 = 77°, which must be supplied to it again by the
fire, before it reaches the respirator, in order not to cool down its
lower extremity, and 25° in addition, to make up for the loss, on
account of the imperfect action of the same.

The air issues into the atmosphere at a temperature 25°
above that of the upper extremity of the respirator, or at
111 + 25 = 136° Fahr., being 136 — GO = 76° hotter than when
iD entered.

44 THE SCIENTIFIC PAPERS OF

Therefore —

The total heat supplied to the air = 77 + 25 = 102° Fahr.
The sensible heat carried off = 136 — 60 = 76 „

There remains the heat absorbed by being con-
verted into effect 26° Fahr.

26
Or the Ericsson engine produces the effect of y^~- 770= 196 Ibs.

lifted 1 foot high, for every (water) unit of heat expended.

This proves, that the Ericsson engine realises, theoretically,
nearly one-fourth the effect of a perfect engine, and would possess
a considerable advantage over any of those before considered, but
for the following serious imperfections : —

1st. Its gross working pressure has been demonstrated to be
3 Ibs. per square inch, but the engine being single-acting, the true
average pressure is only \\ Ib. per inch, and supposing the engine
will move with \ Ib. pressure per inch upon the working piston,
its mere friction will absorb one-third of the whole power.

2nd. The working piston has to move air-tight in a heated
cylinder, which by former attempts has been proved to be
attended with great practical difficulties. These are no doubt
reduced, by the air being heated in a smaller degree than had been
'attempted before ; but the temperature still remains sufficient to
carbonise the lubricating material, and by affecting the shape of
the cylinder, to cause leakage.

3rd. The available heating surface of the engine is confined to
the bottom surface of the working cylinder, and to the passage
leading to the regenerator.

Taking into account the intermittent action and slow heat-
absorbing power of the air, the heating surface of an air-engine
should, in the opinion of the author, not be less than 6 superficial
feet for 1 Ib. of coal consumed per hour, or about seven times
larger than in the engines of the " Ericsson."

4th. The weight, bulk, and first cost of Ericsson's engine are
inversely proportionate to its low working pressure and slow
speed.

5th. Incidental losses of heat, by radiation from the large
exposed surface of the heated cylinder, necessitate a very con-
siderable addition to the expenditure of fuel.

S/R WILLIAM SIEMENS, F.R.S. 45

The engines of the " Ericsson " are said to consist of four
working cylinders of 14 feet diameter and 6 feet stroke, making
upwards of 14 strokes per minute.

. 4 x 21108 x84 x8

Their collective indicated H.P. is - = 67G,

00,000

from which must be deducted 83 per cent, for friction of pistons
alone, and say 27 per cent, for friction of general machinery and
of the air in rushing through the regenerator ; in all CO per cent.,
leaving 271 actual horse power.

There still remains one distinct class of engine for consideration,
namely, the combined steam and ether engine. This consists of
an ordinary steam engine with a tubular condenser. Instead of
the cold water, which is usually admitted into the chamber sur-
rounding the tubes, ether, or chloroform, is substituted, which, it
is well known, boil at a temperature far below the boiling-point of
water, and therefore will generate their own vapours, under a
considerable pressure, by the heat given off by the steam in the
act of condensation. The vapours of ether, or chloroform, are
made to give motion to a second engine, and are in their turn
condensed by very cold water.

It would seem, at first sight, that by this ingenious arrange-
ment the power obtained for a given quantity of fuel was doubled,
if compared with the performance of the steam engine alone ;
but the preceding investigations will have proved, that the heat
imparted to the ether must fall short of that given out under the
steam boiler, in proportion as the heat is changed into the dynamic
effect obtained from the steam engine. The additional effect of
the ether engine might indeed be obtained at once from the steam,
if the expansive action of the steam was sufficiently extended ;
considering, however, that an engine, in which the steam is ex-
panded at one- third part of the stroke, absorbs only about one-
eighth portion of the entire heat of the steam, and considering
also that it is very inconvenient to extend the rate of expansion
much further, in rotary engines, there remains, at present, a
considerable advantage in favour of the combined steam and ether
engine, if the practical difficulties involved, such as the tightness
of the joints, are not taken into account. Supposing the dynamic
equivalent, per unit of heat obtained by the engine, to be
90 foot-lbs., that of the ether engine may be taken at two-thirds,

46

THE SCIENTIFIC PAPERS OF

or 60 foot-lbs., making a total of 150 foot-lbs., — which nearly
equals the performance of the best Cornish engines.

The following table is intended to convey a more distinct idea
of the comparative merits of the different steam and air engines
referred to.

Actual

Theoretical

Actual Performance

Description of Engine.

Performance

Performance in Ibs. of Coal

in Foot-lbs.

in Foot-lbs. per H. P.

per hour.

A Boulton and Watt condeiis- \
ing engine, low pressure . j

51-8

29- 8-00

The best Cornish engine .

158-8

82-

2-38

Combined steam and expan- ")
sive ether engine . . j

150-0

75-

3-09

The expansive air-engine . . 91-0

35-

6-68

Stirling's engine . . . 180-0

65-

3-57

Ericsson's engine . . . . 196-0

65- 3-57

A perfect engine

770-0

385-

0-60

The statements of the actual performance of air-engines must
be considered as only rough approximations, as it is not possible
to calculate losses of heat, with any degree of precision. The
actual performance of the " Ericsson " engine may be deemed too
low, considering its theoretical superiority ; but the author con-
siders the disproportion to be fully accounted for, by the extra-
ordinary losses of useful effect, arising from the exposure of the
heated cylinder, and the low working pressure.

In Stirling's engine, the heated cylinder is closed and sur-
rounded by flues and brickwork, in consequence of which, its
economical effect is thought to be equal to that of Ericsson's
engine, although it is in theory inferior.

Thirdly. " On the necessary characteristics of a perfect engine."

In the first part of this paper it has been shown, that an engine
would be theoretically perfect, if all the heat applied to the elastic
medium was consumed in its expansion behind a working piston
(or its substitutes, such as a disc, a flexible bag, &c.) leaving no
portion of it to be thrown into a condenser, or into the atmo-
sphere. In the second part, several actual engines have been
examined, with a view to test their degree of theoretical and
practical perfection. Such an inquiry should have for its end,

SfA WILLIAM SIEMENS, I'.R.S. 47

the attainment of some more perfect result than could hitherto be
obtained. The author will therefore attempt to state his views of
the characteristics of a perfect engine.

1st. All the elastic material employed should actually enter the
working cylinder (or its substitute), and produce its full value of
effective displacement of piston, without deduction for the
resisting pressure, or the working of pumps.

2nd. The production of the elastic material, previous to its
entering the working cylinder, should not require a continuous
expenditure of heat, or in the case of the steam-engine, the latent
heat expended in the boiler should he recovered.

3rd. That working material is the best which is capable of
receiving the largest possible quantity of heat in a given space.
Its temperature and pressure should be raised to the highest
point which the vessel containing it will admit of, but on leaving
the working cylinder, the temperature should be reduced to a
minimum. This may be accomplished, either by infinite expan-
sion, or practically by the application of a regenerator.

4th. Losses of heat by radiation and leakage, should be reduced
to the smallest possible amount, by working at high pressure and
velocity, and by covering all heated surfaces with non-conducting
materials. These losses being, proportionally, more to be appre-
hended in a perfect, than in an imperfect engine.

5th. Large and compact heating surface and considerable body
of material are essential, to attain a high temperature, without
rapid destruction of the vessels.

6th. No working part of the engine should be brought into
contact with highly-heated material.

The respirator, or regenerator, is undoubtedly a useful agent,
for recovering the free, or otherwise unproductive heat of a
caloric engine, and the following experimental investigations on
its action, by the author, may not be thought devoid of
interest.

The annular space between two concentric cylinders was fitted
with 750 brass strips, each 5 feet 9 inches long, and held TLth of
an inch apart from each other, by projecting ribs upon every
alternate strip. The internal cylinder contained a piston, with an
enlarged hollow piston rod, passing through a stuffing-box, and
was worked to and fro by an engine. The lower extremity of the

48 THE SCIENTIFIC PAPERS OF

external cylinder was heated by a fire to 050° Fahr., as indicated
by the pyrometer, and was maintained at that temperature.

A second, or charging cylinder was provided, which by the
motion of its working piston, alternately withdrew and returned
the same air to the first cylinder. The capacity of the charging
cylinder was 24 cubic feet, and its piston made 18 strokes per
minute : about two-thirds of its contents, or 10 cubic feet of air
of atmospheric pressure, passed with each stroke, to and fro
through the respirator, and all the heat carried away was absorbed
by the sides of the cylinder. After 2f hours' working, the tem-
perature of the charging cylinder was raised from 60° to 110° Fahr.
(50°). Its capacity for heat had been previously ascertained, by
suddenly admitting steam, and weighing the condensed water
obtained (in heating it from (50° to 210° Fahr.) (150°), which
amounted to about 541bs. The quantity of air which passed
from the respirator into the cylinder was 43,200 cubic feet, and
its weight 3360 Ibs. : this, if multiplied by its specific heat 0'2G7,
is equal to 897 Ibs. of water-power of absorbing heat. The heat

given off was-— 1000 = 18000 units, and consequently the
air left the respirator, each time, at a temperature of — - = 20*01°

OtJ /

nigher than that at which it entered. Adding to this, for loss of
heat by radiation during the experiment, which according to
established rules may be taken at 5°, the entire loss of heat by
the respirator amounts to 25° Fahr. when air is employed. In
using steam it does not exceed 10° Fahr., owing to the greater
conducting power of that fluid.

The regenerator condenser, which was designed and executed
some years ago by the author, is an illustration that water may
be also subjected to respirative action.

The paper is illustrated by a series of diagrams from whence
Plate 5 is compiled.

MR. SIEMENS explained, that the diagrams were chiefly intended
to illustrate the peculiar functions of the "respirator," or as
Mr. Ericsson had termed it, the " regenerator." Very conflicting
opinions had been expressed regarding this most essential element
in Ericsson's engine. Some thought that, by its agency, the heat

S/A' WILL1 'AM SIEMENS, F.K.S. 49

used to effect a stroke of the engine could be wholly recovered,
except accidental losses, and that, theoretically, it involved the
accomplishment of a perpetual motion. Others, on the contrary,
contended that the regenerator was only an obstruction to the
passage of the air, and of no utility whatever. He had endea-
voured to prove in his paper, that neither the one extreme view,
nor the other was correct ; that, indeed, the respirator might be
usefully employed, to recover that portion of heat which presented
itself at the exhaust part of the engine, in the form of free, or
sensible heat, but that neither the respirator, nor any other
possible contrivance, could recover the heat that was lost in the
expansion of the air behind the working piston. He had adopted
the new " dynamical theory of heat " for his argument, because
that theory enabled him to calculate the absolute quantities of
heat, that must inevitably be sacrificed, to produce a given
mechanical effect, and to separate the same from the other and
much larger quantity, that served only to form the elastic
medium behind the working piston, and which might be
recovered, by means of a respirator, unless, as he had shown in
the paper, it was all converted into power, by the expansive
action being carried to its last (but impracticable) limits.
Ericsson himself seemed to incline to the idea, that he could
recover the whole of the heat by means of his " regenerator," for
it would be difficult otherwise to account for the extraordinary in-
sufficiency of heating surface he had provided. Mr. Siemens could
speak confidently as to the mechanical efficiency of action of the
respirator, having applied a precisely similar contrivance to a
steam-engine of his design, some years previously.

Mr. Siemens agreed with Mr. Hawksley's proposition, that an
engine of his proportions could not give out any power. The
valve between the two cylinders in Ericsson's engine was as
necessary as it was between the boiler and cylinder of an ordinary
steam-engine, and could not be replaced by the regenerator,
which had a very different office to fulfil. Ericsson's engine
might indeed be compared to a steam-engine, in which the boiler
was represented by the air-chamber between the two cylinders,
and the feed-pump by Ericsson's pumping cylinder. Ericsson
had the disadvantage of sacrificing two-thirds of his power to
move the pump, but had the advantage of expending no latent

VOL. I. E

50 THE SCIENTIFIC PAPERS OF

heat to form his elastic medium. But it must be borne in mind,
that Ericsson had to add to his air a larger amount of sensible
heat, of which only a small proportion was really expended in
expansion, and the remainder would go to waste, unless it was
recovered by the regenerator. Nevertheless, the drawbacks to
Ericsson's engine, on account of the great resistance of the pump,
the small working pressure, the insufficiency of heating surface,
and the working of a piston in a heated cylinder, were so great,
that he thought no beneficial results could be expected from it.

ON A REGENERATIVE STEAM-ENGINE.
BY C. W. SIEMENS, Esq., C.E.*

THE application of the steam-engine to our various purposes of
manufacture and locomotion is of very recent date, although the
elastic force of steam was known even by the ancients ; for we read
in Hero of Alexandria, on Pneumatics (translation by Woodcroft)
that the Egyptian priesthood made use of it for the somewhat
undignified purpose of performing pretended miracles before an
ignorant population. The first suggestion of its useful application
for raising water is due to the Marquis of Worcester, and dwelt
upon in his " Century of Inventions."

The idea was taken up by Papin, Savory, and Newcomen, who
added important elements towards its practical realization ; but to
James Watt belongs the merit of having laid down a comprehensive
principle of the steam-engine, and of having devised means to
render the same capable of performing the rudest as well as the
most delicate operations.

If any proof were wanting of the great genius of Watt, it would
be sufficient to observe that the steam-engine of the present day is,
in point of principle, still the same as it left his hands half a cen-
tury ago, and that our age of material progress could only affect

* Excerpt Minutes of Proceedings of the Royal Institution of Great Britain
Vol. II. 1856, pp. 227-236.

S/A WILLIAM SIEMENS, F.R.S. 51

its form. Great honour, however, is due to Fulton, Stephenson,
Nasmyth, and others, for having adapted the same to the most
important purposes.

The steam-engine of Watt was composed of four organic parts,
which were pointed out on a working model before the meeting,
namely : — 1. The furnace, or chamber of combustion, with its
flues and chimney. 2. The boiler, or steam generator. 3. The
Mi .mi-vessel, or cylinder, wherein the elastic force of the steam is
imparted to the piston, or other first moving parts of the
machinery. 4. The condenser, where the elastic force of the
steam is destroyed by abstracting its latent heat, by injection of
cold water, or by exposure of cooled metallic surfaces. In the
case of high-pressure engines, it would seem that the condenser
was suppressed ; but it might be said, that this class of engines
makes use of one great common condenser, namely, the atmo-
sphere ; the separate condenser possessing only the advantage
of relieving the working piston of the opposing atmospheric
pressure. The only essential improvement of the steam-engine
that has been introduced since the time of "Watt consists in
working the steam expansively, whereby a considerable economy
has been attained ; but it is well known that Watt foresaw the
advantages that would be realized in this direction, and was pre-
vented only by insufficiency of the mechanical means at his
disposal from realising the same.

The lofty superstructure proved the soundness of the foundation
Watt had laid ; and it would seem hopeless to change the same,
unless it could be proved that the very principle regarding the
nature of heat, whereon Watt had built, had given way to another
more comprehensive principle. The engine of Watt was based
upon the material theory of heat that prevailed at his time, and
almost to the present day. According to this theory, steam was
regarded as a chemical compound of water and the supposed im-
ponderable fluid " heat," which possessed amongst others the pro-
perty of occupying under atmospheric pressure nearly 1700 times
the bulk of the water contained in it. The Boulton and Watt
condensing engine took the full advantage of this augmentation of
volume, which effected a proportionate displacement of piston, and
the condensation of the steam obviated all resisting pressure to
the piston.

E 2

52 THE SCIENTIFIC PAPERS OF

In the course of the last few years our views of the nature of
heat had however undergone a complete change ; and, according to
the new " dynamic theory," heat, as well as electricity, light, sound,
and chemical action, are regarded as different manifestations of
motion between the intimate particles of matter, and can be ex-
pressed in equivalent values of palpable motion and dynamic
effect. In support of this theory, he (Mr. Siemens) could not do
better than refer to the able discourses, recently delivered in the
Eoyal Institution, by Mr. Grove and Professor Thomson.

Viewed from the position of the new theory, the heat given out
in the condenser of a steam-engine, represented a loss of mechanical
effect, amounting to if part of the total heat imparted to the
boiler ; and the remaining -^ part was all the heat really converted
into mechanical effect. The greater proportion of the lost heat
might be utilized by a perfect dynamic engine. A vast field for
practical discovery was thus opened out ; but it might yet be
asked whether it was worth while to leave our present tried and
approved forms of engines, to seek for economy, however great, in
a new direction, considering the vast extent of our coal fields.
The reply to this objection was, that the coal in its transit from
the pit to the furnace acquired a considerable value, which, for
this country, might be estimated at £8 per horse-power per annum
(taking a consumption of 13| tons of coal, at an average expen-
diture of 12 shillings per ton).

Estimating the total force of the stationary and locomotive
engines employed in this country at one million nominal horse-
power ; it followed that the total expenditure for steam coal
amounted to eight million pounds sterling per annum, of which at
least two-thirds might be saved. In other countries, where coal is
scarce, the importance of economy becomes still more apparent ;
but it is of the highest importance for marine engines, the coals
whereof had to be purchased at transatlantic stations, at a cost of
several pounds per ton, to which must still be added the indirect
cost of its carriage by the steamer itself in place of merchandise.

These observations, Mr. Siemens thought, might justify him in
bringing before the Institution an engine, the result of nearly ten
years' experimental researches, which he thought to be the first
practical application of the dynamic theory of heat, of which he
was proud to call himself an early disciple. Others, more able than

.S7A- U'lI.LIAM SIEMENS, l-.R.S. 53

himself, might probably have arrived sooner at a practically useful
iv>ult, but he might claim for himself at least that strong con-
viction, approaching enthusiasm, which alone could have given
him strength to combat successfully the general discouragement
and the serious disappointments he had met with.

The following illustrations, proving the imperishable nature of
physical forces and their mutual convertibility, were made use of
to indicate more clearly the principles his engine was based upon.

A weight falling over a pulley, to which it was attached by a
siring, would impart rotary motion to a fly wheel, fixed upon
the same axis with the pulley, and the velocity imparted to the
wheel would cause the string to wind itself upon the pulley, till
the weight had reached nearly its original elevation. If the
friction of the spindle and the resistance of the atmosphere could
be dispensed with, the weight would be lifted to precisely the
same point from whence it fell, before the motion of the wheel
was arrested. In descending again, it would impart motion to the
wheel as before, and this operation of the weight, of alternately
falling and rising, could continue adinfiniium. If the string were
cut at the instant when the weight had descended, the rotation of
the wheel would continue uniformly, but it might soon be brought
to a stop by immersing it in a basin filled with water. In this
case the water was the recipient of the force due to the falling
weight, residing in the wheel ; and by repeating the same ex-
periment a sufficient number of times, we could find an increase of
temperature in the water, a fact discovered by Joule, in 1843,
which first proved the identity of heat and dynamic effect, and
established their numerical relation. If the weight falling over a
pulley were one pound, and the distance through which it fell one
foot, then each impulse given to the wheel would represent one
foot pound, the commonly adopted unit of force ; and if the water
contained in the basin weighed also one pound, it would require
770 repetitions of the experiment of arresting the wheel in the
water, before the temperature of that water was increased by one
degree Fahrenheit.

Another illustration made use of, was that of a hammer falling
in vacua upon a perfectly elastic anvil. The hammer would, under
these circumstances, rebound to precisely its original elevation, and
granting the perfect elasticity of both hammer and anvil, neither

54 THE SCIENTIFIC PAPERS OF

sound nor heat would be produced at the point of concussion.
If a piece of copper were suddenly introduced between anvil and
hammer, the latter would not rebound, but would make the copper
the recipient of the expended force. If the hammer were now
lifted again and again by an engine, and the piece of copper were
turned about on the anvil, so that at the end of the operation it
had precisely the same form as at the commencement, then no
outward effect would be produced by the force expended, but the
piece of copper would be heated perhaps to redness ; and if the
engine employed to lift the hammer were perfect, then the heat
produced within the copper should be sufficient to sustain its
motion.

A familiar instrument for converting force into heat was the
fire-syringe. The force expended in compressing the air imparted
a sufficient temperature (about 600° Fahr.) to the same to ignite
a piece of German tinder. When the plunger of the syringe was
drawn back, it might be observed that the temperature of the
enclosed air was again reduced to its original degree, because the
heat developed in compression of the air had been spent again in
its expansion behind the piston. If the expansion of the heated
and compressed air had been without resistance, no reduction of its
temperature could have taken place, because no force would be
obtained ; a fact which had been recently proved by Eegnault, and
which was perhaps the strongest point in favour of the dynamic
theory of heat that could be brought forward. If the heated and
compressed air in the fire-syringe could be produced by some ex-
ternal cause and be introduced behind the plunger after it had
descended freely to the bottom, then the force imparted to the
plunger in the expansion might be turned to some useful purpose,
and a dynamically perfect engine might be obtained. But although
the elevated temperature might be readily supplied by means of a
fire, it would not be possible to give a sufficient density to the air,
except by an expenditure of force in its compression. If, how-
ever, heat were applied to a drop of water confined below the
plunger till its temperature was raised sufficiently to effect its
conversion into steam of the density of the water itself (Gaignard
de la Tour's state of vapours), and then allowed to expand below
the plunger till its temperature was reduced to zero, a dynamically
perfect engine would be obtained. The impracticable nature of

.S7A1 WILLIAM SIEMENS, I-.K.S. 55

such an engine was however manifest, if it was considered that
steam of the density of the water producing it, would exert a
pressure of probably several hundred atmospheres, which pressure the
moving part of the engine must be made strong enough to bear at
a temperature of more than 1000° Fahr., and that the capacity of
the working cylinder must be sufficient to allow of an expansion of
the steam to several thousand times its original volume. It was
therefore necessary to look for other means of obtaining from heat
its equivalent value of force, which means, it was contended, were
furnished by the "regenerative steam-engine."

This engine, of which several diagrams and a model were
exhibited, consisted of three essential parts, namely, the furnace ;
the working cylinder, with its respirator and heating vessel ; and
the regenerative cylinder. It consisted also of a boiler and con-
denser, (unless the steam were discharged into the atmosphere,)
but these were not essential to the working of the engine, although
of great practical utility. The regenerative cylinder had for its
object alternately to charge and discharge two working cylinders,
and the action of its piston might be compared to that of a hammer
oscillating between two elastic anvils. The regenerative cylinder
communicated at its one extremity with one working cylinder, and
at the other extremity with another and similar working cylinder, and
these communications were not intercepted by valves. The working
cylinders were so constituted that their capacity for steam of
constant pressure was the same, no matter where the working
piston stood. Each consisted of a cylinder of cast iron, open at
both ends, which was completely enclosed in another cylinder or
heating vessel, one end of which was exposed to the action of
a fire. Within the inner cylinder was a large hollow piston,
filled with non-conducting material, to which was attached a
long trunk or enlarged hollow piston rod of nearly half the
sectional area of the piston itself. This trunk was attached to
the working crank of the engine in the usual manner. The
trunk of the second working cylinder stood precisely opposite,
and was connected with the same crank. The piston of the
regenerative cylinder was also connected with the same crank;
but stood at right angles to the two working cylinders. The
consequence of this arrangement was, that while the two working
trunks made their strokes (the one inward and the other out-

56 THE SCIENTIFIC PAPERS OF

ward) the piston of the regenerative cylinder remained com-
paratively quiescent upon its turning or dead point, and vice
versa. Around the two heating vessels boilers were disposed,
which received the heat of the fire, after it had acted upon the
former. The steam generated within the boilers was introduced
into the engine by means of an ordinary slide valve (of com-
paratively very small dimensions) at short intervals, and when
the piston of the regenerative cylinder was in its extreme posi-
tion. The admission of the steam, which was of high pressure,
took place on that side of the regenerative cylinder where
compression by the motion of its piston had already taken place,
and at the same instant a corresponding escape of expanded
steam on the other side of the regenerative piston was allowed to
take place into the atmosphere. The quantity of steam freshly
admitted at each stroke did, however, not exceed one-tenth part
of the steam contained in the working cylinders of the engine,
and served to renew the same by degrees, while it added its
own expansive force to the effect of the engine. The compres-
sion of the steam into either of the working cylinders took
place when its hollow piston stood at the bottom. While in
this position the steam occupied the annular chamber between
the working trunk and the cylinder, besides the narrow space
between the cylinder and the surrounding heating vessel. The
pressure of the steam being the same above and below the hollow
piston, but the effective area below being equal to twice the
area above, the working trunk, attached to the piston, would be
forced outward through the stuffing box, while the steam of the
annular chamber above the piston passed through the narrow
space intervening, into a space of twice the capacity of the
annular chamber below the hollow piston. During its passage
the steam had to traverse a mass of metallic wire gauze or
plates, the respirator, presenting a large aggregate surface, which
reached at one end sufficiently downward into the heating vessel
that its temperature was raised from 600° to 700° Fahr., while
its other extremity remained at the temperature of saturated
steam, or about 250° Fahr. In consequence of the addition of
temperature the steam received on its passage through the re-
spirator, its elastic force was doubled, and it therefore filled
the larger capacity below the hollow piston or displacer without

SIR WILLIAM SIEMENS, F.R.S. 57

loss of pressure. When the effective stroke of the working trunk
was nearly completed, the regenerative piston commenced to re-
ivdr, and the steam below the hollow piston expanded into the
regenerative cylinder, depositing on its regress through the respi-
rator the heat it had received on its egress through the same, less
only the quantity that had been lost in its expansion below the
working piston, which was converted into dynamic effect or
engine-power, and had to be supplied by the fire. The expansion
ami simultaneous reduction of temperature of the steam caused
a diminution of its pressure from four to nearly one atmosphere ;
and the working trunk could now effect its return stroke with-
out opposing pressure, and while the second working trunk made
its effective or outward stroke impelled by a pressure of four
atmospheres.

The respirator, which was invented by the Rev. Mr. Stirling, of
Dundee, in 18 l(i, fulfilled its office with surprising rapidity and
perfection, if it were made of suitable proportions. Its action
was proved at the end of the lecture by a working model. It had
been applied without success to hot-air engines by Stirling and
Ericsson, but failed for want of proper application ; for it had
been assumed (in accordance with the material theory of heat) that
it was capable of recovering all heat imparted to the air, and, in
consequence, no sufficient provision of heating apparatus had been
made. It having been found impossible to produce, what in effect
would have been a perpetual motion, the respirator had been
discarded entirely, and was even now looked upon with great
suspicion by engineers and men of science. Mr. Siemens had,
however, no doubt that its real merits to recover heat that could
not practically be converted by one single operation into mechani-
cal effect, would be better appreciated. The rapidity with which
the temperature of a volume of steam was raised from 250° to
650° Fahr. by means of a respirator, was indicated by the fact
that he had obtained with his engines a velocity of 150 revolutions
per minute. The single action of heating the steam occupied
only a quarter the time of the entire revolution of the engine, and
it followed that it was accomplished in one-tenth part of a second.
But, in explanation of this phenomenon, it was contended, that
the transmitting of a given amount of heat from a hotter to a
cooler body, was proportionate to the heating surface multiplied

58 THE SCIENTIFIC PAPERS OF

by the time occupied, and that the latter factor might be reduced
ad libitum, by increasing the former proportionately. The air-
engines . of Stirling and Ericsson had failed also, because their
heated cylinders had been rapidly destroyed by the fire ; but the
cause of this was, that an insufficient extent of heating surface
had been provided, and it was well known that even a steam-
boiler would be rapidly destroyed under such circumstances. Mr.
Siemens was led by his own experience to believe that his heating
vessels would last certainly from three to five years, and being-
only a piece of rough casting, that could be replaced in a few
hours, and at a cost below that of a slight boiler repair, he con-
sidered that he had practically solved the difficulty arising from
high temperature. It was however important to add, that all the
working parts of his engine were at the temperature of saturated
steam, and therefore in the condition of ordinary steam-engines ;
whereas in Ericsson's engine, the hot air had entered the working
cylinder. In surrounding the heating vessel with the boiler, an
excessive accumulation of heat was prevented from taking place,
and the pressure of the steam in the boiler became the true index
to the engine-driver of the temperature of the heating vessel.
Another essential property of the heating vessel was, that all its
parts should be free to expand by heat without straining other
parts, which was accomplished by a free suspension, and by
undulating its surface. Lastly, it should be massive, to withstand
the fire with impunity, for iron was, strictly speaking, a com-
bustible material. The pyropherus, or finely-divided metallic iron,
took fire spontaneously on exposure to the atmosphere, a chip of
iron was ignited in flying through the flame of a candle ; an iron
tea-kettle was destroyed by exposing it (unfilled with water) to a
kitchen fire ; whereas, in forging a crank shaft, the solid mass of
iron withstands the white heat of the forge fire for several weeks
without deteriorating. A heating vessel, properly constructed and
protected, might be heated with safety to 700° Fahr., at which
temperature it would be almost as able to resist pressure, as at the
ordinary temperature of the atmosphere, the point of maximum
strength of iron being at 550° Fahr., as had been proved by
experiments made for the Franklin Institution. The construction
of a heating vessel combining these desiderata was of paramount
importance for the success of Mr. Siemens's engine, and had not

SfK WILLIAM SIEMENS, F.K.S. 59

been accomplished without combating against considerable practical
difficulty.

Although heat may be entirely converted into mechanical effect,
it would nevertheless be impossible to construct an engine capable
of fulfilling this condition without causing at the same time a
portion of heat to be transferred from a hotter to a cooler body,
and which must ultimately be discharged. This necessity has
been generally proved, and in a very elegant manner, by Professor
Clausius, of Zurich, and implies at least the partial truth of
" Garnet's theory." In the " regenerative steam-engine," pro-
vision had been made for absorbing this quantity of heat, arising
in this case from the circumstance, that the saturated steam enters
the respirator in a state of greatest density or compression, and
returns through it (expanding into the regenerative cylinder) at
a gradually diminishing density, although the temperature of the
extreme edges of the respirator remains proportionate to the con-
densing point of the steam of greatest density, by providing
water chambers about the cover of the working cylinder, and around
the regenerative cylinder, which are in communication with the
steam- boiler. The heat absorbed from the slightly superheated
steam is thus rendered useful to generate fresh steam.

Objection had been raised by casual observers against the re-
generative steam-engine, on account of its apparent similarity in
principle to the "air-engines" of Stirling and Ericsson, implying
similar sources of failure. The apparent similarity in principle
arose from the circumstance that both Stirling and Ericsson, as
well as himself, had employed the respirator and high tempera-
tures ; but these were but subordinate means or appliances, that
might be resorted to in carrying out a correct as well as an
erroneous principle.

In the winter of 1852-53, when Ericsson was engaged upon
his gigantic experiment in America, the speaker had had occasion
to read a paper to the Institution of Civil Engineers, entitled " On
the conversion of heat into mechanical effect," wherein he had
endeavoured to set forth the causes of probable failure of that
experiment, and to guard against a sweeping condemnation on
that account of some of the means Ericsson had employed.

According to the dynamic theory of heat, the elastic medium
employed in a perfect caloric engine was a matter of indiffcr-

6O THE SCIENTIFIC PAPERS OP

ence, and air had been resorted to, because it was perfectly elastic,
and always at hand. In practice, however, the elastic medium
employed was a matter of very great importance, and he (Mr.
Siemens) had given the decided preference to steam, and for
the following reasons : —

1. The co-efficient of expansion of saturated steam by heat
exceeded that of air in the proportion of about 3 : 2, but decreased
with an increase of temperature. This was not in accordance
with the established rule by Gay-Lussac and Dalton, but was the
result of his own experiments (described in a paper, " On the expan-
sion of isolated steam, and the total heat of steam," communicated
to the Institution of Mechanical Engineers, in 1852),* and had
been borne out by his practical experience on a large scale. Mr.
Siemens had been first induced to undertake these experiments in
consequence of an observation by Faraday, that the elastic force
of the more permanent vapours gave way rapidly, when by abstrac-
tion of heat their points of condensation were nearly attained.
He conceived that gases and vapours would expand equally by
heat, when compared, not indeed at the same temperature, but at
temperatures equally removed from their points of condensa-
tion.

2. When saturated steam was compressed (within the regenera-
tive cylinder), its temperature would not rise considerably (as the
fire-syringe evinced in respect of air), because, as Regnault had
proved, the total heat of steam increased with its density, and
consequently the heat generated in compression was required by
the denser steam to prevent its actual condensation. Without this
fortunate circumstance, the steam would be heated already by
compression to such an extent, that it would be difficult indeed
to double its elastic force by the further addition of heat in the
respirator.

3. Steam exercised no chemical action upon the metal of the
heating vessel and respirator, because the oxygen it contained was
engaged by hydrogen, which latter had the stronger affinity for
it until a white heat was reached ; whereas the free oxygen of
atmospheric air attacked iron and brass at much lower tempera-
tures.

4. The specific gravity of steam was only about one-half that

* Vide p. 17, ante.

S/K U'll.l.IAM .S7/-:.1//-:.Y\, l-.R.S. 6l

of atmospheric air at equal temperature and pressure ; moreover
it was a far better conductor of heat, and both circumstances
qualified it for rapid respirative action.

5. The fresh steam required for starting and sustaining the
power of the engine was generated by heat that would otherwise
be lost. No air-pumps, &c., were required, and the management
of the engine became as simple as that of an ordinary high-
pressure steam-engine.

In conclusion, it was stated that at present there were several
regenerative engines in constant practical operation, in this
country (at the works of Messrs. Newall & Co., at Gateshead), in
France, and in Germany, varying from five to forty horse-power,
which had proved the practicability of the principle involved,
although they were still capable of improvement. Several other
engines were now in course of construction at establishments
celebrated for precision of execution, and with the advantage of
Mr. Siemens's increased experience in designing them. He had
been fortunate to meet with men of intelligence and enterprise,
lately joined together in a public company, whose co-operation
insured a more rapid development of his invention than indi-
vidual effort could produce. The benefit he had hoped to derive
from his discourse, incomplete as it necessarily was, would be
realized, if those men, eminent in science, whom he saw around
him, would accept his labours as an earnest towards the prac-
tical realization of the dynamical theory of heat, and hasten its
triumphs by their own researches. It was impossible to over-
estimate the benefits that mankind would derive from a motive
force at one-third or one-fourth part the cost and iucumbrancc
of the present steam-engine. The total consumption of coal would
certainly not diminish ; but our powers of locomotion and pro-
duction would be increased to an extent difficult to conceive,
tending to relieve men from every kind of bodily toil, and hasten
the advent of the hoped for period-of general enlightenment and
comfort.

62 THE SCIENTIFIC PAPERS OF

ON A NEW CONSTRUCTION OF FURNACE,

PARTICULARLY APPLICABLE WHERE

INTENSE HEAT IS REQUIRED.

BY MR. C. WILLIAM SIEMENS, Mem. List. M.E.*

THE high importance of the stores of combustible material
which are distributed upon the surface of the earth renders their
wasteful expenditure and rapid diminution in quantity in many
parts a serious subject for consideration ; and in the writer's
opinion there is no object more worthy of the earnest attention of
engineers and men of science generally than that of causing the
generation and application of heat to be conducted upon scientific
and economical principles. Our knowledge of the nature of heat
has been greatly advanced of late years by the investigations of
Mr. J. P. Joule, of Manchester, and others ; which have enabled
us to appreciate correctly the theoretical equivalent of mechanical
effect or power for a given expenditure of heat. We are enabled
by this new dynamic theory of heat to tell, for instance, that in
working an engine of the most approved description we utilise at
most only one-sixth to one-eighth part of the heat that is actually
communicated to the boiler, allowing the remainder to be washed
away by a flood of cold water in the condenser. If we investigate
the operations of melting and heating metals, and indeed any
operation where intense heat is required, we find that a still larger
proportion of heat is lost, amounting in some cases to more than
90 per cent, of the total heat produced.

Impressed by these views the writer has for many years devoted
much attention to carrying out some conceptions of his own for
obtaining the proper equivalent of effect from heat : some of the
results he has obtained are known to the members of the Institu-
tion, amongst which are the regenerative steam-engine and con-
denser, the regenerative evaporator, and an apparatus for the
economic production of ice. The regenerative principle appears
to be of very great importance and capable of almost universal

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1857, pp. 103-111.

SIR WILLIAM SIEMENS, f.K.S. 63

application ; and the object of the present paper is to describe an
application of this principle to furnaces of every description.

The invention of the regenerative furnace is due to the writer's
In-other, Mr. Frederick Siemens; and it has been matured and
variously applied by the writer within the last few months. The
result has in all cases been a large saving in fuel over the plans in
rnmmon use, amounting to from 70 to 80 per cent, of the total
quantity of fuel hitherto consumed. The apparatus employed is
moreover of a very simple and permanent description, and com-
bines economy of fuel with other advantages, amongst which are
the total prevention of smoke and a general improvement in the
quality of the work produced.

Figs. 1 to 4, Plate C, represent the new furnace in the form
applicable to piling iron, or heating iron, steel, or other substances.

Fig. 1 is a longitudinal section of the furnace, and Fig. 2 a
sectional plan ; Figs. 3 and 4 are transverse sections.

The furnace consists of the heated chamber A, and of two fire-
places or solid hearths B and 0, communicating respectively with
the two regenerators D and E. Each regenerator consists of a
series of walls of firebrick, laid in open Flemish bond, in such a
manner that the pigeon-holes of each wall are opposite the solid
parts of the succeeding wall, the object being to form a number
of zigzag or tortuous passages through the regenerators, leading to
opposite sides of the valve F, shown dotted in Fig. 1, at the
bottom of the chimney G. The valve F consists of a rectangular
box of iron open at the two sides to the two regenerators D and E,
at the bottom to the atmosphere, and at the top to the chimney
G. A spindle passes through the centre of the two remaining
close sides of the box, and carries a rectangular flap or moveable
plate, fitting the box sideways and bearing against one of its
upper and one of its lower edges, according to the position of the
tumbling lever and weight H which are fixed upon .the spindle
outside.

When the valve is in the position shown dotted in Fig. 1, the
atmospheric air entering from below proceeds in the direction
indicated by the arrows, passing through the regenerator D, over
the fireplace B, through the heated chamber A, over the fire-
place C, through the regenerator E, and by the valve F into the
chimney G.

64 THE SCIENTIFIC PAPERS OP

A fire having been lighted upon the hearth B through the side
opening K, the flame passes through the furnace and through the
regenerator E to the chimney Gr. In its passage through the
regenerator E, the first perforated wall that the flame strikes
against will be heated to a considerable degree, the second wall to
a lower degree, and so on in succession, the heat of the current
being thoroughly exhausted by the time it reaches the chimney.

After about one hour's work the position of the valve F is
reversed and fuel is supplied through the opening L to the second
fireplace C, which is then acted upon by a current proceeding in
the opposite direction to that indicated by the arrows. The cold
atmospheric air comes in contact first with the least heated wall
of the regenerator E, and then with the more heated walls succes-
sively, acquiring thereby a degree of temperature approaching the
temperature of the heated current which previously entered the
same regenerator. The heat thus imparted to the fresh air
greatly increases the temperature of the flame which is now being
produced upon the hearth C, and consequently the nearest end of
the regenerator D will be heated also to an increased degree, the
current reaching the chimney comparatively cool.

When the valve F is again reversed, the fresh air will be heated
nearly to the increased temperature of the hot end of the regene-
rator D, and will produce a still hotter flame with the fuel supplied
to the hearth B. It is evident that by a continuation of this
process an accumulation of heat to any degree may be produced
within the furnace, provided only the heat produced in combustion
is greater than the heat lost by radiation and the heat absorbed by
the metal or other substances in the heating chamber.

In the regenerative furnace now described, the temperature at
which the heat is communicated to the materials does not affect
the quantity of fuel requisite, except so far as increased radiation
is concerned ; for the products of combustion pass away in all
cases at a temperature not above 200° or 300° Fahr. This new
principle of furnace is therefore applicable with the greatest
advantage in cases where intense heat is required. It has been
applied to furnaces for reheating steel and iron, at the works of
Messrs. Marriott and Atkinson at Sheffield. One of these furnaces
has now been in constant work for nearly three months ; and
according to a statement received from Mr. Atkinson it has worked

SIR WILLIAM SIEMENS, F.R.S. 65

quite satisfactorily, and the result of a careful comparison has
shown a saving of 79 per cent, to have been effected over the old
furnace in heating the same quantity of metal. Mr. Atkinson
has also applied this principle of furnace for melting cast steel,
and has ohtaiued a still larger saving, although the new melting
furnace has not yet been rendered entirely satisfactory for the
workman.

The regenerative furnace has also been applied to the purpose
of puddling iron ; and though the new puddling furnace has been
completed and worked only for a few days at the works of Messrs.
Rushton and Eckersley at Bolton, the writer is able to state that
it converts a charge of 480 Ibs. of pig metal into wrought iron
with an expenditure of only 160 Ibs. of common coal, as compared
with C cwt. required in the ordinary furnaces : the net yield of
wrought iron is higher than that of the ordinary puddling furnace,
and the quality of the iron produced seems also to be superior. It
is also worth mentioning that the chimney of this puddling furnace
may be watched for hours, and no trace of smoke be seen issuing
from it. Several other applications of this principle of furnace
are contemplated by the writer, which it would be premature to
enter upon on the present occasion.

In the discussion of Hie Paper,

MR. SIEMENS said he considered the durability of the brickwork
might be fully accounted for by the circumstance that, on attain-
ing the very high degree of temperature, the air had first to heat
the ends of the first wall of bricks that it met with to a very
intense degree, but the next wall to a less degree, and so on
gradually diminishing towards the chimney. It might be expected
at first that the cold air entering would have a tendency to crack
the hot bricks ; but it should be remembered that the air became
gradually heated as it advanced, coming in contact at first with a
comparatively cool wall, and afterwards by degrees with hotter
and hotter walls, taking up only 100° or 200° of temperature from
each successive wall, each one being only a little hotter than the
air on reaching it. There was consequently no sudden difference
of temperature at any point between the air and the brickwork
with which it came in contact, and ah1 risk of sudden chilling and
VOL. i. r

66 THE SCIENTIFIC PAPERS OF

cracking of the brickwork was thus avoided. The alternating
principle of the furnace gave a remarkable result in the extremely
high degree of temperature that could be ultimately attained, as
each repetition of the process added something to the previous
temperature of the furnace, the heat continually accumulating ;
and if the surface were sufficiently covered to prevent loss of heat
by radiation, it would be possible by continuing the alternation to
attain a temperature far higher than in any existing furnace, and
limited only by the materials employed.

Mr. Siemens had not yet had any long experience of the
application of the plan to puddling, but from the results already
obtained he thought there was full reason to anticipate its entire
success. This case was a peculiar one from the necessity of
keeping the puddling door open during the process, the furnace
being consequently exposed to the open air ; and some modifica-
tions were therefore required in the details, which only the experi-
ence of further trials could enable them to get fully matured ;
there were also difficulties to be overcome in the working on the
part of the men, from the great contrast to the accustomed work-
ing of puddling furnaces, so much smaller quantity of fuel and so
much less draught being required. Besides the saving in fuel, he
expected a decided advantage both in the quantity and the quality
of the iron made, by removing the violent draught produced by
the intense heat of the chimney in the ordinary puddling furnaces ;
the comparatively cold chimney in the new furnace, with the
damper lowered to within two inches of the top of the chimney,
gave so small a draught that the flame did not cut the iron and
waste it as in the ordinary furnace, a higher temperature being
obtained with a more quiescent atmosphere. There was also the
advantage that no flame or particles of fuel were carried over the
surface of the iron, bringing sulphur and other impurities from
the firegrate.

Ashes had. to be cleared out frequently from the regenerators,
but they were readily removed through the holes provided for the
purpose at the back of the furnace, and the puddler simply raked
them out on putting a fresh charge into the furnace. The new
furnace was found to puddle the iron in about the same time as
the ordinary ones, a charge being brought out about every two hours
if the fires were sufficiently attended to. It took rather longer to

WILLIAM SIEMENS, F.R.S. 67

get up the heat at first, but the furnace retained the heat longer
and a higher temperature was obtained ; one night's stoppage
required about the same time as the ordinary furnace to get up
tin heat again in the morning, but a longer stoppage would cause
a somewhat longer time to be required. The principal difficulty
experienced in working was from the men putting on too much
coal at once, being accustomed to the necessities of the old furnace,
and thus choking the fire and retarding instead of accelerating the
working of the furnace.

The experience with the new puddling furnace was too limited at
present to determine fully the difference in the time of the pro-
cess ; in the furnace now working at Bolton it was found that the
iron was about 10 minutes longer in melting down, but was rather
quicker in coming to nature than in the ordinary puddling
furnaces ; and there was no decided difference in the whole time
of working. Some difference in management from the ordinary
furnace was of course required, which caused a difficulty at first
in carrying out the new plan ; but the principle of the furnace
was quite simple, and no more skill and care was needed than was
to be looked for in the men concerned. The quantity of fuel
consumed and amount of firing required were much less ; and by
keeping the fire low a large proportion of oxygenated air was
allowed to enter during the puddling process, which was supplied
at a very high temperature, preventing fluctuations in the tempe-
rature of the furnace ; and the very low draught of the chimney
was an important point in greatly reducing the quantity of cold
air unavoidably drawn in at the puddling door whilst open during
the process.

Mr. Siemens said he had not attempted the application of the
principle at present to steam-engine furnaces, the trials having
been first made in those kinds of furnaces where the highest
temperature was employed, and where consequently the greatest
economy could be effected by the regenerative plan ; but even in
the case of steam boilers, a considerable portion of the fuel was
generally wasted, by the air being allowed to escape into the
chimney at a higher temperature than was required for the
purpose of maintaining the draught.

v 2

68 THE SCIENTIFIC PAPERS OF

In the discussion of the Paper

"DESCRIPTION OF TWO PAIR OF HORIZONTAL
PUMPING ENGINES," by Mr. E. A. COWPER,

ME. C. TV. SIEMENS * said he was acquainted with the
pumping-engines at the Berlin "Waterworks, where there were
twelve powerful beam-engines, coupled in pairs at right angles
with fly-wheels, and working vertical pumps under a pressure of
160 feet head of water. The present double ring valves allowed
an easy and unrestricted motion of the water, as the joint cir-
cumferences of the four openings gave a large area of passage with
a little lift of the valves.

With regard to the employment of the crank and fly-wheel for
pumping-engines, in place of the Cornish system of lifting the
water by the weight of the plungers, it was true that a large
fly-wheel with a great weight in the rim, running at a considerably
higher speed than the weight in the Cornish engine, had the
advantage of more vis inertice ; but only a limited fluctua-
tion in its speed was available, and the weight was never stopped
as in the Cornish engine, where the whole weight of the pump-
rods was stopped and started again at each stroke. It was there-
fore to be expected that a greater economy could be obtained by
the latter plan, in case of very great lifts, a larger proportion of
moving power being absorbed at the commencement of the stroke,
allowing the expansion to be carried to a higher degree by ad-
mitting of greater extremes of pressure in the stroke. He
accordingly thought there were cases, such as pumping from deep
pits, where the Cornish plan would prove the most economical,
the pump rods falling gradually by their own weight, and being
then lifted by the free action of the steam in the cylinder ; but in
many cases, such as low lift pumps, the horizontal direct con-
struction was preferable. He observed that the engines described
had not been designed for such a high degree of economy as was
attained in the Cornish engines, since the steam was cut off only

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1858, p. 57.

WILLIAM SIEMENS, F.R.S. 69

at erne-third the Btroke ; and there did not appear to be a
steam-jacket to the cylinders. He had found a great advantage
in employing a jacket not only round the sides of the cylinders,
but also at the ends, and it was more important to have a jacket
at the ends than round the sides ; for on the steam first entering
the cylinder, a much larger extent of surface was offered to it by
tin,1 end than by the small portion of the sides then exposed.
"When the cylinder was not protected by a jacket, the end of the
cylinder being cooler than the steam on first entering, caused a
portion to be condensed, which was partly evaporated again in
the latter part of the stroke, when the pressure and consequent
temperature of the steam was lowered by expansion. The effect of
this generation of steam in the cylinder was to increase the pressure
at the end of the stroke beyond that due to the expansion ; and the
line of the indicator diagram was thus raised in the latter part of the
figure above the regular curve of expansion. But this did not appear
to be the case in the indicator diagram from the Crystal Palace
engines, which showed the steam entering the cylinder at a total
pressure of about 34 Ibs. per square inch, cut off at one-third of
the stroke, and expanded down to a final pressure of 11 Ibs. due to
the expansion.

In the discussion of the Paper

"ON THE PERFORMANCES OF THE SCREW STEAM
SHIP ' SAHEL,' fitted with Du Trembley1 s combined-vapour
engine, AND OF THE SISTER SHIP ' OASIS,' fitted with
steam-engines worked expansively, and provided with partial
surface condensation" by JAMES WAEDROP JAMESON,

MR. SIEMENS * said, when treating of air-engines, in a paper
which had been read at the Institution in the Session 1852-53, | he

* Excerpt Minutea of Proceedings of the Institution of Civil Engineers,
Vol. XVIII. Session 1858-59. pp. 256-259.

t Vide Minutes of Proceedings of the Institution of Civil Engineers, Vol. XII.
p. 571, and p. 29, ante.

7<D THE SCIENTIFIC PAPERS OF

had adverted to the combined-vapour engine. The results then
arrived at, from calculation, were very similar to those now
recorded by the author. The comparison of the performances of
the combined engine and of the ordinary expansive steam-engine,
as shown by the diagrams, was very remarkable. There was a
difference in the consumption of fuel, per indicated H. P. per hour,
as between 3'1 Ibs. in the combined engine, and 7'4 Ibs. in a direct
acting engine, of nearly the same power, working steam expan-
sively. That would naturally seem to indicate the vast superiority
of the ether engine. But before accepting such a fact, which
was the result of experiment, as a proof of the superiority of the
principle involved, he thought the conditions under which the
experiments were made should be carefully inquired into, and
an estimate be formed of the causes, either in construction or in
working, which might have influenced the result. For instance,
it was well known, that an unprotected cylinder condensed a
large quantity of steam, if worked expansively. He had known
cases, where the result was modified to the extent of 50 per cent.,
by that cause alone. On the other hand, the ether-engine seemed
to have been under some disadvantage. There was a bad
vacuum, of 5 Ibs. only, in the ether cylinder, which was, no
doubt, in a great measure, to be attributed to the difficulty of
condensing ether, the condensing point being very low ; but he
thought it might also be due, partly, to the exhaust ports being
of insufficient area, for ether steam being five times heavier
than water steam, the ports for its discharge should be more
than twice the usual area. To judge of the merits of the inven-
tion now brought forward, it would be necessary to revert to
broader principles, instead of limiting the inquiry simply to the
facts stated. Perhaps he might be allowed, shortly, to develop
the relative merits, in a theoretical point of view, of the combined
ether-engine, and of a well arranged expansive steam-engine. He
believed he was the first to advocate, in the Institution, the
dynamic theory of heat. According to that theory, heat and
power were identical. The particles of a heated body were in a
more agitated condition than in a cooler body, the motion of the
particles being greatest in elastic fluids, and their pressure upon
the sides of the vessel such, that if one side yielded to the impact,
or pressure, as was the case with a working piston, the reaction

SfK WILLIAM SIEMENS, F.R.S. 7 1

would be less than the impact. Hence the motion of the particles
would be diminished, in proportion as the piston was urged for-
ward, and the result would be a reduction of temperature. Ex-
pansion might be carried to a point, where the elastic force was
entirely exhausted, at which point the elastic fluid must have
turned the liquid, or solid state. Commencing with steam of the
density of water, and expanding it down, until a perfect vacuum
was obtained, would form, in point of principle, a perfect engine,
\vlii.-h nothing could supersede. But the practical difficulty of
carrying expansion to its utmost limits, was the great strength and
size of the cylinders required to withstand the enormous pressure
of steam of that density, and to be at the same time of sufficient
dimensions to allow of the expansion of the steam down to that
extraordinary degree. All the various air and other engines,
designed to supersede the steam-engine, might be said to aim at
one point, that of carrying the expansive action to a farther
degree than could be done with the steam-engine, without having
to resort to cylinders of great size and strength. In the combined
ether-engine, the analogy was evident. Instead of expanding
the steam down to a minimum pressure, the expansive action in
the steam cylinder was stopped at a certain point, and the steam
was made to impart its remaining heat to another liquid, which
evaporated at a lower temperature, and the vapour of which occu-
pied less room than steam at a given temperature, and could
consequently continue the expansive action in a cylinder of smaller
dimensions than would otherwise be required. This was all the
advantage that could be claimed in favour of the combined ether
engine. It was a matter of calculation, whether that advantage
was sufficient to balance the disadvantages of a necessarily im-
perfect vacuum, in both the steam and the ether cylinders, of the
complication of parts, and of other drawbacks which had been
mentioned in the paper. There was an omission of one fact in
the paper, which perhaps the author could supply, — that was, the
quantity of ether evaporated, for a given amount of steam con-
densed. No mention had been made of this ; nor was he
acquainted with any experiments on the subject. It was a well-
ascertained fact, that the vapour of ether was about four times
heavier than steam at the same pressure, and the question was,
how much of that ether vapour would be produced for a given

72 THE SCIENTIFIC PAPERS OF

amount of condensed steam. In the absence of direct experi-
ments on that point, he thought it might be assumed, that a given
volume of steam would produce an equal volume of ether vapour
of the same pressure. That would appear to be the result of
general theory. The practical question was, how far the extra
power obtained by the ether-engine might be realised, by working
steam to a very high degree of expansion. The unprotected
cylinder of a marine engine would not give a favourable result,
because the steam generated in the boiler could not be dealt with
as a permanently elastic fluid, as had been too much the practice,
but would lose much of its elastic force through condensation in
the cylinder. The old engines of Watt were not simply lined
with a conducting material, as was now thought sufficient by many
Engineers, but were supplied with a steam-jacket kept filled with
steam of higher tension than the working pressure, to make up for
the loss of heat sustained in expansion. By properly protecting
the cylinder, by means of a complete steam-jacket, against the
enormous loss sustained by condensation of the high-pressure
steam upon the sides of the cylinder, and re-evaporation from
the same surfaces, after expansion had taken place, he thought
results might be obtained, equal to those claimed for the ether-
engine. There was no doubt, however, that the combined ether-
engine might, in its turn, be still greatly improved, if the practical
disadvantages of the greater complication, and of the danger
arising from the leakage of the ether vapour, should not prevent
its extended application.

In ihe discussion of the Paper

"ON THE APPLICATION" OF SUPERHEATED
STEAM," by Mr. JOHN N. RYDER,

MR. C. W. SIEMENS* said he had seen the trial of superheated
steam that had been referred to at Messrs. Hoyle's works ; and

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1860, pp. 30-33.

SIR WILLIAM SIEMENS, F.R.S. 73

though the result had not proved favourable in that case, he thought
that the failure might be fully accounted for by the imperfect
way in which the experiment had been tried : any single cases
however of want of success could not be admitted, he considered,
as permanent objections to the introduction of a system. There
was much difficulty in making an experiment complete and
arriving at the correct result, and the value of the result depended
upon the mode in which it had been conducted and a perfect
knowledge of all the circumstances involved. Amongst the cir-
cumstances to be considered in any experiments on superheated
steam were the actual size and construction of the engine, whether
the cylinders had steam jackets or were well or imperfectly
clothed, the degree of exposure and length of the steam pipes, and
the construction, form, and size of the boilers : the particulars of
all these circumstances should be carefully noted in any experi-
ments to determine the practical value of superheating the steam.
The advantages to be attained by superheating, most of which
were referred to in the paper, were — firstly, entirely preventing the
passage of priming water with the steam, by completely evaporating
this water in passing the steam through the superheating apparatus ;
secondly, obtaining a greater bulk of steam from the same water,
by its expansion with the increase of temperature ; and thirdly,
preventing any condensation of the steam by contact with the
cooler sides of the cylinder, whereby the steam could be kept in a
perfectly dry state throughout the entire stroke. The evaporation
of the priming water was a clear gain, as it was difficult if not
impossible to prevent priming altogether, and in many cases the
amount of priming was very considerable, causing serious risk of
injury to the engine as well as waste of heat. In regard to the
advantage to be obtained by adding heat to the steam to increase
its bulk, this involved the theoretical question of the rate of
expansion of steam by heat. It had been generally assumed that
isolated steam expanded under all circumstances at the same rate
as air, namely l-490th of its bulk at 32° for each degree Fahr. : but
he had shown by the results of experiments given in a paper at a
former meeting* (Proceedings Inst. M.E. 1852, page 131) that its
rate of expansion was considerably greater when near the con-

* See p. 17, ante.

74 THE SCIENTIFIC PAPERS OP

densing or boiling point, being in the case of atmospheric steam
an average of 5 times the rate of air up to 230° and 3 times up to
260° ; diminishing in fact in a ratio that could be represented by
a hyperbolic curve, approaching gradually at higher temperatures
to the same uniform rate of expansion as air which would be
represented by the asymptote to the curve. On this assumption
it followed that there would be a certain advantage gained in
applying heat direct to expand the bulk of steam, instead of
employing the same heat to generate an additional quantity of
ordinary steam ; but the extent of gain that could be attained was
not large, since the rise of temperature was practically limited to a
moderate range, and the specific heat of steam near the point of
saturation was proportionately great.

The most important source of economy in superheating steam
was no doubt to be found in preventing condensation in the
cylinder ; for the consequence of condensation of any portion of
the steam on entering the cylinder was the total loss of the power
of that steam during the remainder of the stroke. But in order
to effect this object completely he considered that it was not
enough to supply extra heat sufficient to heat the metal of the
cylinder up to the temperature of saturated steam of the maximum
pressure, whether by superheating the steam or by a steam-jacket
round the cylinder ; for although this were done, there must still
be condensation in the cylinder during the expansion of the steam,
on account of the loss of heat accompanying the development of
moving power from the pressure of the steam on the piston ; since
there would be as much heat lost from the steam (or so to speak
rendered latent in the form of power) as was equivalent to the
total work done by the steam, according to the accepted theory of
the mutual convertibility of heat and power, by which according
to Joule's results one unit of heat (the quantity of heat required
to raise the temperature of I Ib. of water 1° Fahr.) was equivalent
to or became converted into a power of 770 ft.-lbs. (770 Ibs. lifted
1 foot). This amount of heat lost during the expansion of the
steam must consequently be supplied by superheating the steam
originally admitted to the cylinder in order entirely to prevent
condensation ; and this appeared about the point to which super-
heating could be carried advantageously, and generally speaking
required an addition of about 100° to the natural temperature of

SIR WILLIAM SIEMENS, F.R.S. 75

the steam, varying of course with the degree to which the steam
was worked expansively. The circumstance that this amount of
heat was got rid of during the expansion of the steam in the
cylinder went far to account for the statement given in the paper,
that the quantity of injection water required to condense the steam
was found to be proportionate to the total weight of steam, or of
water evaporated, whether the steam were superheated or not : for
the heat added to the steam by superheating had probably been
hardly sufficient to replace the loss arising from expansion and
from the cooling by conduction and radiation from the metal
of the cylinder. He believed that by superheating the steam
judiciously a saving of from 15 to even 20 per cent, of fuel might
be effected, even upon a properly constructed engine ; but he was
equally satisfied that superheated steam must eventually give way
to regenerated steam, by which alone the full equivalent of motive
power could be obtained from heat, diminished only by the un-
avoidable losses from radiation, &c.

In the use of high temperatures there was much difficulty in
getting the joints to stand steam-tight, which he had experienced
in superheating steam in his regenerative steam-engine ; and after
the failure of various cements and copper rings, he had succeeded
in making a joint that stood even a red heat and remained quite
steam-tight. The cement he used was composed of red lead and
oil mixed with as much dust of cast iron as could be worked into
it. He showed a specimen of the joint fresh made, and one that
had been exposed to a red heat in which the cement had become
nearly as hard as iron itself.

76 THE SCIENTIFIC PAPERS OF

In the discussion of the Paper

"ON COMBINED STEAM,"
By JOHN WITHERED, United States,

MR. C. W. SIEMENS * said that he had, for nearly the last fifteen
years, occupied himself with this question. When he first turned
his attention to the subject, he endeavoured to ascertain the advan-
tage of superheating steam, and he made some rather elaborate
experiments to determine the rate of expansion. He found, that
Gay-Lussac's law of an uniform rate of expansion of elastic fluids
by heat, was not applicable, but that steam, near its point of
saturation, expanded in a greater ratio than afterwards. A curve
representing the rate of expansion was rounded at the beginning,
but gradually approached a straight line ; it was, in fact, a hyper-
bola, the asymptote of which ran parallel to a line representing
the uniform progression of expansion of air. So that at a tem-
perature considerably removed from the boiling point, air and
steam expanded, practically, at the same rate ; whereas at first,
steam expanded four times, or five times more than air, for the
same increase of temperature. He also found, that steam near
its point of saturation, possessed great capacity for heat ; and
therefore, by simply superheating steam, no great economy could
be produced. The results of recent practice showed, however, in
many instances, a very large saving, which might be ascribed to a
secondary cause. Steam, in expanding behind a working piston,
lost a portion of its heat, which was converted into mechanical
effect. This heat was entirely lost, and a portion of the steam
must condense, or form water. This water re-evaporated when
the pressure had become reduced by expansion, and cooled the
sides of the cylinder, thereby producing condensation in the fresh
steam from the boiler ; and thus an action was established on the
sides of the cylinder, (resembling that of a sponge, which, under
the influence of an alternating pressure, absorbed and emitted
water,) by which a certain amount of steam passed, at each stroke,

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XIX. Session 1859-60, pp. 479-480.

5Y/? WILLIAM SIEMENS, F.R.S. 77

through the cylinder, without exerting its elastic force upon the
piston. To prevent this action, it was necessary to sufficiently
increase the temperature of the steam, that it might be expanded
through the stroke, without condensation. According to the
present well-established law, expressing mechanical force by the
quantity of heat, he found, that an increase of temperature of
100° Fahr. would suffice to prevent condensation in an ordinary
expansive engine. Increasing the heat beyond that point would
not be accompanied with any marked economy, and would en-
gender many difficulties ; lubrication would become difficult, and
friction would be produced. Great stress had been laid upon the
peculiar properties of mixed steam, and the advantages which it
possessed over superheated steam, but he had never seen any re-
markable results from its use. In an engine which he designed, he
virtually used the two steams together, inasmuch as, at each stroke,
he added to a volume of superheated steam, a certain proportion of
saturated steam ; but he never observed any spontaneous increase
in the bulk. There was, however, one advantage in the author's
system, — by mixing ordinary saturated steam with superheated
steam, he had the means of regulating the temperature. If, then,
a compensating rod was introduced into the steam pipe, so as to
limit the admission of steam in proportion to the temperature,
beneficial results might be obtained. He was convinced, however,
that in simply superheating steam, the ultimate degree of economy
would not be reached. The chief advantage of superheating was
to prevent condensation in the working cylinder, whereas Mr. Sie-
mens's object had been, and still was, to prevent the loss of the
latent heat of the steam, by means of the regenerative system.
Although he had advanced but slowly, owing to the difficulties
and expense attending experiments of this description, he had two
engines at work, which gave him every confidence of ultimate and
complete success.

78 THE SCIENTIFIC PAPERS OF

In the discussion of the Paper

"ON GIFFARD INJECTOR FOR FEEDING STEAM-
BOILERS," by Mr. JOHN ROBINSON,

ME. C. W. SIEMENS * said he had made some calculations as to
the rise of temperature in the feed-water by its passage through
the injector, on the assumption that the steam supply carried
the water along with it by impact, taking no account of the
friction of the water ; that is, that if 1 Ib. of steam in motion
were mixed with 2 Ibs. of water at rest, the result produced
would be 3 Ibs. put in motion at one-third the original velocity
of the steam.

Now since the velocity of water or steam issuing into the atmo-
sphere from the same boiler was equal to that acquired by a
falling body in falling through the height of a column of the same
water or steam giving the same effective pressure, and since the
velocity acquired by a falling body was proportional to the square
root of the height through which it fell, it followed that the
velocity of the water and of the steam would be proportional to
the square roots of the relative volumes ; and as the volume of
steam with one atmosphere effective pressure was 860 times that
of water, it would issue with -v/860 or 29 times the velocity of
the water from the same boiler. Hence the steam issuing would
just balance 29 times its own weight of water trying to issue from
the boiler ; and, therefore, assuming the total heat of the steam to
be 1200°, and the original temperature of the feed 100°, the rise

of temperature of the feed would be 12j 110_P = 37°. And

Zi) + 1

calculating the rise of temperature in the same way for the
higher pressure of steam there would be

with 1 atmosphere effective pressure 37° rise of temperature

55 '•" 1i 5> J> " ,, ,,

5> " 55 55 55 *0 ,, ,,

A KK°

55 55 55 5> » ,,

55 10 „ „ „ 80° „ ,,

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers
1860, p. 78.

SIX WILLIAM SIEMENS, F.R.S. 79

Comparing these theoretical results with the experiments given
in Table V., it appeared that in practice with steam of 51 Ihs.,
or about 3 atmospheres effective pressure, the rise of temperature
was 74°, while the calculation gave only a little more than 50° ;
but then it was assumed that there had been no loss of power by
friction, and .the quantity of feed-water propelled into the boiler
had been supposed to be the maximum amount theoretically
possible. The calculation accordingly gave the minimum rise of
temperature when the steam was only just able to balance the
pressure of the water tending to escape from the boiler ; and
consequently in practice the actual rise of temperature must
always be greater than that obtained by the calculation ; but the
two results differed not more, he thought, than might be expected,
if all losses of effect were taken into account. The table agreed
moreover with the calculation in giving a greater rise of tempera-
ture at higher pressures of steam ; although the losses of effect
must necessarily increase with the pressure, or rather with the
increase of velocity of the jet. Much must also depend upon the
proper adjustment of the instrument to make the quantity of
water injected a maximum.

In the discussion of the Paper

"ON AN APPLICATION OF GIFFARD'S INJECTOR
AS AN ELEVATOR FOR THE DRAINAGE OF
COLLIERY WORKINGS," by Mr. C. W. WARDLE,

MR. C. W. SIEMENS * said, that although the injector was very
beautiful and economical in action where water had to be
raised and also to be heated, as in feeding a boiler, it was remark-
ably deficient in respect of economy when employed simply as an
elevator for raising water, where the water was not required to be

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1861, p. 227.

8O THE SCIENTIFIC PAPERS OF

heated. This was shown in the experiment which had been men-
tioned, where the water was raised only 36 feet high in being
heated 26° ; but the perfect equivalent of heat as established
definitely by Joule's investigations and others subsequent, was
that the heat required to raise 1 Ib. of water 1° in temperature
would raise 1 Ib. a height of 772 feet, and 1 Ib. heated 26° would
raise 1 Ib. a height of 772 x 26 or 20,072 feet ; hence the economy
of the elevator was as 36 to 20,072, or only -^th of the theore-
tical perfect duty of the heat. A very good pumpiug-engine
realized £th of the theoretical effect, and ordinary steam-engines
realized T\-th to TVth, and were consequently 40 to 56 times
superior in duty to the elevator. The elevator was therefore
economically applicable only where fuel was no object, or as an
injector where the heat came in again usefully, as in feeding a
steam boiler.

The injector, indeed, although inferior to a pump in mere
propelling power, he considered the most perfect instrument for
feeding boilers, so long as the supply-water was cool enough to
allow it to work ; for then all the heat imparted to the water
was returned into the boiler without any waste, whereas in
using steam to work a pump the larger part of the heat was
wasted by being thrown away with the exhaust steam. The in-
jector, however, would not be economical if the supply water could
be heated by other means free of expense, since its action required
the supply water to be kept cool.

The amount of condensation in the long steam pipe mentioned
in the paper seemed small for such a length of pipe, and he thought
a good deal of water must be carried across the depositing box by
the current of steam and pass through into the elevator. He had
seen lately in France a simple and efficient contrivance by M. Le
Chatelier, for freeing the steam from water, by making the steam
pipe from the boiler descend vertically into the depositing box,
surrounded by a cylindrical and concentric casing, from the top of
which the steam was taken off for the engine ; the wet steam
rushing into the depositing box in a vertical current, carried for-
ward all the water it contained down to the bottom of the box by
the velocity imparted to it, while the steam itself turned sharp
round the bottom of the steam pipe and ascended through the
annular space, passing off dry, and the water was allowed to drain

S/fi WILLIAM SIEMENS, F.R.S. 8 1

back into the boiler. This apparatus was applied to a boiler which
previously consumed 8 Ibs. of water per Ib. of fuel, and the result
was that the consumption was reduced afterwards to only 6 or
6f Ibs. of water real evaporation per Ib. of fuel, while the engine
went, much faster in conse.quence of having drier steam. But
where the steam simply shot across the top of the depositing
vessel, as in the apparatus that had been described, he was satisfied
that a considerable quantity of water must be lost by being carried
across with it.

OX A REGENERATIVE GAS FURXACE, AS APPLIED
TO GLASSHOUSES, PUDDLING, HEATING, ETC.

BY MR. C. WILLIAM SIEMENS.*

THE arrangement of furnaces about to be described is applicable
with the greatest advantage in cases where great heat has to be
maintained : as in melting and refining glass, steel, and metallic
ores, in puddling and welding iron, and in heating gas and zinc
retorts, &c. The fuel employed, which may be of very inferior
description, is separately converted into a crude gas, which in
being conducted to the furnace has its naturally low heating
power greatly increased by being heated to nearly the high
temperature of the furnace itself, ranging to above 3000° Fahr. ;
undergoing at the same time certain chemical changes whereby
the heat developed in its subsequent combustion is increased.
The heating effect produced is still further augmented by the air
necessary for combustion being also heated separately to the same
high degree of temperature, before mixing with the heated gas in
the combustion chamber or furnace ; and the latter is thus filled
with a pure and gentle flame of equal intensity throughout the
whole chamber. The heat imparted to the gas and air before
mixing is obtained from the products of combustion, which after
leaving the furnace are reduced to a temperature frequently not

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1862, pp. 21-36, and 40-44.

VOL. I. G

82 THE SCIENTIFIC PAPERS OF

exceeding 250° Fahr. on reaching the chimney, whereby great
economy in fuel is produced, with other advantages.

The transfer of heat from the products of combustion to the air
and gas entering the furnace is effected by means of regenerators,
the principle of which has been recognised to some extent since
the early part of the present century, but has not hitherto been
carried out in any useful application in the arts, unless the respira-
tor invented by Dr. Jeffreys be so considered. The discovery of
this principle is ascribed to Rev. Mr. Stirling of Dundee, who in
conjunction with his brother, James Stirling, attempted as early
as the year 1817 to apply it to the construction of a hot-air
engine : their engine did not however succeed, nor did Captain
Ericsson's later attempts in the same direction lead to more satis-
factory results. The economical principle of the regenerator
having attracted the writer's attention in 1846, he constructed in
the following year an engine in which superheated steam was used
in conjunction with the regenerator : many practical difficulties
however prevented a realisation of the success which theory and
experiments appeared to promise ; but it is gratifying to find that
one principle then adopted, that of superheating the steam, has
since received the sanction of an extended application.

The employment of regenerators for getting up a high degree
of heat in furnaces was suggested in 1857 by the writer's brother,
Mr. Frederick Siemens, and has since been worked out by them
conjointly through the several stages of progressive improvement.
The results obtained by the earlier applications of the principle
were communicated by the writer in a paper read at a former
meeting of this Institution (see Proceedings Inst. M.E., 1857,
page 103) * : and two or three of the furnaces then described,
employed for heating bars of steel, remain still in operation. In
attempting however to apply the principle to puddling and other
larger furnaces, serious practical difficulties arose, which for a con-
siderable time frustrated all efforts ; until by adopting the plan of
volatilising the solid fuel in the first instance, and employing it
entirely in a gaseous form for heating purposes, practical results
were at length attained surpassing even the sanguine expectations
previously formed.

* Vide ante, p. 62.

S/fi WILLIAM SIEMENS, F.R.S. 83

1 11 the early form of the regenerative heating furnace, which
has been in continuous work during the lost three years for heating
Icirs of steel at Messrs. Marriott and Atkinson's Steel Works,
Sheffield, and also at the Broughton Copper Works, Manchester,
there is a single fireplace containing a ridge of fuel fed from the
top ; and two heating chambers, in which the bars of metal to be
licatrd are laid, with a regenerator at the end of each chamber, by
which the waste heat passing off from the furnace is intercepted
on its way to the chimney, and transferred to the air entering the
furnace. Each regenerator is composed of a mass of open fire-
bricks, exposing a large surface for the absorption of heat, through
which the products of combustion are made to pass from the
furnace, and are thus gradually deprived of nearly all their heat
previous to escaping into the chimney : the end of the regenerator
nearest the furnace becomes gradually heated to nearly the tempera-
ture of the furnace itself, while the other end next the chimney
remains comparatively cool. The direction of the draught being now
reversed by means of a valve, the air entering the furnace is made
to pass through the heated regenerator in the contrary direction,
encountering first the cooler portions of the brickwork, and ac-
quiring successive additions of heat in passing through the re-
generator, until it issues into the first chamber of the furnace at a
very high temperature, and traversing the ridge of fuel produces
a flame which fills the second heating chamber ; whence the
products of combustion passing through the second cold regenera-
tor deposit their heat successively in the inverse manner, reaching
the chimney comparatively cool. By thus alternating the current
through the two regenerators, a high degree of temperature is
maintained constantly in the furnace. This arrangement of fur-
nace is evidently applicable only in exceptional cases where two
chambers are to be heated alternately, nor does it admit of being
carried out upon a large scale.

In heating a single chamber the expedient was resorted to of
providing two fireplaces to be traversed in succession by the
heated air, with the heating chamber placed between, as in the
furnace shown in the drawings accompanying the previous paper
(Proceedings Inst. M.E., 1857, Plate 118). Here the difficulty
arose that the air, the oxygen of which was already combined with
carbon (forming carbonic acid) in traversing the first fireplace,

o 2

84 THE SCIENTIFIC PAPERS OF

took up a second equivalent of carbon (forming carbonic oxide) in
traversing the second, so that the fuel of the second fire was con-
sumed to no purpose. In order to dimmish this loss and also
avoid impairing the draught by a double resistance, the ridges of
fuel were discontinued and the coal was fed into the furnace from
the sides, resting on a solid hearth, to be there volatilised by the
heated air passing over it. By frequently stirring the first fire its
combustion was favoured until the current was reversed, when it
was left undisturbed until the next change, and so on alternately.
It was found very difficult however to maintain an active and
uniform combustion and to burn the purely carbonaceous substance
that was left in the fireplace after the gaseous portion of the fuel
had been volatilised ; and it had frequently to be raked out in
order to make room for fresh gaseous fuel. This circumstance led
to the first step towards the employment of fuel in the form of
gas, by providing a small grate below the heap of fuel, through
which a gentle current of air was allowed to enter, forming car-
bonic oxide, which afterwards further combined with oxygen on
meeting with the hot current of air entering the furnace from the
regenerator. The two fireplaces of alternating activity were how-
ever attended with considerable practical inconvenience : the
furnacenien in particular disliked the idea of attending two fire-
places instead of one, and being little interested in the saving of
fuel, took no pains to work the furnace in a satisfactory manner.

It therefore became necessary to devise a plan of heating a
single chamber continuously by one fireplace, in combination with
the alternate reversal of currents through the regenerators, but
without reversing the direction of the flame. This was accom-
plished by means of double reversing valves, and was practically
carried out in a puddling furnace that worked for a considerable
length of time at the ironworks of Messrs. K. and W. Johnson
near Manchester. The two regenerators were placed longitudinally
side by side, with a flue between, underneath the puddling chamber,
and the fireplace was put at one end of the puddling chamber, as
in an ordinary puddling furnace, and fed with fuel from above.
The heated air from the first regenerator was brought up at the
back of the fireplace, and meeting there with the fuel produced
the required flame in the puddling chamber ; whence the products
of combustion passed down at the end of the chamber, and were

SSK WILLIAM SIEMENS, F.R.S. 85

•il buck along the flue below to the hot end of the second
Tutor, through which they made their way to the chimney.
For reversing the currents through the regenerators two valves
led, connected by a lever, one at the hot end of the
Mcrators near the fire, and the other at the cool end next the
chimney ; whereby the heated air was made to enter the fireplace
by the same passage as previously, and the direction of the flame
through the puddling chamber was not changed. By this arrange-
ment the regenerative furnace was assimilated as nearly as could
be to an ordinary puddling furnace in form and mode of working.
The few furnaces constructed in this manner produced a great
heat with little more than one-half the consumption of fuel of
ordinary furnaces in doing the same amount of work. A con-
siderable saving of iron was also effected in puddling, owing to
the absence of strong cutting draughts, a mild draught being
found sufficient to produce the necessary heat. There still re-
mained drawbacks however which prevented an extensive appli-
cation of this form of furnace : the fire required frequent attention,
and it was difficult to maintain a uniform volume of flame in the
furnace ; the reversing valve at the hot end of the regenerators
was moreover liable to get out of order, and the furnace was
costly to erect.

The most important step in the development of the regenerative
furnace has been the complete separation of the fireplace or gas
producer from the heating chamber or furnace itself. When a
uniform and sufficient supply of combustible gas is ensured, it can
evidently be heated just like the air, by being passed through a
separate regenerator before reaching the furnace, whereby its
heating power is greatly increased. The difficulty of maintaining
a uniform flame in the furnace is thereby certainly removed, and
there is no longer any necessity for keeping the flame always in
the same direction through the furnace, since the gas can be
introduced with equal facility at each end of the heating chamber
in turn, and the periodical change of direction of the flame through
the furnace tends only to make the heat more uniform through-
out : whereas in the previous plan of employing solid fuel for
heating in the furnace, the relative position of the fireplace and
heating chamber being fixed and unchangeable required tne direc-
tion of the flame to be kept always the same, unaltered by the

86 THE SCIENTIFIC PAPERS OF

reversal of currents through the regenerators. The new plan of a
separate gas producer has now been successfully earned out in
practice, and there are already a considerable number of the
regenerative gas furnaces in satisfactory operation in this country
and on the continent, applied to glasshouses, iron furnaces, &c.
In the neighbourhood of Birmingham, at Messrs. Lloyd and
Summerfield's Glass Works, a flint glass furnace constructed upon
this plan has now been in continuous operation for nearly twelve
months, and affords a good opportunity for ascertaining the con-
sumption of fuel of the regenerative furnace as compared with the
previous furnace performing the same work. At the Glass Works
of Messrs. Chance Brothers and Co. near Birmingham, the regene-
rative gas furnace has been under trial for the same length of
time, and has latterly been adopted for the various purposes in
crown and sheet glass making upon a very large scale. Messrs.
James Russell and Sons, Crown Tube Works, Wednesbury, are also
applying the furnace to the delicate operation of welding iron
tubes, and in a short time will probably employ no solid fuel for
any furnaces at their works. Another flint glass furnace erected
by Messrs. Osier in Birmingham, and several puddling furnaces
erected by Messrs. Gibbs Brothers at Deepfields, and by Mr.
Richard Smith at the Round Oak Iron Works, are amongst the
latest applications of the regenerative gas furnace, the designs
having in all cases been furnished by the writer and carried out
under his brother's immediate superintendence.

The gas producer is shown in Figs. 1, 2, and 3, Plates 7 and 8 :
Fig. 1 is a longitudinal section, Fig. 2 a front elevation and
transverse section at the front, and Fig. 3 a transverse section at
the back. The producers are entirely separate from the furnace
where the heat is required, and are made sufficient in number and
capacity to supply several furnaces. The fuel, which may be of
the poorest description, such as slack, coke dust, lignite, or peat,
is supplied at intervals of from 6 to 8 hours through the covered
holes A, Figs. 1 and 2, and descends gradually on the inclined
plane B, which is set at an inclination of from 45° to 60° accord-
ing to the nature of the fuel used. The upper portion of the
incline B is made solid, being formed of iron plates covered with
firebrick ; but the lower portion C is an open grate formed of
horizontal flat steps. At the foot of the grate C is a covered

S/K WILLIAM SIEMENS, F.R.S. 87

water trough D, filled with water up to a constant level from the
small feeding cistern E, supplied by a water pipe with a ball tap.
Tlii- large opening under the water trough is convenient for draw-
ing out clinkers, which generally collect at that point. The small
stoppered holes F F at the front and G- G at the top of the pro-
ducer are provided to allow of putting in an iron bar occasionally
to break up the mass of fuel and detach clinkers from the side
walls. Each producer is made large enough to hold about 10 tons
of fuel in a low incandescent state, and is capable of converting
about 2 tons of it daily into a combustible gas, which passes off
through the opening H into the main gas flue leading to the
furnaces.

The action of the gas producer in working is as follows : the
fuel descending slowly on the solid portion B of the inclined plane,
Plate 7, becomes heated and parts with its volatile constituents,
the hydro-carbon gases, water, ammonia, and some carbonic acid,
which are the same as would be evolved from it in a gas retort.
There now remains from 60 to 70 per cent, of purely carbonaceous
matter to be disposed of, which is accomplished by the slow current
of air entering through the grate C, producing regular combustion
immediately upon the grate ; but the carbonic acid thereby pro-
duced, having to pass slowly on through a layer of incandescent
fuel from 3 to 4 feet thick, takes up another equivalent of carbon,
and the carbonic oxide thus formed passes ofi' with the other
combustible gases to the furnace. For every cubic foot of com-
bustible carbonic oxide thus produced, taking the atmosphere to
consist of £th part by volume of oxygen and £ths of nitrogen, two
cubic feet of incombustible nitrogen pass also through the grate,
tending greatly to diminish the richness or heating power of the
gas. All the carbonaceous portion of the fuel is not however
volatilised on such disadvantageous terms : for the water trough
D at the foot of the grate, absorbing the spare heat from the fire,
emits steam through the small holes I under the lid ; and each
cubic foot of steam in traversing the layer of from 3 to 4 feet of
incandescent fuel is decomposed into a mixture consisting of one
cubic foot of hydrogen and nearly an equal volume of carbonic
oxide, with a variable small proportion of carbonic acid. Thus
one cubic foot of steam yields as much inflammable gas as five
cubic feet of atmospheric air ; but the one operation is dependent

88 THE SCIENTIFIC PAPERS OF

upon the other, inasmuch as the passage of air through the fire is
attended with the generation of heat, whereas the production of
the water gases, as well as the evolution of the hydro-carbons, is
carried on at the expense of heat. The generation of steam in
the water trough being dependent on the amount of heat in the
fire, regulates itself naturally to the requirements ; and the total
production of combustible gases varies with the admission of air.
And since the admission of air into the grate depends in its turn
upon the withdrawal of the gases evolved in the producer, the
production of the gases is entirely regulated by the demand for
them. The production of gas may even be arrested entirely for
12 hours without deranging the producer, which will begin work
again as soon as the gas valve of the furnace is reopened ; since
the mass of fuel and brickwork retain sufficient heat to keep up a
dull red heat in the producer during that interval. The gas is
however of a more uniform quality when there is a continuous
demand for it, and for this reason it is best to supply several
furnaces from one set of producers, so as to keep the producers
constantly at work. The opening H leading from each producer
into the main gas flue can be closed by inserting a damper from
above, as shown in Fig. 7, in case any one of the producers is
required to be stopped for repairs or because part of the furnaces
supplied are out of work.

It is important that the main gas flue leading to the furnaces
should contain an excess of pressure however slight above the
atmosphere, in order to prevent any inward draughts of air
through crevices, which would produce a partial combustion of the
gas and diminish its heating power in the furnace, besides causing
a deposit of soot in the flues. It is therefore necessary to deliver
the gas into the furnace without depending upon a chimney
draught for that purpose. This could easily be accomplished if
the gas producers were placed at a lower level than the furnaces,
but as that is generally impossible, the following plan has been
adopted. The mixture of gases on leaving the producers has a
temperature ranging between 300° and 400° Fahr., which must
under all circumstances be sacrificed, since it makes no difference
to the result at what temperature the gas to be heated enters the
regenerators, the final temperature being in all cases very nearly
that of the heated chamber of the furnace or say 2500° Fahr.

SfX WILLIAM SIEMENS, F.R.S. 89

The initial heat of the gas is therefore made available for produc-
ing a plenum of pressure by making the gas rise about 20 feet
above the producers, then carrying it horizontally 20 or 30 feet
through the wrought iron tube J, Plate 7, and letting it again
descend to the furnace, as shown by the arrows in Fig. 1. The
horizontal tube J being exposed to the atmosphere causes the gas
to lose from 100° to 150° of temperature, which increases its
density from 15 to 20 per cent, and gives a preponderating weight
to that extent to the descending column, urging it forwards into
the furnace.

The application of the regenerative gas furnace as a plate glass
melting furnace is shown in Plates 9 to 12, which represent a
melting furnace now in course of erection at the British Plate
Glass "Works near St. Helen's. This furnace does not differ
materially from the regenerative gas furnaces previously erected
and at work at Messrs. Chance's and Messrs. Lloyd and Summer-
field's, but is selected in preference because it is the most improved
in details of construction. Plate 9 shows a longitudinal section of
the furnace, Plate 10 a transverse section, and Plate 11 a sectional
plan above and below the bed or " siege " as it is termed of the
furnace. Figs. 7, 8, and 9, Plate 12, show the detail of the gas
and air valves.

The heating chamber A of the furnace, Figs. 4 and 5, contains
twelve glass pots B, which are got out through the side doors
when the glass is ready for casting upon the moulding table.
Underneath are placed transversely the four regenerators C C,
composed of open firebricks built up on a grating, which are
arched over at the top and support the bed or siege D of the
furnace. The regenerators work in pairs, the two under the right
hand end of the siege communicating with that end of the heating
chamber, while the other two communicate with the opposite end,
as shown in Fig. 4. The gas enters the chamber through the
three passages E, Figs. 5 and 6, and the air through the two
intermediate passages F, whereby they are kept entirely separate
up to the moment of entering the furnace, but are then able
immediately to mingle intimately, producing at once an intense
and uniform flame in the heating chamber. The siege D is built
of firebrick, with a number of transverse channels, shown black in
Figs. 4 and 8, through which the cold entering air is made to

9O THE SCIENTIFIC PAPERS OF

pass on its way into the air flue G-, as shown by the arrows in
Fig. 5 ; by this means the siege is kept comparatively cool, so
that no fluid glass can pass through crevices into the regenerators.
Any melted glass that may fall from the heating chamber through
the apertures at the ends of the siege does not get into the regene-
rators, but falls into the pockets M, Fig. 4, whence it can be
removed through the opening at the bottom. The passage N,
Fig. 5, by which the air enters, affords the means of getting at the
regenerators through an opening at the end of each.

From the air flue G-, Fig. 8, the entering air is directed by the
reversing valve H into the air regenerator, as shown by the arrows,
and there becomes heated ready for entering the furnace ; at the
same time the gas entering from the gas flue I, Fig. 7, is directed
by the reversing valve J into the gas regenerator, where it becomes
heated to the same temperature as the air. Similarly the products
of combustion on leaving the opposite end of the furnace pass
down through the second pair of regenerators, as shown by the
arrows in Fig. 4, and after being here deprived of their heat are
directed by the reversing valves H and J into the chimney flue K.
When the second pair of regenerators have become considerably
heated by the passage of the hot products of combustion, and the
first pair correspondingly cooled by the entering air and gas, the
valves H and J are reversed by the hand levers, as shown dotted in
Figs. 7 and 8, causing the currents to pass through the regenera-
tors and the heating chamber in the contrary direction, whereby
the hot pair of regenerators are now made use of for heating the
gas and air entering the furnace, while the cool pair abstract the
heat from the products of combustion escaping from the furnace.
The supply of air and gas to the furnace is regulated by the
adjustable stop valves L, whereby the nature and volume of the
flame in the furnace may be varied at pleasure ; whilst the chim-
ney damper is used to regulate the amount of pressure in the
furnace in relation to the atmosphere, so as to allow the opening
of working holes.

The construction of furnace above described may be varied in
many ways to suit local circumstances. The regenerators are in
some instances not placed immediately under but at the side of
the furnace ; but it is important that they should always be placed
at a lower level than the furnace, in order that the air and gas

WILLIAM SIEMENS, l-.R.S. 9!

may rise naturally into the heating chamber, forming there a
plenum of pressure.

Plates 13, 14, and 15 show the application of the regenerative gas
furnace as a round flint glass furnace. Plates 13 and 14 show
vertical sections of the furnace taken at right angles to each other,
and Plate 15 a sectional plan above and below the siege. The
round form of furnace is found convenient in flint glass-houses,
affording the greatest amount of accommodation to the glass-
blowers. The four regenerators C are here arranged below the siege
as before, and the air and gas from the hot regenerators enter the
annular heating chamber A at one side of the furnace, as shown
by the arrows in Fig. 10, and pass all round it, the products of
combustion escaping at the opposite side into the cold regenerators.
The direction of the current is reversed at intervals exactly as in
the plate glass furnace already described, by means of the re-
versing valves H and J in the air and gas passages, Figs. 11 and
12. Furnaces of this construction have lately been got to work at
Namur in Belgium and at Montlucon in France, and several
others of the same description are in course of erection at the
present time. The furnace just started by Messrs. Osier in
Birmingham also partakes of this form, being made semicircular.

In setting out each individual furnace, the heating effect required,
the quality of the fuel employed, and the particular nature of the
process to be performed, have to be considered. The amount of
heat required determines the capacity of the regenerators ; and the
gas regenerators require fully as large a capacity as the re-
generators, and sometimes even a greater. This would perhaps
hardly be expected, but will be seen to be the case from the
following considerations. The gases proceeding from the gas
producers are a mixture of olefiant gas, marsh gas, vapour of tar,
water and ammoniacal compounds, hydrogen gas and carbonic
oxide ; besides nitrogen, carbonic acid, some sulphuretted
hydrogen, and some bisulphuret of carbon. The specific gravity
of this mixture averages 0'78, that of air being TOO ; and a ton
of fuel, not including the earthy remnants, produces according to
calculation nearly 64,000 cubic feet of gas. By heating these gases
to 8000° Fahr. their volume would be fully six times increased, but
in reality a much larger increase of volume ensues, in consequence
of some important chemical changes effected at the same time.

92 THE SCIENTIFIC PAPERS OF

The olefiant gas and tar vapour are well known to deposit carbon
on being heated to redness, which is immediately taken up by the
carbonic acid and vapour of water, the former being converted
into carbonic oxide and the latter into carbonic oxide and pure
hydrogen. The ammoniacal vapours and sulphuretted hydrogen
are also decomposed, and permanently elastic gases with a pre-
ponderance of hydrogen are formed. The specific gravity of the
mixture is reduced in consequence of these transformations to
0'70, showing an increase of volume from 64,000 to nearly 72,000
cubic feet per ton of fuel, taken at the same temperature. This
chemical change represents a large absorption of heat from the
regenerator, but the heat is given out again by combustion in the
furnace, enhancing the heating power of the fuel beyond the
increase due to elevation of temperature alone.

The chemical transformation is also of importance in preventing
" sulphuring ; " for it is believed that the sulphur in separating
from its hydrogen takes up oxygen supplied by the carbonic acid
and water, formiug sulphurous acid, a firm compound, which is
not decomposed on meeting with metallic oxides in the furnace.
This view is so far borne out by experience that glass containing a
moderate proportion of lead in its composition may be melted in
open crucibles without injury, instead of requiring covered pots
for the purpose as in ordinary furnaces. In dealing with the
highest quality of flint glass, however, it is found necessary to
retain covered pots ; but every other description of glass is melted
in open pots. In all branches of glass manufacture, saving
of fuel is of relatively small moment as compared with the im-
provement effected in the colour and general quality of the glass
by the use of the regenerative gas furnace, owing to the absence
of dust and cinders and the high degree of temperature which
may with safety be maintained throughout the heating chamber.

These advantages of the regenerative gas furnace are of equal
value in the case of puddling and welding iron. Plates 1C and 17
represent a puddling furnace constructed on this plan. Fig. 13 is
a longitudinal section of the furnace, Fig. 14 a sectional plan of
the puddling chamber, and Fig. 15 a sectional plan of the re-
generators ; Fig. 16 is a transverse section at the end of the
furnace, and Figs. 17 and 18 are vertical sections through the
gas and air passages.

WILLIAM SIEMENS, F.K.S. 93

The four regenerators C are in this case arranged longitudinally
underneath the puddling chamber A, which may be of the usual
form. In order to complete the combustion of the gas and air in
passing through the comparatively short length of the puddling
chamber, it is necessary to mix them more intimately than is
requisite in the large glass furnaces previously described. For this
purpose a mixing chamber 0, Fig. 13, is provided at each end of
the puddling chamber, and the gas and air from the regenerators
are made to enter the mixing chamber from opposite sides, as
shown in Fig. 16 ; the gas aperture E is moreover placed several
inches lower than the air aperture F, so that the lighter stream of
gas rises through the stream of air while both are urged forward into
the puddling chamber, and an intense and perfect combustion is
produced. The mixing chambers 0 are sloped towards the furnace,
as shown in Fig. 13, in order to drain them of any cinders which
may get over the bridge. The reversal of the current through
the furnace is effected about eveiy hour by the reversing valves
H and J in the air and gas flues, the arrangement of which
is exactly similar to that already described in the glass furnace :
the supply of gas and air is regulated by the throttle valves L,
and the draught through the furnace by the ordinary chimney-
damper.

This same arrangement, with obvious modifications, may be
applied also to blooming and heating furnaces, the advantages in
both cases being a decided saving of iron, besides an important
saving in the quantity and quality of the fuel employed.
The space saved near the hammer and rolls by doing away
with fireplaces, separate chimney-stacks, and stores of fuel, is
also a considerable advantage in favour of the regenerative gas
furnace in ironworks. The facility which it affords for either
concentrating the heating effect or diffusing it equally over a long
chamber, by effecting a more or less rapid mixture of the air and
gas, renders the furnace particularly applicable for heating large
and irregular forgings or long strips or tubes which have to be
brought to a welding heat throughout. It has already been applied
to a considerable extent in Germany for heating iron, having been
worked out there under the direction of the writer's eldest brother,
Dr. Werner Siemens, who has also contributed essentially to the
development of the system. The furnaces at the extensive iron and

94 THE SCIENTIFIC PAPERS OF

engine works of M. Borsig of Berlin are being remodelled for the
adoption of this system of heating, as have also been those at the
imperial factories at "Warsaw.

Another important application of the regenerative gas furnace
is as a steel melting furnace, in which the highest degree of heat
known in the arts is required, presenting consequently the greatest
margin for saving of fuel. Plate 18 represents a regenerative steel
furnace which has been in satisfactory operation in Germany for a
considerable length of time, being worked with lignite, a fuel little
superior to peat in heating power. This application of the re-
generative gas furnace is indeed rapidly extending in Germany,
but has not yet practically succeeded in Sheffield where it was
also tried : it is however in course of application at the Brades
Steel works near Birmingham. Fig. 19 is a longitudinal section
of the furnace, Fig. 20 a transverse section, and Fig. 21 a sectional
plan.

The two pairs of regenerators C C, Figs. 19 and 21, are situated
at the ends of the long melting chamber A, in which the steel
melting pots B B are arranged in a double row. The chamber
is covered with iron cramped arch-pieces P, any of which can be
readily removed for getting at the pots. The arrangement of the
reversing valves and the air and gas flues is similar to that in the
glass furnace previously described.

Other applications of the regenerative gas furnace are being
carried out at the present time : among which may be mentioned
one to brick and pottery kilns for Mr. Humphrey Chamberlain
near Southampton, for Messrs. Cliff of Wortley near Leeds, and for
Mr. Cliff of the Imperial Potteries, Lambeth ; also to the heating
of gas retorts at the Paris General Gas Works, and at the Chartered
Gas Co.'s Works, London. The description already given however
is sufficient to show the facility with which this mode of heating
may be adapted to the various circumstances under which furnaces
are employed. The important application of the regenerative
system to hot-blast stoves for blast furnaces by Mr. E. A. Cowper
has already been separately communicated to this Institution (see
Proceedings Inst. M. E., 18GO, page 54).

The experience hitherto obtained with the regenerative mode of
heating shows that it is attended with the greatest proportionate
advantage in localities where good coal is scarce but where an

WILLIAM SIEMENS, F.R.S. 95

inferior fuel abounds. This applies most forcibly to the South
Staffordshire district, where the best coal in lumps is worth 12s. Gd.
per ton, whereas good slack can be had at 3s. or 4s. per ton. The
question gains moreover in importance when it is considered that,
according to the best authorities, the thick coal of the district is
coming to an end, while millions of tons of coal-dust have accu-
mulated, of no present commercial value, which on being converted
into gas in the manner described by means of the gas producers
would acquire a heating value equal at any rate to the same weight
of the best coal in the manner in which it is at present used.
Considering also the proximity of the pits to the ironworks in this
district, it may be suggested whether the gas producers being of
very simple construction might not with advantage be placed near
the banks of fuel above or even under ground, the gas being con-
veyed to the works by a culvert so as to supersede carting of the
fuel. Such an arrangement might notably contribute to per-
petuate the high position which South Staffordshire has so long
maintained as an iron producing district.

In the discussion of the Paper

MR. SIEMENS observed that the essential features of the regene-
rative gas furnace described in the paper, as now matured and
carried out in practice, were the separate gas producers, in which the
solid fuel was converted into a gaseous form for use in the furnace,
and the regenerators, in which the gas and air were each raised to
a high degree of temperature previous to their mixture and com-
bustion in the furnace, whereby the heat produced by the combus-
tion was very greatly increased. In the gas producers the fuel
underwent a slow digestion, and the whole of the combustible
constituents were drawn off into the furnace in the form of gas.
while the incombustible ash or valueless portion of the fuel was
left behind. The gas produced was of a crude nature in its
original state, and much inferior to common gas for illuminating
or heating purposes, and if burnt only with ordinary air would
give very poor results : but it underwent a further change in the
regenerator where it was heated up to about 3000° Fahr., at
which temperature the several gaseous compounds contained in it
became decomposed, the rich carburetted hydrogen depositing

g6 THE SCIENTIFIC PAPERS OF

carbon which was at once taken up by the vapour of water present,
producing carbonic oxide and hydrogen ; so that there was then
present the greatest amount of free hydrogen, which had three or
four times the heating power of any other gas. The air used for
burning the gas was also heated by the regenerator up to about
3000° and then mixed with the gas at the same temperature,
producing perfect and most intense combustion : the regenerative
system thus presented the means of attaining an almost unlimited
degree of temperature. At the same time there was no great
current or draught through the furnace, since the chimney draught
was not required in this furnace to urge the combustion as in
ordinary furnaces heated by solid fuel, and the cutting draughts
destructive of ordinary furnaces were therefore entirely avoided.
There was no difficulty in keeping up an abundant supply of gas
if there were enough gas-producers and if the passages to the
furnace were large enough ; the puddling chamber could then be
completely filled with flame at any moment, or the flame could be
as instantly stopped, by means of the regulating valves and
chimney damper. By having separate air and gas valves the
chemical nature or heating power of the flame could also be
regulated to any desired degree, by altering the proportion of air
admitted with the gas, so as to produce any required effect from a
smoky flame to a pure bright flame. In the furnace for flattening
the cylinders of sheet glass a quantity of bright clean flame was
required for softening the glass without melting it ; but in the
melting furnace, on the contrary, an intense soaking heat was
wanted, with very little variation : and both sorts of heat were
obtained in the new furnace from the same gas main, by simply
regulating the quantities of air and gas admitted. Of the
puddling furnaces two were now just being started in the South
Staffordshire district, but about twenty puddling and heating
furnaces had been at work in Germany for some months already
with complete success. They had not yet obtained any absolute
results with the regenerative puddling furnaces in this country,
but at present the time of working a heat was about the same
as in the ordinary puddling furnaces. The puddling furnaces
working on this plan near Wolverhampton were not yet in a
complete state for operation as they had been expected to be
before this time, on account of a defect in the chimney flue and

SfK WILLIAM SIEMENS, F.R.S. 97

in the drainage of the premises. No permanent difficulty could
be anticipated in carrying out the new puddling furnace in
practice, on account of the large amount of experience gained in
the application of the regenerative furnace for other purposes.
There was only one chimney to a number of furnaces, and the
draught was simply regulated by a separate damper to each
furnace. The chimney was not required to be lined with fire-
brick, but was built of red bricks, as the heat passed off from
the puddling chamber was all arrested in the regenerators and
the chimney was always cool. The chimney damper regulated
only the force of draught in each furnace, but the quantity and
quality of flame were regulated by the gas valve. No furring-
up had taken place in the regenerators, because the steam mixed
with the gas would volatilise any carbon that might otherwise
be deposited on the walls of the regenerator : it was for this
purpose that care was taken to supply an excess of vapour of
water from the water trough in the grate of the gas producer
A regenerative heating furnace had been three years in constant
work at Messrs. Marriott and Atkinson's Steel Works, Sheffield,
heating" the steel for the rolling mill, and no inconvenience had
been experienced from vitrification of the materials of the fur-
nace. Mr. Atkinson was prevented from being present at the
meeting himself as he had wished, but had sent "a letter ex-
pressing his satisfaction with the furnace, as especially advan-
tageous for heating steel on account of the uniformity of the
heat obtained in it. The glass-melting furnace at Messrs. Lloyd
and Summerfield's had also been in constant work for nearly a
year without sustaining any injury from the great heat employed
in it. The temperature in the furnace was at least a full welding
heat of iron, as a bar of iron held in it dropped melted in half
a minute ; the hot end of the regenerator had a temperature of
about 3000° Fahr., and the heat in the furnace must of course be
greater. For measuring such high temperatures he made use of the
pyrometer described at a previous meeting,* consisting of a well
protected vessel containing a measured quantity of water, and a piece
of copper or platinum of definite size, which was exposed for a suffi-
cient length of time to the heat to be measured and was then dropped

* See Proceedings of the Institution of Mechanical Engineers, 1860, p. 59.
VOL. I. H

98 THE SCIENTIFIC PAPERS OF

into the water ; the rise of temperature produced in the water
showed the degree of heat upon a thermometer scale graduated
in the proper proportion, and the results thus obtained must
certainly be correct within 10° or 15" Fahr. Rather larger yields
were obtained with the regenerative puddling furnace, but it had
not been long enough at work yet to give any definite results.
The puddling furnaces at work in Germany however showed at
least 4 or 5 per cent, increase in the yield of iron, and this result
was indeed to be expected, because the new furnace was free from
the cutting action of the flame produced by a strong draught, and
the ball was surrounded on all sides by an equally hot flame. As
regarded cost, Mr. Siemens said that the alteration of present fur-
naces would be attended with a considerable expense, as there was
all the extra bottom brickwork of the regenerators, besides the
separate gas producers and the valves and mains ; but the separate
chimneys for each furnace were saved, and the cost of maintenance
was greatly reduced, judging from the condition of the glass
furnaces that had been twelve months at work. In new works
however the cost of construction of the regenerative puddling
furnaces would not much exceed the total cost of the present
furnaces complete ; and the new furnaces had the advantage of
occupying only their own space, without requiring room for a
coal pen to each furnace ; they could thus be built closer together
and consequently more could be brought within reach of one
hammer.

Mr. Siemens thought a pair of the new puddling furnaces would
cost about £300 complete. The regenerative glass furnaces
hitherto erected had been very expensive in construction, having
heavy iron plates in the sieges and a great deal of ironwork in the
fittings of the furnace and in the gas producers, much of which
had been greatly reduced or dispensed with in the furnaces sub-
sequently put up. In starting the new plan of furnace he had
thought it best to keep all the work very substantial, to be on the
safe side for strength and durability ; and the gas producers had
also been provided each with a separate gas tube and valve, so that
each could be shut off from the furnace if desired, to avoid risk of
the furnace being interfered with in its working by a defect at
any particular point.

S/A' WILLIAM SIEMENS, F.R.S. 99

In Hie discussion of the Paper

"OX THE DISTILLATION OF COAL, THE UTILIZA-
TION OF THE RESULTING PRODUCTS, AND THE
MANQFACTURE OF COKE," by M. PERNOLET,*

MR. C. W. SIEMENS said he had lately paid considerable attention
to the subject of coke-making ; and he believed that the coke-ovens
generally used, especially in the north of England, were very
primitive. The admission of air into them caused a portion of
the coke to burn, and the gases from the remaining coal, being
driven out, went to waste. He thought it hardly possible to
conceive anything more rude than that system, because those
gases, possessing heating powers of nearly equal value to the
remaining coke, were allowed to escape freely into the atmosphere,
involving not only a great loss of fuel in the most valuable form,
but also a serious nuisance in populous districts. As regarded the
quality of the coke produced in the ordinary oven, it had been
satisfactorily proved to be inferior to that produced in the close
oven when properly arranged, although it was, perhaps, more
lustrous in appearance. The reason of its real inferiority was,
that the coke first acted upon by atmospheric air was partially
burned ; but the carbonic acid formed by this combustion passed
through the remaining mass in the oven, and robbed it of such an
amount of carbon as was necessary to reduce the carbonic acid to
carbonic oxide. Hence the earthy constituent of the coke was
relatively increased, and, as was stated in the paper, the coke pro-
duced in close ovens contained, for the same amount of earthy matter
present, 1 5 per cent, pure carbon. This, with a poor description of
coal, was a most important feature, irrespective of the value of the
additional yield, both of coke and of the secondary products.

As regarded the conversion of the ordinary coke-oven into a
close oven — which he understood to be the subject of this paper —
he considered that it was practicable, and that the kind of oven
now brought before the notice of the Institution, had many ex-

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Yol
XXIII., Session 1863-64, pp. 452-3.

H 2

100 THE SCIENTIFIC PAPERS OF

cellent qualities, but, that it was not the only one possessing im-
portant improvements. The plan had been tried, indeed, at
Lambeth, with this difference, that coal was burnt to effect the
carbonization, the gases being used for illuminating purposes.
Those ovens were being pulled down, because the gas produced in
them was found to be deficient in brilliancy. Those most in use
on the Continent were of two different descriptions. The first, was
a semi- cylindrical horizontal oven, closed at the ends with
movable doors, and provided with a covering arch forming an
annular chamber, through which the gases generated in the oven
circulated, and were consumed there by the admission of atmo-
spheric air, in order to continue the distillation of the coal inside.
That was an improvement, inasmuch as the gases did not go
to waste ; and he believed ovens of that description were used at
the Ebbw Yale works, with great effect. In South "Wales excellent
coke was formed by mixing anthracite with bituminous substances,
and by exposing the mixture to great heat in close ovens.

Another description of coke-oven which he had seen in France,
and which seemed to be a favourite in many establishments, was
in the form of a number of upright retorts, and was known as the
" Four Appold." Each retort was slightly conical, and was filled
by shooting the coal from carts at the top. The gases evolved,
passed through the sides of the retort, and were burnt there by
the admission of atmospheric air. After a distillation of 24 hours
or 40 hours, a trap-door at the bottom was opened, and the whole
of the coke, amounting to several tons, was dropped into railway
trucks, or carts, and carried away. That was another instance
where the full value of the carbonaceous material was obtained.
He might mention, also, an oven he had lately tried, which was
based upon the same principle, as regarded the general arrangement.
It was an upright retort, but, instead of being simply heated by
the gas issuing from the coal, a poorer kind of gas was used, which
had been previously heated by passing through a regenerator ; the
air necessary to combustion being also passed through a regenerator
placed at the side of the retorts, so that only a portion of the gases
evolved was required to heat the coal, the rest being applied to
other uses, either for illuminating purposes or for the heating of
working furnaces, &c. He had noticed the remarkable fact,
that when the gas was drawn from the bottom of the retort,

S//? WILLIAM SIEMENS, F.R.S. IOI

instead of from the top, a much higher illuminating power was
obtained.

Operating on Staffordshire coal, which, in the ordinary way,
yielded gas of about 10 candle-power, he had obtained, by this
method, an illuminating power of eighteen candles, or twenty
candles. Very little tar, or other secondary products were formed
because, in passing over the heated surfaces, they were decomposed
and turned into gas, or combined with the coke, which latter was
thereby increased in hardness. It might be considered a drawback
that there were no secondary products, but it was in reality not
BO ; for in most places the tar and ammoniacal liquor were difficult
to dispose of, and if used to improve the coke they were applied to
the best advantage.

He thought the question of coke making had oeen treated too
lightly. The present tendency of engineers was to burn coal in
locomotive engines, because coking, as generally performed, was
a wasteful process, in which the mosfi valuable part of the fuel
was lost. But by improving the coke ovens, and by making
judicious use of the gases, both for illuminating and heating
purposes, coke might be produced at a reduced rate, to be used for
locomotive engines and blast-furnaces, and a larger aggregate
saving of coal might be effected.

In the discussion of the Paper

" ON GIFFARD'S INJECTOR,"
By JOHN ENGLAND, M. Inst. C.E.,

MR. C. "VV. SIEMENS* said he thought that the cause of action of
this instrument should be better understood than it appeared to be,
on account of the physical principles involved, which were, without
necessity, at present invested with mystery. He could not agree
in the view, expressed by Mr. Phipps, that the action of the
Injector was due to the circumstance, that the impact produced

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXIV. Session, 1864-65, pp. 222-225, and 236-238.

IO2

by a jet on a solid surface was equal to double the pressure due to
the area which produced it ; but he thought it might be shown,
that the jet produced by the steam possessed a very much higher
velocity than a jet of water produced by the same pressure which
would come out to meet it from the boiler. Thus the steam from
the boiler issued with a velocity v = 8 Jh, or nearly so, neglecting
the error arising from the circumstance of its being an elastic
fluid ; while the water of the boiler would issue with a velocity
represented also by the formula v' = 8 Jti ; but the pressure being
the same in the two cases, it followed that the height of the column
h was superior to h' in the proportion of the relative densities of
steam and water. Hence the velocity v of the steam would exceed
the velocity v' of the water in the proportion of *J/t : *J~R, or in the
case of steam of four atmospheres' pressure, v\v'= Jl : ^47 7,
water being 477 times denser than steam of that pressure, or
v : v' = 1 : 21-8.

It was therefore evident, that every particle of steam rushed
forward with a velocity nearly twenty-two times greater than that
with which the water could meet it, only if it were to rush forward
in its elastic state, it was evident that each particle of steam would
have to encounter 477 particles of water within the same volume,
and must be forced back notwithstanding its superior velocity.

But if the steam were imagined to be condensed between the
point of issue and that of its meeting with the jet of water, if, for
instance, its elastic force could be destroyed by making it pass
through an infinitely cold space, then the result would be a very
contracted jet of water passing forward with the extraordinary
velocity of v' = 8 *Jh' = 8 ^/477 x 3 x 32 = 1712 feet per second,
meeting a column of v — 8 Jh = 8 *J3 x 32 = 78 feet per second.

The effect would be, that the column of water resulting from
the condensation of the steam would pierce the comparatively
sluggish jet of water meeting it, as though it were a steel wire
forcing its way into the boiler. This was, however, an impossible
position, for the steam could not be condensed by itself spon-
taneously. But the injector effected the spontaneous condensa-
tion of the jet of steam by incorporating with it a sufficient
quantity of cold water, which was so regulated, that the mean
velocity of the united jet of steam and water still exceeded the
velocity of 78 feet per second with which the opposing jet from

WILLIAM SIEMENS, F.R.S. 103

the boiler would meet it. Supposing the initial temperature
of the water to be about 70" Fahr., the weight of the condensing
water should exceed the weight of the condensed steam about ten
times to effect perfect condensation ; the velocity of the united

1 712
jet or sheaf would be in this case = -r^ — = 155 feet per

second, or nearly doubly sufficient to force its way into the boiler.
It might here be remarked, that the author of the paper was
dourly in error when he stated that the steam was only partially
condensed in the sheaf. Uncondensed steam encountering water
could not drive it back ; but a column of water encountering
another column of inferior velocity must drive it back. It was
essential that water of sufficiently low temperature should be brought
into contact with the jet of steam to effect its entire condensation,
without reducing its velocity below the limit indicated by the
imaginary velocity of the opposing column. In the case assumed,
the weight of condensing water, or of the water injected into the

1 712

boiler, must not exceed the weight of condensed steam by -— -

78

= 22 times, or making allowance for losses of effect, say by 18
times. Taking the total heat of steam at 1,150 units, it followed
that the rise of temperature of the injected water would be

- = 64° Fahr., and assuming that the final temperature of
18

the sheaf must not exceed 190°, in order to effect entire con-
densation, it followed that the initial temperature of the injected
water mustjiot exceed 190 - 64 -= 126° Fahr. If it should be
attempted, under the circumstances, to inject water of a tempera-
ture exceeding 120° Fahr., the sheaf must break up, and the action
of the instrument be paralyzed. On the other hand, the greatest
quantity of water which each pound of steam would be able to
take depended upon the relative velocity of steam and water when
acted upon by the same pressure, and might be expressed by

1 712

-^wrr- = 22 Ibs. That would be the greatest quantity of water

which each pound of steam could urge forward into the boiler, or
could urge forward if there were no loss from friction or by the
vortices formed. In reality it would probably not deliver much
above half that quantity into the boiler.

104 THE SCIENTIFIC PAPERS OF

The subject had been considered irrespective of the question of
heat, but that, in his opinion, was a fallacious method, for, it
would be seen, that without complete condensation the instrument
could not act ; neither could its effect, nor the limits of its action,
be determined, without taking into account the amount of heat
contained in the steam and in the water, as well as their proper
velocities when acted upon by pressure.

Mr. Siemens thought the injector could only be advantageously
applied as a motor, in cases where the water had not only to be
raised but also to be heated, and its legitimate application would
probably remain confined to the feeding of steam boilers. Its
practical advantages were not overstated in the paper, where feed-
water of sufficiently low temperature was used. The minimum
temperature, to obtain an advantageous result, would be about
100° Fahr. In the case of a high-pressure engine, it was a
question whether a heater and a pump would not be more advan-
tageous than this instrument. The author had stated that 2'3
cubic feet of steam would be expended to pump one cubic foot of
water into the boiler. But even this calculation, unfavourable as
it might seem, still left a great margin in favour of the pump,
inasmuch as the water could be raised by waste heat to nearly
212°. He believed that the injector would be found applicable
with the greatest advantage in the case of marine engines, or of
condensing engines in general, where there were no means of
heating the feed-water to above 100°. Under such circumstances
the power expended to work the feed-pump could be entirely
saved, because although the injector would expend heat on the
water, it would return the same to the boiler ; whereas the pump
used the steam to work it, and gave nothing in return.

Mr. Siemens said, the areas of the different jets of water and
steam, which had been stated to be adopted in practice, agreed
very well with his own calculations.* The areas were determined
by the quantity of steam necessary to condense in order to impel
the water. Taking the case of a pressure of 60 Ibs., or an
effective pressure of three atmospheres, the steam was 477 times
lighter than water, while its velocity was only in the ratio of
J i77, or nearly 22 times greater than that of the water ; the

* Vide Minutes of Proceedings of the Institution of Mechanical Engineer*,
1860, pp. 48-78.

SJR WILLIAM SIEMENS, F.R.S. 105

consequence was, that the area must be larger to allow a given
weight of steam to issue from an orifice, than if water issued under
the same pressure, in the proportion of ^477 : «/ 1, or about
21 times the area. But, inasmuch as the quantity of water had
to be increased ten -fold, it followed that -^th of 21 times, or 2'1
times, ought to be the area of the steam orifice as compared with
the orifice in the boiler. This calculation agreed accurately with
the proportions it had been found expedient to adopt ; but he was
surprised to learn that the same proportions of area were given to
the injector, although the conditions of its application, especially
as regarded the pressure in the boiler, might be very different.
With a pressure of twenty atmospheres, for instance, the density
of the steam was 110 times less than that of the water ; therefore,
the relative velocity would be as *J 110 : J 1, or as 10'5 : 1, and
about double the area of steam would be required to supply the
necessary quantity to impel the water forward into the boiler.

He felt convinced many of the partial failures they had heard of
might be due to this, — that the injector had been calculated, or
determined experimentally, for ordinary high-pressure engines, but
if applied to low-pressure, or very high-pressure engines, it failed
to work satisfactorily. It could even be shown, that the injector
was not applicable at all if the pressure exceeded certain limits.
As, with a pressure of twenty atmospheres, the velocity of the steam
was only 10 '5 times greater than that of the water, and as the steam
must take up about 10 times its own weight of water in order to
condense, it would consequently possess only just power sufficient
to impart to that amount of water the requisite velocity, when that
point would be reached, where the injector would no longer work
with water of ordinary temperature.

The author had referred to the dynamical theory of heat, and
had said, that a sudden change in the sheaf took place, when there
was an instantaneous conversion of heat into work, a vortex being
found in the sheaf, producing this result. Mr. Siemens had been
one of the earliest disciples of this new theory of heat ; and in the
year 1853,* a paper by him was read at the Institution in which
the caloric engines were passed in review by the light of that
theory. Taking the case of a pressure of GO Ibs., the force repre-

* Vide Minutes of Proceedings of the Institution of Civil Engineers, Vol. XII.
p. 571, and p. 29, ante.

106 THE SCIENTIFIC PAPERS OF

scnted by a jet of steam issuing out into the atmosphere was
simply the product of the column in feet representing that pressure,
multiplied by the weight of steam in Ibs. issuing in a given time.
The column representing the pressure of steam of four atmo-
spheres total pressure was, according to the table, 45,792 feet ;
— that was, steam of a pressure of 60 Ibs. would be produced by a
column of its own weight 45,792 feet high. Therefore, 45,792
foot-lbs. was the mechanical work due to 1 Ib. of that column
of steam. Now the quantity of heat contained in 1 Ib. of steam
was 1,150 units. These 1,150 units of heat produced in the
injector 45,792 foot-lbs. of work, or for every unit of heat

45 792

— - = 39'8 foot-lbs. The full mechanical equivalent of a unit
l,loO

of heat was however = 772 foot-lbs., or units of force, and therefore

39 *8
the maximum useful effect of the injector did not exceed _^j,or

/ la

about the twentieth part of the theoretical effect due to the heat in
the steam. A good expansive steam-engine would give as much
as one-sixth part of the full theoretical effect due to the heat
expended, and supposing that there was 33 per cent, loss in an
ordinary pump, it followed that water could be raised with an

33

expenditure of 6 + y^; = 8 times the theoretical expenditure of

heat, instead of 20 times that quantity which the injector required.
Consequently, the injector could only be advantageously applied,
when the water was required to be both heated and raised ; and
it could never compete with an engine and pump for simply rais-
ing and propelling water.

WILLIAM SIEMENS, F.R.S. 1 07

ON UNIFORM ROTATION.
BY C. W. SIEMENS, F.R.S.,* Mem. Inst. C.E.

AMONGST the means at our disposal for obtaining uniform rota-
tion, there is none for which the same degree of accuracy can be
claimed as that which distinguishes the vibrating pendulum, or
the oscillating spring-wheel of the common watch ; yet there are
many purposes, both in physical science and in the mechanical
arts, for which smaller subdivisions of time than the period of one
oscillation are matter of considerable importance, and which can
only be measured by uniform rotation.

CONICAL PENDULUM. — The apparatus by which continuous
rotation of the greatest regularity has hitherto been obtained, is
the conical pendulum, which was first applied by James "Watt to
regulate the speed of his engines, and which has since received
further development in the instrument known as the Chronometric
Governor.

On examining into the principle involved in the conical pen-
dulum, it will be found that the time t of its rotation is dependent
upon its length I, and on the angle a which it makes with its
vertical axis of rotation, which dependence is expressed by the
formula t = cj> *Jl cos a, the coefficient <£ being a function due to
gravitation. The value of t being dependent upon a, it follows
that " uniform rotation of a conical pendulum cannot be obtained
except on condition that the angle of its rotation remains constant"

WATT'S GOVERNOR. — In the case of Watt's centrifugal governor,
the angle of rotation of the pendulum varies with every change in
the relative condition of power and load on the engine, and the
change of angle is, indeed, taken advantage of to close or open the
steam-supply valve. In order to close the steam- (or throttle-)
valve the angle of rotation has to be increased, which necessitates
a corresponding increase of the engine's velocity ; on the other
hand, an increase of the valve-orifice must be consequent on a
reduction of the speed of the engine.

Considering this dependence of the action of the instrument

* Excerpt Philosophical Transactions of the Royal Society, 1866, pp. 657-670.

108 THE SCIENTIFIC PAPERS OF

upon permanent change of speed of the engine, the name of
" governor " seems inappropriate, the instrument being, in fact,
only a " moderator " of the amount of fluctuation to which the
engine would be subjected without its agency. The amount of
these fluctuations depends, in a great measure, upon the mechanical
construction of the instrument, which is, generally speaking, very
objectionable, inasmuch as the pendulous arms are mostly sus-
pended at points beside the common axis of rotation, giving rise
to an increased variation of time for a given change of angle ; the
reason being that the true pendulous length has to be measured
from the point of intersection of the pendulums or rods with the
axis of rotation, which point descends as the angle increases,
causing the pendulous length to diminish at both extremities. A
better result would be obtained if each pendulous rod were
suspended from the point past the axis of rotation (as shown by
dotted lines in Plate 19, Fig. 1), causing the two rods to cross in
the line of the axis at a point which would rise with increase of
angle, and render the true pendulous length approximately uniform
between certain limits.

The governor of Watt is, however, subject to another defect
which does not admit of an easy rectification, and which consists
in its want of power to operate on the steam-valve at the moment
when the equilibrium between the power and load on the engine
is disturbed. It will be seen that while the engine proceeds
uniformly, gravitation and centrifugal force must be in equilibrium
as regards the governor-balls, but at the moment when, for instance,
a portion of the load on the engine is removed, steam-power will
be set at liberty for accelerating the fly-wheel. This acceleration
proceeds in accordance with the well-known gravitation laws until
the increase of centrifugal force imparted to the governor-balls
suffices to overcome the friction of the valve and its mechanical
connexions, which are not inconsiderable. In the mean time the
speed of the engine will have been increased to an extent con-
siderably beyond what is required in order to maintain the valve
in its new adjustment ; the action of the governor, when it does
take place, will therefore be excessive, and a series of fluctuations
in the speed of the engine must follow before the proper readjust-
ment of its valve can be effected.

CHRONOMETRIC GOVERNOR. — Impressed with these imperfections

WILLIAM SIEHfENS, F.R.S. 109

in Watt's centrifugal governor, I proposed, twenty-three years
ago, in connexion with my brother, Werner Siemens, a governor
based also on the conical pendulum, which in this case was to be
independent in its action of change in its angular rotation, and,
moreover, was to be provided always with a store of power ready
to overcome the resistance of the valve at the first moment when
the balance between the power and load of the engine was dis-
turbed. This instrument (the chronometric governor) has been
applied with success in many cases where engines were required to
work with great regularity under varying conditions of load ; it
has also received an interesting application by the Astronomer
Royal for regulating astronomical and chronographical instru-
ments, proving the high degree of precision of which it is capable.
The leading idea involved in this governor consisted in providing
a conical pendulum in uniform rotation, and in establishing a
differential motion between it and the wheel driven by the engine,
which differential motion was made to act upon the source of
power of the latter. If this idea could be carried out, it was
evident that the engine, or machine to be governed, must suit its
motion closely to that of the independent rotating pendulum,
because the least retardation on its part must immediately result
in an enlargement of the valve-orifice (it being understood that an
increase of available power is always obtainable), and the least
motion in advance of the pendulum must automatically reduce the
source of power. This action would be effected instantly, notwith-
standing a considerable resistance in the valve, because' the weight
of the pendulum in rotation represents, in this case, a store of
power, inasmuch as its angular velocity would have to be suddenly
checked or increased as the case may be, unless the valve obeyed
to the first appearance of a differential motion. In order to realize
these conditions, it was necessary to maintain the conical pendulum
in question at a uniform angle of rotation, notwithstanding the
changes of driving- or sustaining-power which must necessarily
arise through its action upon the regulating valve of the engine ;
and this could only be accomplished by providing the means of
destroying or absorbing any excess of driving-power beyond what
was necessary to overcome the friction of the instrument, and which
must tend otherwise to increase its angle of rotation. On the
other hand, a surplus of driving-power had to be provided to

110 THE SCIENTIFIC PAPERS OF

prevent an occasional collapse of the pendulum in opening the
valve. The sustaining- or driving-power of the pendulum was
obtained in taking advantage of the differential motion, a weight
being attached to a horizontal lever upon the throttle-valve
spindle, which, in exerting a pressure of the differential wheel
against the two principal wheels, caused an impulse to be given
to the one connected with the pendulum, whereas the third wheel
was driven by the engine in the direction proper to raise the
weight continually. The excess of driving-power imparted to the
pendulum was destroyed by calling into action a friction between
two solid substances at the moment when the angle of rotation had
reached its intended maximum ; or a liquid resistance was intro-
duced, such as a fan, rotating with the pendulum, being dipped
into a bath of mercury or other liquid, at the moment when the
extension of the pendulum was attained (see Plate 19, Fig. 2).

The governors constructed on this principle are remarkable for
their instantaneous action upon the supply-valve of the engine,
when a sudden disturbance of the balance between load and power
takes place ; and they possess also, to the fullest extent, the power
of maintaining the regulated machine at the same speed, when the
load reaches its maximum, as when it is at its minimum. In the
hands of the Astronomer Royal the uniformity of motion obtained
by the use of this governor approaches indeed that of an ordinary
chronometer, yet it is not free from objections resulting from the
delicate adjustment and frequent attention requisite to maintain
it in good working condition. In its application to the regulating
of physical apparatus it has, moreover, been found that the conical
pendulum is apt to fall into elliptical rotation, whereby the sub-
divisions of time below the half second become inaccurate.

FOUCAULT'S GOVERNOR. — Monsieur Foucault, of Paris, has
treated the same question in a different manner, substituting for the
friction between solids, or of solids against fluids which we
employed to fix the angle of rotation of the conical pendulum,
the varying resistance of an air-propeller, according to the amount
of area for the escape of air, which area he regulates by means of
the angular elevation of his conical pendulum. The conical
pendulum he employs is mounted in the manner of a Watt's
governor of the best construction, which renders his apparatus
portable, and therefore convenient for general purposes, while, on

SSK WILLIAM SIEMENS, F.R.S. Ill

the other hand, it appears to depend for its correct action upon
the perfect condition of many mechanical details.

Some months since an idea suggested itself to me which, while
it furnishes the elements of a very general and complete solution
of the problem under consideration, appears to possess also a
separate scientific interest, which chiefly induces me to bring the
subject before the Royal Society.

LIQUID IN ROTATION. — If an open cylindrical glass vessel or
tumbler containing some liquid be made to rotate upon its vertical
axis, the liquid will be observed to rise from the centre towards the
sides to a height depending on the angular velocity and the
diameter of the vessel. As soon as the velocity has reached a
certain limit the liquid will commence to overflow the upper edge
of the vessel, being thrown from it in the form of a fluid sheet in
a tangential direction. If the velocity remain constant from this
moment, the overflow of the liquid will be observed to cease,
although the liquid remaining in the vessel will continue to touch
the extreme edge or brim. Supposing that the velocity of the
vessel be now diminished, the liquid will be observed to sink, but
will rise again immediately to its former position when the rotation
returns to its previous limit of angular velocity. This velocity is
the result of the balance of two forces acting on the liquid
particles, namely, gravity and centrifugal force.

It is a well-known fact that the curvilinear surface produced by
a liquid in rotation is that of a paraboloid, the parameter of

which is expressed by -^, and the curve itself therefore by the
formula

x signifying vertical distance from the apex, y the corresponding
horizontal distance from the axis of rotation, w the angular
velocity of rotation, and g acceleration by gravity in one second.

In this formula there is no factor denoting the density of the
liquid, which proves that the point to which the liquid is raised by
a given angular velocity is independent of the specific gravity of
the liquid employed.

By substituting for y the radius r of the rotating cup at the
brim, and for x the height h of the brim above the lowest level

112 THE SCIENTIFIC PAPERS OF

of the liquid, we may write the foregoing general formula as
follows : —

rzw~
A-^-, ....... (II.)

and

(III.)

two convenient formulae for determining the height to which liquid
will rise when the diameter of the rotating vessel and its angular
velocity are given, or for determining the angular velocity due to
a given height and diameter, it being understood that the values
of g, r, and h, must be expressed in the same unit of length.
If it is desired to determine the number of revolutions per

w
second n, instead of the angular velocity w, we may put n = ^- , and

Z7T

in putting this value for w ( = 2nri) into the last formula, we

have w = lSr ...... (IV°

These formulas are applicable to vessels of every possible exter-
nal form rotating upon their vertical axis, and to every possible
liquid, but they are based upon the supposition that the liquid is
raised by rotation to the upper edge or brim of the vessel, a
condition which it may appear at first sight difficult to realize for
practical purposes.

If the production of uniform rotation be the object in view, it
is necessary that the liquid in the rotating vessel should always be
at the point of overflow, although the driving-power might vary
considerably, and that all surplus driving-power should be absorbed
by other work than that of accelerating the rotating vessel, also
that the stock of liquid within the vessel should never diminish.
I hope to prove by the following demonstration that these various
conditions can be fulfilled by suitable mechanical arrangements
resulting in the construction of a simple and efficient instrument,
which is represented by Plate 20, Fig. 1, and which has sur-
vived the ordeal of experimental proof.

LIQUID PYROMETER. — The rotating vessel consists, in this case,
of a cup C open both at the top and bottom, but widest at the
top, the sides being made to close in towards the bottom in a
parabolic curve analogous to that formed by the liquid in rotation.

SIR WILLIAM SIEMENS, l-.R.S. I 1 3

This cup contains upon its inner surface three or four radial ribs
which unite in a central boss, by which the cup is supported upon
the spindle S. This support is not an absolute one, but the spindle
is armed with a screw-thread of rapidly ascending path, into which
the screw-threads upon the inner surface of the boss are made to
fit ; a fixed connection between the driving-spindle and the cup C
is established by means of a spiral spring E, one end of which is
fastened to the projecting end of the spindle S, and the other to
the cup. Before this spring is fixed, it is drawn out longitudinally
to such an extent as to balance the weight of the cup, which latter
may therefore be said to float upon the screw-threads without
exercising any pressure upon the same. The upper support of the
spindle S is a boss projecting from the bottom of a cylindrical
vessel B of glass sides and glass-domed top, which completely
enclose the cup C, while it renders its action visible : this outer
vessel is filled with liquid to such a height as to submerge the
lower edge of the cup. Rotation of the liquid in the outer
vessel is prevented by radial ribs upon its bottom surface ; and
upon the external surface of the rotating cup C two concentric
projections are provided, one at the upper edge, and the second
near the surface of the outer liquid, for the purpose of throwing
off some liquid which would otherwise be apt to adhere to the
external surface of the cup, in defiance of centrifugal force, and
interfere slightly with its proper action.

Rotation being imparted to the shaft S and the cup C by clock-
work or from any other source, the liquid at the bottom of the
cup will be acted upon by centrifugal force and rise upon its
inner sides, while additional liquid will enter from without and
maintain the apex of the liquid curve nearly on a level with the
surrounding lake. At the moment when the liquid in rotation
touches the upper edge of the cup, the speed should be such as is

determined by the formula n = ~-^-, in which h may be taken

£TlT

for the height of the brim of the cup above the lake surface ; but
considering that the power necessary to maintain the cup at its
velocity, after the liquid has been raised to its upper edge, is
exceedingly small, because no fresh material has to be put into
motion, it would be practically impossible to prevent further
acceleration. In order to make sure that the liquid will not fall
VOL. i. i

114 THE SCIENTIFIC PAPERS OF

below the brim of the rotating cup, an excess of driving-power
must be applied, and that excess must be disposed of otherwise
than in producing further acceleration of the cup. This is accom-
plished by means of a continual overflow of a thin sheet of liquid,
which is projected against the sides of the outer vessel and falls
back into the lake at the bottom, whence a similar quantity of
liquid penetrates into the cup, to be also raised by its rotation and
projected over its edge.

The power absorbed in raising and projecting the liquid must
be strictly proportionate to the quantity of liquid so acted on in a
given time, or to the thickness of overflow, provided the condi-
tion of uniform velocity be realized ; but the velocity of the cup
depends upon the height to which the liquid has to be raised, and
must therefore be increased in order to raise the liquid column
above the brim of the cup. The. necessary increase of velocity to
produce such an overflow is not great, considering that the height
of the liquid column increases in the square ratio of the velocity,
and may be neglected in all cases where only approximate results
are required, or where variations in the driving-power are com-
paratively small ; but no high degree of accuracy could have been
claimed for this instrument unless the following compensation
action had suggested itself : —

AUTOMATIC DIP OF CUP. — We have assumed hitherto a rigid
connexion between the cup and its driving-spindle, and unless the
cup does overflow, we are, indeed, justified in this assumption, the
spring E being too rigid to yield to the resistance of the cup in
motion when no work is performed. With an increase of power
the resistance of the cup also increases, and an overflow of liquid
proportionate to the power is produced ; the connecting spring E
must yield, at the same time, proportionately to the torsional
resistance thus created, and in the same ratio the cup will descend
upon the helical surface which serves for its guide. While, there-
fore, on the one hand, the h of our formula increases with the
overflow, it is diminished, in the same ratio, by the descent of the
cup, both depending directly upon the driving-power. In so ad-
justing the stiffness and length of the spring that the one action
equals the other for any given increase of power, it must
equal it also for other amounts of increase within reasonable
limits, and strictly uniform rotation must le the result. The

S7K WILLIAM SIEMENS, F.R.S. 115

" reasonable limits " to this automatic adjustment are imposed by
the restricted orifice through \vhich the liquid has to penetrate
into the cup, and also by the range of action of the spring, for
which the law of Mariotte is applicable. Experiments, to be
hereafter described, have shown that the driving-power may be
varied between wide limits without producing any sensible varia-
tion of speed. The final adjustment of the instrument to the
normal velocity required is, moreover, easily effected by raising or
lowering the cup while it is running, for which purpose the lower
end of the upright spindle S is supported in the axis of an
adjusting screw E, as will be seen by inspection of Plate 20,
Fig. 1.

RANGE OF POWER INCREASED. — The range of power through
which uniform rotation can be obtained, may be further increased
by an arrangement which is represented in Plate 20, Fig. 2, and
Plate 21, Fig. 6, and which consists in arresting the liquid projected
over the edge of the cup by a belt of fixed vanes M, whence it drops
through the zone of rotating radial vanes L, which again impart
tangential motion to it at the expense of the superfluous driving-
power of the cup of which they form part. It is hardly necessary to
add that these fixed and rotating vanes only increase the range
of power of tJie instrument, ivitfwut in any way affecting its rate of
rotation, and that a second set of fixed and rotating vanes might be
added with the same effect as the first. Although the rotating cup
represented in Plate 20, Fig. 2, and Plate 21, Fig. G, is not pro-
vided with the automatic dip, it is equally evident that the fixed
and moveable vanes do not preclude that arrangement, which is
only dispensed with in such cases where great uniformity of motion
is not required.

An interesting application of such a " Liquid Gyrometer," as
this instrument may appropriately be called, would be that of
obtaining synchronous motion at different places connected by a
telegraphic wire, for philosophical or telegraphic purposes ; but in
order to test its fitness for such purposes, I have constructed a
clock of which it constitutes the regulating principle, the moving
power being obtained by electro-magnetism.

CLOCK REGULATED BY LIQUID IN ROTATION. — This clock is

represented by Plate 22, Figs. 7 and 8, and consists of three
principal parts : —

i 2

Il6 THE SCIENTIFIC PAPERS OF

1. The pedestal containing a battery of two " Marie Davy's
elements " suitably arranged.

2. The body of the clock, with sides formed of plate glass, con-
taining an electro-magnet, by which rotatory motion is imparted to
an iron bar or keeper fixed upon the vertical main axis which
passes into

3. the regulating chamber, consisting of a close cylindrical glass
vessel with domed glass top containing the rotating cup and a
certain quantity of paraffin oil, which fluid is particularly appli-
cable on account of its perfect fluidity and non-affinity for the
materials composing the regulator.

The regulating cup is in this instance formed of vulcanite, and
is suspended from the top of the vertical axis by means of a spiral
spring, which, being fixed at both ends, not only supports the
weight of the cup but acts also as a torsional spring, enabling the
cup to descend upon its helical central guide whenever an increase
of driving-power calls into existence its equivalent of torsional
resistance.

The rings of stationary and rotating vanes are dispensed with in
this instance, because no great variations in the driving-power are
contemplated. The electro-magnet acts by attraction of the
armature during a small portion of its rotation, and one contact
only is required, which is so arranged that no destruction of the
metallic surfaces can arise through the discharge of extra-current
sparks, which latter are received by an elastic point of platinum
slightly in advance of the proper contact surface and moved by
the same eccentric. By this simple arrangement the usual diffi-
culty attending dry contacts is avoided, and a continued action of
the instrument ensured. A train of reducing-wheels communi-
cates the motion of the cup-spindle to hands upon the face of the
clock, which record hours and minutes in the usual manner.

The diameter of the rotating cup being = 0'040 metre, and the
height of its edge over the surface of the liquid = 0-034 metre, the
number n of its rotations per second in accordance with our
formula

metre x -038 metre=6.9 reyolutions gecon(L

Experiment gave, on the contrary, a speed of 7*5 revolutions

.s/A' WILLIAM SIEMENS, F.K.S. Ii;

per second, or *G revolution per second more than was indicated by
t lift »ry, a result which seemed to stamp the action of the instru-
ment with uncertainty, when it was recollected that no allowance
had been made for the aperture at the bottom of the cup, leaving
a portion of the rotating liquid without an external support.

CORRECTION FOR LOWER ORIFICE OF Cur. — If we assume, for
instance, that the sides of the cup were cylindrical and merely
descended below the surface of the liquid without closing in at the
bottom, we should find that by rotation of this cylinder, supposing
the inner surface to be rough or armed with radial projections, the
liquid would rise on the circumference above the external level of
surface, but would also form a depression or vortex in the centre of
rotation. The surface of the rotating liquid would be that pro-
duced by the rotation of a vertical parabola, as represented in the
diagram (Plate 19, Fig. 3). In order to maintain the hydrostatic
equilibrium between the rotating column with the external liquid,
it is necessary that its mean height of column must remain the
same, or that the plane of external liquid level must divide the
solid figure comprised between the cylinder and the curve of
rotation into equal parts. In order to realise these conditions
experimentally, it would be necessary to provide the lower portion
of the rotating cylinder with an easy fitting and balanced piston>
without which circular currents would be produced within the
rotating liquid (descending by force of gravity at the sides, and
rising again in the centre) and mar the result.

The body comprised between a paraboloid of the height ti and
a cylinder is, however, divided equally by a horizontal plane cutting

it at the distance of — - -— from the edge, or at ti — -j- = -293# from

\['£ »J i

the apex. The form of the curve depends upon the angular
velocity alone, and would remain the same if the upper diameter
of the tube were to be increased, only the curve would in that case
have to be continued until it met the sides (as shown by dotted
lines), and to the extent of these prolongations the liquid would
be raised higher above the external level without producing a
corresponding depression in the centre of rotation.

It follows that the aperture at the bottom of the rotating cup
causes a vortex or depression of the apex of the curve of rotation
below the external liquid to the extent of '293 part of the height

Il8 THE SCIENTIFIC PAPERS OF

due to its own diameter, and to the correct number of rotations
ri per second.

it is

2m
and also n'=^2ffh'

2pir
if p is the radius of lower aperture of cup ; therefore

2rir

Or *' = V--298p>;

and in substituting this value for h' into the first formula, we have

n =

2rir
or

for our corrected formula to determine the velocity of cup with
continuous overflow).

In applying this formula to the clock, taking for p its value of
8 millimetres and the other values also in millimetres in order to
avoid fractions, we find

19600 *881+

400

-16-05)

= 7*4 revolutions per second,

or '1 revolution per second less than the actual speed. This
remaining excess of the actual over the calculated velocity of the
cup is rather more than what may be fairly attributed to error of
measurements, and appears to be due to adhesion of a film of
liquid (which may be estimated at nearly '25 millimetre thickness)
to the inner sides and edge of the cup, whereby its effectual dimen-
sions are proportionately reduced. For a uniform speed this error
must be a constant quantity, which cannot affect the working of
the instrument injuriously, so long as the other conditions are such
as to produce uniform rotation. The discrepancy is diminished
by increasing the dimensions of the cup, and its amount is such

WILL/AM SIEMENS, F.R.S.

119

that the compensation is effected, under all circumstances, by one
or two turns of the regulating screw supporting the vertical spindle.

\Ylien the regulation of the cup is once effected, it continues to
rotate at a remarkably uniform rate. Change of temperature
affects the density of the liquid considerably, but does not in-
fluence the rate of the cup otherwise than inasmuch as it affects
the level in the cup-chamber, which will rise with increase of
temperature proportionately to the depth of liquid it contains, and
which is inconsiderable. The lineal dimensions of the cup, and
the length of the suspending spring, will also increase, all tending
to lower the rate with increase of temperature ; but, on the
other hand, the length of the upright cup-spindle increases with
increase of temperature, and in regulating the length and com-
position of that spindle properly, entire compensation for change
of temperature is effected. The cup-chamber being entirely closed
against the atmosphere, no fault can arise through evaporation or
dispersion of the liquid within moderate periods of time.

In the case of the clock under consideration, no compensation
for change of temperature was provided, nor is the cup entirely
balanced by the spring, yet its rate is as uniform as that of the
common clock with which it has been compared ; but in order to
test the regulating-power of the instrument, the driving-power
was varied by the introduction of artificial resistances into the
galvanic circuit, when the following results were observed : —

Time of
Common
Clock,
in Seconds.

Mercury units
of Resistance
of Clock
and Battery.

Units of
Resistance
put into
Circuit.

Relative
Value of
Driving-
power.

Time of
Liquid Gyro-
meter Clock,
in Seconds.

Variations
in Time
produced,
in Seconds.

720

110

0

100

736

16 slow

720

110

10

92

734

14 „

720

110

20

85

734

14 „

720

110

30

78

733

13 „

720

110

40

73

732

12 „

720

110

60

69

731

11 „

720

110

60

64

729

9 „

720

110

70

61

726

6 ,,

720

110

80

58

723

3 „

720

110

90

55

720

correct

720

110

100

52

723

3 fast

720

110

110

50

726

6 „

720

110

120

48

725

6 „

720

110

130

46

724

4 „

720

110

140

44

722

2 „

720

110

150

42

729

9 slow

when the overflow had ceased.

120 THE SCIENTIFIC PAPERS OF

This result shows that the spring employed was decidedly too
weak, producing a decrease, of speed with increase of power.

A more careful adjustment of the spring would improve these
results, which suffice, however, to prove the capabilities of the
instrument.

It appears at first sight as though the friction of the cup upon
the threads of the screw must interfere with its automatic adjust-
ment ; but this is practically not the case, owing to the circumstance
that small fluctuations in the resistance continually occur, causing
torsional oscillations of the cup, the mean of which must be its
true position notwithstanding friction, which friction moreover is
reduced to a minimum, owing to the suspension of the weight by
the spring.

Another interesting quality of the gyrometric cup is its com-
parative indifference to a vertical position ; it may, indeed, be
tipped very considerably without interfering with the uniform
overflow all round, and the time of its rotations is diminished only
in the ratio of the square roots of the vertical mean heights, or
it is ri : n = JOT: >Jh cos /3, or for a tipping angle /3 = 3°,
ri : n = 1 : 1*0007, showing that no particular care is requisite
to place the instrument upon a horizontal foundation.

G-YHOMETRIC GOVERNOR. — The most useful practical applica-
tion of this instrument is that of regulating the power and velocity
of steam-engines. A cup of very large dimensions, provided with
several belts of check-vanes and with the automatic dip arrange-
ment, might be conceived which, in being connected by gearing
with the main shaft of an engine, would limit its velocity by
absorbing directly all its surplus power. This surplus power
would appear in the cup-chamber in the form of molecular motion
or heat, and would have to be got rid of by the application
of cooling agents, if the natural dispersion of heat by radiation
would no longer suffice to keep down the temperature. This
would, however, be a wasteful proceeding, and it becomes necessary
to operate, not upon the power produced, but rather upon the
source of power, by rendering it always equal to the accidental
resistance or load in order to maintain uniform velocity. The
arrangement adapted for this purpose is represented by Plate 20,
Fig. 2, and Plate 21, Figs. 3, 4, 5 and 6.

S is the main shaft of the engine, which imparts motion to the

WILLIAM SIEMENS, F.R.S. 121

upright spindle H carrying an inverted wheel E. The axis A of
the regulating cup, being concentric with the spindle H, carries a
pinion F, and both the wheel and the pinion gear into two inter-
mediate or planet-wheels G and G', which latter are loose upon
studs and are suspended from a rocking-frame I, the latter being
free to turn upon its central support. The rocking-frame I is
connected by a rod R to a bell-crank lever 0, which is fastened
upon the spindle of the steam-regulating or throttle-valve V ; but
the horizontal arm of the lever 0 carries a weight P which, being
acted upon by gravity, tends to open the valve and at the same
time to force the rocking-frame I, with its planet-wheels G and G',
round the vertical axis. This pressure is resisted, on the one
hand, by the teeth of the pinion F, which can only yield in the
ratio imposed by the rotating cup C, and on the other hand, by
the teeth of the inverted wheel E, which latter, being driven
round by the engine in the direction contrary to the effort pro-
duced by the weight P, causes the latter to be continually raised.
If the engine should succeed in raising the weight P, say 100
millims., while the regulating cup yields also to the extent of
100 millims., then the lever 0, and with it the regulating valve V,
will retain their relative position ; but if the engine should raise it,
say 95 or 105 millims., while the cup yields 100 millims., then the
valve will be either opened or closed by 5 millims., and the engine
will be urged or checked, as the case may be, to such an extent as
to make its revolutions coincide again absolutely with those of the
regulating cup.

The driving-power of the cup is limited by the weight B, and
may be considered as a constant regulated quantity (although it is
derived indirectly from the engine) ; but whenever the valve has
to be opened or closed, a resistance arises which goes either in
diminution of or in addition to the weight P, and the power of
the cup must be such that its speed remains sufficiently uniform
under the influence of these occasional variations.

By means of the automatic dip arrangement it would not be
difficult to obtain perfect uniformity, notwithstanding these irregu-
larities in the driving power ; but in the case of steam-engine
governors this arrangement is dispensed with for the sake of a
more immediate action upon the valve at the moment when a
differential velocity arises. Supposing that the greatest amount of

122 THE SCIENTIFIC PAPERS OF

occasional variation of speed is to be 1 per cent., and that the
dimensions of the cup are as follows,

Diameter at upper rim . . . = 200 millims.

Diameter of rim above the liquid . . = 200 millims.

Diameter of orifice at bottom of cup . = 90 millims.
we find the available power of the instrument in the following
manner : —

The normal speed of a cup of these dimensions is, according to
our formula.

= 3-31 revolutions per

second.

The height to which the liquid is raised by rotation increases in
the square ratio of the speed of rotation, the increase of height
due to 1 per cent, increase of speed would therefore result from
the following proportion, 100 v : 101 v = ,Jh : JR, or the height
due to the increased speed, ti = 204 millims., that is to say, the
liquid would be raised 4 millims. above the brim of the cup,
which, being an unbalanced column, will produce an upward flow
of the liquid in the cup, as expressed by the well-known formula,
v = ftgh ; or h being in this case = 4 millims., we have v = *280
metre flow per second, which, if multiplied by the least sectional
area at the entrance into the cup = "0057 square metre, gives
the quantity of liquid '280 metre x '0057 square metre = '0016
cubic metre, or r.6 litre of liquid raised 204 millims. high and
projected with a velocity of \%^ x 3*31 x 27ir = 2*1 metres per second
over the brim of the cup, to be stopped by the stationary wings
and once more accelerated by the rotating wings, which, being
one-fifth more in diameter than the cup itself, impart to the
liquid a velocity of f x 2'1 = 2'5 metres per second. These accele-

vn-
rations represent a power which, according to the formula h =-x-,

*9
is equal to the same liquid being lifted

for v = 2*1 metres . to '225 metre height,

. and for v = 2'5 . . to "320 metre height.

In the cup it was lifted to '200 metre height,

making a total lift . to '745 metre height.

WILLIAM SIEMENS, F.R.S. 123

If the liquid employed is water, the 1'6 Hire represents 1'6

kilogram., and the resistance produced by 1 per cent, increase of

ity is 1'G x '745 = l"2 kilogrammetres per second, which is

equivalent to forcing a valve rod or lever through one decimetre

space in fen seconds against a resistance of 120 kilograms.

This force exceeds by far what is usually required of a governor,
and fits the new instrument for the accomplishment of objects
which could hitherto not be attempted, such for instance as regu-
lating the power of engines by the link motion or other variable
expansion gear (whereby a considerable saving of fuel may be
effected), or moving the gate of a water-wheel.

The most striking feature of this governor is, however, the
rapidity with which a readjustment between power and load of an
engine is effected ; experiment has proved that two-thirds of the
Mai load upon an engine may be suddenly thrown off without pro-
ducing any visible change in its rotation. It must be borne in mind
that the variation of 1 per cent., which was assumed in calculating
the power of the instrument, applies only to the short period of ten
seconds, when the readjustment of the valve is effected, after which
the cup and, with it, the engine returns to its normal rate of rota-
tion ; but an increase or decrease of speed of 1 per cent, during ten
seconds is so small a total amount that it is not perceptible to the
eye, and may be regarded as non-existing for practical purposes.

A uniform velocity of engines, such as is secured by this
instrument, is of considerable practical importance in spinning,
grinding corn, and other manufacturing operations, because it not
only prevents breakages and irregularities in the quality of the
work produced, but it enables manufacturers to attain to the
maximum speed at which their work can be produced.

It has been shown that the rotating cup is not sensitive to
tipping, a circumstance rendering the gyrometric governor also
applicable to marine engines, which are much in need of a means
for maintaining a uniform speed, particularly where the screw
propeller is used, which, in being raised by waves above the water,
allows the engine to fly off at a dangerous speed.

The gyrometric cup appears therefore to be equally applicable
for maintaining powerful machinery at a nearly uniform velocity
and for obtaining the higher uniformity of rotation requisite for
philosophical and telegraphic instruments.

124 THE SCIENTIFIC PAPERS OF

In the discussion of the Paper entitled

"DESCRIPTION OF THE STEAM STEERING ENGINE

IN THE 'GREAT EASTERN' STEAMSHIP,"

By Mr. J. MCFARLANE GRAY,

MR. 0. "W. SIEMENS * observed that the present paper was the
first they had had upon a subject which he thought was a very
important one ; for to see a large number of men at the helm of
a big ship, trying to bear her round, certainly showed great imper-
fection in the present mechanical arrangements for steering. The
apparatus now described as applied to the steering of the " Great
Eastern " gave evidence of a great deal of thought ; and the
details of the arrangement for starting the steering engine, and
making it obey signals given from a distant part of the ship, were
highly ingenious, and the results of the working in practice
showed, that all the requirements of the case had been duly con-
sidered beforehand. He inquired whether any difficulty had ever
been experienced in starting the engine after it had been standing
for a time, owing to accumulation of water in the cylinders ; and
whether a difficulty did not arise from the same cause whenever
the rudder had to be held by the engine for any length of time in
the same position ; and he asked what provision was made for
getting rid of water in the steam cylinders.

Another question that had occurred to him, though somewhat
beyond the immediate subject of the paper, was whether the
steering engine during the time it was standing might not be
made serviceable for aiding in the propulsion of the ship by means
of a Ruthven re-action propeller, discharging a jet of water from
the stern in the opposite direction to the ship's motion. He did
not advocate such a plan of propulsion in substitution for the
screw and paddles ; but he thought it might be advantageously
applied in conjunction with these, for utilizing a moderate auxiliary
power from a separate engine like that employed for the steering-

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1867, pp. 279-281.

WILLIAM SIEMENS, f.K.S. 125

r, which was not required to be constantly at work for steering.
The stream of water projected backwards from the stern might
either be discharged in a line parallel to the keel of the vessel, or
by means of deflectors it might readily be diverted to either side
for changing the ship's course. Such an arrangement, he thought,
would have several advantages over any plan of steering by a
rudder, because the action of a rudder depended upon the onward
motion of the ship, and without onward motion a rudder could
produce no effect ; whereas by a jet of water issuing laterally from
the stern the ship could be steered round while she was not moving
forwards or moving very slowly, which in some circumstances was
an important consideration. In managing a ship for the purpose
of picking up submarine telegraph cables, he had himself found
very great difficulty in turning the ship round by means of the
rudder, while obliged to remain stationary upon the same spot for
the performance of the necessary operations. Moreover, if such a
plan of steering by a water jet were thoroughly carried out, the
rudder, which was the most vulnerable portion of a ship, might be
dispensed with.

In the discussion of the Paper

"ON LIQUID FUEL," by BENJAMIN H. PAUL,

THE Chairman * (Mr. C. "W. SIEMENS) said that the discussion
had elicited quite a conflict of opinions, while those held on each
side were apparently based on independent facts ; but he believed
it was not impossible to reconcile many of the statements which
were apparently contradictory, by properly discriminating between
theoretical and practical results. These two ought to be kept
distinctly apart. The practical result would ultimately approach
to the theoretical, if all the conditions of the various operations
were perfect, but it could never quite attain it ; and, unfor-
tunately, in the consumption of solid fuel, the practical results did

* Etccrpt Journal of the Society of Arts, Vol. XVI. 1867-1868, pp. 409-410.

126 THE SCIENTIFIC PAPERS OF

not nearly approximate to the theoretical ones. He believed that
some years hence any engineer who looked at the furnace of one
of our present marine boilers would be ashamed of it. The mode
of throwing the fuel on a kind of volcano, giving off a large
proportion of valuable substance in the form of thick smoke,
was very objectionable. As far as he could follow the calculations
given, he thought Dr. Paul had rather overstated than otherwise
the theoretical evaporating power of liquid fuel as compared with
coal. He gave the evaporating power of hydrogen at 64 Ibs., which
agreed with the statement laid before the United Service Institu-
tion by Professor Macquorn Rankine, but there was a correction
to be made in that. Professor Rankine gave the evaporating
power of hydrogen at 64*2, and of carbon at 16 '05, but he con-
sidered, in this calculation, the hydrogen as existing in a gaseous
state, and the carbon in that of a solid. He made an allowance
for the chemical affinity between the hydrogen and carbon when
in the form of marsh gas, but he did not allow for the hydro-
carbon as being in a liquid condition, in which state it was when
in the form of oil. There must be a correction made on that
account, though he could not precisely say what it should be.
There was another correction to be made with regard to the latent
heat of the steam resulting from the combustion of hydrogen, by
which, no doubt, the results would be very sensibly modified.
However, if they went from theory to practice, they had certainly
great allowances to make in favour of the oil. For instance, coal
contained not only alkaline matters, but also a great deal of water,
which oil did not ; then, a certain portion of oxygen was already
absorbed by the coal, and there was a great waste by smoke.
Further, it was impossible, whilst burning fuel upon a grate, to
obtain that regularity of proportion between the air and the
material consumed, which was necessary to produce an economical
result. All these points were arguments in favour of the liquid
fuel. He must say, that he quite agreed with several of the
speakers, that volatile liquid fuel was totally inapplicable, and
would be one of the most dangerous things imaginable on board
ship ; they must, therefore, consider the question as confined to
the use of heavy oils, which might, no doubt, be employed with
advantage. They admitted of better stowage than coal, occupied
much less bulk, and would save on board ship a great deal of

Sffi WILLIAM SIEMENS, F.R.S. 127

labour, which meant space, of course, as less men would be
required. Then there was another great advantage — there was no
smoke. This, in the case of men-of-war, was very important,
because a fleet of steam vessels at present could be seen while they
were many miles below the horizon. For the mercantile marine,
however, the question would reduce itself to one of price ; and if
the oil were £5 a ton, and the coal £1, no doubt the advantage
would be in favour of the latter.

Captain Selwyn said he was now using oil at Id. a gallon, or not
quite £1 Is. a ton.

The Chairman said, that so long as the oil could be obtained at
anything like the price now mentioned, no doubt it would be a
most valuable fuel ; but the question was, would the price remain
so favourable to the consumer if the demand should increase ? Of
that he must say he had considerable doubt. If they had to distil
the oil specially for the purpose from coal it must be expensive,
and they must therefore fall back upon the natural supplies, or
those which were incidental to other manufactures, which supplies
must necessarily be limited. As to the use of water for burning,
he was quite sure that no one acquainted with the subject would
attribute any special evaporating power to water itself. . Water
might be usefully applied sometimes in conveying heat from one
place to another, as, for instance, the introduction of a jet of
steam under a grate on which anthracite coal was burning pro-
duced a gaseous fuel, the heat from which might be readily con-
veyed to a considerable distance, but as to getting heat out of
water it was absolutely impossible. He would conclude by moving
a vote of thanks to Dr. Paul, to which he was sure they
would feel he was fully entitled for his able and carefully
written paper.

128 THE SCIENTIFIC PAPERS OF

In the discussion of the Papers

"ON MACHINES EMPLOYED IN WORKING AND
BREAKING DOWN COAL, so AS TO AVOID THE USE
OF GUNPOWDEE," by SAMUEL PAEKEE BIDDEE, Jun. Assoc.
Inst. C.E. ; and

"ON COAL GETTING MACHINERY, AS A SUBSTITUTE

FOE THE USE OF GUNPOWDEE," by CHAELES JOHN CHUBB,

ME. C. W. SIEMENS*" remarked that it was of the greatest
interest to all that coal should not only be got economically, but
without involving human suffering. Although it had been as-
serted that the casualties caused by the explosion of powder were
not so numerous as might have been supposed, nevertheless it was
acknowledged that there was a serious loss of life every year.
Independently, however, of that, there was the question of annoy-
ance and injury to health that might be produced by noxious
vapours arising from the explosion of powder in confined spaces.
Moreover, it had been shown that coal could be got more
economically by wedging than by blasting. The quantity of coal
released by wedge action must be greater than that released by
explosions, inasmuch as powder threw out the cone of least
resistance, whereas a wedge would exert its power upon a larger
range of coal. The question of undercutting the coal was no
doubt important. He had been endeavouring to calculate what
force would be expended in each case, but he had not been able to
obtain a satisfactory result. If the coal was holed, or undercut, a
disc of it had to be dealt with, and with the wedge the line of
cleavage would be contained in that disc. He expected that the
force expended would increase with the depth at which the force
was applied ; so that if the machine was inserted into a hole 4 feet
deep, it would expend four times the force that it would if the
machine was put in only 2 feet deep. But the question became
much involved if the coal was to be torn out of the face, because

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXVIII. Session 1868-1869, pp. 141-142.

S/R WILLIAM SIEMENS, F.R.S. 129

then the top and bottom rocks formed an element which he
thought no calculation could reach. If the working was carried
out in a mass of coal of infinite extent, then one might be
disposed to consider the resistance to the circumference of the
semicircular or shallow cylinder and the sphere which would have
to be separated if the force were applied in a mass ; but inasmuch
as the layer was confined between sides of rock, and that rock was
pressed by a superincumbent weight of untold amount, the
question was incapable of solution. On general grounds he
believed it would be impossible for the Institution to arrive at
any definite conclusion with regard to the best machine. The
machines of Mr. Chubb and of Mr. Jones were similar in
principle, and were both no doubt powerful though limited in
range of expansion ; but, while such limited range was sufficient
to separate the coal, he believed that the largest quantity of coal
could be got with a minimum expenditure of force, because all
the force of that short range was directly applied. It was, how-
ever, necessary that the coal should be sufficiently hard to allow of
a perfectly cylindrical hole being drilled or bored, and that it
should not yield more than the length of the stroke of the pistons.
Mr. Bidder's machine, on the other hand, was capable of greater
range, and by it greater force could be brought to bear upon one
point. It seemed also to have the advantage that a smaller hole
sufficed for its action ; but, on the other hand, that hole must be
deeper than was necessary in the case of Mr. Chubb's machine.
He subscribed to the objection urged by Mr. Mallet, that a pro-
portion, not easily ascertained, of the force expended must be lost
by pressing the front of the wedge against the mass of dead coal.
The wedges pressed not against the coal that was going to be
loosened, but against the dead mass of coal further on ; and it was
a question, which would puzzle a mathematician to determine,
what proportion of force was expended in compressing the coal,
and what proportion in cracking it away from its position. With
Mr. Bidder's machine the loss of a portion of the pressure by the
compression of the coal suggested to Mr. Siemens a modification
which was this : — where coal was naturally tender, and where it
was advisable to work it in comparatively thin layers, there might
be applied a machine composed of three strong steel bars, arranged
like a pair of scissors, with a hydraulic ram driving at the ends
VOL. i. K

130 THE SCIENTIFIC PAPERS OP

into gudgeons, so as to press out the ends of the scissors ; the ends
of the bars being provided with little surface-plates to distribute
the pressure over a certain extent of surface. Such a machine, in
coal sufficiently tender, would work rapidly, because with one
piston in operation a range of at least 9 inches could be obtained,
and the fissure in that case would be in aid of the separation of
the mass.

In the discussion of the Paper

" ON THE PROGRESS OF LIQUID FUEL," *
By Captain J. H. SELWYX,

MR. CHARLES W. SIEMENS said, as the Chairman has called
upon me, I may say I go so far with the author of the paper, in
the able manner in which he has put it forward, that in burning a
fuel, a hydro-carbon, such as oil is, we may expect very superior
results to those which we obtain from evaporating water by carbon
or ordinary coal. No doubt, the calculation which Captain Selwyn
has put before us is the correct one, according to the best facts
with which we are acquainted. Only I would observe that if the
6'4G per cent, of non-combustible matter in the oil is oxygen
there should be a tare allowed on account of it.

Captain Selwyn. I think I gave that.

Mr. Siemens. It appeared to me that Captain Selwyn did not
make any distinction between oxygen and mere foreign matter.

Captain Selwyn. I took both cases and gave you the result of
both cases.

Mr. Siemens. Well, in the case of oxygen, of course combus-
tion takes place without generation of heat, inasmuch as the fixed
oxygen in combining with fixed carbon produces a gaseous com-
pound, carbonic acid, without generating heat at all. Therefore,
for the equivalent of the oxygen, carbon ought to be taken off the
result. It may be asked, why should oil give such a superior result

* Excerpt Transactions of the Institute of Naval Architects, Vol. X. 1869, p. 37.

.S7A' WILLIAM SIEMENS, F.R.S. 131

to coal, which, if you analyze it, contains very often not more than
<; |> r cent, of foreign matter. I think an explanation maybe
attempted in this way, that in burning coal we feed the furnace in
an exceedingly irregular manner, charging the grate at one time
with an excess of hydro-carbon, which flies off" and finds an in-
sufficient amount of oxygen in the atmospheric air that enters the
Lrnu<- to burn it ; and at other times the air that rushes through
the grate has not time to take up its equivalent of carbon to pro-
duce carbonic acid. The result is, that in examining the products
of combustion of an ordinary furnace, we find as much as 25 or
30 per cent, of pure oxygen on one hand, and we find a consider-
able amount of unconsumed carbonaceous matter going into the
chimney along with it. This must be attributed not only to the
unequal feeding of the coal into the grate, but it must also be
attributed to the cooling effect exercised by the boiler, in checking
combustion before the chemical action is completed. The first
consideration in constructing a suitable grate ought to be to main-
tain the. temperature of the flame for a sufficient length of time
to complete the chemical action between the oxygen and car-
bonaceous matter. That, no doubt, can be obtained in a much
more perfect manner in dealing with a liquid fuel like oil. This
liquid in itself would be rather difficult to deal with, inasmuch as
the surface it presents in the liquid state to the oxygen of the
atmosphere would be very limited ; but in blowing in it, as Captain
Selwyn does, by a jet, you get an extreme subdivision of the oil,
and in consequence you may obtain a perfect combustion. From
the results brought before us by Captain Selwyn that seems to be
the case ; we get an extremely perfect combustion, and, therefore,
we get an extremely great duty performed. The practical question
then arises, what are the sources of supply of this oil ? I do not
know whether the author of the paper has gone into that point, but
it is a question of very great practical importance. As long as we
have liquid fuel, there can be no doubt that, weight for weight, it
is superior to solid fuel. These heavy oils are free from the great
objections that apply to light hydro-carbons, on account of their
volatile qualities. But the question with me is, whether we can
obtain such a supply of heavy oils at a moderate cost as will
justify their employment upon a large scale ; for instance, will
justify their application to our marine boilers.

K 2

132 THE SCIENTIFIC PAPERS OP

In the discussion of the Paper

"ON RECENT IMPROVEMENTS IN REGENERATIVE
HOT BLAST STOVES FOR BLAST FURNACES,"

By EDWARD ALFRED COWPER, M. Inst. C.E.,

MR. C. W. SIEMENS* said, the apparatus described by Mr.
Cowper for heating the blast differed from the ordinary hot blast
stove in this essential respect, — that the same surface of refractory
material that absorbed the heat from the products of combustion
served also to impart heat to the incoming blast. It was neces-
sary, under these circumstances, that there should be two stoves,
or air-tight receivers, filled with the refractory materials : the one
to give off its heat to the incoming blast on its way to the furnace,
and the other to be heated up for continuing the operation when
the heat of the first-named stove was exhausted, valves being pro-
vided for reversing the currents from time to time, as described
in the paper. By these means the same surface of refractory
materials was made to take up the heat from the burning gas and
to impart it to the blast ; and as a consequence it followed, that
the degree of heat that might be imparted to the blast was not
limited by the point of safety of the iron tubes composing the
ordinary stoves, but was limited only by the point of fusion of the
refractory material's composing the regenerator of the new stoves.
For heating blast to a moderate degree only, this stove possessed
some advantages over the ordinary tube-stove, as the use of re-
generators was always attended with a saving of fuel : but when-
ever a high degree of heat was desired exceeding 1,000° Fahr., then
the tubular stove altogether failed, and the regenerative stove
must be used.

The question of importance connected with this subject was,
whether it was desirable to use blast of such very high tempera-
ture. Some years ago it was the opinion of practical ironmasters,
that the quality of the pig metal produced was deteriorated when

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXX. Session 1869-70, pp. 315-317.

5Y/? WILLIAM SIEMENS, F.R.S. 133

mperature of the blast exceeded 500° or 600°. Now, how-
i-vrr, ironmasters saw the advantage of raising the temperature of
the lilast from G(>»>° to 800° or 1,000° ; and in using this regenera-
tive stove, the temperature had been raised as high as 1,400°,
without any deterioration of the quality having been observed.
At the same time, there were ironmasters of high intelligence who
asserted, that though the quality of the metal was maintained, yet
that there was very little or no economy whatever in increasing the
temperature of the blast above 1,000° ; but although this view had
been gravely advanced lately, in discussions that had taken place
before another Institution, he felt convinced that the conclusion in
question was based upon a misapprehension, and would share the
fate of the former objections to the introduction of hot blast
generally.

In inquiring into the chemistry of the blast furnace, it would
be difficult to separate carefully the operations going on at each
stage within the vast mass of incandescent matter ; but if the
problem was to decide upon the goodness or the efficiency of a
mechanical contrivance for transmitting power, consisting of a
number of wheels, cams, and differential pulleys, there would be no
very minute inspection of the effect produced by each crank, wheel,
or pulley ; but the conditions of power applied to the machine
would be taken, and the power judged by the motion of the load
at the other end, taking it for granted that, according to the law
of the conservation of force, nothing could be added or taken away
except by friction. In the same way he looked upon blast furnaces.
It was known that carbon, fluxing material, and ore, consisting for
the most part of sesqui-oxide of iron and of silica, were put into
the furnace, and atmospheric air was blown in at bottom. It was
known that atmospheric, air on coming into contact with carbon in
a state of incandescence formed one of two combinations — viz.,
carbonic acid if the combustion was perfect, and carbonic oxide
if the combustion was imperfect. It was known that when a
pound of carbon was burned into carbonic acid 14,000 units of
heat were liberated ; whereas if carbonic acid was formed, only
4,000 units of heat were liberated. It depended, therefore, entirely
upon the composition of the gases resulting from the combustion
in the blast furnace what would be the effective duty performed.
Analyses of the gases issuing from the top of the furnace proved

134 THE SCIENTIFIC PAPERS OF

them to consist of about -ifchs of carbonic oxide and -J-th of car-
bonic acid, and if this proportion was invariable, it would be
possible to say with great precision what would be the amount of
heat developed by a given amount of blast of a given temperature.
There were several duties to be performed by the heat thus
developed. The iron oxide had to be deoxidized or reduced,
and the iron had to be melted. The " gang " and flux had
to be combined and melted; and the products of combustion
themselves escaped at the top at a certain temperature, which im-
plied a loss of heat. But, without going into a priori calculations
of these separate duties, he found that the necessary consumption
of coke to smelt a ton of iron from Cleveland iron-stone, contain-
ing 40 per cent, of iron, could not be less than 19'6 cwt. This limit of
economy could not be surpassed, however much the capacity of the
furnace might be increased ; it was the limitation which could be
approached but never exceeded, except by raising the temperature
of the blast to a higher point. In doing that there was heat im-
parted from without the furnace, which necessarily saved a similar
amount of heat that had to be developed inside the furnace, by com-
bustion of coke by means of the blast. In continuing this argument
it might be thought, at first sight, that the burning of 1 pound of
fuel in the stoves would save 1 pound of coke in the blast furnace ;
and this would really be the case, if the products of combustion
of the blast furnace were carbonic acid, as in the stove ; but in-
asmuch as the escaping gases were a mixture of carbonic acid with
carbonic oxide, in the proportion of 1 to 4, the number of units of
heat produced within the furnace per Ib. of fuel would be reduced in
the proportion of 14,000 : $ 4,000 + £ 14,000 or of 14,000 : 6,000 ;
that was to say, the number of units of heat produced by each
pound of carbon consumed in the blast furnace did not exceed
6,000 ; whereas each pound of carbon burned in the stove would
produce 14,000 units. It was evidently, therefore, important to
raise as much heat in the stove as possible, in order to save more
than double the amount in the blast furnace ; and when it was
borne in mind, that the fuel in the stoves was obtained for
nothing by burning the waste gases from the top of the blast
furnace, it was obvious that there was the strongest argument for
getting as much heat as possible in that way. The argument in-
sisted upon by some ironmasters, that it would be no use to go

Sffi WILLIAM SIEMENS, F.R.S. 135

beyond 1,000°, was based upon a misconception. They showed
that each degree of extra heat in the blast produced a diminishing
amount of saving in coke per ton of iron ; but they lost sight of
this circumstance, that the quantity of blast used per ton of
iron was greatly reduced when raised to a high temperature, —
the quantity of blast used being proportionate to the amount of
carbon burnt in the blast furnace. In effecting, therefore, a re-
duction of the coke from 24 cwt. to 20 cwt., the amount of blast
used per ton of iron was diminished in the same proportion, and
1" of temperature added to the blast did not represent the same
amount of heat as 1° of heat communicated to blast of a lower
temperature, when relatively a larger quantity was used ; but if
the effect of 1 unit of heat added to the blast in the stove at
various degrees was taken into account, it would be found that as
much saving of coke was effected in the blast furnace by adding to
the blast above 1,000° as there was by adding 1 unit of heat below
that limit. There was the further important saving of fuel and
engine-power, in effecting an increased duty with a given amount
of blast, which had been neglected in these calculations.

It would take up too much time to go further into this some-
what complex question, as it would be necessary to take into
account the increased proportions of the carbonic acid gas and the
lower temperature of the waste gases which resulted from high
temperature of blast ; but the question resolved itself into a simple
one if it was looked at from a general point of view — that each
unit of heat given to the blast before entering the furnace must
save fuel equivalent to between 2 and 3 units of heat in the furnace,
and moreover, that a much cheaper fuel was available in the stoves
than within the blast furnaces.

136 THE SCIENTIFIC PAPERS OF

In the discussion of the Paper

"ON THE WARSOP AERO-STEAM ENGINE,"
By Mr. RICHARD EATON,

THE CHAIRMAN (MR. C. WILLIAM SIEMENS) * remarked that
the subject of the paper was a very interesting one, both on
account of the theoretical questions involved, and also in conse-
quence of the practical results that had been arrived at, which
seemed to be of a decidedly favourable character.

The main question that presented itself in regard to the use
of the air injection was, what was the cause of the advantages
ascribed to it. These advantages were of two kinds : — first, the
advantage in the boiler of preventing sediment and preventing
priming by the injection of heated air into the water ; and secondly
the advantage in the engine of getting more work done by a given
volume of the mixed steam and air than would be performed by an
equal volume of steam alone at the same pressure.

With regard to the advantage in the boiler, consequent upon the
injection of divided currents of heated air into the water, he thought
there was reason to, anticipate a marked effect from such a process ;
for each bubble of heated air rising through the mass of hot water
would naturally become saturated with steam by its contact with
the water, increasing in volume many times, and would thus relieve
the evaporating surface of the boiler of part of the work it woukj
otherwise have to perform. Moreover in ordinary boilers, the steam
being generated only in contact with the evaporating surfaces of
the boiler-plates and tubes, each bubble of steam at the moment of
its generation left behind it a liberated particle of solid matter,
which had previously been contained in the water ; and this being
liberated in immediate contact with the boiler surface, it was natural
and probable that it should attach itself at once to that surface in
the form of a solid incrustation. If however the required evapora-
tion in a boiler could be produced otherwise than in contact with the

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1870, pp. 245-248.

SIX WILLIAM SIEMENS, F.R.S. 137

N it appeared likely that the sediment in the water would form
its'-lf into mud composed of an aggregation of separate small
particles ; and from the facts brought forward in the paper this

•d to be the action that really took place with the injection
of the heated air. With regard to priming, an interesting point
to ascertain in a boiler working with the air injection would be the
temperature of the water with reference to the working pressure of
the mixed steam and air : because when two gases co-existed in the
same space, each exerted its pressure independently of the other,
and therefore it seemed reasonable to infer that the temperature of
the water would be lower than the temperature corresponding to the
pressure in the boiler, in the proportion of the original pressure
of the steam alone to the actual pressure of the combined steam and
air in the boiler. Thus supposing that the steam in the boiler
were at a total pressure of three atmospheres, the water being at
the corresponding temperature of 275° Fahr., and that an equal
volume of air at atmospheric pressure were compressed and forced
into the boiler, the compressed air would expand again to atmo-
spheric pressure in the boiler during the transit of the air-bubbles
through the heated water, and each bubble of air would take up
steam by vaporisation until the point of saturation was reached.
The pressure of the mixed steam and air in the boiler would then
be four atmospheres, while the temperature of the water would
correspond to the density of the steam irrespective of the air
present, and would consequently remain that corresponding to a
pressure of only three atmospheres, the atmosphere of air permeating
the whole steam space in equilibrium, in accordance with the law
of the mutual diffusion of gases. It would be interesting to observe
whether this were really the fact ; and if it were so, he thought it
would explain the circumstance which had been mentioned in the
paper, that the moment the air was turned off violent ebullition
and priming took place. For as soon as the air was turned off, the
mixture of air and steam over the water in the boiler would no
longer be replenished by a similar mixture of gases ; and the
temperature of the water being below that corresponding to the
pressure of the mixed air and steam, a sudden collapse of the steam
would necessarily take place ; and as a consequence the engine
would require more steam, the throttle valve would open wide, and
the larger volume of steam suddenly called for would of itself

138 THE SCIENTIFIC PAPERS OF

naturally induce priming, independently of the further circumstance
that the liberation of the steam from the water would no longer be
facilitated by the passage of bubbles of heated air through the mass
of the water.

In reference to the other question — why the mixture of steam
and air should do more duty in the engine than an equal volume of
steam alone at the same pressure — he did not think this could be
because the air-compressing engine which was united with the
steam-engine was more economical than the steam-engine itself.
On the contrary, mere air-engines, wherever they had been tried,
had not given such favourable results as good steam-engines ; and
there were theoretical reasons why they could not give better results
than a proper steam-engine. It had however occurred to him as an
explanation which might be suggested of the economy found to
result from the use of mixed steam and air in the engine, that the
presence of air in the steam cylinder would certainly prevent the
condensation of steam against the sides of the cylinder : and the
loss from this cause he believed to be much greater than was usually
supposed. If the cylinder was uncovered, the loss by condensation
was very great, and was shown by the form of the indicator dia-
grams taken from such engines ; and even when the cylinder was
carefully clothed, there was a considerable condensation from the
high-pressure steam coming in contact with surfaces that had been
cooled by previous exposure to the condenser or to the exhaust steam
at atmospheric pressure. This condensation he believed took place
even when the steam cylinder was thoroughly steam-jacketed all
round and at the ends, although to a much less extent, and in that
case it arose from the slow conducting power of the thick metal sides
of the cylinder allowing the inner surface to continue considerably
cooler than the outer surface heated by the steam in the jacket ;
and though the steam thus condensed in the cylinder was afterwards
evaporated a£ \n in the latter part of the stroke, yet its mechanical
effect was aln ^<st entirely lost. The presence of air mixed with the
steam, it appeared to him, would have the effect of preventing
this condensation from taking place after the first moment of con-
tact of the steam with the cool metal, because at that moment the
condensation of the particles of steam which first touched the cool
metal would leave a lining of air in contact with the metal, consist-
ing of the portion of air that had been mixed with the steam before

in 1. 1. 1 AM SIKMENSi F.K.S. 139

its condensation ; and a veil composed of the non-condensible and
non-conducting portion of the mixed gases would thus remain to
protect the rest of the steam in the cylinder from contact with the
cool metal surfaces, and would thereby prevent further loss from
condensation. He believed this was a point of considerable im-
portance in favour of the use of the combined steam and air in an
t-n^ine with unprotected cylinders. The application of the same
plan to a condensing engine was one that he had not been acquainted
with previously ; and from the statements which had been given
it appeared that the introduction of so large a quantity of air with
the steam had not been found to prevent the formation of a
tolerable vacuum.

The use of the air injection was a subject which he hoped would
be more fully investigated ; and from the perseverance which had
been shown in working it out thus far there was reason to expect
the plan would be developed to a still further extent. He moved a
vote of thanks to Mr. Eaton and Mr. Warsop for the paper, which
\\as passed.

In the discussion of the Paper

" OX A SIMPLE CONSTRUCTION OF STEAM-ENGINE

GOVERNOR HAVING A CLOSE APPROXIMATION

TO PERFECT ACTION," by Mr. JEREMIAH HEAD,

MR. C. W. SIEMENS * remarked there could be no doubt that
by the plan of crossing the suspending arms of the governor, so as
to cause the balls to expand in a parabolic curve within certain
limits, as described in the paper, a real chronometric action was ob-
tained ; but such an action would not be practically applicable to
regulate the speed of a steam-engine, because, as had been ex-

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1871, p. 226.

140 THE SCIENTIFIC PAPERS OF

plained, there would be no tendency for the balls to stop in their
movement if the governor was perfectly chronometric, but they
would fly at once from one extremity of their range to the other under
the slightest alternations of speed. The common governor was the
one that was least affected by change- of speed, and required the
greatest amount of change to bring it into action ; but it had the
advantage of not allowing the engine to " hunt " or beat about
between the extremes of speed before settling down to the proper
rate ; it was thus a safe governor in this respect, though a bad one
in others. The original invention of Watt, which had also this ad-
vantage, and, moreover, allowed of only a slight variation in speed,
occupied a sort of mean position between the common governor
and a parabolic one, and was therefore the best practical governor
hitherto applied ; and the same might now be said of the modified
construction of the crossed-arm governor, in which the spiral
spring was added as described in the paper. The addition of the
spiral spring, however, did away with the true chronometric
character of the governor, because the spring acted in conjunction
with gravity to depress the. balls, without being influenced by the
cross-suspension ; and therefore a great increase of speed was still
required to raise the balls, and the governor was thus rendered a
practical one, instead of being a mere theoretical abstraction, as
would be the case without the spring. A farther advantage of the
spring, when aiding gravity to depress the balls, was that the
centrifugal force on the balls was thereby increased ; if, for
instance, a speed of 40 revolutions per minute was sufficient to
maintain the balls at their mean angle when no spring was used,
and if with the spring 60 revolutions per minute were requisite to
keep them in the same position, the centrifugal force residing in
the balls in these two cases would be in the proportion of the
square of 40 to the square of 60, or as 4 to 9, and the regulating
power for a given variation of speed would be increased in the
same proportion. A smaller relative variation in speed therefore,
in the governor provided with the spring, would be sufficient
for enabling the balls to overcome the constant resistance of
the valve-levers and valve ; and this he believed to be one
of the reasons why the crossed-arm governor described in the
paper acted so well. The only kind of governor that was capable
of acting instantly upon the throttle-valve at the moment of the

£/A' WILLIAM SIEMENS, F.R.S. 141

change of speed was the differential governor ; but short
of the differential governor the arrangement of the crossed-arm
governor with spiral spring appeared to him a very useful and
valuable one.

ON A STEAM JET FOR EXHAUSTING AIR, &c., AND
SOME RESULTS OF ITS APPLICATION,*

BY C. WILLIAM SIEMENS, Esq., D.C.L., F.R.S., Pres. Inst. M.E.

THE steam jet, although it has long been used with such
remarkable success for producing a current of air to form an
artificial draught in locomotive and portable engine boilers, has
hitherto been extremely limited in its application to other pur-
poses ; for though its simplicity for the propulsion of air is a
great recommendation, the effect realised from a given expenditure
of steam has been extremely unsatisfactory, in comparison with
the results of the same steam in working an engine and air pump ;
nor has sufficient pressure of air been obtained from the steam jet
to render it applicable for pneumatic propulsion or the production
of furnace blast. The form and application of the steam jet
having remained hitherto essentially the same as in the original
steam blast of the locomotive in 1829, it occurred to the writer
that much might be done to improve its effect by a judicious
arrangement of the parts, so as to avoid eddies in the combined
current of steam and air, and also to utilise more completely the
initial momentum of the steam. In carrying out this idea, the
first results obtained about a year ago were sufficiently encourag-
ing ; and by gradually recognising more fully the points of essen-
tial importance, the writer succeeded in constructing a steam jet
exhausting apparatus capable of producing a vacuum of as much
as 24 inches of mercury, with steam of only three atmospheres
effective pressure ; and a useful effect has been obtained from the

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1872, pp. 97-116.

142 THE SCIENTIFIC PAPERS OF

steam, equal to that of the ordinary fan blast, and even of the
steam engine and pump.

This improved steam jet is shown half full size in the longi-
tudinal and transverse sections, Figs. 2 to 5, Plates 23 and 24.
A very thin annular jet of steam is employed, in the form of a
hollow cylindrical column, discharged from the annular orifice
between the two conical nozzles A and B, the steam being supplied
from the pipe 0 into the space between the two nozzles. The
inner nozzle A can be adjusted up or down by the hand screw D,
so as to diminish or increase the area of the annular orifice
between the two nozzles, for regulating the quantity of steam
issuing. The air to be propelled by the steam jet is admitted
from the pipe E through an exterior annular orifice surrounding
the steam jet, and also through the centre of the hollow jet. The
tube G, into which the steam jet issues, is made of conical shape
at the bottom, so as to form with the outer nozzle B a rapidly
converging annular passage for the entrance of the air ; and the
width of this air passage is regulated by adjusting the nozzle B
by means of the nut H at bottom. The tube Gr continues to
converge very gradually for some distance above the jet orifice,
the length of the convergent portion increasing with the width of
the outer annular air orifice, and also with the steam pressure
employed ; the most advantageous length varies from 12 to 20
times the width of the annular air orifice, the object being to
ensure the complete commingling of the steam and air within the
length of the mixing chamber G-, beyond which the tube gradually
increases in diameter in a parabolic curve to the upper end, as
shown in Fig. 1. A tapering spindle I is sometimes fixed in the
centre of the inner nozzle A, and carried up through the mixing
chamber G, for the purpose of preventing reflux through the
centre of the combined current.

The rationale of this arrangement is as follows. First, by
gradually contracting the area of the air passages on approaching
the jet, the velocity of motion of the entering air is so much
accelerated before it is brought into contact with the steam, that
the difference in the velocity of the two currents at the point
where they come together is much reduced, and in consequence
the eddies which previously impaired the efficiency of the steam
jet are to a great extent obviated, and a higher useful result is

s//t \\-ii.i.iA.\i XII-:MKNS% F.K.S. 143

il. S-condly, by the annular form of the steam jet the
extent of surface contact between the air and the steam is greatly
increased, and the quantity of air delivered is by this means very
murh augmented in proportion to the quantity of steam employed ;
iilsn the great extent of surface contact tends to diminish eddies.
Thirdly, by discharging the combined current of steam and air
through the expanding parabolic delivery funnel of considerable
length, in which its velocity is gradually reduced and its momen-
tum accordingly utilised by being converted into pressure, the
degree of exhaustion or compression produced by the steam
jet is very materially increased under otherwise similar circum-
stances.

The results of a long series of experiments with this form of
steam jet, both for exhausting and compressing air, have led to the
following conclusions : —

First, that the quantity of air delivered per minute by a
steam jet depends upon the extent of surface contact between
the air and the steam, irrespective of the steam pressure, up to
the limit of exhaustion or compression that the jet is capable of
producing.

Second, that the maximum degree of vacuum or of pressure
attainable increases in direct proportion to the steam pressure
employed, other circumstances being similar.

Third, that the quantity of air delivered per minute, within the
limits of effective action of the apparatus, is in inverse relation
to the weight of air acted upon ; and that a better result is there-
fore realised in exhausting air than in compressing it.

Fourth, that the limits of air pressure attainable with a given
pressure of steam are the same in compressing and in exhausting,
within the limit of a perfect vacuum in the latter case.

The principle of action of the steam jet had received but little
attention until the time of the interesting question raised in 1858
by the invention of the Gifl'ard injector, by means of which water
can be forced into a high-pressure boiler by a jet of steam of the
same pressure, or even of greatly inferior pressure. The physical
explanation of this remarkable fact, which was first attempted the
writer believes in a discussion of the subject in this Institution,*

* See Proceedings of the Institution of Mechanical Engineers, January, 1860,
p. 39.

144 THE SCIENTIFIC PAPERS OF

is based on the principle of conservation of momentum in the
combined jet of steam and water ; but although the source of
power in both cases is a steam jet, the mode of action in the water
injector differs essentially from that of the steam jet applied to
propulsion of air, as in the former case the steam is' condensed by
contact with the water, and ceases to be an elastic fluid at the
moment of issue, while in the latter the steam forms with the air
a combined elastic stream.

A very elaborate investigation of the ordinary steam jet applied
to propulsion of air was given by Professor Zeuner of Zurich
in 1863, showing the effects produced by varying the relative
areas of inflow of the air and steam, and varying the steam
pressure employed. These theoretical enquiries, which were
supported by elaborate experiments, have since been consider-
ably advanced by Professor Rankine, who has shown that in
the combined stream a considerable portion of the total mo-
mentum is lost, and that the relative proportion of this loss
increases with the difference of velocity between the component
streams.

The form of steam jet employed in Zeuner's experiments con-
sisted of a contracted steam orifice, directed upwards in the line of
the axis of a vertical delivery tube, but terminating a little below
the base of the tube ; the length of the tube relatively to its
diameter was found to be only of minor importance. Exhaustion
of air was effected at the lower end of this delivery tube, or com-
pression of air at the upper end. The greatest extent of exhaus-
tion or compression that was maintained with steam of two
atmospheres effective pressure amounted to 7 inches of mercury ;
and 100 volumes of steam measured at atmospheric pressure
were expended in compressing only 7 volumes of air to the
above pressure of about \ atmosphere. "With a reduced orifice
for the steam jet, a compression of 3^ inches of mercury was
maintained, and 37 volumes of air were compressed by the
expenditure of 100 volumes of steam measured at atmospheric
pressure.

In a corresponding experiment made with the improved steam
jet shown in Plate 23, maintaining a vacuum of 3£ inches of
mercury, 137 volumes of air were removed by the expenditure of
100 volumes of steam, both measured at atmospheric pressure ;

.S7A1 \\~II.1.IAM Sir.MENS, F.R.S.

'45

which gives a result of nearly four times the useful effect that was

ii'-il in Zenner's apparatus.

The following Table I. gives the results obtained with the steam
jet arranged as an exhauster for drawing air out of a closed vessel
having a capacity of 225 cubic feet. The four last columns show
the vacuum produced in the vessel, measured in inches of mercury,
when the exhauster had continued in action for the length of time
indicated in the first column. The pressure of steam in the boiler
was 45 Ibs.. per square inch, and the sectional area of the annular
steam orifice of thu jet was varied from 0'05 to 0'20 square inch.

TABLE I.

EXPERIMENTS WITH STEAM-JET EXHAUSTER EXHAUSTING AIR FROM
A CLOSED VESSEL.

Time of
action.

ARI:A OF AJTNIILA.B ORIKICK OK SIKVM .11:1

o-n.-i sc|. in.

0-10 8(i. in.

0-10 s<|. in.

o-J.i si|. in.

Vacuum.

Vacuum. Vacuum.

Vaouiim

Minutes. Indifs of

Inche.s of

Indies of

Inches of

Mcrrurv.

Mercury.

Mercury.

Mercury.

1

!(

Id

iU 84

2

18

lft|

isl

13

3

14£

18

i<>!

15J

4

l.H

154

17J

17

f>

1 .>-,•

155

18

17f

(i

184

m

7

18J

184

Another set of experiments has been tried with the steam jet
used as a blower for compressing air into the same closed vessel
of 225 cubic feet capacity. For this purpose the conical delivery
funnel of the steam jet was connected with the vessel by means
of a 3-inch pipe containing a stop-cock ; and into this pipe was
delivered a small jet of cold water through a pipe of | inch
diameter under a pressure of 40 Ibs. per square inch, with the
object of condensing the steam from the blower, the water being
turned on simultaneously with the starting of the steam jet. The
sectional area of the outer annular air passage in the blower was
0'20 square inch, and of the inner air passage in the centre of the jet
T 1 1; square inch, making a total of 0'36 square inch of air section.
The sectional area of the annular steam jet was 0'07 and (fl 2 square
VOL. i. L

146

THE SCIENTIFIC PAPERS OF

inch in the different experiments, the results of which are given in
the following Table II. : —

TABLE II.

EXPERIMENTS WITH STEA'M-JET BLOWEK COMPRESSING AIR INTO
A CLOSED VESSEL.

AREA OF ANNULAR ORIFICE OF STEAM JET.

Time of

0'07 sq. in.

0-12 sq. in.

0'07 sq. in.

With

With

Without

Condensing

Condensing

Condensing

Water.

Water.

Water.

Pressure.

Pressure.

Pressure.

Minutes.

Inches of

Inches of

Inches of

Mercury.

Mercury.

Mercury.

1

6

5

9

I

10

!)

11

i

12

12

14

H

13

16

15

if

13£

19

15J

2

14

21i

16

3

15

24

16£

4

15i

25

16J

5

15

24i

12

8

14|

24i

11

7

14^

233

.io|

8

14|

23£

10

Final temperature
of air in vessel.

113° Fahr.

158° Fahr.

At the end of the fourth minute the stop-cock was closed, shutting
off the steam-jet blower, and at the same time the condensing water
was shut off ; it will be seen that the pressure in the vessel then
fell slightly, owing to the condensation of the small residue of the
steam from the jet. In the third experiment, in which no con-
densing water was used, the fall of pressure was more considerable,
and the final temperature of the air in the vessel was higher. The
steam pressure in the boiler was 50 Ibs. per square inch.

The following are some of the applications which have already
been made or are in course of being made of this improved form of
steam jet.

PNEUMATIC DESPATCH TUBES. — First, to the working of pneu-
matic despatch tubes. For this purpose the steam jet has a great

SfK WILLIAM SIEMENS, F.R.S. 147

advantage over an ordinary steam-engine and air-pump in first
cost and simplicity, and also in taking up much less space : the
latter advantage being of the greatest value when it is required to
work pneumatic despatch tubes in crowded localities where it is
di Hi cult to obtain room for steam machinery. A pneumatic tube
% inches in diameter, to be worked on the circuit system arranged
by the writer, has been laid down by the postal authorities in
London, from the Central Telegraph Station in Telegraph Street
to Charing Cross and back, with intermediate stations at the
General Post Office and near Temple Bar. The length of tube
forming the whole of this circuit is 6,890 yards, or nearly four
miles. This line was designed to be, and is, as a rule, worked by
means of an air-pump A, driven by a steam-engine, as indicated
in the diagram, Fig. 6, Plate 25. The pump draws air out of a
vacuum vessel V, and forces it into a pressure vessel P, both
vessels being in connection with the pneumatic tube T, as indi-
cated by the arrows. These vessels are introduced in order to
prevent the pulsations of the engine from being felt in the work-
ing of the tube, and are of course unnecessary when the tube is
worked by means of the steam jet. The piston carriers, which
are propelled through the tube by the vacuum produced, are of
the cylindrical form shown in Figs. 7 to 9, Plate 26 ; they consist
of a gutta-percha case covered with drugget or felt, and are made
an easy fit in the tube, so as to slide freely through it.

This line has been experimentally worked by means of three
steam-jet exhausters, similar to the one shown in Fig. 2, arranged
in the manner shown at E E E, Fig. 6, so that all three draw air
out of the same tube F, the steam being supplied to them by the
steam pipe G. When working with these three exhausters, the
mean speed of a piston carrier travelling through the tube from
Charing Cross to Telegraph Street was 14| miles per hour. The
vacuum maintained in the tube was equivalent to 10 inches of
mercury, the steam pressure being 40 Ibs. per square inch ; and
the quantity of coal consumed under the boiler was 5G Ibs. per
hour. In the case of long lines of pneumatic tubes, it will be
better to work them with an exhauster at each end.

For placing the piston carriers into the tube or taking them out
of it at the different stations S S, Fig. 6, without interrupting the
current of air, the intercepting apparatus shown in Figs. 10 to 12,

L 2

148 THE SCIENTIFIC PAPERS OF

Plates 20 and 27, is employed. This consists of two short tubes B
and C, fixed side by side in a rocking frame, each of which can be
brought into line with the circuit tube T at pleasure. Each end of
the rocking frame is faced, and works against the faced side of a
boss on the end of the circuit tube. Three annular grooves are
turned in the faced side of the boss round the end of the circuit
tube, for the purpose of preventing the leakage of air between the
ends of the rocking frame and the bosses. One of the tubes B in
the rocking frame is used as the sending or " through " tube, and
is simply a hollow cylinder of the same internal diameter as the
circuit tube T ; when this is in line with the circuit tube, a earner
can pass through the instrument without being stopped, and this
tube is also used for putting carriers into the circuit. The other
or receiving tube C has a perforated diaphragm at its down-stream
end, so as to arrest the carriers when it is placed in line with the
circuit tube, as in Figs. 11 and 12. This receiving tube is D-
shaped in section, with a flat cover, which can be taken off if
required ; as for instance, to remove carriers, in the event of two
arriving at once and so preventing the rocking frame from being
moved. The flat cover is furnished with a pane of glass, to
enable the attendant to see when a carrier has arrived. In order
to prevent the continuous flow of air in the whole circuit of tube
from being impeded by the receiving tube being left in the circuit
after it has caught a carrier, a by-pass F for the air is provided,
which communicates with the circuit tube T at both ends of the
instrument. A sliding rod H, Fig. 10, held on suitable supports,
is supplied for pushing the carriers out of the receiving tube,
when intercepted and brought out of the circuit. The manipula-
tion for sending and receiving the carriers is exceedingly simple ;
and a treadle is provided for moving the rocking frame with the
foot.

RAISING OF WATER. — A second application of the improved
steam jet is to the raising of water. For lifts not exceeding 20 feet
a steam-jet exhauster could be used with advantage in situations
where the erection of an engine and pumps would be attended
with considerable cost and inconvenience, or where the work to be
done was of short duration or of an occasional character, such as
in draining lands, &c. When employed for this purpose the
exhauster would be applied in the manner shown in Fig. 13,

.sy/i1 WILLIAM SIEMENS, F.R.S. 149

Plate L>8, where A and B are two closed air-tight chambers fixed
at a height of from 16 to 20 feet above the level of the water to be
raised; inlet valves C are provided in the bottom of the chambers
for the water to enter from the suction pipe I), and outlet valves (T
for the discharge of the water raised. The exhauster E, supplied
with steam by the pipe H, exhausts the air from the chambers
A and H through the pipe F, which is provided with a reversing
vuKv L for placing the exhauster in communication with each
of the two chambers alternately ; and the delivery end of the
rxhaiisU-r, which may be closed by a valve, also communicates
with the chambers through the pipe K and the same reversing
valve L. The chamber A is provided with a float M, the rod of
which works by tappets the tumbling lever N, and this throws
over the weighted lever of the reversing valve L by means of the
looped rod R.

The action of the apparatus is as follows. While the exhauster
is drawing the air out of the chamber B through the pipe F,
as shown in Fig. 13, thus causing the water to rise into the
chamber through the suction pipe D under the pressure of the
atmosphere, the discharged jet of combined steam and air from
the exhauster passes through the pipe K into the top of the other
chamber A, which is full of water. The water in this chamber
will consequently flow out through the bottom outlet valve G,
its discharge being aided by the forcing action of the entering
current of steam and air from the exhauster ; and this forcing
action may be regulated by closing the top of the delivery funnel
of the exhauster, in which case the discharge pipe for the water
from the outlet valve Gr may be raised above the level of the
chamber A. As the water descends in the chamber A, the float M
sinking with it moves over the tumbling lever N into the vertical
position, from which when the chamber is emptied the lever will
fall into the reversed position, thereby throwing over the reversing
valve L. The action of the exhauster upon the two chambers is
now reversed, the air being exhausted from the emptied chamber A,
and the combined jet discharged into the chamber B, which during
this time has become filled with water ; in this way the two
chambers become alternately filled and emptied, and a continuous
delivery of water is thus obtained. As the steam contained in the
combined jet entering either chamber from the exhauster becomes

150 THE SCIENTIFIC PAPERS OF

gradually condensed by contact with the water, a partial vacuum
will be already formed in the chamber as soon as emptied, whereby
the work to be done by the exhauster in then exhausting the
chamber will be diminished. In order to prevent the noise which
would be caused by the combined jet of steam and air issuing
from the open top of the delivery funnel of the exhauster, a
" sound killer " S may be placed on the top of the funnel, consisting
of a cylindrical metal vessel containing a series of perforated
wooden diaphragms ; this contrivance has been found by experi-
ment to be very efficient in preventing noise.

The following are the results of a preliminary experiment made
with a rough apparatus for raising water, arranged as shown in the
diagram, Fig. 14, Plate 28. The exhauster E was connected by a
2-inch pipe F to a closed vessel A capable of holding 10*3 cubic
feet, into which the water was raised from varying depths through
the suction pipe D of 2 inches diameter. The sectional area of
the outer annular air passage in the exhauster used in this
experiment was 0'35 square inch, and of the inner air passage
in the centre of the jet 0'16 square inch, giving a total of 0'51
square inch of air section. "With the sectional area of the
annular steam orifice adjusted to (VOS) square inch, and with a
steam pressure of 60 Ibs. per square inch in the boiler, the exhauster
raised 10'3 cubic feet of water 12 feet high in 40 seconds, and the
same quantity 17^ feet high in 75 seconds. With the same area
of air section, but with the area of the steam orifice adjusted to
0'08 square inch, and with 50 Ibs. steam in the boiler, 10'3 cubic
feet of water were -raised 15 feet high in 40 seconds. The height
of lift attainable depends upon the pressure of steam employed ;
and the quantity of water raised depends within certain limits
upon the magnitude of the jet.

EVAPORATION OF SUGAR. — A third application of the improved
steam jet is to the evaporation of sugar. In consequence of the
remarkable results obtained with this steam-jet exhauster, it is
proposed by Mr. R. A. Robertson of London to apply it to sugar-
boiling in the West Indies ; and he has communicated to the writer
the following notes on the subject.

The steam-jet exhauster has been employed experimentally
with considerable success in exhausting vessels for evaporating
liquids in vacuo, and its application for this purpose promises to

SIR WILLIAM SIEMENS, F.R.S. 151

become of great value in the colonies for evaporating cane juice,
principally on account of the simplicity of the arrangement. The
Lrivut, loss and deterioration consequent on the high temperature to
which cane juice must be exposed for evaporating it on the old
system in open pans are well known, and many ingenious plans
li;t\' IHVII invented and to some extent worked for evaporating at
lo\v temperatures ; but, on account of the improved plans being
i-itlitT very much more costly or requiring much more skilled
attention to work them, the greater part of the sugar produced is
still made on the old and wasteful plan. Of all the contrivances
fur evaporating at low temperatures the ordinary vacuum pan
exhausted by pumps is at present the best, when carefully designed ;
and in it under favourable circumstances very rapid evaporation
can be produced at low temperatures. In the sugar-growing
colonies however almost every circumstance is unfavourable for
this mode of working ; the water required for condensing the
vapour from the evaporating pan is warm, and consequently must
be used in large quantities, thereby necessitating large pumps for
its removal, and for exhausting the pan ; the pumps and motive
power thus form together with the pan a very costly apparatus,
which requires a considerable amount of skilled attention, often
not obtainable, and frequent repairs.

On the contrary, a vacuum pan exhausted by the steam jet
exhauster in the manner shown in Figs. 15 and 16, Plate 29,
becomes a very simple apparatus, only requiring a supply of steam
at a moderate pressure for the jet A, which exhausts the vapour
given off by the boiling solution of sugar or other liquid in the
pan B ; and the steam and vapour together are then passed through
the heating tubes D of the pan, thereby producing evaporation.
The area of the steam jet is regulated by the hand-wheel C, the
steam being supplied by the pipe E ; and the course of the current
of steam and vapour is indicated by the arrows. By this arrange-
ment the costly vacuum pumps and the steam-engine or other
motive power are dispensed with, as well as the condenser and its
supply of condensing water, the latter being in many places a
consideration of vital importance ; and in their stead is sub-
stituted the steam-jet exhauster, a comparatively cheap and simple
apparatus, requiring little or no attention, and not liable to get
out of order.

152 THE SCIENTIFIC PAPERS OF

Experiments on this mode of evaporating have been so successful
that a vacuum pan is now being constructed as above described,
capable of evaporating 50 cubic feet of water per hour, in which it
is estimated that a vacuum of from 18 to 20 inches of mercury
will be maintained. The form of pan shown in Fig. 16 is the
simplest arrangement in which the steam jet exhauster can be
applied for the purpose ; but the exhauster can with equal advan-
tage be applied in those cases where vacuum pans are worked
on the systems known as double and triple effect. It has been
successfully employed to exhaust a vacuum pan having a condenser
placed high enough above the ground to discharge by gravitation
against the vacuum the condensing water in which the vapour
from the pan was condensed ; there was thus left only the air and
a very small quantity of vapour to be removed by the exhauster,
which in this case was a very small one, not using sufficient steam
to produce evaporation.

The steam-jet exhauster is further expected, on account of its
cheapness and simplicity, to prove very useful in the colonies for
draining the molasses from the sugar, by exhausting the air from
below the perforated bottom of a strainer containing the undrained
sugar, the pressure of the atmosphere then driving the syrup or
molasses through the sugar and perforated bottom. By this means
the crude and imperfect mode of draining by gravitation, and also
the more elaborate but costly and troublesome centrifugal strainers,
can be superseded with advantage.

BLOWER FOE GAS-PRODUCERS. — A further application of the
improved steam jet is to gas-producers for heating purposes. The
author has had occasion to make numerous applications of the
steam jet arranged as a blower for accelerating the distillation of
fuel in his gas-producers, as shown in Figs. 17 and 18, Plate 30.

The blower B is built into the side wall of the producer, and the
combined current of air and steam delivered by it issues through
an opening A into the space C underneath the fire-grate, which is
closed by doors 1). The small proportion of steam that enters
together with the air is just sufficient to assist beneficially in the
production of the combustible gas, inasmuch as in passing through
the incandescent fuel the steam becomes converted into hydrogen
and carbonic oxide. The steam is admitted to the blower through
the branch pipe E from the main steam pipe F, which supplies a

WILLIAM SIEMENS, F.R.S. 153

number of blowers in a series of gas-producers. A valve <i is
provided in the brunch pipe E, for shutting off the steam from the
lilu\\i-r \\lr ii not in use ; and a small tap J serves for discharging
any water that may collect in the pipe by condensation of the
steam.

The advantages found to result from applying these blowers to
gas-producers are, that coal dust of the most inferior description
ran l>e used, and that the gas production of each producer in con-
suming small fuel is raised from l£ tons to 8 tons of gas per 24
hours ; while at the same time the quality of the gas is improved,
owing to the generation of hydrogen from the steam which enters
intermingled with the air.

These applications of the improved steam jet suffice to illustrate
its scope and value. Other useful applications will readily suggest
themselves, where work may be accomplished by exhausting or
compressing air or gases.

In conclusion the author would remark that although the steam
jet is not a new mechanical agent, and has been applied before for
various purposes and in a variety of forms, yet the mechanical
conditions under which it is capable of developing the greatest
amount of useful effect have not previously been laid down or
practically realised in such a way as to produce results at all
comparable with those obtained from a steam-engine working an
air-pump ; but this it is maintained a steam jet may do, if care be
taken that the available elastic force is changed into onward
motion of the combined jet, and again into onward pressure, with-
out undue loss of effect by eddies or by development of sound or
heat, which have constituted the principal results in the case of an
ordinary steam jet.

The PRESIDENT said he had been led to investigate the subject of
the steam jet in consequence of having been engaged in con-
structing a circuit of pneumatic despatch tube for the Central
telegraph office in London ; the tube itself had cost £3000, and
fully as much had been spent on the engine and air pump. As
there was moreover great difficulty in finding room for this
machinery in so crowded a locality, it had occurred to him that a
more direct means of propelling the piston carriers through the
tube would be a matter of some importance ; and in considering

154 THE SCIENTIFIC PAPERS OF

the conditions of a steam jet he had been led to the conclusion
that, if the elastic force of the steam could be all employed for
giving velocity to its particles, and if each particle of steam so
accelerated could be brought into direct and immediate contact
with the air to be impelled, an average speed of the combined
current could be obtained, which should represent nearly the
whole of the elastic force originally existing in the steam. If this
combined current were then sent through a long expanding mouth-
piece or tube, enlarging gradually in a parabolic curve, so that the
velocity of the current should be gradually converted into pressure
by bringing the particles nearly to rest, it had appeared to him
that very favourable results might be obtained ; and the experi-
ments he had made had proved that these anticipations had been
correct.

There were some remarkable results connected with this steam
jet. The quantity of work done by it, measured by the weight of
air delivered per minute, was proportionate absolutely to the
amount of surface contact between the steam jet and the air to be
impelled, irrespective of the pressure of steam employed. Thus in
discharging the combined current into the atmosphere, the same
weight of air per minute would be thrown out with a steam
pressure of 5 Ibs. as with GO Ibs., the only difference being that
with the lower pressure of steam the limit of exhaustion or com-
pression would be sooner reached than with the higher. The
experiments also showed that the degree of exhaustion or com-
pression capable of being produced by the steam jet, that is the
difference of pressure between the air at the jet and that in the
vessel from which the jet was exhausting or into which it was
compressing the air, was exactly proportionate to the steam
pressure ; so that with 100 Ibs. steam above the atmosphere the
degree of exhaustion or compression attainable was double what it
was with 50 Ibs. steam. A third result was that, in exhausting
the air from any vessel by the steam jet, it took as long a period
of time to exhaust the first 1 Ib. of pressure, from the atmospheric
pressure of say 15 Ibs. down to 14 Ibs., as it did to reduce the
pressure 1 Ib. in any lower part of the scale, say from 8 Ibs. to
7 Ibs. ; this was explained by the consideration that the weight of
air, which had to be put into motion with such a velocity as would
force it through the contracted opening at the base of the expanding

.s/A' WILLIAM SIEMENS, F.R.S. 155

delivery tube, was the measure of the working capacity of the
instrument. These were very distinctly marked characteristics of
a good steam jet ; and if they were attended to in the construc-
tion of the jet apparatus for any particular application, it would
be found that very economical results could be realised for the
steam expended.

He exhibited a full-size specimen of one of the steam-jet ex-
hausters employed to exhaust the pneumatic tube connecting the
iclfirraph offices in London, so as to maintain a current of air
flowing continuously through the tube in the same direction.
Tin- tube formed a complete circuit, and the piston carriers,
whether put in at the commencing station of the circuit or at
any intermediate station, all travelled in the same direction ;
the result was that a tube of only 3 inches diameter was sufficient
for a great amount of work, because there might be ten or more
carriers in the tube at a time, all flowing towards the end station,
but capable of being intercepted and taken out at any intermediate
station.

In answer to various Speakei-s —

The PRESIDENT replied that the expanding form of the delivery
tube from the steam jet was of considerable importance, though
not so much so in the case of air as where a denser fluid had to be
dealt with, as in the water jet of the Giffard injector. From
experiments that he had made with a parallel delivery tube, in
comparison with the expanding form, he had found that, in
exhausting a vessel to the extent of producing a vacuum of
'20 inches of mercury, the expanding tube rendered the jet about
10 per cent, more effective than it was with the parallel tube ; but
in exhausting to only a small extent, the greater density of the
fluid would cause the expanding tube to be as much as 20 per
cent, more effective than a parallel one. The form which he had
adopted for the expanding delivery tube of the steam jet was a
parabolic curve ; but the difference in effect between this and an
ordinary straight cone, with the same areas of passage at the two
extremities, would be immaterial. The mixing chamber, which
had to be of a definite length in proportion to the breadth of the
annular air passage, was made to diminish gradually in area,

156 THE SCIENTIFIC PAPERS OF

somewhat in the form of a " vena contracta," the rate and extent
of contraction being determined in each individual case according
to the velocity of the combined issuing current. But the point of
chief importance in the instrument was the mode of bringing the
entering air into contact with the steam jet. As the working
power of the jet was proportionate to the extent of surface contact
between the steam and the air, the annular form of jet suggested
itself as the most suitable ; and the thickness of the annular film
of steam should be very small, about -gV^1 inch having been
found by experiment to be the maximum thickness of film that
was consistent with economical results. If the thickness of the
jet were increased beyond this amount, the particles of steam
inside the thickness of the jet would flow on at a greater rate
than the outer particles forming the skin of the jet. the latter
being retarded by contact with the air ; and eddies within the
atmosphere of the steam itself would be the result. Again, if the
steam jet were simply made to issue into a plain open tube,
through which the air had to be propelled, the result would be
very inconsiderable, because the eddies formed between the air and
the steam would then attain their maximum amount ; the steam
issuing with great velocity and the air being nearly stationary, the
latter would not be so much impelled in a steady current as set in
rotation and rolled forwards by the power represented by the
difference of velocity between the two fluids. It was consequently
requisite that the air should be accelerated as nearly to the full
velocity as practicable before coming into actual contact with the
steam ; and this was the most important point connected with the
apparatus. The area of the air passages was therefore gradually
contracted for some distance as they approached the issuing
orifices ; and the combined areas of the central air aperture inside
the steam jet and of the annular aperture outside the jet were
made to be together rather less than the area of the issuing orifice
at the outer extremity of the mixing chamber, through which the
combined current issued ; this arrangement ensured the air
attaining the full velocity of the combined jet, whereby the
amount of eddies and the consequent loss of power were reduced
to a minimum.

The PRESIDENT considered that the suggested application of the

A/A' \\-II.IJ.\M .S7/.M//..\'.V, I-.R.S. 157

D jet for removing dust from the casing of millstones daring
the grinding might be attended with in><><l ivsnlis ; hut it must

-••ni'd that the jet acted at a disadvantage when producing
only a small degree of exhaustion or compression, though this
tlisadvanta^'- might to a certain extent be obviated by reducing
the area of the jet orifice when the work was below the capabilities
of the full jet. For as the surface of contact between the steam
and air determined the work done, and the quantity of steam of a
iriven pressure in the jet in proportion to the air determined the
of vacuum or compression produced, it followed that, in
to work economically, the area of the steam orifice ought
to increase gradually as the vacuum or compression increased.
Under any circumstances however the jet would be to some extent
less advantageous when producing only a small degree of vacuum,
because the difference of velocity between the steam and the air,
which was productive of eddies, was then greater than when an
equal weight of air was discharged from a higher degree of vacuum,
a larger proportion of the velocity of the steam being utilised in the
latter case for overcoming the greater excess of pressure of the
external atmosphere. The steam jet had not yet been applied for
producing the air blast for millstones, but no doubt it might be
employed advantageously for that purpose.

In the discussion of the Paper

" ON THE EJECTOR CONDENSER FOR STEAM-ENGINES

DISPENSING WITH AN AIR-PUMP,"

By Mr. ALEXANDER MORTON,

The PRESIDENT (MR. C. WILLIAM SIEMENS) * considered that,
ingenious as the ejector condenser was, its efficiency might be
materially augmented by increasing the extent of surface of the
water jet for condensing the steam. The condensing surface was

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1872, p. 270.

158 THE SCIENTIFIC PAPERS OF

at present limited to the small cylindrical surface exposed to the
steam by the short length of central jet intervening between the
water nozzle and the receive nozzle of the discharge pipe. This
seemed to him an extremely small surface for absorbing such a
large amount of heat as had to be taken up by the steam ; and he
thought that better results in condensation of the steam, together
with some economy of condensing water, would be obtained, if
the surface of the water jet could be largely increased, without
altering the apparatus in other respects. Such an increase of
surface he suggested might be obtained by making the water jet
annular, and introducing the exhaust steam by a central jet inside
the annular water jet, as well as outside it as at present. The
steam would then be condensed both by the internal and external
surface of the annular water jet, instead of only by the external
surface of the present solid jet ; and judging from the experiments
he had made with an annular steam jet for the propulsion of air,
he believed the condensing action of the annular water jet would
be much more efficacious and prompt. He inquired whether any
trial had been made of the annular water jet.

The President observed that in the experiment which had been
described with a small tube inserted up the centre of the dis-
charged jet, it was to be expected that, if the height of the top of
the tube were such that it encountered the water of the jet after
this had had its motion accelerated by the propelling action of the
steam, the great momentum of the water moving at so high
a velocity would produce a high pressure by impact upon the
orifice of the tube ; but if the tube orifice Avere raised a little
higher, it would be clear of the water, and would be exposed only
to the central jet of boiler steam, at the point where that steam
was itself undergoing condensation by the water jet, and where
consequently there would be more or less of a vacuum, as had
been mentioned to be the case. He did not think, however, that
there could be any piercing of the water jet by the steam in the
condenser, but only an interchange of particles between the colder
water in the centre of the jet, and the hotter external portion,
which was essentially a gradual process; and the larger the
diameter of the solid jet, the greater would be the difference of
temperature between the internal and external portions, and the
greater would be the loss of condensing water.

SIR WILLIAM SIEMENS, F.R.S. 159

/// the discussion of the Paper

"ON THE ALLEN GOVERNOR AND THROTTLE-
VALVE FOR STEAM-ENGINES,"

By Mr. FREDERICK "W. KITSON,

The I'KKSIDKXT (MR. C. W. SIEMENS)* said he considered the
»;«• \vrnor described in the paper was remarkable chiefly for its
simplicity, and it appeared to have been found to answer well in
jirartice. Reference having- been made to his own liquid governor
in (•( innection with that now described, it was to be observed that,
though both of them dealt with liquid resistance, they did so in a
different manner.

In the governor now described the power to act upon the throttle-
val\v was obtained in an indirect way ; the rotating paddle-wheel
did not act directly upon the valve, but impelled the oil against the
corrugations in the casing containing it, and the impact tended to
make the casing rotate in the same direction ; the casing however
was held back either by a dead weight, or, as had been suggested,
by a spring, or really by a combination of a dead weight and a
spring, because a weight alone would overrun itself if acting at a
constant leverage. When therefore the velocity of the rotating
paddle-wheel was so proportioned to the weight as just to hold
the latter suspended, a balance was established ; but as soon as
the engine exceeded its normal speed, an additional amount of
impact was created in the oil casing, which accumulated until it
had sufficient power to overcome the resistance of the throttle-
valve and of the stuffing-box on the valve spindle. This power
however to move the valve was not large, in comparison with the
total force acting to support the governor weight at the normal
speed ; if for instance 100 revolutions of the paddle-wheel per
minute sufficed to balance a weight of 10 Ibs., then a variation in
I of two or three revolutions per minute would affect the weight
to the extent of only a small fraction of a pound, which would

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers
1873, pp. 60-62.

160 THE SCIENTIFIC PAPERS OF

accordingly be the limit of the force available for moving the valve.
It was therefore an object of primary importance that the f rictional
resistance in the valve and stuffing-box should be reduced as much
as possible ; and this appeared to have been accomplished success-
fully by the construction of throttle-valve now described. If this
delicacy of action could be maintained, the governor would be
applicable no doubt to engines subjected to frequent and sudden
alterations of load. His own liquid governor, described at a
former meeting some years ago,* consisted of a cup of parabolic
section revolving upon a vertical spindle within a vessel partly
filled with oil or water ; and by the rotation of the cup, which
was open at top and bottom, the liquid was caused to rise up the
sides of the cup, but did not overflow the edge until the speed of
rotation had reached a certain limit. Up to the moment of the
cup overflowing, it acted only as a flywheel, but at the moment it
overflowed it became a pump, drawing in liquid through the
central aperture at the bottom and discharging it over the top
edge ; the external surface of the cup and the interior of the
vessel in which it revolved were provided with a series of radial
vanes, and the overflowing stream of liquid from the cup impinged
successively upon the stationary vanes and upon those on the
revolving cup, thus presenting a practically uniform resistance to
its rotation. The cup was driven by the engine through differ-
ential gearing, with which wyas also connected the weighted lever
of the throttle-valve, this constant weight acting always to main-
tain the uniform rotation of the cup. Although a weight was
thus employed both in his own and in the Allen governor, there
was an essential difference of action between the two, inasmuch as
in the Allen governor the throttle- valve had to be moved by only
a fractional portion of the suspended weight ; whereas in his own
governor the difference between the uniform rotation of the cup
and the varying speed of the engine acted direct upon the valve
through the differential gearing, the uniformly rotating cup
serving as a fulcrum or abutment, while the actual amount of the
weight upon the throttle-valve lever was immaterial, except as
regarded the original determination of the frictioual resistance for
the cup. It would thus be seen that there was indeed more simi-

* See Proceedings of the Institution of Mechanical Engineers, 1866, p. 19 and
p. 112 ante.

WILLIAM SIEMENS, F.R.S. l6l

laritv in appearance between the two governors than really existed
in their modes of action ; and it was clear that the throttle-valve
ilfsi-i-ilicil in connection with the Allen governor must b'e
lookod upon as an essential part of the governor, the prompt
action of the governor depending upon the ease with which the
valve could be moved with a slight amount of force. Owing to
the irroat simplicity of this governor, and the careful manner in
which the mechanical details had been worked out, he had no
do'ibt it would meet with success in its application.

II*1 proposed a vote of thanks for the paper, which was passed,
to Mr. Ivitson, who he regretted was prevented by illness fro:n
bi-ing present at the meeting.

fn the discussion of the Paper

"ON WENHAM'S HEATED-AIR ENGINE,"
By Mr. CONBAD W. COOKE,

The PRESIDENT (MR. C. WILLIAM SIEMENS) * said that many
years ago he had given much attention to the question of obtain-
ing from heat a larger proportion of mechanical effect than had
previously been realised.

With regard to the best medium to be employed for the purpose,
although on theoretical grounds this was immaterial so long as no
heat was thrown away, there were many important considerations
in favour of steam, which had a higher rate of expansion by heat
than air, and did not involve the employment of an air-pump for
producing a supply under pressure. By the application of the re-
generative principle he had obtained in small steam engines satis-
factory results upon the whole ; fifteen or sixteen engines altogether
had been made on that plan, of from 5 to 10 horse power, and had
worked for a series of years with very fair results. One of them

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1873, pp. 82-85.

VOL. i. M

1 62 THE SCIENTIFIC PAPERS OF

had been put to work in Paris, and had been examined by

G-en. Morin and M. Tresca, and in a whole day's working the

consumption of coal had been found to be 1/67 kilos, or 3'68 Ibs.

per break horse power per hour. But on attempting to carry out

the same plan in engines of larger size, of 20 horse power or more,

difficulties had arisen in every direction, which had compelled

him to relinquish the endeavour. One difficulty had been that

mentioned in the paper with regard to the joints ; and he had

found that the best joint to resist a high heat was made by turning

a number of concentric grooves in the faces of the flanges, and

filling them with a cement composed of fine dust of cast iron

mixed with white and red lead previously mixed up with linseed

oil, and worked into a very stiff cement that set as hard as iron

when exposed to heat ; by that means he had succeeded in making

tight joints. But what had discouraged him at that time was the

result he had arrived at in preparing a paper * to which reference

had been made, " On the conversion of heat into mechanical

effect." In considering what would give a proper conception of a

really perfect engine, he came to the conclusion that if steam were

generated at such a pressure as to occupy only the same bulk as

the water itself, and were then expanded down until it was all

condensed through expansion, the utmost effect theoretically

possible would thereby be obtained, because the whole of the heat

would have been converted into mechanical effect. In Fig. 2, Plato

32, the diagram of a portion of the dynamical expansion curve of

steam, the shaded rectangle showed the utmost power that could

be obtained in a condensing engine with perfect vacuum using

* (See Proceedings of the Institution of Civil Engineers, 1852-53, p. 571).
In the Steam expansion curve shown in Fig. 2, Plate 32, the horizontal scale
gives the volumes of steam compared with the volume of the water from which
it is produced ; this is the correct dynamical expansion curve of saturated st' am
allowing for the loss of the heat that is converted into mechanical effect when
the steam is expanded behind a working piston ; the shaded area, or any other
rectangle drawn to the curve, represents the power obtained from a condensing
engine with perfect vacuum but without expansion. In the Air expansion curve
shown in Fig. 1, Plate 31, the horizontal scale gives the volumes of air compared
with the volume of an equal weight of water ; the dotted curve is the hyperbolic
expansion curve of constant heat, representing Marriotte's law that the pres&ure
and volume are inversely proportionate to each other ; and the full curve represents
the actual rate of expansion, showing the reduction of temperature during expan-
sion, and the consequent contraction of volume ; this curve is in accordance with
the observed fact that, when air at any pressure and at the temperature of 32°
Fahr. is compressed to double its original pressure, its temperature is raised 70°
Fahr.

S/K WILLIAM SIEMENS, F.R.S. 163

steam without expansion, and the result was found to be that
\\ hrivver this rectangle was taken its area amounted to only -j^th
of the area of the curve expressing the power due to the ultimate
usion of the same steam ; but a good high-pressure expansive
engine using steam of four or five atmospheres total pressure,
cutting off at about one tenth of the stroke and working down to
a good vacuum, realised a very considerable proportion, something
like one fourth or one fifth of the theoretical maximum which
could ever be obtained. The theoretical minimum consumption of
1'iu'l in a perfect steam engine he had calculated would be (taking
14,000 as the total units of heat developed by the complete com-

33,000 ft.-lbs. x 60 mins.
bustion of 1 Ib. of carbon) ^m unitg heafc x ._, ffc .^ = 0«2 Ib.

or one fifth of a pound of carbon per horse power per hour ; or one
fourth of a pound of coal, taking into account impurities.

In the case of the air engine, it was apparent that both Stirling
and Ericsson had over-estimated the real value of the regenerator
under a misconception of its true action ; they had imagined that
it was possible to absorb and give back the whole of the heat
originally put into the air, with the exception only of accidental
losses, and had overlooked the fact that a portion of the heat
became entirely used up by being changed into mechanical effect ;
both their engines had accordingly been deficient in heating power,
and had failed to give permanently , satisfactory results. Even
supposing perfect re-absorption of heat from the exhaust air,
theoretical considerations showed that an air engine was necessarily
very imperfect as a means of developing power from heat ; because
although a volume of highly compressed and highly heated air, if
expanded down to atmospheric pressure and discharged at no
higher temperature than that of the external atmosphere, would
yield the full result for the heat absorbed in expansion, yet an
equal weight of air would then have to be taken up again and
compressed to the original pressure, thereby generating a great
amount of heat, which would all be wasted because it was
generated in the air before its expansion by heat took place and
when it should occupy the least volume, the power of the engine
being dependent upon the increase of volume. In the best air
engines, therefore, even supposing perfect absorption of the escaping
heat, it would not be possible to realise anything like so much as

1 64

THE SCIENTIFIC PAPERS OF

one fourth or one fifth of the theoretical maximum of mechanical
effect due to the heat put into the air. Another drawback was
that in most air engines, and particularly in Stirling's, owing to
the low conducting power of air and insufficient amount of heating
suri'ace, the cylinder or vessel in which the air was heated by the
fire was found to get fully red-hot, so that the products of com-
bustion reached the chimney at that elevated temperature. This
source of loss was obviated in the engine described in the present
paper, by causing the products of combustion to pass through the
working cylinder, the air being heated by direct contact with
them ; and if the expansion in the cylinder could be carried far
enough, no doubt the whole of the heat in the products of
combustion might in this case be utilised ; but it was clearly
impossible to carry expansive action very far in this engine, owing
to its low working pressure, and moreover the working of the air-
pump constituted a very heavy loss of useful effect. The loss of
sensible heat escaping at the exhaust might be remedied by the
application of a regenerator ; but this could not be done except at
a sacrifice of the simplicity of construction which appeared to him
to constitute the chief recommendation of the engine. It was
impracticable he believed to carry out the principle of the hot-air
engine on a scale sufficient to give any large amount of power ;
but a question of much practical importance was to produce a
safe engine of small power, which could be put up anywhere, in
any room, because requiring no boiler, and therefore necessitating
no increased rate of insurance against fire. This object had been
already accomplished by various constructions of gas engines, and
was also effected by the hot-air engine now described, which he
hoped would be so far perfected in its details as to give an
effective power of as much as 4 or 5 horse power ; and even
though the consumption of coal were not reduced below 8 Ibs.
per horse -power per hour, there were no doubt many cases in which
such a source of motive power could be advantageously employed.
He moved a vote of thanks, which was passed, to Mr. Cooke for
his paper, and also to Mr. Wenham for the additional information
he had kindly given.

S/A WILLIAM SIEMENS, F.R.S. 165

In the discussion of the Paper

" ON MODERN LOCOMOTIVES, designed luith a view
to ECONOMY, DURABILITY, and FACILITY OP
REPAIR, toiji'tlicr ti-ith some particulars of the DUTY
PERFORMED, ami of the COST OF REPAIRS," by JOHN
ROBINSON, M. Insfc. C. E.,

M u. SIEMENS * said he was one of the members of the jury in
Vienna, where the subject of the contre-vapeur was discussed, and
the exceptional reward of a Diploma of Honour was awarded to
M. Le Chatelier for that remarkable invention, which was fully
appreciated throughout the Continent. It was a contrivance that
should always be used ; otherwise the result would be a failure.
He imagined that on the London and North-Western line it had
not been adopted as a system, but had been put into the hands of
the engine driver to be employed occasionally ; sometimes it
might have been used properly, whereas at other times probably
the water was not put on when reversing. In M. Le Chatelier
not only France but Europe had lost one of her ablest engineers.
He was a man of very extensive information, of a genial nature,
and a true-hearted friend ; and he had never known any one
more ready to investigate and to appreciate any new idea. M. Le
Chatelier had suggested several important improvements in steam
economy. He was the first to draw attention to the great advan-
tages of perfect balancing and also of perfect jacketing in using
steam expansively. He had, with his friend M. Deville, devised a
practical process for the production of aluminium, and had made
many valuable suggestions in regard to metallurgical processes.
The earnest activity of his mind was such that he could give
attention to these matters, and to the high education of his
children, while holding the responsible appointments of engineer-
in-chief to many large undertakings.

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXXVII. Session 1873-74, p. 27.

1 66 THE SCIENTIFIC PAPERS OF

In the discussion of the Paper

" ON PEAT FUEL MACHINERY," by JOSEPH
MCCARTHY MEADOWS,

ME. SIEMENS * said, in the course of the discussion, the use to
which the peat was to be applied appeared to have been left out
of sight. If it was to be worked into peat charcoal, it was of
course necessary that it should be thoroughly masticated and
compressed, in order to make it of great density. Such peat
charcoal was of considerable value, and in blast furnaces he
thought it would be better than the best coke that could be
obtained, because it was generally free from sulphur. This,
however, was not always the case, for at Swansea he had analysed
peat, and had found it to contain fully \ per cent, of sulphur,
owing, no doubt, to the presence of the Copper Smelting Works.
Again, if peat was required for locomotive work it must be ren-
dered dense, or it would fly out in large quantities from the
chimney ; in which case a machine like the Messrs. Clayton's
might be extremely useful. If, however, peat was only to be used
for the production of heat, in stationary steam boilers or in fur-
naces, it would be entirely useless to subject it to manufacturing
operations such as had been described. He agreed with Mr.
Cowper in thinking that it was desirable to separate each piece as
much as possible from its neighbour, so as to let the wind act
upon it, rather than to dry it in sheds by artificial heat, because
in the latter case a large proportion of it had to be burnt to
produce the heat, and the operation was actually retarded because
only a limited quantity of air could be brought in contact with
each piece. Air-dried peat, containing 40 or 50 per cent, of
water, was capable of producing any degree of heat if properly
burnt ; and he considered the proper way of burning such fuel
was to convert it into a gaseous fuel. The gas was as rich if the
peat contained 50 per cent, of moisture as if it contained none ;

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXXVIII. Session 1873-74, pp. 265-66.

WILLIAM SIEMENS, F.R.S. 167

lint care should be taken to condense the water out of it before the
gaseous fuel was used. He had employed peat so treated for forg-
ing i'liniact's and steel melting furnaces, and found it the best fuel
that could be used. He should prefer it even to the best coal for
litMting purposes, and he believed the time would come when the
stores of peat in Ireland, in Scotland, and in the north of England
would form a valuable addition to the stock of fuel for all opera-
tions for which coal was at present used. Such employment would
('•ml more than anything else to the development of iron industry
in Ireland.

/// the discussion of the Paper

"ON THE McCARTER CONDENSER WITHOUT AIR-
PUMP FOR STEAM-ENGINES," by Mr. F. PRESTON,

MR. C. "VV. SIEMENS* enquired about the amount of steam in the
jet necessary to work the auxiliary condenser. A calculation had
been given of the quantity of water which this condenser required
as compared with the ordinary condenser, and he supposed there
would not be any material difference in the quantity, but he had
considerable doubt as to the economy of working this plan. This
condenser worked upon the Newcomen principle, boiler-steam
being admitted in the lower part of the condenser, which stood in
lieu of a cylinder in the Newcomen engine ; and the hot steam
was brought into contact with the cold surface, in order to expel
the water and air accumulated there. Now the quantity of steam
so accumulated must be considerably more than the steam that
would be necessary to work the air-pump, because the air-pump
was worked on as economical a principle as was involved in the
engine itself. Therefore the expulsion of a certain amount of
water and a certain amount of air from the upper part of the con-
denser was accomplished in this condenser not in an economical
manner. There was a compensating advantage in doing away

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1876, pp. 311-12.

1 68 THE SCIENTIFIC PAPERS OF

with one complication of the condensing engine, namely, the air-
pump ; but whether that advantage was not purchased at too
great a price was a question which he should like to see more fully
elucidated than was done in the paper at present.

He should also like to know what vacuum could be regularly
maintained by this condenser ; there seemed to him a difficulty in
that respect. When the vacuum was produced in the auxiliary
condenser no doubt the water in the upper chamber by its own
gravity fell readily into the lower chamber ; but only a small
portion of the air that was accumulated in the upper chamber
would follow it ; and in proportion as the lower chamber was
small or reduced in comparison with the upper chamber, or main
condenser, in the same ratio the vacuum would be reduced in the
upper chamber. It would be interesting to know, therefore, what
was the vacuum that could be usually obtained with this con-
denser, and what was the amount of steam necessary to work it.

In the discussion of the Paper

"ON THE COMBUSTION OF REFUSE VEGETABLE
SUBSTANCES, such as STRAW, EEEDS, COTTON STALKS,
BRUSHWOOD, MEGASS, &c., UNDER STEAM BOILERS,"
By JOHN HEAD, Assoc. Inst. C. E.,

DR. SIEMENS * said his experience of the form of engine under
discussion was confined to what he had seen, as one of the jury,
at the Vienna Exhibition a few years ago. Messrs. Ransome,
Sims, and Head's apparatus was then worked very satisfactorily,
as was also Mr. Garrett's. The jury, he believed, decided that
both makers were entitled to an award, not only for their straw-
burning engines, but for their exhibits generally. With regard
to the relative merits of the two methods, he thought Mr. Head's

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XLVIII. Session 1876-1877, pp. 99-100.

SfX WILLIAM SIEMENS, F.R.S. 169

apparatus deserved the first place, because it contained all the
dements of continuous, equal, and systematic feeding ; whereas
in Mr. Garrett's much depended upon the skill with which the
operation was can-led on. Mr. Garrett had stated that each time
the dried straw was introduced into the furnace an explosion took
place. That surely meant that, at the moment when the straw
entered the furnace, the combustible matter met an excessive
aniuimt of oxygen which had previously accumulated there. No
doubt, however, a skilled workman could produce excellent results
with Mr. Garrett's apparatus. With reference to the figures
brought forward by Mr. Head, which had been contested by
Mr. Garrett, he had just made a calculation which might throw
some light on the subject. Mr. Head had stated that 1 Ib. of
coal was equal in heating power to 3| Ibs. of dried straw ; and he
had given analyses of the coal and of the straw. From this it
appeared that the combustible constituents of Newcastle coal
were, carbon 80, and hydrogen 5 ; and of straw, carbon 35, and
hydrogen 5 ; making a total of combustible matter in the one
case of 85, and in the other 40 : but the hydrogen had a greater
calorific power than the carbon, and for comparison might be
doubled ; so that the figures would be 90 and 45. According to
that analysis, straw ought to possess, roughly speaking, half the
calorific effect of coal, and it would have that effect if the com-
bustion of both materials were perfect. In the case of straw
there was one deduction to be made which did not apply equally
to coal ; it contained a greater amount of ash in the form of
silica ; and that ash was the real difficulty to be contended with.
There was also a greater amount of water in the straw than in the
coal, but that was a matter of less difficulty. In the case of straw
15 per cent, of water had to be evaporated, and in the case of coal
only 1*3 per cent. Taking those elements at their utmost value,
he considered that Mr. Head's estimate, that 3| Ibs. of straw
were equal in calorific effect to 1 Ib. of coal, was not excessive ; it
appeared to him a reasonable amount, provided that something
like perfect combustion were obtained. The apparatus, so ad-
mirably explained and illustrated, showed that the question had
been thoroughly gone into, and the assumption that complete
combustion wap obtained had been verified. When straw and other
similar vegetable matter were consumed, the important point was

1 70 THE SCIENTIFIC PAPERS OF

to remove the ash so as to continue the combustion under
favourable circumstances, and this difficulty has apparently been
successfully overcome. He thought the thanks of the Institution
were due to Mr. Head for the admirable manner in which he had
investigated the subject and brought it before the members.

In the discussion of the Paper

"ON EXPERIMENTS RELATIVE TO STEAM
BOILERS," by Mr. WILLIAM BOYD,

DK. C. "W. SIEMENS* remarked that the paper was exceedingly
gratifying to himself, inasmuch as it contained results arrived at
by an engineer who had evidently gone very systematically to work
in testing material, and had not refrained from bringing the
results he had obtained before that Institution. The first news
he had had of this application of Landore steel was unfortunate.
He was told that the steel had in one instance at least entirely
failed to stand the test. Mr. Boyd had now brought before the
meeting the particular circumstances under which his apparent
failure arose. A test plate had been fastened between two plates
of iron ; and when the tensile strain was applied, the steel, instead
of elongating 20 or 25 per cent, as was expected, and of breaking,
as was also expected, across the rivets, took its own course and
broke through the fastening, along a line of fracture some 20 per
cent, stronger than the line of least resistance. It was inferred
from that experiment that such steel after all must be an unreliable
material ; but when he saw the details of the experiments he
immediately suggested that the cause of failure would probably be
found to lie in the mode in which the fastening had been made.
Mild steel or homogeneous iron was a material that yielded very
much before rupture, provided the tensile strain was applied fairly

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1878, pp. 233-236.

SfK WILLIAM SIEMENS, F.R.S. 1 71

over the whole section. This very property of yielding before
breaking made it necessary that it should be fastened along the
\vlmle line of its section. In the particular fastening referred to,
t wi > larger rivets stood forward, and naturally would take nearly the
whole of the strain that fell upon the plate. The other four rivets
stood back to such an extent that, before they would receive any
considerable portion of the strain, the two forward rivets would be
loaded to such an extent as to cause a partial yielding of the metal,
and, as they were near the edge, tearing action would set in. Now
a material, however strong it might be, could be torn without
much effort. Thus if a strip of india-rubber were nicked, it tore
with little stress ; whereas if there was no nick, it would bear
almost any amount of elongation. In like manner the mild steel
had to be treated differently from iron ; and experiments like the
present gave better information than any learned essays on the
subject could possibly supply. When the strips of the same plate
•were tested in the usual way, they stood the test as well as every
other portion of the material supplied.

The use of iron rivets had been advocated, and was still advo-
cated by many, for riveting these mild steel plates together. He
could not too strongly protest against that practice. It was
utterly against nature to stick together material like this mild
steel with another material, iron, which behaved quite differently
as to elongation and yielding. Again, when iron rivets were used,
the rivet-holes must necessarily be larger than they would need to
be for steel rivets ; being larger they would be put a greater
distance apart, and the distance practically chosen was the same
as was in ordinary use for iron plates of 20 per cent, greater
thickness. Thus the thinner plate was held together at points
relatively further apart ; and the result was that the metal between
hole and hole could not be so securely held as it should be, and
tearing action set in. He was very glad to see that Mr. Boyd
had resisted the temptation of using iron rivets, and had adopted
steel riveting in these boilers.

He did not quite agree with the author that punching neces-
sarily diminished the strength of a steel plate something like 33
per cent. No doubt his experiments had been very carefully made,
but there was a great difference between punching and punching.
His belief was that on punching a cylindrical piece or " burr " out

THE SCIENTIFIC PAPERS OF

of that plate it would be found that this piece was not as thick as
the plate, but on the contrary considerably thinner. What then
had become of this metal ? It had not gone to increase the
diameter of the burr, although its lower part might be a little
bulged out in that way ; but while the punch was being driven
into it that metal had accumulated somewhere in the plate.
There was thus a ring of highly compressed material formed
round the hole, and this compression caused the surrounding
portion of the plate to be under high tension. The result must
be that, if pressure was applied, the portion already under high
tension received additional tension, and tearing action began.
Hence the great loss of strength in punching. He held in his
hands the results of some experiments, made with great care by
Mr. James Riley at the Landore works, which proved precisely
what he had stated. There however plates, which before punch-
ing bore 29'33 tons per sq. in., bore after punching 30*7 tons per
sq. in. without annealing. The reason of this anomaly was, the*
die in these cases was made considerably larger than the punch,
and thereby the compression was practically avoided.* Another
kind of punch had lately been introduced from America, which
was well worth the attention of the meeting. It was a helical
punch, which, instead of driving out the metal before it, cut it out
in a spiral form. This mode of punching he thought very pro-
mising, and he believed by adopting such a method the weakening
effect that had been experienced in punching thick plates would
be entirely got over, without unduly increasing the size of the die.
In punching thin plates the author found there was but very
slight reduction of strength, and that the punched plate when
annealed resumed its former strength exactly. It had also been
found that by riming out a punched hole the strength of the metal
was entirely restored : showing that the cause of weakness was in

* Mr. J. GL Mair has since communicated to the Secretary the results of some
experiments made on this point by Mr. Henry Sharp, of Bolton. Four pieces of
\ in. steel plate, 2^ in. wide, were punched with an \± in. punch, but with a die
g in. in two cases and f in. bare in the other two. The results were as under,
showing 25 per cent, more strength for the greater clearance : —

I in. die. f in. bare die.

1st experiment . . . 32 '30 25 '92

2nd experiment . . . 32 '85 26 '08

32 '5 tons per sq. in. 26 '0 tons per. sq. in.

SIX WILLIAM SIEMENS, F.R.S. 173

immediate vicinity of the hole, and did not extend to any
distance through the metal, which was in accordance with the
explanation he had suggested.

The author had added an important item of information with
iv_rard to the staying of flat surfaces. The addition of the nuts
to the stays showed a remarkable increase of strength, and he
hoped that mode of staying would be generally adopted. It was
a question however whether for flat stay-plates this very mild
steel should be used ; it would probably be more advantageous to
use for flat surfaces steel containing perhaps 0*4 per cent, of
carbon ; they would then get a material of great stiffness as well
as great strength, and still of sufficient ductility. Steel with
about 0'4 per cent, of carbon would not indeed elongate 25 per
cent, before breaking, but it would elongate probably 12 per
cent. ; and in the case of a flat plate that ought to be sufficient.

He might mention that lately he had witnessed two experiments
made with a view of bursting a steel boiler. These were made at
Swindon by Mr. Dean of the Great Western Railway ; and Mr.
Parker, the Chief Surveyor at Lloyd's for the boiler department,
was present at the second experiment. Both experiments failed,
i.e., they could not burst the boiler ; and he believed it was
impossible to burst a steel boiler. It might be expanded and the
joints racked to the extent of introducing an amount of leakage
that would prevent further accumulation of pressure ; but it
would never be burst. That condition of things was certainly a
satisfactory one.

In the discussion of the Paper

"ON A NEW REVERSING AND EXPANSIVE
VALVE-GEAR," by Mr. DAVID JOY,

DR. C. W. SIEMENS * said that, listening to the very excellent
paper which had been brought before them, and to the observa-

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers
1880, pp. 444-445.

174 THE SCIENTIFIC PAPERS OF

tions of the speakers who had taken part in the discussion, one
result seemed to his mind to be quite certain, namely, that the
link-motion was doomed. He must say he did not feel the same
regret on that score which appeared to animate Mr. Head. The
link-motion had no doubt been a way out of a difficulty ; but it
was correct only within very narrow limits, and the moment
those limits were exceeded it did not produce the expansive action
desired. A motion had now been brought before them which
challenged comparison with the best expansive gear that could be
mentioned. There was a clean cut-off ; the steam was put on at
once in ample quantity, and the exhaust also was opened promptly
at the right time. There was also the means of adjusting the
action so as to make the up-stroke and the down-stroke perfectly
alike, which, as they all knew, was not the case with the link-
motion. And, as Mr. Webb had beautifully illustrated by his
locomotive, the new gear had the advantage of giving a large
useful space on the main shaft for increasing the length of the bear-
ings. These were very important advantages, which Mr. Joy
might claim for his motion. But they had also been put in pos-
session of two other motions, which seemed to be as perfect as
Mr. Joy's. Still they need not regret that result. There were
points of difference between the gears, though they all aimed at
the same result, a very perfect cut-off and a perfect mode of re-
versing the engine. The object was achieved by different
mechanical details, all of which he considered were superior, both
theoretically and practically, to the old link-motion.

With the criticism offered by Mr. Boyd, in regard to a portion of
Mr. Joy's motion, he thoroughly agreed. The slotted disc, which
was turned into a different angular position, with the slide-block
grinding up and down in the slot, was not, he thought, a desirable
detail ; and he would recommend Mr. Joy to do away with it. He
observed it had actually been done away with in one arrangement
shown. All the friction was there reduced to the friction on the
pins, which must be preferable to friction of sliding surfaces.
With that exception Mr. Joy's motion appeared to him perfect, and
one that would no doubt receive the most earnest attention of
mechanical engineers.

S7A WILLIAM SIEMENS, F.R.S. 175

In ////• ili*-i;*sion of (ha Paper

"ON GAS-GENERATING FURNACES,"
By Mr. F. LIRMANN,

Du. Si KM FA'S* said, Mr. Liirraann was so well known as a
gentleman who had originated practical ideas which had been
hiiLTt-ly used, that anything coming from him deserved much con-
sideration. The gas-producer now advocated by Mr. Liirmann
\v;i< l»ased, as might be supposed, upon a sound philosophical prin-
ciple— that of concentrating heat as much as possible in the
operation of converting the solid fuel into gas. He had, however,
two criticisms to make on Mr. Liirmann's paper, the one regarding
the historical record with which he prefaced his arguments, and
the other as to the technical value of the particular arrangements
proposed by him. It was now more than thirty years since he
had constructed the first gas-producer applied to the regenerative
gas furnace, and he might add, that the subject of fuel conversion
had engaged his attention ever since. Under these circumstances
he should have expected to find in the historical record of the
author some allusion to his labours in this direction. In the years
1863 and 1864 he patented several varieties of gas-producers, in
which the coal passed towards the fire-grate through retorts, in a
vertical or inclined direction, which retorts were heated either by
the gas resulting from the operation itself or from extraneous
sources. Indeed, the hopper arrangement, described in the paper
as Price's gas-producer, might, amongst others, have been taken
from one of his specifications of that early date. He might
further mention that in the year 1863 he brought a bill before
Parliament to supply Birmingham with heating gas, and that the
gas-producer then put forward comprised both the principle of re-
tort-heating and that of the open grate for the final conversion of
coke into carbonic oxide, whilst by another arrangement the coke
was withdrawn from the apparatus to be used separately as breeze.
This arrangement, he believed, was still in operation at Messrs.

* Excerpt Journal of the Iron and Steel Institute, 1880, p. 583-585.

176 THE SCIENTIFIC PAPERS OF

Russell & Co.'s tube works near Birmingham. He held in his
hand the specifications of the two patents alluded to, and it might
not be without interest to the Institute to learn exactly what was
known and claimed regarding the proposed method of producing
gas as early as 1863 and 18G4. In the patent of 18G3 it is said :
" Thus, although we much prefer to make use of regenerators for
producing the intense heat required, yet heating apparatus of
different construction may be applied, and the form and construc-
tion of the heated chambers, as well as the means adopted for
charging and discharging them, may be greatly modified ; also
the chambers or retorts may be arranged in a somewhat inclined
position instead of being placed vertical," &c. ; and he claimed
" constructing furnaces in which coal or other carbonaceous material
is subjected, whilst descending gradually in upright recipients or
retorts, to the action of intense heat resulting from the application
of regenerators with reversal of currents, the result being either
the total conversion of the carbonaceous matter into combustible
gases when a grate is applied at the bottom of such recipients, or
the production of both combustible gases and coke of superior
quality, when admission of atmospheric air to the bottom of the
recipient is prevented." And in 1864 the first two claims run as
follows : " First, constructing apparatus for converting carbona-
ceous matter into combustible gases, in which such carbonaceous
matter is first caused to descend gradually through one or more
vertical or inclined retorts, from the lower end of which it passes
in a thick layer upon a fire-grate, the heated combustible gases
resulting from the imperfect combustion of the fuel upon the
latter being made to pass around or against the outer surface of
the said retort or retorts. Second, constructing gas-producers
(which are chiefly applicable for the conversion of binding coal
into combustible gas) in such a manner that the fuel descends
through an inverted funnel, acting more or less as a retort, upon
a solid moderately-inclined plane supporting the column of fuel,
such inclined plane having one or more openings at the top for
the admission of atmospheric air, as well as for the introduction
of bars to break up the masses of coke, whilst one or more other
openings are provided at bottom for the removal of clinkers or
ashes, and for the introduction of a further quantity of air or
steam or both, in order to complete the conversion of the car-

WJ 1.1.1 AM SIEMENS, F.R.S. 177

bonaceous matter, substantially as hereinbefore described." He
made these quotations simply by way of addendum to the historical
rrrord brought before them by the author, his personal interest in
the matter having ceased \\ith the expiry of the patent through
lapse of time. The reason for not persevering in the direction
ndirated was due to the retort action being found too slow
for the production of combustible gas in such quantities as were
required for working large furnaces, and he expected that Mr.
Liu-maun would be met with the same kind of difficulty when he
attempted to give his gas-producers an extended application. His
endeavours had since been directed to concentrate as much latent
heat in the gas as possible, in order that it might travel, and
t Ii-Telbre he had always used the spare heat for drawing up vapour,
and to convert the water in this form into its two elements. In
the form of producer now put forward, the enrichment of the gas
was simply in the direction of obtaining a higher temperature, and
this higher temperature could be utilized if the gas-producer was
put into close connection with the furnace. But in the regenera-
tive system that high heat was secured through the regenerators,
and therefore was not required previously. In one form of the
gas-producer, which was also largely used, a steam blast was
employed, with a closed grate. He had since gone a step further
in that direction, and he had obtained certainly a very remarkable
result.

ON THE USE OF COAL GAS AS A FUEL.
BY DR. C. WILLIAM SIEMENS.

DR. C. WILLIAM SIEMENS : * Mr. President, among the subjects
suitable for discussion before this Society there is, perhaps, none
more so than that of combustion, or rather that of the production
of chemical energy, and its application. It has devolved upon
to bring this subject before you, and although I should have

* Excerpt Journal of the Society of Chemical Industry, 1881, pp. 39-46.
VOL. I. N

178 THE SCIENTIFIC PAPERS OF

preferred that the choice had fallen on a gentleman who could
have given more time to the actual preparation of such a paper
than I have been able to bestow upon it, yet having for nearly
thirty years given much attention to the subject of combustion,
and the utilization of gaseous fuel, it was perhaps natural thnt
your President called upon me to open out the subject before this
new Society.

The fuel at our disposal, generally speaking, is coal, and in
most applications coal, in its raw condition, is still used for the
production of heat. I have put a table before you, showing the
amount of heat energy residing in a pound of fuel. If that fuel
is pure carbon, 14,544 units of heat are producible. If coke were
used, and burnt entirely to carbonic acid, 12,500 heat units would
be produced. If coal of the average quality is employed, 11,000
heat units may be obtained. But, if instead of using solid fuel
we should employ gaseous fuel, we immediately jump to much
higher figures. Olefiant gas gives in its combustion 21,344 heat
units ; marsh gas, 23,500 ; and hydrogen, 62,000 units, or very
nearly six times as much as raw coal. Then carbonic oxide, not
counting the oxygen already taken up, but simply the carbon in
the carbonic oxide, gives 10,100. These figures may serve to
show the great advantage in the use of gaseous fuel ; and if by
some means or other we could impart to the fuel as it is consumed
the greatest possible power for the production of heat energy, it
would be clearly to the advantage of the user, because not only
does this better fuel produce for the same weight of material a
greater result, but it produces that result with a smaller expendi-
ture of oxygen, that is, with less draught.

The principal reason why gaseous fuel, especially hydrogen,
produces such a large amount of heat energy, consists in the heat
energy already expended in expanding it to its volume, and the
ultimate result, therefore, depends less upon combustion than it
would do in the case of coal where a large proportion of the heat
produced is expended in converting the solid into gaseous fuel to
begin with. But gaseous fuel, when used instead of solid fuel,
possesses several other advantages. It is much easier to produce
perfect combustion in using gaseous than in using solid fuel.
Solid fuel, when first heated, flies off to a great extent into
products of distillation, and the supply of oxygen can hardly keep

WILLIAM .S/A'.J/A.Y.s; F.R.S. 179

with such a sudden increase of production of gaa. The
result is smoke, and smoke is waste, whichever way it is viewed.
.Moreover, it is a public nuisance, and as such it should be done
away with.

Then, in using gaseous fuel, the supply can be kept constant,
and when the supply of air is once regulated, it can be employed
in the exact proportion to produce the maximum result. This is
an important argument in favour of gaseous fuel. Then in
•us fuel we have already a product of manufacture, and we
have it free of ashes, and those drawbacks which attach to solid
fuel, making disturbance and dirt not only in the atmosphere, but
in works or dwellings where it is employed. Again, by the use
of gaseous fuel, labour is saved to a very great extent. We have
not to cany coal, but the fuel is supplied through tubes to the
place where it is used, and the labour of producing the gas being
concentrated labour, involves very much less expense than it
would if solid fuel had to be dealt with.

Lastly, gaseous fuel admits of the attainment of results such as
cannot possibly be produced with solid fuel. You can heat gas to
any degree almost within ordinary furnace temperatures before it
enters into combustion, and I have shown by means of furnaces
which I have been constructing for the last twenty years, that air
can be heated to a very high temperature before entering into
combustion with gas, and there is thus produced a compound,
regenerative, cumulative action, which enables us to obtain by
means of gaseous fuel the highest temperature which can be
attained by combustion at all. You are all aware that the
temperature attainable by combustion is limited by the point of
dissociation, as Bunsen and St. Claire Deville have shown in their
beautiful researches. I may mention that in applying a gas
furnace for the fusion of steel on the open hearth, the point of
dissociation is practically very nearly attained.

Although this furnace is sufficiently well known by this time,
I may just shortly refer to its leading details. It comprises si
gas-producer which is presented in a somewhat novel form ; I
shall refer to it again presently.

The furnace itself consists of a furnace bed, and of brick
chambers filled with refractory material piled up loosely so
us to admit the circulation of air or gas through the same ; gas

N 2

l8o THE SCIENTIFIC PAPERS OF

from the gas-producer is directed upwards through one of these
chambers, and cold air is directed upwards through the adjoining
chamber. Both air and gas meet at the entrance of the combus-
tion chamber, and if the brickwork contained in the two chambers
is heated by any process, the air and gas flowing through it will
also be heated to very nearly the temperature of the materials
which impart that heat. They will therefore meet in combustion
at a high platform, so to speak. If the temperature of the brick-
work be at the upper surface, say 1000°, the air and gas will
be heated practically to 1000° before they combine, and the result
of that combustion must be a temperature of at least 1000° higher
than it would have been if the gas had not been so heated. The
products of combustion flow away at the other side through pas-
sages, every alternate passage leading into one or other of the two
remaining chambers, and in flowing downwards over the surfaces
of the brickwork they impart heat to those surfaces, raising their
upper layers to nearly the temperature of the products of combus-
tion, which after depositing their heat, reach the chimney compara-
tively cool. By reversing the direction of the gases, say every
hour, or every half-hour, as the case may be, and admitting gas
fuel at the place where one portion of the products of combustion
left, and air at the point where the other portion of the products
of combustion left, the currents of coal gas and air will pick up,
so to speak, the heat which was kept there in deposit, and reach
the point where they enter into combustion heated to a higher
degree by another 1000° ; and the consequence of this combustion
will be the attainment of a still higher degree of heat than was
obtained in the previous operation ; and so on, and on, until either
there is an equilibrium produced by the fusion of the materials
to be heated, at which point, of course, the excess of heat will be
absorbed in work done, or it will go on accumulating until com-
bustion itself will cease. Such a point will be reached at about
2500° Cent., the temperature of dissociation, and in heating one of
these furnaces up to the highest point which it would be safe to
attain in them, the flame, which at one portion of the operation is
a short white flame, becomes gradually extended into a long
attenuated flame of somewhat bluish appearance, clearly showing
that combustion is earned on under great difficulties. For any
temperature exceeding these limits we should have to resort to

S/fi WILLIAM SIEMENS,, F.R.S. l8l

difU'ivnt means altogether, such as the electric arc, because com-
bustion, however carried on, would not give us a higher tempera-
ture, even if we found materials that would resist the heat. These,
thru, are some of the advantages which may be claimed in favour
of gaseous fuel.

Fuel may be dealt with in two ways in order to produce gaseous
fuel ; it may be placed in a retort, and the retort heated by the
external application of heat in order to drive out the volatile con-
st .imciits. This is the process, as you are aware, generally followed
in gas-works. The use of gas for illuminating purposes dates
from the end of the last century. It was in the year 1792 that
Murdoch, of Soho, erected his gas- retort. After him Lampadius
published, in 1801, his investigation, very crude according to our
present ideas, on the subject of gas, and gas illumination, but it
was not until 1815 that gas illumination came to be practically
used, and it was applied first of all, I believe, in London. This
mode of producing gas by heating a retort, gives rise to two prin-
cipal products, the gas or gaseous fuel, and the residue, the car-
bonaceous matter, or coke. Both these constituents are very
suitable for heating purposes. Coke has this advantage over raw
coal, that it burns without developing smoke, and it is, therefore,
applicable for many purposes where this rapid development, or
change of the solid fuel into gaseous constituents, would absolutely
impede the operations to be performed, such as blast furnaces, or
locomotive engines. The gaseous fuel produced in the retort is
also applicable for heating purposes, but has hitherto been used
chiefly as an illuminating agent. I shall presently refer to the
working of a gas retort, with a view of showing that, although
we treat the gas coming from the retort as illuminating gas, as an
entity, it really consists of many constituents, and that the quality
of the mixture of gas coming from such a retort at one portion
of the process is very different from that which comes from it at
other portions. If the purpose before us is the heating of a fur-
nace, and our object is to obtain from the solid coal as much
gaseous fuel as possible, then it would not be desirable to have
resort to a retort. The retort process is a slow one, and the coke
produced, although it can be used also as fuel, would be undesirable
at works where no such fuel is required, where the only object is
to heat furnaces for melting glass, for making steel, for heating

1 82 THE SCIENTIFIC PAPERS OF

iron, or whatever it may be. For such purposes the whole of the
solid coal ought to be converted into gaseous fuel, and that has
been accomplished for many years. It is accomplished in the blast
furnace, where the combustion of coke under the influence of an
insufficient amount of blast gives rise to a product at the top,
consisting, in a great measure, of carbonic oxide gas mixed with a
large proportion of nitrogen, which necessarily comes in with the
air at the bottom. It is due chiefly to Mr. Ebelmen, a French
physicist of some distinction, that attention was called to these
gases in the early part of the century, but no practical application
was made of gas for a number of years, nor were the gases coming
from the blast furnace practically utilized until within the last
twenty-five years.

In the year 1861, I, in connection with my brother Frederick,
succeeded in the construction of a regenerative gas furnace in
which gas is used resulting from the total destructive distillation
of coal, and I now wish to place before you an apparatus of an
improved character for effecting that purpose. The gas-producer
which we have hitherto used was effective, so far as it goes, in
producing a combustible gas, but it did not fulfil all the conditions
necessary to produce this chemical operation to the best advan-
tage. One of the essential conditions for the total conversion of
coal into combustible gas is a maximum degree of heat at one
portion of the process. Where air enters, or is forced into a
mass of incandescent fuel, the temperature of that fuel is raised,
and the result of perfect combustion is carbonic acid. In order
to produce carbonic oxide, which is the gas fuel producible from
the carbonaceous constituent of the coal, it is absolutely necessary
that carbonic acid, or the result of perfect combustion, should be
produced in the first instance, and only by passing this intensely
heated carbonic acid over an additional mass of incandescent
carbon will it take up another constituent of carbon, and con-
stitute itself carbonic oxide ; unless the temperature is high
enough, this reconversion, or backward conversion from carbonic
acid into carbonic oxide, will be effected only partially and
imperfectly.

The apparatus, Plate 33, which I have lately tried, and which pro-
mises to be a great improvement on the former arrangement, consists
of a cylindrical chamber truncated towards the bottom, which is

WILLIAM SIEMENS, F.R.S. 183

filled with coal through a large hopper at the top ; this iron casing
is covered on the interior with a lining of refractory material, and
in this lining passages are arranged all round for the exit of the
; a large opening at the bottom admits of the removal of ashes
and clinkers that may be formed in the combustion of the fuel,
and a blast, which by preference should be heated, is directed
riirlit into the very heart of the mass of fuel. The result is a very
high temperature in the centre of the mass, and no heat is lost at
that point, which loss would interfere with the due conversion
back into carbonic oxide. At the same time, water, which is
admitted by a continual streamlet into a pan near the bottom, is
evaporated, and currents of steam are directed or drawn into the
apparatus, passing through the zone of highest temperature
towards the centre of the mass. They there become converted
by contact with the incandescent fuel into carbonic oxide, and
hydrogen gas, which add greatly to the calorific effect of the gas
produced.

Much has been said of late years of water-gas, and one has
sn 'ii statements in the papers calling attention to the enormous
store of fuel we possess in the ocean ; but a very little considera-
tion will show the utter fallacy of such a proposition, as getting
actual heat energy from water, which is already the product of
complete combustion. Water can only usefully assist in the
operation of gas production if you have spare heat to expend.
If in the gas-producer no water at all were admitted, you would
obtain a current of carbonic oxide, mixed with the other products
of distillation of the raw fuel that enters the apparatus ; and the
heat that would result from this combustion, even after a portion
of it had been expended in the backward reconversion of the
carbonic acid into carbonic oxide, and in the work of distillation
of the raw coal into hydro-carbons, would be sufficient to impart
to the gas at the point of exit a temperature of about 1,000°
F., or a little more. This heat must be considered as a loss,
because if the gas had to be carried to any reasonable distance,
this free heat would be dispersed before it reached the furnace,
and it is with the view of utilising this heat, which would other-
wise be lost, that water can be had recourse to advantageously.
But it must be administered in judicious proportions. If too
much water is admitted the chemical action which is intended to

1 84

. THE SCIENTIFIC PAPERS OF

be carried into effect in the gas-producer would cease, and the
temperature would never attain the degree necessary to convert
carbonic acid back into carbonic oxide. There would result a
poor vapoury gas, which on analysis would show a very large
proportion of carbonic acid still present, and probably free
oxygen, which never had combined at all. This is a point to
which I have wished particularly to call the attention of this
Society (which has for its object to combine strictly scientific
investigation with its practical applications), viz., the degree to
which water can be advantageously used for effecting this decom-
position, and its re-combustion.

When this process is properly conducted in the old form of
producer, the result of the total combustion is a gas of moderate
temperature, perhaps 400° F., found on analysis to contain of
carbonic oxide 24 per cent., hydrogen 8 per cent., carburetted
hydrogen 2 per cent., carbonic acid 4 per cent., nitrogen 61 per
cent. ; whereas in a well-organised gas-producer of the new form,
the proportions of hydrogen and carbonic oxide could be materially
increased, and the amount of nitrogen diminished proportionately.
I may also mention with regard to this form of producer, that the
earthy constituents of the coal are fused, or brought into a semi-
fused condition, and are carried down in that condition by the
weight of the fuel itself ; the object being to produce this semi-
fused mass of clinkers within the mass of the fuel without giving
them the opportunity of attaching themselves to the furnace, and
thus to save labour.

Another matter I wish to call attention to is the supply of gas
that can be carried to a distance. After all, gas produced by total
distillation in the gas-producer can only serve at works where it
can be used at once, but for the multifarious purposes of daily life,
for small works, and for domestic purposes where heating gas will
become from day to day used to a larger extent, there it is necessary
to send through pipes a gas of a much richer character than that
coming from a gas-producer. It must be a pure combustible gas,
such as marsh gas, or hydro-carbons, or hydrogen. This may be
produced in retorts of the ordinary kind, or, as I proposed eighteen
•years ago, when I suggested the supply of heating gas to the
Town Council of Birmingham, by the use of large vertical retorts
heated by a regenerative furnace, and furnished with a supply of

SIR WILLIAM SIEMENS, F.K.S. 185

water at the lower extremity, in order to get a very active distilla-
tinn, irrespective of the illuminating quality of the gas produced.
P.iu the (lill'uulty which presented itself was one rather of estab-
lishment. It was proposed to build works simply for the purpose
of supplying heating gas ; and although the Town Council of
Birmingham went to Parliament for a Bill to empower them to
carry out this suggestion, they lost their Bill in the Committee of
tin ilouse of Lords. Since then the public desire for the use of
heating gas has been gradually increasing, and another mode of
supplying it has suggested itself to me. It may have been thought
of by others, and it is one that recommends itself, I think, by its
simplicity to your consideration. It consists in separating the
produce of the distillation of coal in the gas retorts into two parts :
one to be set aside for heating, and the other for illuminating pur-
poses. The result of such a separation is illustrated by the diagram,
Plate :;i, which is based on some experiments made many years
ago by M. Ellissen, the Chemist of the Paris Gas Works, working
in connection with M. Regnault, the great physicist. They wanted
to determine by accurate observation and experiment what was
the least time to be allowed for each distillation. Instead of
carrying on the operation through six hours, they came to the
conclusion that it should be hastened, and completed in four hours,
and for this purpose the Paris Gas Works first adopted the
regenerative gas furnace as a means of heating retorts, and they
have used it very largely ever since. Through the kindness of
M. Ellissen I have been able to construct this diagram, showing
the quantity and the illuminating quality of the gas leaving the
retorts at each portion of the four hours during which the opera-
tion is carried on. The cross line shows the illuminating power
of the gas at the different portions. The abscissae give the time,
and each line represents a quarter of an hour, and the ordinates
give the illuminating power on the curve and the quantity on the
other curve. The diagram shows at once that during the first
twenty-five minutes or half-an-hour, the gas coming from the
retorts has a very low illuminating power. They are chiefly
occluded gases, consisting in a great measure of marsh gas. At
the end of the first half-hour, the maximum illuminating power is
reached, and for about an hour and a half this high illuminating
power is pretty wrell maintained. It then gradually diminishes,

1 86

THE SCIENTIFIC PAPERS OF

until at the end of the fourth hour it is less than half what it was
at the earlier portion of the process.

Now, if the gas produced in the retort could be directed by a
simple reversing valve into one main or the other the illuminating
power would rise from 18| to 18 caudle-power, and produce thus
all the advantages which the public now desire, a gas of increased
illuminating power ; whereas the heating gas constituting the
other half, or even two-thirds, of the total produce, would be all
the better qualified for heating purposes than the gas which is now
sent into our mains. It is well known that marsh gas produces
a higher calorific effect weight for weight than olefiant gas.
Therefore, if the marsh gas, and the hydrogen, which has still
higher calorific power, could be supplied separately, we should
get a better illuminating gas, and a better heating gas. It may
be objected that this would involve the laying down of a separate
set of mains ; but in answer to that, I contend that in many towns
we have already two sets of mains, because at one time or another
we have had opposing Gas Companies, each with their set of
mains ; but even if they do not exist, the question of the supply of
heating gas is such a large one, that the laying down of additional
mains should not stand in the way. If gas heating is to come
into general use, we shall soon require perhaps three or four times
the amount of gas which can possibly be supplied through the
existing mains. If at the same time such a great advantage can
be realised as the production of a gas of a higher illuminating
power, and with it a heating gas at lower cost, we should have an
advantage every way. But with regard to the question of cost, of
course, the public would be willing to pay a somewhat increased
price for a gas of 40 or 50 per cent, increased illuminating power,
whilst the heating gas, on the other hand, would not require the
same amount of care and expense in purifying as is necessary for
illuminating gas.

Then for this Society the question has another important aspect.
If by the supply of heating gas the total consumption can be vastly
increased the bye-products would be increased in the same pro-
portion, and these bye-products have now attained an im-
portance only second to that of the value of the gas itself.
Therefore, if by the carrying out of such a plan the total pro-
duction of gas can be very greatly increased, an impulse would

.S7A' WILUAM SIEMENS, F.R.S. 1 87

ivi'ii to chemical manufactures such us I need not specify
i,* ihis Society.

There are other questions which I should like to have touched
upon, regarding certain applications of heating gas, and certain
UB of intensifying illuminating gas, so as to bring it more
niMi-ly up to the standard of brilliancy of its great rival, the electric
light ; but this would, I fear, be travelling outside my subject,
which was to bring before, the Society the question of gas as a
lieu ting agent.

Tin I 'resident (Professor Koscoe) : We have all listened with
interest to this suggestive communication which Dr. Siemens
has made to the Society.

1 feel sure that it is a subject in which every one takes an
interest, and it has been treated in his masterly manner. I am
sure it will not fail to please all the members who are present.
The suggestions he makes are certainly bold enough, and let us hope
they may receive that attention from corporations, and from the
gas companies, which their importance and the originality of the
suggestions warrant. I am sure all the members of the Society
would wish to hear, if Dr. Siemens will consent to give us in a
few words, the mode of application, which 1 see from some models
he has here, he has already thought out. I think it does form a
very important part of his communication — the method in which
lie proposes to use the gases of low illuminating power for heating
purposes, and with your consent, gentlemen, I would ask him to
give us an account of the method which he proposes to employ
before we begin the discussion.

Dr. Siemens : I did not wish to extend my subject too much,
but willingly comply with the desire expressed by the President,
and will describe shortly two applications, one of heating gas, and
the other of illuminating gas, I have had occasion to make lately,
;ui(l which have given very encouraging results. In using gaseous
fuel in an ordinary h're-place, it becomes necessary to introduce
some substance which shall become incandescent in order to
radiate the heat into the apartment. We have substance of this

t description in use in the ordinary gas-grates, such as pumice-
stone, or other purely refractory material, but coke or anthracite
naturally suggests itself also as a suitable substance for taking up
heat and radiating it out ; it joins in the incandescence, and

1 88 THE SCIENTIFIC PAPERS OF

through its owii slow combustion aids the gas, which per s& is at
the present time at all events the more expensive fuel. If gas
could be supplied at Is. a thousand feet, it would no longer be
more expensive than coal, and the time will come, I believe,
when we shall have good heating gas supplied to us at that price.
Wherever gas fuel is used, there ought to be some means of
accumulating the effect of heat — some mode of regeneration,
and the question occupied my mind in my leisure hours, how
could the regenerative principle be applied to an ordinary open
fire-grate ? After some time the idea occurred to me, that in
taking advantage of the conducting power for heat of some metals,
I might be able to pick up heat at the back of the fire-grate,
and bring it down to a system of surfaces below the fire-grate, in
order there to communicate it to currents of air supporting
combustion in front of the grate.

Such an action is carried out in the grate of which I have a
model here, Plate 35. It has the ordinary fire-bars, to the bottom-
most one of which an inclined plate is attached, but instead of the
ordinary horizontal grate, an angular casting of iron with ribs
projecting from its lower surface supports the fuel. In front of
this casting, and behind the lower grate-bar, a gas-pipe is fixed,
having a limited number, (from 8 to 12) of gas orifices, of about
one-sixteenth of an inch diameter. The gas issuing from these
openings towards the back of the pipe, gives rise to flames percola-
ting through the coke oranthracite, or other similar substances which
fill the grate. Each gas flame in percolating up through the coke
produces very much the appearance of an ordinary coal fire in its
best condition, and the coke in front of the grate soon becomes
incandescent, giving out a very considerable amount of radiant
heat. When this point has been reached, the gas may be nearly
altogether turned off, as the glow of heat will be maintained
almost entirely by the hot air currents which by this time set in,
and produce intense combustion in front of the grate. I have, I
think, fourteen such grates in my office and house, including all
the principal rooms, and I have succeeded gradually in winning the
lady portion of the household over in favour of the improvement.
At first they said it was a dead fire, and you could not poke it,
but gradually even the housemaids, who at first thought it was an
abomination, have come round. They see that it gives them less

S/A' WILLIAM SIEMENS, F.R.S. 189

t n>u I ile ; that in lighting it is very agreeable to turn the gas on or
off as needful, and to be able to keep up the fire for half a day
without touching it. Many friends when they have come into the
room have said, "What a magnificent fire."

The cost also I have been able to ascertain by this time. For
tin- first quarter of the present year, in the five large grates,
including the drawing-room, my own study, and the dining-room,
where a fire is maintained nearly the whole day, the consumption
of gas amounted per grate to 6,HOO cubic feet, being of the value
of one guinea, or of one ton of coal when at its cheapest in
London. I am quite sure I have saved more than a ton of coal
in each grate, even after debiting myself with the consumption of
coke, which amounts to about one-third the consumption of coal.
Taking it all in all, it is a cheap fire, and certainly it is perfectly
smokeless. I am glad to say that several of the leading grate-
builders have taken up the construction of this grate, and I believe
at this present moment there are several hundreds of them in
actual use.

I have adopted the same principle for intensifying the light of
an ordinary argand gas-burner, Plate 35. Through the centre of the
burner a stem of copper projects, which is bulged out cup-like, at its
upper extremity, and is connected at its lower extremity with a
cylindrical box of copper, of high conductive power. The
cylindrical vessel is perforated all over, and round it ten layers
of copper-wire gauze are wound. The bottom also is closed by
ten discs of wire gauze. The burner is mounted in the ordinary
w.ay with a glass chimney, but at the point of the flame the
bulged portion of the rod, filled up with fire-clay, gathers up
heat, and conducts it downwards to the wire gauze, whilst the
current of air is drawn in by the flame itself, and is heated in its
transit through the wire gauze, to a temperature of from about
500° to 700° Fahr. By this simple arrangement the intensity of
the gas flame can be very nearly doubled. In fact, according to
one experiment, it was rather more than doubled, measured
photometrically ; and at the same time a flame of much whiter
character is obtained.

TJie President: Do you remember what is the illuminating
power of such a burner ?

Dr. Siemens : In using ordinary gas we get from this burner

190 THE SCIENTIFIC PAPERS OP

about 25 candles. I should mention that my brother, Frederick
Siemens, has worked independently on this same subject, and pro-
duced a burner which has been tried lately by some of the gas
authorities, and has given an economy superior even to what I
have stated. But the construction of my brother's burner is
essentially different, the gas flame being drawn downward through
heating apparatus.

The President : Like Faraday's ? Faraday arranged gas-burners
of that kind.

Dr. Siemens : I am not acquainted with Faraday's arrangement,
but although my brother has obtained very excellent results, I
think that the. plan I described is certainly a more simple arrange-
ment, and for ordinary applications commends itself on that
account. I should have mentioned that, in order to prevent the
oxidation of the copper conductor, and also in order to have a
reflecting surface, I have covered the copper conductor with a
very thin coating of enamel or platinum.

In 1he discussion of the Paper

"ON THE COMBINATION SYSTEM OF STEAM

HEATING FOR TOWNS AND VILLAGES,"

By CAPT. DOUGLAS GALTON,

The CHAIBJIAN (DE. C. W. SIEMENS),* in proposing a vote of
thanks to Capt. Douglas Galton, said this question was not a new
one in the abstract, for heating by steam had been carried out to
his knowledge forty years ago. But there was a novelty in the
proposal to generate steam in large volumes, at high pressure, at
certain centres ; and to send it at that high pressure for miles
under the streets. It was also perhaps a novelty to divide steam
by regulating and stopping valves, and supply it to individual

* Excerpt Journal of the Society of Arts, Vol. XXX. 1881-1882, pp. 93-94.

WILLIAM SIEMENS, F.R.S. 191

houses, so that it might be used for tho various purposes men-
tiuiu-d. The great question which presented itself was, would
such a pipe be trustworthy ; could it be laid and maintained for
this purpose without danger and without trouble. The paper
described an elaborate method of insulating the pipe to prevent
Mve loss by radiation, and from the figures given it seemed
to have been very successful ; but it appeared to him doubtful
whether, in our streets, a pipe so swelled out by insulators, and
logs of wood, and planks, could be laid. The space below our
streets was already so fully occupied, that the establishment of
such pipes would present very great difficulties. As Sir Frederick
Brain well remarked, if ,^hey wanted to bring heat into houses,
there was another means already established, and in full opera-
tion ; and from a little calculation he had just made, he found
that a cubic foot of coal-gas brought with it about 750 heat units,
which would be about the same amount as would be supplied by
8j cubic feet of steam, at a pressure of 6 atmospheres. The ques-
tion, therefore, would reduce itself to this, whether the one pipe or
the other could be maintained with the least expense. He, for
one, must adhere, for the present, to the opinion that it would be
easier to supply gas than steam. The paper spoke of various
modes in which steam heat could be used, and the radiator was
alluded to, but he did not think that apparatus was very happily
named. A steam-heated surface in a room no doubt radiated heat,
but not under the same conditions as a fire-place. Probably, nine-
tenths of the heat given off from that hot surface would be given
off to circulating currents of air, which would pass again and
again through this radiator, and a very small portion would pass
by radiation into the room. The rays from such a low source of
heat would hardly be felt, as radiant heat, at any sensible distance
from the source. Now, he looked upon an open fire-place not only
as a thing endeared to them by long habit, but he attributed to it
very high sanitary qualities. The open fire-place communicated
absolutely no heat to the air of the room, because air being a per-
fectly transparent medium, the rays of heat passed clean through
it. It gave heat only by heating the walls, ceiling, and furniture ;
and here was the great advantage of the open fire. If the air in
the room were hotter than the walls, condensation would take
place on them, and mildew and fermentation of various kinds

192

THE SCIENTIFIC PAPERS OF

would be engendered ; whereas if the air were cooler than the walls,
the latter must be absolutely dry. That was the great advantage
of the open fire, and it endeared it to them, though many did not
know exactly where and how the comfort arose. The Americans
were very ingenious, but he would not take their word with regard
to the comfort of their rooms ; when he was there he suffered
severely from the closeness of the atmosphere of the dwellings, and
when he travelled at night in the sleeping-cars he was obliged to
get up four or five times and step out on the foot-board to get
thoroughly cool. The only explanation he could offer was that
the attendants were black men, whose standard seemed to be the
interior of Africa. But although he could not agree absolutely
with the proposal, yet the question of doing away with the smoke
and nuisance which now resulted from open fire-places was so
important, that any suggestion, backed by a considerable amount
of experience was welcome ; and Captain Galton well deserved
their thanks for the able manner in which he had brought the
subject forward.

In the discussion of the Paper

"ON AIR REFRIGERATING MACHINERY AND
ITS APPLICATIONS,"

By JOSEPH JAMES COLEMAN, F.I.C., F.C.S., &c.,

DR. SIEMENS * said the question that had been brought before
the Institution was one of considerable engineering and public
importance, and he was glad to see it brought forward in such a
clear and efficient manner. The subject was not one of recent
date, but had occupied the minds of many men for a considerable
number of years. The author had been good enough to mention
his name in connection with a proposal which he had made twenty -

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
LXVIII. 1882, pp. 176-186.

ir/I./.fAM X/EMENS, F.R.S. 193

five years ago to the effect of superadding to the simple play of
.so jiu nips — the compressing and the re-expanding pump —
an intercli;ui'_ror of temperature. That suggestion, as it presented
to him, was published only in a provisional specification
of a patent which had never been specified finally ; but more
thiiii that had been accomplished, and perhaps it might interest
the members to know exactly what had been done. Dr. Gorrie,
an American, had patented and brought to England a certain
scheme, which had been taken up by Mr. Wollastou Blake and a
fr\\ other gentlemen in London, who erected plant which was
perfect in all its details. The engine had a 25-inch cylinder and
5 feet stroke ; the compressing and re-expanding pumps were made
by James Watt & Co., and had been erected somewhere in the
north of London ; but when it came to be tried it was impossible
to get a depression of temperature exceeding 20° Fahrenheit. That
being the case, after several ineffectual attempts to improve the
result, he was asked to examine the machine and report to the
proprietors. He had lost sight of the report, but Mr. Wollaston
Blake had sent him a copy with the drawings, and suggestions
which ho had made for improving the efficiency of the engine, and

lese would be appended to his remarks (see p. 195 et seq.~) The

eport was of some interest, as it went very fully into the principles
underlying the question of refrigeration, and the causes of the
ion-realisation of those results which should have been effected.
It seemed curious that, from the & priori examination and calcula-
tions which he made, he had come to the conclusion that the

jmperature could not be depressed more than 20° by the machine,
that when that reduction of temperature was reached the loss of
affect would exactly balance the beneficial result; and the actual trial

jrroborated that statement. The author had. perhaps, hardly given
lira credit for the labour which he had spent upon the subject. He

id not only devised the temperature interchanges, which was the
most essential feature in the whole process, and without which no
ufficient results could be obtained, but he had fully recognised
the antagonistic effect of aqueous vapour in the air dealt with.

"he loss of effect " by throttled passages " he had given at

LO'08 HP. That required some explanation. Dr. Gorrie evidently

wished to avail himself of the expansive force of the compressed

and cooled gases before they were discharged ag lin ; but he did

VOL. i. o

IQ4 THE SCIENTIFIC PAPERS OF

not conceive that the whole effect of the machine depended upon
the heat (utilized negatively) in that final expansion ; and in
order to be sure, as he thought, doling out as it were the com-
pressed air to the expanding cylinder, he introduced a throttle-
valve into the compressed and cooled air ; and instructions were
given to wiredraw that valve as much as possible on the sup-
position that after all the safe plan of gaining refrigeration would
be to expand air through the throttle-valve, that was through a
contracted orifice. Dr. Siemens showed that contraction was
the worst thing that could happen — that by merely expanding
air without exacting work from it, no depression of temperature
would ensue. By friction and useless agitation of the fluid, the
loss was 3 HP., by condensation of vapour, TOO HP. That, of
course, was a fruitful source of loss. If the air was compressed
and then not sufficiently refrigerated, the vapour which it con*
tained would, on re-expansion and final cooling in the expansive
cylinder, not only condense, but be converted into ice. The
vapour which went in as steam would come out as ice, and the
whole of the latent heat taken up in that double process of con-
version was entirely lost. That loss in some of the earlier
attempts was fully equal to the whole useful effects that could be
expected. But the greatest loss of all was incurred in allowing
the cold air to leave the apparatus at the minimum temperature.
Dr. Siemens suggested that a current exchanger should be supplied ;
that the cold air, after it had done its negative work, so to speak,
should pass up through spaces surrounding a number of vertical
tubes, and that the compressed air should pass through the tubes
themselves in the opposite direction, and exchange temperatures.
In that way 1\ HP. could be saved. It was shown in the report
that a very economical result might be obtained by observing
these precautions. Attached to the report were diagrams taken
from the three cylinders, and a sketch showing the kind of
apparatus which was then proposed, and which, he thought, in all
material points was similar to the machine brought before the
members by the author of the paper. There was, however, one
point in which he still dissented from the knowledge and experience
of the present day. He had stated in the report that as ice-making
was really negative work, there could be no absolute limit to the
effect to be produced by heat in that direction. It was true that

SfK WJLLIAAf SIE.\fENS, F.R.S. 195

great men like Dr. Clausius, Sir William Thomson, and Dr. Joule,
had established a certain equivalent, an equivalent which was
perfectly true if the heat generated in the mechanical process of
t-< impression were treated as waste. But by means of the re-
trciit-rator it was possible to recover, at any rate, a large proportion
df the heat so produced at the minimum temperature attained in
run pressing the air. The ice produced was, therefore, only
i-viilcnce of a transfer of heat from the fresh water to air or water
raised to a high temperature. And in this respect the operation
of freezing differed very essentially from that of power-producing,
when a given amount of heat must absolutely disappear. The
report might perhaps be interesting, as it gave the particulars of the
views which he then entertained, and of the theoretical conclusion
at which he arrived — a conclusion which, he had little doubt,
though it was not now fully admitted, would be admitted before
the subject was done with. The paper dealt with the question as it
stood at the present day in a very able and proper manner, and the
applications which would flow from a cheap and ready mode of
lucing temperature to any desirable point could not be easily
overrated.

tBPORT OX THE PERFORMANCE OF NEWTON'S PATENT REFRIGE-
RATING APPARATUS, WITH SUGGESTIONS FOR IMPROVING

THE SAME.

Having examined your apparatus for the artificial production of
; on your premises at Camden Town on the 8th inst, and having
examined carefully into the principles involved, and into the
rases of its present lack of success, I have to report thereon as
allows : —

" Description. — The apparatus consists of a horizontal condens-
ing steam-engine of 32 inches diameter of cylinder, of 5 feet
Bngth of stroke, in excellent condition, by Messrs. James Watt
id Co., which imparts motion to a crank-shaft with two
Iditional cranks, which are connected with the pistons of two air
cylinders of 4 feet 6 inches stroke, and of 21 inches and 19 inches
diameter respectively. The piston rods of the air cylinders are
carried horizontally through the cylinder bottoms, and are each of

o 2

196 THE SCIENTIFIC PAPERS OF

them attached to a second piston 4 inches in diameter, working in
the barrels of double-acting pumps, which inject at each stroke
their fluid contents into one extremity and the other of their
respective air cylinders. The larger air cylinder (of 21 inches
diameter) compresses atmospheric air into an upright receiver,
where it is allowed to separate from the injected water, the water
issuing from the bottom through a ball-tap, and the comparatively
dry air proceeding through a pipe leading from the tap into a
larger horizontal receiver or reservoir, which is provided with a
mercurial pressure-gauge indicating 55 Ibs. effective pressure per
square inch when the apparatus is at work.

"A copper pipe (of 3 inches diameter, and provided with a stop-
valve) leads from the receiver to the second air cylinder of 19
inches diameter, wherein the compressed and cooled air is allowed
to expand again down to atmospheric pressure, and being mixed
in the expanding cylinder with a spray of injected brine (by
means of the second injection pump) it is expelled into a large
cistern containing about 500 gallons of brine. The injected brine
here subsides, and the expanded air escapes freely into the atmo-
sphere. It appears from records of experiments made on the 28th
and 29th of April, and on the 13th of May last, that the water
injected into the compressing cylinder was raised uniformly from
2^° to 3° Fahr. in temperature, whereas the brine in the cistern
(500 gallons) fell gradually in temperature from 2° to 3° per hour at
first, but diminishing gradually, as the temperature descended gradu-
ally below that of the atmosphere, until it reached the temperature of
18'5°, which appears to be the lowest degree ever attained. The
speed of the main shaft was, during the first-named experiment,
twenty revolutions per minute, when the injected water was raised
3°, and the brine in the cistern reduced only 2° per hour at the
commencement of operations, and during the second experiment
the speed varied from fourteen to sixteen revolutions per minute,
when the injection water was raised only 2^°, whereas the brine in
the cistern was lowered 3° Fahr. per hour, until it descended corre-
spondingly below atmospheric temperature. The water to be
frozen had been filled into cans of copper and dipped into the
brine, after the greatest depression of temperature had been
attained, which caused a rise of temperature from 18'5° to 22°
or 23°.

.V7A' \V1LUAM SIEMENS, F.R.S. 197

" When I examined the apparatus I found the speed to average
fifteen revolutions per minute, and the following are the mean
;tvs and horse-power indicated upon the three cylinders,
viz.: —

MEAN HORSE

PBESSUBE. POWER.

The steam cylinder according to data

obtained from Mr. Blake . . 8'8 Ibs. 32

Comiiressing cylinder . . . . 26'75 „ 37'85

Expanding cylinder with injection . 18*25 ., 2T1

Kxpanding cylinder without injec-
tion 14-22 „ 16-5

" Theory. — Before entering into a consideration of the merits
and defects of the apparatus in its actual condition, it will be of
advantage to establish generally the principles whereon its effect
depends. The heat which it is purposed to abstract from the
water is communicated to expanding air, or, in the language of
modern science, is converted into mechanical effect, and it can be
proved philosophically that a caloric engine may be devised which
derives its power from the latent heat of water in causing it to
congeal. In fact, the heat communicated by the injected solution
to the expanding air is a source of power to the machine, which
would yield both effective power and ice but for the compressing
cylinder, which in the arrangement described has to be worked by
an engine. The power expended in compressing the air produces,
however, its equivalent of heat, which heat would be sufficient, in
a philosophically perfect arrangement, to reproduce the power
necessarily expended by the engine, and a surplus of effective
power proportionate to the heat abstracted from the water would
be realised.

" This view of the principle involved in your machine is novel,
I believe, and somewhat startling ; but I am confident it will
be acquiesced in by those conversant with the dynamic theory
of heat, and although the conditions of the ideal machine will
probably never be realised, it is important for our purpose to
know that the production of ice does not necessarily involve an
expenditure of power.

" Starting, then, with the admission that the heat produced in

198

THE SCIENTIFIC PAPERS OF

the compressing cylinder is practically unavoidable, the question
arises, — how great is the necessary expenditure of engine-power, or,
in other words, the difference between the power absorbed in the
compressing cylinder and the power obtained in the expanding
cylinder ? The difference of power arises in consequence of the
abstraction of heat from the compressed air, and is directly pro-
portionate to the number of degrees of heat abstracted or generated
in compression and afterwards lost in expansion. The quantity of
heat lost in the expanding cylinder is, however, proportionate to
the power produced in expansion, and is entirely independent of
the limits of expansion which, as has just been shown, determines
the degree or quality of cold produced. It follows that low rates
of compression and expansion of the air are favourable to the
production of cold with the least expenditure of engine-power,
and I find by calculation that the power value of a given amount of
heat abstracted (or cold produced) increases in the ratios of the
squares of the number of degrees of reduction of temperature, re-
quired.

" It is therefore highly advantageous to compress and expand
large volumes of air to a moderate extent, consistent with the
requisite energy of action of the machine.

"Adhering for the present to the rate of compression obtained in
working your apparatus, namely 55 Ibs. per square inch effective
pressure, the following result might be obtained if no incidental
losses of any kind were obtained, and taking the speed at 15 revo-
lutions per minute.

H.P. indicated in compressing cylinder
„ ,, expanding cylinder

engine ....

Units of heat produced in compression
minute .......

Injection water raised in compression cylinder

Units of heat produced in expansion

Ice produced per hour from water of 60°

600 gallons of water cooled per hour

The diameter of expanding cylinder Lo be .

per

35-0

25-2
9'8

1,500
2-5°
10-80
381 Ibs.
10-8°
17| ins.

In comparing these results with those actually obtained, it will

SIR WILLIAM SIEMENS, F.R.S. 199

be observed tho compressing cylinder did its work remarkably
well, whereas the duty obtained from the expanding cylinder falls
manifestly short of the results indicated by theory. It was
necessary then to inquire carefully into the causes operating
a.u'ainst the performance of the expanding or freezing apparatus,
and the following are the results of that enquiry : —

" 1. The power developed in the expanding cylinder should be
•_'.")•!' H.P., whereas only 16*5 H.P. was actually indicated (when
no solution was injected) owing to —

"a. A great loss of pressure between the cylinder and the

receiver.
" b. Excessive expansion beneath the atmospheric line, which

involves a loss of power on the return stroke.
" c. Imperfect adjustment of the admission valves.

"2. When solution was injected the indicated power rose to
21*15 H.P., or G'93 H.P. higher than before, which excess of
power may to a great extent be fictitious, being caused by the
penetration of the solution into the indicator, but it proves that
considerable force must be expended in forcing and agitating the
solution, and all the power so spent will produce its equivalent of
heat and must be deducted, besides the power lost in friction of
piston, from the indicated 16'5 H.P. in estimating the refrigerating
effect to be produced.

" 3. The compressed air being cooled by injection of cold water,
it will certainly be saturated with vapour, which vapour will be
condensed and converted into ice during expansion, and produce a
white fog pervading the expanded air.

" The quantity of vapour so condensed and congealed, supposing
the temperature of the compressed air in the reservoir to be 68°
Fahrenheit, amounts to O'OG Ib. per minute, yielding about 1330
units of heat, which represents a loss = 1'9 H.P.

" 4. The expanded air leaves the apparatus at the temperature
of the cistern, or say at 20°, which is 6 8° -20° = 48° below the
temperature at which it entered the compressing cylinder,
which represents a loss of effect = 312 units = 7'25 H.P. at
least.

" 5. The cistern containing the solution, the expanding cylinder,
the pump and the connecting pipes, present an aggregate surface

200

THE SCIENTIFIC -PAPERS OF

of about 200 superficial feet of heat-absorbing surface 6 5° -20° =
45° cooler than the surrounding atmosphere on an ordinary
summer's day. The machinery especially being ill-protected, and
of a dark colour, I estimate the loss of effect from this source at
about 3'6 H.P., assuming that the covering prevents two-thirds of
the radiation which would take place if the surface were unpro-
tected. Assuming that the solution in the cylinder has been
reduced to 20° Fahrenheit, the total losses are as follows : —

By throttled passages, &c.
,, agitation and friction
,, condensation of vapour
,, cold air leaving cistern .
., absorption of heat from the atmo-
sphere

see paragraph 1

2 say
„ 3
4

H. P.

10-08
3-00

Total loss .

Total power to be obtained

3-60

25-83
25-2

It follows that when the temperature of the cistern is reduced
to 20°, the losses of cold produced equal or actually exceed the
power of the apparatus of producing the same, which result is
corroborated sufficiently well by the experience obtained, although
it has been the result of independent calculation as regards para-
graphs 1, 3 and 4, but only assumed as regards paragraphs 2
and 5, which, however, are the less important.

" The next question is whether, and to what extent, these
different sources of loss can be obviated ; and I will at once
proceed to describe the fresh arrangements I propose for your
adoption, and which are illustrated by the sketch shown
on Plate 36.

" The proposed expanding and freezing apparatus consist of a
cylinder of cast iron of 17 inches diameter, instead of 19 inches,
and 4 feet 6 inches stroke. The pipes and valves connected with
this cylinder to have a sectional area of not less than 15 square
inches (instead of 7 square inches), and no stop-valve to be
introduced. The admission valves to cut off at one-fourth part
of the stroke. The expanding cylinder is immersed in the cistern
containing the freezing solution, and the expanded air issues from

.s/A' WILLIAM SIEMENS, F.R.S. 2OI

the cylinder into chambers a, a, forming the inner sides of the
cistern, which chambers are perforated towards the bottom to
allow the air to bubble through the solution. The cistern to be
about 10 feet long, by 5 feet broad, by 2 feet 6 inches deep. The
t \|i;itided air issues from the cistern into the atmosphere through
a series of vertical tubes contained in a cylindrical air-tight
chamber (C) through which the compressed air circulates on its
way to the cylinder. The compressed air enters the annular
chamber through a pipe &, and issues into the tube chamber (('),
through a number of holes in order to distribute it uniformly. It
then descends gradually, and effects an interchange of temperature
with the expanded air within the tubes, which latter issues at the
top at nearly the temperature of the compressed air on entering,
whereas the compressed air reaches the expanding cylinder nearly
at the temperature of the cold cistern, or below the freezing point.
The annular chamber d, serves to collect the cold compressed air
at the bottom, and to supply it to the admission valves e and /,
by means of pipes not shown. The tube in the centre of the
chamber C is of about 9 inches diameter, and contains a cylindrical
vessel, g, filled with water, which enters the same from above, and
is withdrawn at the bottom in a cooled condition to fill the former
for the production of ice. The tube surface required is about
200 superficial feet.

"The advantages obtained by this exchange of temperatures
are very important. The compressed air is reduced below freezing
point before it reaches the expanding cylinder, and the aqueous
vapour it contained is at the same time condensed upon the tubes.
Indeed the lower the temperature of the cistern descends, the
more will the compressed air be reduced in temperature, and the
cold produced will accumulate in intensity to any desired degree.
Both injection pumps and one of the large air receivers will be
dispensed with, for it makes no difference at what temperature
the compressed air enters the exchanging apparatus. I should
recommend, however, to retain the injection pump of the com-
pressing cylinder till the efficacy of the new apparatus has been
proved by experiment.

"This new apparatus would reduce the different losses of
refrigerating effect in the following proportions, according to my
estimation, viz. : —

202 THE SCIENTIFIC PAPERS 01'

Probable loss expressed in H.P.

1. By throttled air passages .... say 2 '5

2. „ agitation and friction of piston . . . . 1>5

3. „ condensation of vapour in expanding cylinder . O'o

4. ,, cold air escaping ....... 2'5

5. ,, heat absorbed by radiation .... 2*0

9-0
Gross power of expansion ... . . 25'2

Leaves effective refrigerating power . . .16*2
which is capable of producing 258 Ibs. of ice per hour.

" The production of ice may, however, be increased 25 per
cent, in working the apparatus at the rate of twenty revolutions
per minute, which the increased passages, &c., admit of. It is
hardly necessary to draw your attention to the gain of engine
power which the proposed alterations are calculated to effect. In
order to reduce the engine power required to its proper limit, I
should propose considerable modifications in the general arrange-
ment of the apparatus, consisting chiefly in attaching the three
pistons upon one rod, in accelerating the speed of the pistons, and
in compressing the air to only about 30 Ibs. per square inch.

" COST. An apparatus so constructed would be considerably less
complicated and costly than our present apparatus. The ex-
panding cylinder might be made of cast iron, the two injection
pumps and one air reservoir would be suppressed, and in their
stead the comparatively cheap exchanger of temperatures or
reciprocator be added. The engine and moving gear would,
moreover, be greatly diminished. The proposed alterations in the
existing apparatus would not cost above £200 (I consider), taking
into account the value of materials of the present expanding
cylinder and injection pumps. London, 28th July, 1857.

Signed, C. W. SIEMENS.

To WOLLASTON BLAKE, Esq., F.E.S. &c.

Mr. David Thomson said he had been much interested in the
paper, the theoretical part of which he thought was quite correct.
Dr. Siemens had thrown some doubt upon the formula as given in

S/A' ll'/LL/AM SIEMENS, J-'.N.S. 203

the paper, p. 147 ; Mr. Thomson had the honour of laying this
formula before the Institution some years ago, when Mr. Kirk's
paper mi the Mechanical Production of Cold was read and
discussed. At that time Dr. Siemens mentioned that the formula
was due to Clausius, but Mr. Thomson had deduced it from
Rankine's works.

Dr. Siemens observed that Clausius and Rankine worked
independently and simultaneously at the same question, and
arrived at the same conclusion. He did not doubt the correct-
ness of the formula put forward by those eminent mathematicians
find physicists, but he doubted its applicability under all con-
ceivable circumstances to the question under consideration.

In the discussion of the Paper

"ON THE THEORY OF THE GAS-ENGINE,"
By DUGALD CLERK,

DR. SIEMENS * said that one part of the paper dealt with
matters regarding the mechanical arrangement of gas-engines, and
the other with a theoretical question, that of the law of com-
bustion. He would refer to the theoretical part first, because
the author appeared to attach great importance to it, and as
Dr. Siemens had from time to time given a great amount of con-
sideration to the action of negative combustion or dissociation,
it might be of some interest to the members to see how far his
views fell in with those set forth by the author. It was well
known that by combustion no unlimited degree of temperature
could be attained. Thus, in a furnace worked at very high tem-
perature the fuel was not completely burned when it came in
contact with the oxygen of the heated or non-heated air. The
moment a certain comparatively high temperature was reached

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
LXIX. Session 1881-82, pp. 251-4.

204 THE SCIENTIFIC PAPERS OF

the carbon refused to take up oxygen, or the hydrogen refused to
take oxygen, and what had been called by Bunsen, and shortly
after him, by St. Claire Deville, dissociation, arose. The point
of dissociation was not a fixed one ; partial dissociation came
into play at a comparatively low temperature, and went on
increasing at a higher temperature in very much the same
ratio as vapour density increased with temperature. Thus,
if aqueous vapour were passed through a tube at a sufficient
temperature the whole of the vapour would be dissociated, and the
oxygen and the hydrogen would be separated. It was true if these
gases were left to themselves they would, the moment the tem-
perature lowered, again associate or burn ; but if precautions were
taken to cool them rapidly after they had attained that high
temperature they would be found as a mixture of oxygen and
hydrogen simply. The author had stated that the law which
governed these actions was not well known and required research,
but Dr. Siemens would like to know whether he was aware of the
researches of St. Claire Deville on the subject. It might be that
the determinations of St. Claire Deville were not quite correct, but
in the meantime they might be regarded as being so. He found
that at atmospheric pressure the point of half dissociation of
aqueous vapour arose at a temperature of 2,800° Centigrade, and
that of complete dissociation at a much higher temperature. Taking
that law as determined by the French philosopher, it did seem
reasonable to suppose that when a mixture of hydrogen and
oxygen, with or without a mixture of nitrogen, exploded, the point
was reached beyond which the temperature did not increase, and,
according to the author, that point was 1,500° Centigrade. If such
a temperature was reached in the working cylinder complete
combustion would not take place immediately, but only partial
combustion would occur, which would go on as the temperature
diminished by absorption into the cylinder or by expansion, and
that combustion would be completed only in the course of the
stroke. In that way the action which had been described with
reference to the diagrams was reasonable enough. With regard to
the mechanical arrangement of gas-engines, the author distin-
guished between three types. In the first, the mixture of gas
and air drawn in at atmospheric pressure was exploded. In the
second, with which the author had connected Dr. Siemens's name

WILLIAM SIEMENS, F.R.S. 205

ostliat of the first proposer, the combustion was produced gradually ;
the gases were ignited as they flowed into the heating cylinder. In
the third type, the gases, after being compressed and mixed,
admitted into the working cylinder, and suddenly exploded.
With reference to the early engine which Dr. Siemens constructed
in 1860, Plate 87, the author had stated that it combined other
rl. 'incuts, which were entirely wanting in the gas-engines of the
present day. The gas-engine of the present day, taking either of
the three types, was, in his opinion, in the condition of the
steam-engine at the time of Newcomen. The fuel was burnt in
a cylinder which it was attempted to keep cold by passing water
over it, and it was easy to conceive that the heat so generated,
was only partly utilised for maintaining the state of expansion of the
heated gases, the cold sides of the cylinder taking a good half of
it away at once, thus causing a great loss. Then there was another
palj >uble loss in these engines. After expansion had taken place,
after half the heat had been wasted in heating a cylinder which
was intended to be kept cool in order to allow the piston to move,
the gases were discharged at a temperature of 1,000°, or in the
best types about 700°. That amount of heat, representing in one
case one-half and in the other two-thirds of the total heat gene-
rated, was thrown away. This was heat which could be saved and
made useful. Instead of commencing the combustion at a tem-
perature of 60°, if the heat of the outgoing gases were transferred
to the incoming gases, combustion might commence at a tempera-
ture of nearly 1,000°, and the result would be a very great
economy. In the engine which he constructed in 1860 all
those points were fully taken into account. The combustion of
the gases took place in a cylinder without working a piston, and in
a cylinder that could be maintained hot, and the gases, after having
complete expansive action, communicated their heat by means of
a regenerator to the incoming gases before explosion took place.
Although the engine was not worked with ordinary gas used for
illumination, but by a cheaper kind made in a gas-producer, he
then thought that a gas-engine constructed on that principle
would prove to be the nearest approach to the theoretical limits
which could never be exceeded, but which might exceed the limits
of the steam-engine four or five fold. The engine promised to give
very good results, but about the same time he began to give his

206 SCIENTIFIC PAPERS.

attention to the production of intense heat in furnaces, and having
to make his choice between the two subjects, he selected the
furnace and the metallurgic process leading out of it ; and that
was why the engine had remained where it was for so long a time.
But now the time had come when there was a greater demand for
engines of a smaller kind to do their best in houses and in small
works, and when marine engineers especially had become fully
alive to the importance of more economical arrangements. He
therefore looked upon the question before the Institution as one of
first importance to engineers, and he hoped that it would be well
discussed.

METALLURGY.

METALLURGY.

"ON THE REGENERATIVE GAS FURNACE AS
APPLIED TO THE MANUFACTURE OF CAST STEEL."

[A Lecture delivered before Hit Fellows of the Cliemical Society, May 7th, 1868.]

By C. W. SIEMENS,* F.R.S., Mem. Inst. C.E.

IN responding to your call to deliver a lecture to your Society,
on a subject of applied chemistry, I feel that I have undertaken
a very responsible task, a responsibility which is only balanced
by the honorary distinction conferred by your call.

It is a hopeful sign of the advancement of science that your
Society puts itself into intellectual communication with en-
gineers and others, whose mission it is to apply and practise
that pure science cultivated within your body. It would be pre-
sumptuous on my part to reason with you upon a purely chemical
subject, for although many years ago I had the advantage of re-
ceiving instruction f romWohler and Himly, I can in no way lay claim
to be a chemist of the present day. Yet I can safely affirm that
of all the instructions I received in early life, there is none that
has been a more useful guide to me in my professional pursuits.

The subject to which I wish to call your attention this evening
is one that has occupied my mind for many years, and which I hope
may engage your interest, involving as it does the generation of
intense heat by means essentially differing from those in general
use, and the application of that heat to the production of cast steel
in large masses, and directly from the ore or scrap metal.

* Excerpt Journal of the Chemical Society, 1868, pp. 279-310.
VOL. I. P

2IO THE SCIENTIFIC PAPERS OF

The regenerative gas furnace, which is the joint production of
my brother, Frederick Siemens, and myself, is already known
through its application to nearly all those branches of industry in
which furnace heat is employed. It was described by Faraday,
to the members of the Royal Institution, in the last public lecture
delivered by that great philosopher in 1862, and has since been
the subject of several publications. I need not, therefore, bring
before you all the features of this mode of producing intense heat
for the various purposes to which it has been applied, but shall
confine myself to its description as applied only to the production
of cast steel in large masses, which is the principal subject of my
present communication.

NATURE OF STEEL. — Before proceeding with this description I
shall briefly remark upon the nature of steel and the methods by
which it has hitherto been produced.

STEEL. — Steel is generally regarded as a compound of iron and
carbon, possessing the remarkable quality of becoming exceedingly
hard when heated and suddenly cooled. The proportion of carbon
determines the degree of hardness of which the steel is capable, or
what is termed its temper.

The following table of the percentage of carbon in steel suited
to different purposes, is prepared from analyses made in my
laboratory by Mr. A. Willis, unless another authority is men-
tioned.

Description. Carbon per cent. Authority.

Wootz . ' . . . . 1-34 T. H. Henry

Steel for flat files . . . . T2 A.Willis

„ for turning tools . TO „

,, (Huntsman's) for cutters .1*0 „

., for cutters .... '9 „

„ for chisels .... "75 „

Die steel (welding) ... '74 „

Double shear steel .... '7 ,,

Welding steel .... '68 „

Quarry drills *64 ,,

Mason's tools ... '6 „

Ramrods '6 .,

Common steel for stamping . '42

.S7A' WILLIAM SIEMENS, l-.R.S.

21 I

Description.

for magnets

containing
tungsten .

„ spades . . . .

„ hammers .

Bessemer steel for rails . . .

Homogeneous metal armour plates

Vi ry mild steel from open

hearth furnace

Sample before Spiegel was applied
Bessemer iron (pure) .

Carbon per cent. Authority.

•4
•32
•3

•25 to •:
•23

•18
•05
trace

A. Willis

various
Percy

A. Willis
Abel

Cast-steel containing less than 0*3 per cent, of carbon is no
longer capable of being hardened, and should be classed rather
as homogeneous or melted iron than as steel, while on the other
hand an excess of carbon above 1'4 per cent, again deprives the
metal of the quality of taking a temper, and it must then be
regarded as approaching in character rather to white cast-iron.

It is, however, a contested question between chemists whether
the presence of a third substance may not be necessary to produce
steel. Fre*my, in his celebrated controversy with M. Caron before
the French Academy, maintains that nitrogen or cyanogen is a
necessary constituent of steel ; the investigations of other chemists
appear to confirm this theory, and the old Sheffield practice of
mixing leather and other animal substances with the charcoal
used in the converting furnace would also seem to corroborate it.
M. Fremy's theory probably admits of great extension, and it may
be that we should regard steel as a triple combination of iron with
carbon and with another substance, taken from a series comprising
nitrogen, sulphur, phosphorus, silicon, manganese, tungsten,
titanium, tin, silver, and probably many other elementary bodies,
each of which entering, in an exceedingly small proportion, into
combination with the iron and carbon is capable of imparting to
the steel distinctive physical properties.

NITROGEN. — Nitrogen has been found invariably in the stee
produced by the ordinary processes.

SULPHUR AND PHOSPHORUS. — Sulphur and phosphorus are
commonly regarded as the worst enemies to steel, the one render-
ing it " red-short," or incapable of being forged, and the other

P 2

212 THE SCIENTIFIC PAPERS OF

" cold-short," or brittle at ordinary temperatures. It is owing to
the presence of these impurities in nearly all British irons, that
the highest qualities of tool steel are still made entirely from
Swedish or other charcoal-iron of high quality, produced from the
purest ores. If phosphorus exists in steel in quantities exceeding
O'l per cent., its presence is indicated on breaking the sample, by
peculiarly bright and distinct faces of crystallisation on the fractured
surface, but the steel admits of being forged and welded, and
though somewhat brittle when cold, it is remarkable for hardness.
It would, therefore, not be safe to maintain that phosphorus in
small quantities is objectionable in steel under all circumstances.
Steel containing above 0'2 per cent, of sulphur breaks under the
hammer at a low red heat, and is, therefore, only applicable to the
production of steel castings : for this purpose the presence of a
small amount of sulphur is unquestionably a positive advantage,
imparting increased fluidity to the molten metal and toughness
to the casting when cold. This effect is so well understood in
Sweden as regards cast-iron, that in casting ordnance sulphur
is added to the metal.

MANGANESE. — Manganese possesses the remarkable property of
counteracting to a great extent the effect of sulphur (or red-
shortness) in steel. Its application for this purpose is due to
Josiah Marshall Heath (1839), and must be classed among, the
most important discoveries in modern times ; for it is only in
consequence of the addition of manganese (in the form of oxide,
mixed with carbon) to the steel melted in pots, that the Sheffield
melters have been enabled to use English puddled iron for the
commoner qualities of steel in place of the purer and more costly
Swedish iron ; and again it is only in consequence of the addition
(by Mushet) of manganese in the form of ferro-manganese or
spwgeleisen to the liquid bath produced in the Bessemer converter
that the better qualities of English pig-iron have been made avail-
able for the production of malleable Bessemer metal. It is dis-
tressing to think that the author of this invaluable discovery was
deprived of the fruits of his labour by an unjust combination
among the manufacturers chiefly benefited by him, who contested
the validity of Heath's patents upon trivial grounds, relying for
their success upon their financial power to crush his unquestion-
able claims. The specific chemical action of manganese has never

-S7A' WILLIAM .S7AM//;.V\, I-.R.S. 213

been clearly demonstrated. It is certain that only traces of
metallic manganese are ever found in cast-steel, and that the bulk
of the manganese added is found combined with oxygen, silica,
and alumina, in the form of slag ; yet it follows, from experiments
which I have had occasion to make, that on adding manganese to
a bath of liquid steel in an open reverberatory furnace, its beneficial
action ceases if the metal is retained in the furnace, exposed to the
flame for any length of time, say more than half an hour. This is
the case quite independently of any sensible reduction in the pro-
portion of carbon in the steel, and indicates that the effect of
manganese is not due to the removal of sulphur or other substances
with which it may enter into combination, but it is inseparable
from the presence of metallic manganese in the steel, in however
small a quantity. A further confirmation of the view, that the
addition of manganese removes no impurity from the steel, is
supplied by the fact, that if cast-steel from English iron which
has been rendered perfectly malleable at a red-heat, by the ad-
dition of manganese, is re-melted ivitlwut adding manganese, it
becomes as red-short as ever. According to Parry, the presence of
manganese in sensible quantity in steel tends to render it brittle
when cold.

SILICON. — Silicon in small quantity seems to increase the hard-
ness of steel without taking away from its malleability or tough-
ness when cold, and the presence of a trace of silicon appears to
have the singular effect of preventing that violent evolution of
gas from fluid steel at the moment of solidification, which renders
the ingot so frequently unsound. Steel containing above 0'5 per
cent, of silicon breaks up under the hammer if heated above low
redness.

TITANIUM, &c. — The effect of titanium, tin, arsenic, silver, and
other metals upon steel, is stated to produce increased Itardmss,
but I have not myself made any experiments upon this subject,
nor been able to find very reliable observations of others.

TUNGSTEN has a very remarkable effect upon steel, first ob-
served by Dr. Werner Siemens, in 1853, in increasing its power of
retaining magnetism when hardened. Being specially interested
in this question, I have determined, by careful experiments, the
extent of the increase ; and the practical result is, that whereas
a horse-shoe magnet of ordinary steel, weighing 2 Ibs., is con-

214 THE SCIENTIFIC PAPERS OF

sidered of good quality when it bears seven times its own weight,
and the famous Haarlem magnet of the same weight, supports
about 13 times its weight, I am now able to produce a similar
horse-shoe magnet carrying 20 times its weight, suspended from
its armature.

The chief difficulty besetting experiments on the effects of
these various admixtures upon the quality of steel, consists in
the unavoidable presence of other substances in variable propor-
tions ; but the effects of these other substances could be eliminated
if the experiments were accompanied by exhaustive analyses, and
it is impossible to over-estimate the advantages that would result
from such a course.

PROCESSES. — Dr. Percy, in his truly invaluable metallurgical
work, has made us acquainted with the various known processes
for obtaining steel which have been followed from the earliest
times ; but no method of producing steel can be considered
admissible at the present day which does not pass the metal
through the condition of entire liquefaction, for it is only by
fusion that foreign admixtures can be thoroughly separated, and
that flaws and fissures can be avoided ; inasmuch, however, as the
steel obtained by tiiem may be subsequently fused, I shall briefly
refer to them. The principal processes of this class are : —

1. The direct process of making steel, or steely iron, from the
ore in the Catalan forge, by employing a large excess of charcoal.
The steel is obtained without fusion in the form of a ball of
spongy metal, and is drawn out into bars.

2. The cementation process, in which bar-iron is converted into
steel by prolonged contact at a comparatively low temperature
either with liquid cast-iron, as formerly practised in Styria and
elsewhere,* or with crushed charcoal — a method extensively em-
ployed in Sheffield at the present day for the manufacture of
steel for railway and carriage springs, and for the production of
bar-steel from Swedish iron, which, when subsequently melted in
pots, makes the finest quality of steel for tools and cutlery.

3. The decarburization process, in which steel is made from
cast-iron by the removal of part of the carbon in the puddling

* Percy, Metallurgy, II. 790, 807.

A7A' WILLIAM .S7A-.J//:.\.s, l-.R.S. 215

furnace, or by the partial application of any of the older methods
nf decarburetting cast iron in an open In-aith in direct contact
\\ith the fuel. A modified form of the decarhuri/ution process has
••'•(vntly introduced to some extent by Messrs. Heaton and
Hariri-raves. They remove the carbon of the pig-iron by the
action of oxydizing salts, principally nitrate of soda, and either
fuse the crude metal obtained, producing cast steel, or work it up
into blooms under the hammer.

The Bessemer process, now so well known, is also a method of
producing steel by the decarbonization of cast iron : but it
presents a vast advantage over those just named, in the fact that
the steel is obtained in the liquid state, and may be cast, free
from flaws, into homogeneous ingots of any size. As Bessemer
steel is produced much more cheaply than the cast steel obtained
by any other process yet extensively in use, its introduction has
opened out a vast field for the application of steel where iron
alone could formerly be thought of. Thus many parts of steam
and other machinery, railway plant, boiler-plates, and even rails,
are now made of Bessemer steel, and are found to be cheaper in
the end when made of that material rather than when made of
iron, although their absolute cost is fully twice as great.

CAST STEEL. — The remaining methods of producing cast steel,
are those in which it is obtained by the fusion either of steel
already made by other processes, or of its component materials,
iron or iron-ore, on the one hand, and carbon, either as charcoal
or already combined in the form of cast iron, on the other.

The oldest known steel of high quality, the Indian Wootz, is
obtained by the fusion of compact iron and carbonaceous sub-
stances in small crucibles, but it is only in the course of the last
century that the manufacture of cast-steel was first introduced in
Europe.

Reaumur states, in his work on the conversion of forged iron
into steel, published in 1722, that he had succeeded in producing
steel by the fusion together of cast and wrought iron in a common
forge ; but steel melting was first practically carried out in the
latter half of the century by Huntsman, of Sheffield ; the process
he employed, consisting in " the fusion in closed crucibles of steel
already made by cementation in cJiarcoal" is still universally in use
for the production of the finest qualities of steel for tools and

2l6 THE SCIENTIFIC PAPERS OF

cutlery. The Hindoo process has, however, been revived within
the last twenty years, by Heath, Price and Nicholson, Gentle
Brown, Attwood, and others, and large quantities of steel are now
made in Sheffield and elsewhere, by the fusion of puddled iron or
puddled steel with charcoal, or with pure pig-iron (Acadian or
Spiegeleisen), in such proportions as to form steel of the required
quality. The Uchatius process is somewhat similar to these in
principle ; it consists in effecting the partial decarburization of
granulated pig-iron by fusing it in contact with iron-ore, but the
temper obtained is said to be irregular, and, together with the
process just mentioned, it labours under the disadvantage of
involving the expensive operation of fusion in pots.

OPEN HEARTH. — Some method of effecting the fusion of steel
more cheaply than in crucibles, as well as in larger masses, has
long been a desideratum. Heath, the discoverer of the beneficial
action of manganese, was the first (in 1845) to conceive that cast-
steel might be produced in large quantities by fusing wrought
and cast-iron together upon the open hearth of a reverberatory
furnace. The modus operandi he proposed, consisted in melting
pig-metal in a cupola, and running it into the heated furnace.
The wrought-iron was introduced into another part of the furnace,
forming a bank between the bath of fluid metal and the chimney,
to be there heated by the waste heat of the flame, previously to
its being pushed forward into the liquid in order to be dissolved.
Fearing the effect of the ashes from a common fire-place, Heath
proposed to heat his furnace by jets of gas, and there is every
probability that his experiments would have been crowned with
success, if he had possessed the means of imparting to his flame
the intensity of heat and, at the same time, the absence of cutting
draught, which are essentially necessary. Since the date of
Heath's patent, the fusion of steel in an open furnace has formed
the subject of an extensive series of experiments, by Sudre, in
France. The experiments of M. Sudre were made at the Monta-
taire Iron Works, at the expense of the Emperor of the French,
and were superintended, on his behalf, by three members of the
French Institute ; MM. Sainte-Claire Deville, Treuille de Beau-
lieu, Colonel of Artillery, and Caron, Captain of Artillery, who
have made an able report on the subject, showing that it is just
possible to raise the heat of an ordinary furnace by means of a

A/A' WILLIAM .sy/-;.i//-;.VA, /-'.A-.A. 217

fan -I (last, sufficiently to effect the fusion of tool-steel upon the
op.'ii hearth in protecting the metal by a layer of glass, but that
the rapid destruction of the furnace, the cost of fuel, and other
diilk'iilties attending the operation, were such as to render the
process, commercially, of doubtful value.

RKCKNKKATIVI: (IAS FURNACE. — The regenerative gas furnace
is so manifestly suitable for the operation of melting steel, both in
pots and on the open hearth, that my attention was directed from
the first towards this object. The early experiments conducted
by my brother and myself at Sheffield failed, however, partly on
account of certain irregularities, arising from defects in the furnace
which have since been removed, but chiefly in consequence of the
want of determination on the part of the manufacturers and their
workmen to persevere with us to the attainment of the proposed
results.

ATTWOOD. — In 1862, Mr. Charles Attwood took a licence to
apply the regenerative gas furnace to the melting of steel upon
the open hearth in connection with certain chemical processes or
mixtures of his own. I supplied the design of a furnace which
answered the purpose, except that the quality of steel produced
was not such as Mr. Attwood desired. This circumstance decided
him to carry out his process in closed pots heated in the same
furnace.

LE CHATELIER. — In 1863, my friend, M. Le Chatelier, Inge-
nieur en Chef des Mines, elaborated a process for producing steel
from cast-iron by puddling, and melting the hot puddled blooms
in a bath of cast-iron prepared in a regenerative gas furnace, upon
a bed of 'bauxite, of the following composition : —

Silica 13 to 17 per cent.

Alumina GO to 65 „

Peroxide of iron . . . . 4 to 8 „

Water in combination . . . 15 to 17 ,,

This material presents the advantage of being exceedingly in-
fusible and of containing no materials that could impart hurtful
ingredients to the steel. A furnace of great heating power was
constructed by Messrs. Boigue, Rambour and Co., at their works
near Montlu9on, in France, under my superintendence, and would

2l8 THE SCIENTIFIC PAPERS OF

certainly have accomplished the desired object if the company had
displayed the least determination to succeed. The furnace-bottom
of bauxite did not succeed, as it was not solidified by the heat,
and rose to the surface of the liquid bath, — but this defect was
soon rectified by the substitution of a white sand bottom.
Through some carelessness, however, the covering arch of the
furnace was damaged by excess of heat ; and this slight accident,
which proved nothing except an ample sufficiency of heating
power, sufficed to deter the company from pressing on to the
attainment of that success which was so nearly within their reach.
In the meantime I had granted a licence to Messrs. Eraile and Pierre
Martin, of the Sireuil Works, to melt steel, both in pots and on
the open hearth, and a furnace was erected by them in 1864 which
was chiefly intended for a heating furnace, but was at the same
time constructed of such materials (Dinas brick) and in such a
form as to be also applicable for melting steel.

With this furnace, which was really less suitable than those
previously erected, MM. Martin have succeeded in producing
cast-steel of good quality and of various tempers, ard their pro-
duce was awarded a gold medal at the great French Exhibition of
last year. MM. Martin have since patented various arrangements
of their own, such as the employment of particular fluxes to cover
the surface of the molten metal, the application of a separate
furnace for heating the iron before charging it into the melting
furnace, and the employment of particular brands of cast and
wrought-iron, which may be useful under special circumstances
but which form no essential part of the general solution of the
problem.

Having been so often disappointed by the indifference of
manufacturers and the antagonism of their workmen, I deter-
mined, in 1865, to erect experimental or " Sample Steel Works "
of my own at Birmingham, for the purpose of maturing the
details of these processes, before inviting manufacturers to adopt
them. The first furnace erected at these works, is one for melt-
ing the higher qualities of steel in closed pots, and contains 16
pots of the usual capacity. The second, erected in 1867, is an
open bath furnace, capable of melting a charge of 24 cwt. of steel
every 6 hours. Although these works have been carried on under
every disadvantage, inasmuch as I had to educate a set of men

\M' WILLIAM SIEMENS, F.R.S. 2ig

capable of managing steel furnaces), the result has been most
beneficial, in affording me an opportunity of working out the
details of processes for producing cast-steel from scrap-iron of
tn-dinary quality and also directly from the ore, and in proving
these results to others.

I shall now proceed to describe the construction and working ot
the regenerative gas furnaces (similar to those at Birmingham)
wlurh are now at work, or in course of erection, in this country
and abroad for the production of cast-steel, both by the old method
of fusion in pots, and by the new system of making cast-steel on a
large scale and on an open furnace bed, from scrap-iron and from
the ore.

The regenerative gas furnace consists of two essential parts :

The gas producer, in which the coal or other fuel used is con-
verted into a combustible gas ; and

The furnace, with its " regenerators " or chambers for storing
the waste heat of the flame, and giving it up to the in-coming air
and gas.

Any combustible gas might be burned in the regenerative
furnace ; I have used ordinary lighting gas very successfully on a
small laboratory scale, but it is far too costly to be employed in
larger furnaces, and the only gas generally available is that gene-
rated by the complete volatilisation of coal, wood, or other fuel,
with admission of air in a special " gas producer." Any descrip*
tion of carbonaceous matter may be worked in a suitable gas
producer, and will afford gas sufficiently good for the supply of
even those furnaces in which the highest heat is required. Coal
is the fuel chiefly used for gas furnaces in England ; small coke
has been employed in some cases, as in gas-works, where it is to
be had at a cheap rate ; wood is used in France, Bohemia, and
Spain ; sawdust in Sweden, furnishing gas for welding and other
high-heat furnaces ; lignite in various parts of Germany ; and peat
in Italy and elsewhere ; this last being applicable with the greatest
relative advantage.

Plate 38 represents a gas producer suitable for burning non-
caking slack.

In form it is a rectangular fire-brick chamber, one side of
which, B, is inclined at an angle of from 45° to 60°, and is pro-
vided with a grate, C, at its foot. The fuel is filled in at the top

220 THE SCIENTIFIC PAPERS OF

of the incline at A, and falls in a thick bed upon the grate. Air
is admitted at the grate, and as it rises slowly through the ignited
mass, the carbonic acid, first formed by the combination of the
oxygen with the carbon of the fuel, takes up an additional equiva-
lent of carbon, forming carbonic oxide, which diluted by the inert
nitrogen of the air and by a little unreduced carbonic acid, and
mixed with the gases and vapours distilled from the raw fuel during
its gradual descent towards the grate, is led off by the gas flue to
the furnace. The ashes and clinkers that accumulate on the grate
are removed at intervals of one or two days.

The composition of the gas varies with the nature of the fuel
used, and the management of the gas producer. That of the gas
from the producers at the Plate Glass Works, St. Gobain, France,
burning a mixture of f caking coal and \ non-caking coal, is as
follows, by an analysis dated July, 1865 : —

VOLUMES.

Carbonic oxide . 23"7

Hydrogen . . . . . . . . 8'0

Carburetted hydrogen . . < . .2*2

Carbonic acid 4'1

Nitrogen ..... ... 61'5

Oxygen 0'4

99-9

The trace of oxygen present is no doubt due to carelessness in
collecting the gas, or to the leakage of air into the flue, and
allowing for this, the corrected analysis will stand as under : —

VOLUMES.

Carbonic oxide 24'2 \

Hydrogen 8'2 / 34'6

Carburetted hydrogen . . . 2'2 ]

Carbonic acid . . . . . 4'2

•vr-i. fi . > f G5'4

Nitrogen (>1

lOO'O

Only the first three of these constituents, say 35 per cent, of the
whole, are of any use as fuel, the nitrogen and carbonic acid
present only diluting the gas. It is the presence of this large
proportion of inert gases, which must be heated to the full

\/A> WILLIAM SIEMENS, F.R.S. 221

temperature of the flame, that renders it so difficult to maintain a
high heat by gas of this description burned in the ordinary way.
In u>ing such gas in a regenerative furnace the presence of so
large an amount of nitrogen is not objectionable, as the heat it
ies off' is given up again to the air and gas coming in.

The gas as it passes off from the fuel contains also more or less
aqueous vapour, which is got rid of by cooling it, with some tar and
other impurities, and a small quantity of suspended soot and dust.

Any air drawing in unburned through a hole in the mass of
fuel, reduces the value of the gas, by burning the carbonic oxide
again to carbonic acid. To prevent the indraught of air in this
way at the side of the grate, I have found it very advantageous to
set the side walls of the gas producer back, forming a broad step,
about nine or ten inches above the grate ; any air creeping up
along the wall is thus thrown into the mass of fuel and completely
burned. The effect of this feature in the form of the producer on
the quality of the gas has been very striking.

Three-tenths of the total heat of combustion of solid carbon are
evolved in burning it to carbonic oxide ; but in the gas producer,
a small portion only of "this heat is really lost, because it is in a
great measure taken up and utilized in distilling the tar and
hydrocarbon gases from the raw fuel ; and it may be still further
economised, especially in burning a fuel, such as coke or anthra-
cite, which contains little or no volatile matter, by introducing a
regulated supply of steam with the air entering at the grate. This
is effected very simply by keeping the ash-pit always wet. The
steam is decomposed by the ignited coke, and its constituents,
hydrogen and oxygen, are rearranged as a mixture of hydrogen and
carbonic oxide, with a small variable proportion of carbonic acid.
Each cubic foot of steam produces nearly two cubic feet of the
mixed gases, which, being free from nitrogen have great heating
power and form a valuable addition to the gas. The proportion
of steam that can be advantageously introduced into the gas pro-
ducer, is, however, limited, as it tends to cool the fire, and if this
is at too low a heat, much carbonic acid is produced instead of
carbonic oxide, causing waste of fuel.

From the high temperature of the gas, as it rises from the fuel
(1,000° F. to 1,300° F.), and from its comparatively low specific
gravity, it is considerably lighter than atmospheric air, and

222 THE SCIENTIFIC PAPERS OF

ascends into the upper part of the producer with a slight outward
pressure. It is necessary to maintain this pressure through the
whole length of the gas flue, in order to ensure a free supply of
gas to the furnaces, and to prevent its deterioration in the flue,
through the indraught of air at crevices in the brickwork. The
slight loss of gas by leakage, which results from a pressure in the
flue, is of no moment, as it ceases entirely in the course of a day
or two, when the crevices become closed by tar and soot.

Where the furnace stands so much higher than the gas producer,
that the flue may be made to rise considerably, the required
plenum of pressure is at once obtained ; but more frequently the
furnaces and gas producers are placed nearly on the same level,
and some special arrangement is necessary to maintain the pressure
in the flue. The most simple contrivance for this purpose is the
" elevated cooling tube." The hot gas is carried up by a brick
stack to a height of eight or ten feet above the top of the gas
producer, and is led through a horizontal sheet-iron cooling tube,
J, of not less than 60 square feet of surface per gas producer, from
which it passes down either directly to the furnace, or into an
underground brick flue.

The gas rising from the producer at a temperature of about
1100° F., is cooled as it passes along the overhead tube, and the
descending column is consequently denser and heavier than the
ascending column of the same length, and continually over-balances
it. The system forms, in fact, a syphon in which the two limbs
are of equal length, but the one is filled with a heavier fluid than
the other. The height of cooling tube required to produce as
great a pressure in the flue as would be obtained by placing the
gas producers, say ten feet deeper in the ground, may be readily
calculated. The temperature of the gas as it rises from the pro-
ducers has been taken as 1,100° F., and we may assume that it is
cooled in the overhead tube to 100° F., an extent of cooling very
easily attained. The calculated specific gravity, referred to
hydrogen, of the gas of which I have quoted the analysis, being
13*14, we obtain the following data : —

LB.

Weight of the gas per cube foot at 1,100° F. = -022

100° F. = -061

Weight of atmospheric air per cube foot at 60° F. = -076

S/K WILLIAM SIEMENS, F.R.S. 22$

ami from these we have, on the one hand, the increase of pressure
J.:T loot of height, in a flue rising directly from the gas producer,
= -on; — •( '!'•_' = M).H ll>. per square foot; and on the other
hand, tip- excess of pressure at the foot of the downtake from the
cooling tube, over that at the same level in the flue, leading up
from the gas producer (for each foot in height of the cooling
tub.-) = -OG1 — '022 = -039 Ib. per square foot. The height of
the cooling tube above the level of the flue that will be sufficient
to produce the required pressure, equal to 10 feet of heated gas

column, is therefore '. - . 10 ft. = 13' 10", or say 14 feet.

'OoJ

This method of obtaining a pressure in the gas-flue by cooling
the gas, has been objected to as throwing away heat that might be
employed to more advantage in the furnace, but this is not the
case, because the action of a regenerator is such, that the initial
temperature of the gases to be heated has no effect on the final
temperature, and only renders the cooling of the hotter fluid more
or less complete. The only result, therefore, of working the
furnace with gas of high temperature is to increase the heat of the
waste gases passing off by the chimney flue. The complete cooling
of the gas results, on the other hand, in the great advantage of
condensing the steam that it always carries with it from the gas
producer ; and in the case of iron and steel furnaces, in burning
wet fuel, it is absolutely necessary to cool the gas very thoroughly,
in order to get rid of the large amount of steam that it contains
which, if allowed to pass on to the furnace would oxidise the
metal.

There is, undoubtedly, a certain waste of heat, which might
be utilized by surrounding the cooling tube with a boiler, or by
otherwise economising the heat it gives off, as, for instance, in
drying the fuel ; but the saving to be effected is not very great,
for as 100 volumes of the gas require for combustion about 130
volumes of air, including 20 per cent, above that theoretically
required, the heat given off in cooling the gas 1,000° is no more
than would be lost in discharging the products of the complete
combustion of the fuel, at a temperature 435° in excess of the
actual temperature of 200°, and this loss is greatly diminished if a
richer gas is obtained.

In erecting a number of gas producers and furnaces, I generally

224 THE SCIENTIFIC PAPERS OF

prefer to group the producers together, leading the gas from all
into one main flue, from which the several furnaces draw their
supplies. The advantages of this are saving of labour and con-
venience of management, from the gas producers being all close
together, and greater regularity in working, as the furnaces are
seldom all shut off at once ; nor is it likely that all will require at
the same time an exceptional amount of gas.

From the fact that the gas producers may be at any distance
from the furnaces that they supply, if they are only at a lower
level, it would be perfectly practicable to erect them in the very
coal mine itself, burning the slack and waste coal in situ (in place
of leaving it in the workings as is now often done), and distribut-
ing the gas by culverts to the works in the neighbourhood, instead
of carrying the coal to the different works and establishing special
gas producers at each. In rising to the mouth of the pit, the gas
would acquire sufficient pressure to send it through several miles
of culvert.

In the regenerative furnace the gas and air employed are sepa-
rately heated by the waste heat of the flame, by means of what are
termed "regenerators," placed beneath the furnace. These are
four chambers, filled with fire-bricks, stacked loosely together, so
as to expose as much surface as possible ; the waste gases from the
flame are drawn down through two of the regenerators, and heat-
ing the upper rows of bricks to a temperature little short of that
in the furnace itself, pass successively over cooler and cooler sur-
faces and escape, at length, to the chimney flue nearly cold. The
current of hot gases is continued down through these two regene-
rators until a considerable depth of brickwork, near the top, is
uniformly heated to a temperature nearly equal to that of the
entering gas, the heat of the lower portion decreasing gradually
downwards, at a rate depending on the velocity of the current,
and the size and arrangement of the bricks. The direction of the
draught is then reversed ; the current of flame or hot waste gases
is employed to heat up the second pair of regenerators ; and the
gas and air entering the furnace are passed in the opposite direc-
tion through the first pair, and coming into contact, in the first
instance, with the cooler brickwork below, are gradually heated as
they ascend, until, at some distance from the top, they attain a
temperature nearly equal to the initial heat of the waste gases,

.s/A' WILLIAM SIEMENS, F.R.S. 22$

ami, passing up into the furnace, meet and at once ignite, pro-
'1 iii-ing a strong flame, which, after passing through the heating-
chamber, is drawn down through the second pair of regenerators
to the chimney-flue. The temperature attained by the ascending
gas and air remains nearly constant, until the uppermost courses
of the regenerator brickwork begin sensibly to cool ; but by this
time the other two regenerators are sufficiently heated, and the
draught is again reversed, the stream of waste gases being turned
down through the first pair of regenerators, re-heating them in
turn, and the gas and air which enter the furnace being passed up
the second.

By thus reversing the direction of the draught at regular inter-
vals, nearly all the heat is retained in the furnace that would
othenvise be carried off by the products of combustion, the tem-
perature in the chimney-flue rarely exceeding 300° Fahr., whatever
may be the heat in the furnace. The proportion of heat carried
off in an ordinary furnace by the products of combustion is gene-
rally far greater than that which can be utilized, as all the heat of
the flame below the temperature of the work to be heated is abso-
lutely lost. The economy of fuel effected in the regenerative gas
furnace, by removing this source of loss, and making all the heat
of the waste gases, however low its intensity, contribute to raise
the temperature of the flame, amounts in average practice to fully
50 per cent, on the quantity used in an ordinary furnace, and the
saving is greater the higher the heat at which the furnace is
worked. In addition to this economy in the amount of fuel used,
a much cheaper quality may generally be burned in the gas pro-
ducer than could be used in a furnace working at the same heat,
and in which the fuel is burned directly upon the grate in the
ordinary way.

When the heat of the furnace is not abstracted continually by
cold materials charged into it, the temperature necessarily increases
after each reversal, as only a very small fraction of the heat gene-
rated is carried off by the waste gases. The gas and air, in rising
through the regenerators, are heated to a temperature nearly equal
to that at which the flame had been passing down, and when they
meet and burn in the furnace the heat of combustion is added to
that carried up from the regenerators, and the flame is necessarily
lotter than before, and raises the second pair of regenerators to

VOL. I. Q

226 THE SCIENTIFIC PAPERS OF

a higher heat. On again reversing, this higher heat is communi-
cated to the gas and air passing in, and a still hotter flame is the
result.

The temperature that may be attained in this way by the
gradual accumulation of heat in the furnace and in the upper
part of the regenerators appears to be quite unlimited, and the
heat at which a suitably designed furnace can be worked is
limited in practice only by the difficulty of finding a material
sufficiently refractory of which it can be built.

Welsh Dinas brick, consisting of nearly pure silica,* is the only
material, of those practically available on a large scale, that I have
found to resist the intense heat at which steel-melting furnaces
are worked ; but though it withstands perfectly the temperature
required for the fusion of the mildest steel, even this is melted
easily if the furnace is pushed to a still higher heat.

As the gas flame is quite free from the suspended dust which
is always carried over from the fuel by the keen draught of an
ordinary furnace, the brickwork exposed to it is not fluxed on the
surface and gradually cut away, but fails, if at all, only from abso-
lute softening and fusion throughout its mass. A Stourbridge
brick, for example, exposed for a few hours to the heat of the
steel-melting furnace, remains quite sharp on the edges, and is
little altered even in colour ; but it is so thoroughly softened by
the intense heat, that on attempting to take it out,, the tongs
press into it and almost meet, and it is often pulled in two, the
half-fused material drawing out in long strings. It results from
this perfect purity of the flame, that where the heat is not suffi-
cient to effect the absolute fusion of the bricks employed, the
length of time is almost unlimited during which a gas furnace
will work without repairs.

* Analysis of Dinas " clay " from Pont-Neath, Vaughan, Vale of Neath (Percy's
Metallurgy, Vol. I. p. 237).

Silica .

Alumina .

Protoxide of iron

Lime

Potass and soda

Water combined

98-31
0-72
0-18
0-22
0-14
0-35

99-92
The " clay " is mixed with 1 per cen\ of lime in making the bricks.

SfK WILLIAAf SIEMENS, F.R.S.

227

Another advantage in employing the fuel in the manageable
f<>rm of gas is that the rate of combustion may be regulated at
ire to produce an active heating flame of any length, from
little more than two feet, as in the pot steel-melting furnaces, to
thirty feet in the largest furnaces for the fusion of plate glass ;
and the most intense heat may be thrown exactly upon the charge,
the ends of the furnace and the apertures through which the gas
and air are introduced being actually protected from the heat by
the currents of unburned and comparatively cool gases flowing
through them, and only mixing and burning at the very point at
\\hirh the heat is required, and where it is taken up at once by the
materials to be fused or heated. This is of especial importance in
the case of those furnaces in which a very intense heat is employed.

The amount of brickwork required in the regenerators to absorb
the waste heat of a given furnace is a matter of simple calculation.
The products of the complete combustion of one pound of coal
have a capacity for heat equal to that of nearly 17 pounds of fire-
brick,* and (in reversing every hour) 17 pounds of regenerator

* Taking the analysis by Vaux of the celebrated ten-yard coal of South Stafford-
shire (Watts' Dictionary of Chemistry, i. 1081), the exact calculation is as
follows : —

Composition of the coal.
Carbon . 7857

Hydrogen
Sulphur

Nitrogen

( ixy-
Ash

•0529
•0039

Oxygen required.
2-0952
0-4232
0-0039

•0184

•1288 . . less . .

•0103

net oxygen required

1 -0000 20 per cent, excess

Total Oxygen

Corresponding Nitrogen . . 9 '6 16
Nitrogen in the fuel . . -018

2-5223
0-1288

2-3935
0-4787

2-8721

Total Nitrogen . . 9 -634

Oases produced from 1 Ib.

of coal.

Carbonic acid = 2'881

Water (Steam) = 0'476

Sulphurous acid = 0'004

Oxygen in excess = 0 "479

Nitrogen = 9 '634

Specific heats.

•217
•480
•154
•218
•244

Equivalent weight

of water.

•625

•228

•001

•104

2-350

Total equivalent weight of water .... 3 '308
,, ,, firebrick (sp. heat = 0-2). 16-540

Q 2

228 THE SCIENTIFIC PAPERS OF

brickwork at each end of the furnace per pound of coal burned in
the gas-producer per hour would be theoretically sufficient to
absorb the waste heat, if the whole mass of the regenerator were
uniformly heated at each reversal to the full temperature of the
flame, and then completely cooled by the gases coming in ;
but in practice by far the larger part of the depth of regenerator
chequer-work is required to effect the gradual cooling of the
products of combustion, and only a small portion near the
top, perhaps a fourth of the whole mass, is heated uniformly
to the full temperature of the flame ; the heat of the lower
portion decreasing gradually downwards nearly to the bottom.
Three or four times as much brickwork is thus required in the
regenerators, as is equal in capacity for heat to the products of
combustion.

The best size and arrangement of the bricks is determined by
the consideration of the extent of opening required between them
to give a free passage to the air and gas, and by the rule, deduced
from my experiments on the action of regenerators in 1851-2,*
that a surface of six square feet is necessary in the regenerator to
take up the heat of the products of combustion of one pound of
coal in an hour.

By placing the regenerators vertically and heating them from
the top, the heating and cooling actions are made much more
uniform throughout than when the draught is in any other
direction, as the hot descending current on the one hand passes
down most freely through the coolest part of the mass, while the
ascending current of air or gas to be heated, rises chiefly through
that part which happens to be hottest, and cools it to an equality
with the rest.

The regenerators should be always at a lower level than the
heating chamber ; as the gas and air are then forced into the
furnace by the draught of the heated regenerators, and it may be
worked to its full power, either with an outward pressure in the
heating chamber, so that the flame blows out on opening the doors
or with the pressure in the chamber just balanced, the flame some-
times blowing out a little, and sometimes drawing in. The out-
ward pressure of the flame prevents that chilling of the furnace,

* Proceedings of the Institution of Civil Engineers, 1852-3, page 571, " On
the Conversion of Heat into Mechanical Effect. "

SfX WILLIAM SIEMENS, F.R.S. 22$

and injury to the brickwork, from the in-draught of cold air
through crevices, which is otherwise unavoidable in any furnace
worked without blast.

The action of the furnace is regulated by the chimney damper,
and by valves governing the supply of gas and air, and the draught
is reversed by cast-iron reversing valves, on the principle of the
common four-way cock.

FUSION OF STEEL IN CRUCIBLES. — In the application of the
system to the fusion of steel in closed pots or crucibles, the
melting-chamber, containing generally 24 pots, is constructed in
the form of a long trench, 8 feet 6 inches wide at the bottom, and
gathered in to under 2 feet at the top. The sides of the melting-
chamber are arched both horizontally and vertically, to keep them
from sinking together in working, and the work is strengthened
by cross walls at intervals. The pots are set in a double row along
the centre of the melting-chamber, and the flame passes from
side to side, the gas and air from the regenerators being intro-
duced alternately from one side and from the other, opposite to
each pair of pots. The melting-chamber is closed above by loose
firebrick covers, which are drawn partly off in succession by
means of a lever suspended from a pulley above the furnace, when
the pots are to be charged or drawn out. The pots stand in a bed
of finely-ground coke-dust, resting on iron plates. The coke-dust
burns away only very slowly, if it is made of hard coke and finely
ground, and it presents the great advantage of remaining always
in the form of a loose dry powder, in which the pots stand firmly,
while every other material that I have tried either softens at the
intense heat, or sets after a time into a hard, uneven mass, in
which the pots do not stand well.

The process of melting carried out in this form of gas furnace
is the same in all respects as that in the small air furnaces or
melting-holes fired with coke which are commonly employed, but a
great saving is effected in the cost of fuel, and in the number of
crucibles required.

The ordinary consumption of hard coke, costing 22s. per ton in
Sheffield, is between three and four tons per ton of steel fused,
while in the gas furnace the same work may be done by the
expenditure of 15 to 20 cwt. of common coal slack (worth only
5s. to Ss. per ton), at a cost that is of only 5s. against 75s. per ton

230 THE SCIENTIFIC PAPERS OF

of steel melted. There is a further saving in the number of
crucibles required, as they may be used in the gas furnace four or
five, and sometimes even ten times, while in furnaces heated by
coke, two or three casts are as much as are ever obtained. The
lining of the furnace lasts at least 15 to 20 weeks without repair
(in working day and night), while 4 to 5 weeks is the longest
duration of the ordinary coke-fired holes.

FUSION OF STEEL ON THE OPEN BED.— The furnace employed
for the fusion of steel on the open bed is similar in shape to a
reheating or puddling furnace ; the direction of the flame is from
end to end ; and the regenerators are placed transversely below
ihe bed, which is supported on iron plates, kept cool by a current
of air. The air enters beneath the bed plates in front, and escapes
by two ventilating shafts at the back of the furnace near the
ends. This cooling of the bed is very necessary to keep the slag
or melted metal from finding its way through into the regenerator
chambers. The upper part of the furnace is built entirely of
Dinas brick.

There are three doors in the front of the furnace, one in the
centre immediately over the tap-hole, and two near the bridges,
through which the bed can be repaired when necessary, and ingot
ends or other heavy scraps may be charged in. Sloping shoots
are provided at the back of the furnace, through which long bars,
such as old rails, may be conveniently charged, and beneath these
are openings for charging the pig-iron. The upper end of the
shoots is on a level with an elevated charging platform behind the
furnace.

The bottom of the furnace is formed of silicious sand, which
answers exceedingly well if properly selected and treated.

Instead of putting moist sand into the cold furnace, as is usually
done in preparing the bottoms of furnaces for heating or melting
iron or copper, I dry the sand, and introduce it into the hot
furnace, in layers of about 1" thickness. The heat of the furnace
must be sufficient to fuse the surface of each layer, that is to say,
it must rather exceed a welding heat to begin with, and rise to a
full steel-melting heat at the end of the operation, in order to
impart additional solidity to the uppermost layers. Care must be
taken that the surface of the bath assumes the form of a, shallow
basin, being deepest- near the tap-hole. Some white sands, such as

S/R WILLIAM SIEMENS, F,R.S. 231

from Gornal, near Birmingham, will set under these circum-
stances into a hard impervious crust, capable of surviving
from L'<) to 80 charges of liquid steel, without requiring material
repairs. If no natural sand of proper quality is available, white
sand, such as Fontainebleau sand, may be mixed intimately
with about 25 per cent, of common red sand, to obtain the same
results.

In tapping the furnace, the loose sand near the tapping-hole is
removed, when the lower surface of the hard crust will be reached.
The lowest point of this surface is thereupon pierced by means of a
pointed bar, upon the withdrawal of which the fluid metal runs
out from the hottest and deepest portion of the bath into the ladle
in front of the furnace.

M. Le Chatelier now proposes to mix the natural Bauxite, of
which the bottom of the experimental furnace at the works of
MM. Boignes, Ilambourg and Co., near MontluQon was first made,
with about 1 per cent, of chloride of calcium dissolved in water,
to calcine the mixture, and to form it into moulded masses of
highly refractory material.

A hard bottom being thus prepared, and the heat of the furnace
being raised to whiteness, it is ready to receive the materials to be
melted.

If these materials consist of bar iron, or of old iron and steel
rails, they are cut into lengths of about six feet, and are intro-
duced into the furnace through slanting hoppers from the elevated
platform at the back, so that their ends rest upon the sand bottom
forming the bath.

If the capacity of the furnace is such, that charges of 3 tons can
be formed, about 6 cwt. of grey pig-iron is introduced through
the ports or short hoppers, below the main charging-hoppers
before-mentioned. As soon as a bath of pig metal is formed, the
heated ends of the rails or bars begin to dissolve, causing the bars
gradually to descend. By partially closing the mouths of the
charging-hoppers, a regulated quantity of flame is allowed to
escape from the furnace, in order to heat the descending bars of
metal previous to their entry into the melting-chamber, the
object being to maintain the high temperature of the furnace,
notwithstanding the constant introduction of cold metal. The
escaping products of combustion, which are thus withdrawn from

232 THE SCIENTIFIC PAPERS OF

the regenerators, are a positive gain to the heat of the furnace,
because, having been in contact with comparatively cold metal,
they would be at a heat inferior to that of the upper portions of
the regenerators, and would therefore only lower their tempera-
tures.

As the bars sink in the hoppers by their gravity, they are
followed up by additional bars until the metal charged amounts to
about three tons, all of which will be rendered fluid within about
four hours from the time of commencing the charge. The metallic
bath is tested from time to time by the introduction of a bar
through one of the front doors of the furnace, and if the bath
should become thick before the end of the operation, although the
heat has been maintained, it will be necessary to introduce an
additional quantity of pig-metal. All the metal being liquid, a
sample is taken out by means of a small iron ladle, and plunged
into cold water while still red-hot. In breaking this sample upon
an anvil, the temper and quality of the metal may be fairly judged.
Its fracture should be bright and crystalline, betokening a
very small proportion of carbon (not exceeding '1 per cent.),
and the metal should be tough and malleable, notwithstanding
its sudden refrigeration. From 5 to 8 per cent, of Spiegeleisen
(containing not less than 9 per cent, of manganese) is there-
upon charged through the side openings upon the bank of the
furnace, and allowed to melt down into the bath, which is
then stirred and made ready for tapping in the manner before
described.

The amount of carbon introduced with the Spiegeleisen deter-
mines the temper of the steel produced, the manganese being
necessary to prevent redshortness, unless Swedish or Styrian iron
is used.

When old iron rails or scrap of inferior quality are charged, the
addition of manganese does not suffice to effect the necessary
purification of the steel produced ; but the perfectly liquid con-
dition of the bath, together with the unlimited time available for
chemical reaction, offer extraordinary advantages for the introduc-
tion of such materials as may be found to combine with sulphur,
phosphorus, silicon, or arsenic, which are the usual antagonists
to be dealt with.

The experiments which I have been able to institute in this

WILLIAM SIEMENS, F.R.S. 233

direction, are by no means complete ; — nevertheless, I have ob-
tained most beneficial results from the introduction into the bath
of lithariM', in conjunction with oxidising salts containing strong
bases, such as the alkaline nitrates, chromates, chlorates, stannates,
titanates, &c.

The choice of the reagents and the quantity to be employed
depend, naturally, upon the quality and quantity of objectionable
matter to be removed.

By the aid of the process just described, it will be possible to
convert old iron rails into steel rails of sufficiently good quality at
a cost scarcely exceeding that of re-rolling them into fresh iron
rails. The non-expensive nature of the process may be judged by
the fact that extremely little labour is required in conducting it ;
that the loss of metal does not exceed from 5 to G per cent., and
that from 10 to 12 cwt. of coal suffices to produce a ton of cast
steel.

ORE. — Although I have succeeded in producing malleable steel
from ordinary English iron by this process, it would be unreason-
able to expect steel of really high quality, in using those materials
which are already contaminated in the blast-furnace ; and I am
sanguine in the expectation of producing cast steel superior in
quality, and at a low cost, directly from the better description of
ores, such as the hematites, magnetic oxides, and the spathic
carbonates. My experiments in this direction extend over several
years ; and last year I sent a few bars of steel produced
from hematite ore, to the French Exhibition, which had stood
a high test in Kirkcaldy's machine. A " grand prix " was
awarded for this and other applications of the regenerative gas
furnaces.

Having tried various modifications of the furnace, I have
arrived at a form of apparatus (Plates 89 and 40) not dissimilar
to the one just described.

The furnace and tapping arrangements are, indeed, the same,
except that for the slanting hoppers, vertical hoppers over the
middle of the bath are substituted, in which the ore gradually
descends. Each hopper is formed of a cast-iron pipe, supporting
a clay-pipe which is attached to it by means of a bayonet-joint,
and reaches down into the furnace, while the cast-iron pipe rests
with its flange on the charging platform.

234 -THE SCIENTIFIC PAPERS OF

A fire space is provided surrounding each hopper, through
which flame ascends from the furnace, and is allowed to escape in
regulated quantities near the upper extremity of the retort, the*
object being to heat the latter and the ore contained in it to a red
heat. A wrought-iron pipe descends into each hopper from a
general gas-tube above, through which a current of ordinary pro-
ducer gas is forced in amongst the heated ore. The propulsion of
the gas is effected most conveniently by means of a steam-jei
in the gas-tube leading from the main gas-channel to the top of
the furnace, care being taken to effect a total condensation
of the steam by passing the gas finally through a small
scrubber, in which water trickles over pieces of coke. In this
way the gas is at the same time purified from sulphurous acid,
the sulphur of which might otherwise combine with the reduced
ore.

The furnace is charged in the following manner : —

The hoppers and gas-pipes being placed in position, about \ cwt.
of charcoal is charged through each hopper to form a basis for the
ore with which these are afterwards filled.

About 10 cwt. of pig-metal is charged through the ports at the
back or front of the furnace, which, upon being melted, forms a
metallic bath below the hoppers. In the meantime, the ore in
the lower parts of the hoppers, being heated in an atmosphere of
reducing gas, has become partially reduced into metal sponge,
which, in reaching the metallic bath, is readily dissolved in it,
making room for the descent of the superincumbent ore, which is
likewise reduced in its descent and dissolved in due course, fresh
ore being continually supplied on the charging platform. The
dissolution of the reduced ore proceeds with extraordinary
rapidity, but is practically limited by the time necessary to
effect the reduction of the ore in the hopper which occupies
several hours. It is, however, not essential that the ore should
be thoroughly reduced before reaching the bath, because the
carbon contained in the cast metal serves also to complete the
operation.

I prefer to employ a mixture of hematite and spathic ore,
containing the elements for forming a fusible slag, which will
accumulate on the surface of the metallic bath, and may be from
time to time removed through the centre door. If the ore con-

SfK WILLIAM SIEMENS, F.R.S. 235

any silica, it is necessary to add some lime or other fluxing
mat' rials, but it is desirable to employ ores containing little
gan^ue, in order not to encumber the furnace with slag, reserving
the poorer ores for the blast furnace. The ore should, moreover*
be in pieces ranging from the size of a pea to that of a walnut, in
order to be pervious to the reducing gases. If ores in the form
of powder are employed, it is necessary to mix them with about
10 per cent, by weight of light carbonaceous materials, such as
dry peat, wood, or charcoal.

The metallic bath having sufficiently increased in the course of
from three to four hours, the supply of ore is stopped, and that
contained in the hoppers is allowed to sink. Before the hoppers
are empty, a false cover of cast iron, lined with clay at its under
side, is introduced, being suspended from above by a strong wire,
in order to prevent the access of flame to the interior of the empty
hoppers. Charcoal and ore are filled in upon the top of this false
cover, and, on cutting the wire, afterwards form the commence-
ment of the succeeding charge.

When all the ore has disappeared, the metallic bath is tested
as before described in reference to the melting of scrap. If it
should be partially solidified, cast iron is added to re-establish
complete liquefaction ; but if, on the other hand, the bath con-
tains an excess of carbon, oxidising agents may be added as
before described, in requisite proportion. From 5 to 8 per cent,
of spiegeleisen is then added, and the furnace is tapped as already
described.

The quality of the steel produced is chiefly dependent upon the
quality of the ore, but considering that ores of great freedom from
sulphur, phosphorus, or arsenic can be had in large quantities,
this process contains all the elements for producing steel of high
quality.

Having tried a variety of ores, I do not attach much im-
portance to their precise composition, so long as they are
comparatively free from gaugue, and from sulphur and phos-
phorus, the heat being sufficient to reduce the most refrac-
tory. My experience is, however, as yet limited to experimental
working.

I hoped to have been in a position to have given you the
temperature of this furnace, as determined by an electric resistance

236 THE SCIENTIFIC PAPERS OF

pyrometer, which I have constructed for this purpose, but have
not yet been able to obtain satisfactory results, owing to the
destruction of the coil of platinum wire, which has to be exposed
to the heat. My efforts were baffled moreover by the fact,
interesting in itself, that platinum wire produced by fusion in
Deville's furnace, does not increase in electrical resistance with
increase of temperature in the same ratio as that produced by
the old process, owing probably to the presence of carbon or other
alloy in fractional quantities.

Avoiding the use of fused platinum wire, I have measured
temperatures by electrical resistance up to a full welding heat,
which I estimate at 1,600° C. = 2,900° F. ; and in judging
the heat of the steel-melting furnace by comparison of effects,
I should put it at not less than 2,200° C. = 4,000° F.
The effect of this degree of heat may be judged by the
following : —

An ingot of rather hard cast-steel, weighing 6 cwt., was intro-
duced into the furnace to be incorporated with the bath of steel.
The ingot was nearly cold, and was allowed to remain fifteen
minutes upon the bank before it was pushed into the bath, where
it was completely dissolved in fifteen minutes, the time occupied
in heating and melting the ingot being thirty minutes. A cube
of wrought metal of nearly 8 inches, weighing 130 Ibs., was also
introduced cold into the furnace, and allowed to remain upon
the bank during ten minutes to be heated externally to
whiteness, before being pushed into the metallic bath, when
twelve minutes sufficed to render it completely liquid. It must
be borne in mind that these results are produced without a
strong draught, the flame being indeed so mild, as not to
oxidise the unprotected metal, which can be maintained for
several hours as liquid steel in the furnace without adding carbon
in any form.

It may be matter for surprise, that the material composing the
furnace can be made to resist such a heat, and it must be admitted
that best Dinas brick is the only brick capable of resisting for
four to five weeks, by which time the thickness of the arch is
reduced to about 2 inches, by the absolute fusion of the inner
surface ; but this excessive heat is confined to the heated chamber
only, the regenerators being at such a moderate heat, that the

S/A WILLIAM SIEMENS, F.R.S. 237

chequerwork will stand for months, and the arches for years, of
constant working.

In conclusion, I wish to express to you my sense of the dispro-
portion that exists between the magnitude of the task I have
brought before you, and my ability to accomplish it in all its rami-
fications. It may be granted that the regenerative gas furnace
itself has passed beyond its experimental stage, but much has yet
to be done in working out the applications of the system to the
useful arts, and the modifications and improvements in processes
which it frequently involves.

Much, also, has yet to be done before the method of producing
cast steel, in one direct process from the ore, can take its place as
the recognised system of making steel on a large scale from the
purer iron ores, a position that I firmly believe it will assume,
sooner or later. It is my earnest hope that, in bringing the
subject before your Society, I may induce some of its members to
forward this result, by taking up particular branches of the
scientific inquiries upon which I have only been able to touch
lightly in my present communication. I have much pleasure in
acknowledging the assistance I have received in working out
this subject from Mr. William Hackney, my principal superin-
tendent, and Mr. Willis, the chemist in charge at my experi-
mental works.

ON PUDDLING IRON,
BY C. W. SIEMENS, F.R.S.

[Paper read before the Chemical and Mechanical Sections of tlic British
Association at Norwich, 1868, and ordered to be printed in extenso
among the Reports.] *

NOTWITHSTANDING the recent introduction of cast steel for
structural purposes, the production of wrought iron (and puddled
steel) by the puddling process ranks among the most important
branches of British manufacture, representing an annual produc-

* Excerpt Journal of the British Association, 1868, pp. 58-71.

238

THE SCIENTIFIC PAPERS OF

tion exceeding one and a half millions of tons, and a money value of
about nine millions sterling.

Notwithstanding its great national importance and the in-
teresting chemical problems involved, the puddling process has
received less scientific attention than other processes of more
recent origin and inferior importance, owing probably to the mis-
taken sentiment that a time-honoured practice implies perfect
adaptation of the best means to the end and leaves little scope for
improvement.

The scanty scientific literature on the subject "will be found in
Dr. Percy's important work on iron and steel. Messrs. Grace,
Calvert and Richard Johnson, of Manchester,* have supplied most
valuable information by a series of analyses of the contents of a
puddling furnace, during the different stages of the process. These
prove that the molten pig metal is mixed intimately, in the first
place, either with a molten portion of the oxides (or Fettling) which
form the lining or protecting covering to the cast-iron tray of the
puddling chamber, or with a proportion of oxide of iron in the
form of hammer slag or red ore thrown in expressly with the
charge ; that the silicon is first separated from the iron ; that the
carbon only leaves the iron during the " boil " or period of
ebullition ; and that the sulphur and phosphorus separate last of
all while the metal is " coming to nature."

The investigations by Price and Nicholson, and by M. Lan,
confirm these results, from which Dr. Percy draws some important
general conclusions, which have only to be followed up and supple-

* "Phil. Mag.," Sept. 1857. The following Table, from Messrs. Calvert and
Johnson's Paper, includes the chief results of their investigations : —

Time.

Carbon.

Silicon.

Pig iron charged
Sample No. 1

12
12-40

2-275
2-726

2-720
0-915

2. .

1-0

2-905

0-197

3 . .

1-5

2-444

0-194

4. .

1-20

2-305

0-182

5 . .

1-35

1-647

0-183

6 .

1-40

1-206

0-163

7 . .

1-45

0-963

0-163

8 .

1-50

0-772

0-168

Puddled bar 9

0-296

0-120

Wire iron 10 .

...

0-111

0-088

5Y/? WILLIAM SIEMENS, J-.K.S.

239

mented by some additional chemical facts and observations, in
to render the puddling process perfectly intelligible, and to
into relief the defective manner in which it is at present put
into practice, involving, as it does, great loss of metal, waste of
fuel, and of human labour, and an imperfect separation of the two
hurtful ingredients, sulphur and phosphorus.

SILICON. — In forming (by means of the rabble) an intimate
mechanical mixture between the fluid cast metal and the cinder,
the silicon contained in the iron is brought into intimate contact
with metallic oxide, and is rapidly attacked, being found afterwards
in the cinder in the form of silicic acid (combined with oxide of
iron). The heat of the furnace is always kept low during this stage
of the process, and the flame is maintained as reducing as possible.

CARBON. — The disappearance of the carbon from the metal is
accompanied by the appearance of violent ebullition and the
evolution of carbonic oxide, which rises in innumerable bubbles to
the surface of the bath, and burns (in an ordinary puddling
furnace) with the blue flame peculiar to that gas. In puddling in
a regenerative gas furnace this blue flame cannot be observed,
because the flame of this furnace is strictly neutral, and there is no
free oxygen present to burn the carbonic oxide rising from the fluid
mass, a circumstance which by itself explains the superior results
obtained from the gas furnace.

It is popularly believed that the oxygen acting upon the silicon
and carbon of the metal is derived directly from the flame, which
should, on that account, be made to contain an excess of oxygen,
but the very appearance of the process proves that the combination
between the carbon and oxygen does not take place on the surface,
but throughout the body of the fluid mass, and must be attributed
to reaction of the carbon upon the fluid cinder in separating from
it metallic iron ; while as the removal of the silicon is still more
rapid, and is effected under a reducing flame, there is strong
evidence that it also is oxidized rather by the oxygen of the cinder
than by the flame.*

But it has been argued that, although the reaction takes place
below the surface, the oxygen may, nevertheless, be derived from

* At page 258 is appended a Table showing the comparative quantities of
carbon in various kinds of iron and steel.

240 THE SCIENTIFIC PAPERS OP

the flame, which may oxidize the iron on the surface, forming an
oxide or cinder, which is then transferred to the carbon at the
bottom, in consequence of the general agitation of the mass.

This view I am, however, in a position to disprove by my recent
experience in melting cast steel upon the open flame bed of a
furnace, having invariably observed that no oxidation of the
unprotected fluid metal takes place so long as it contains carbon in
however slight a proportion.

But being desirous to ascertain by positive proof, what is the
behaviour of silicon, and carbon, in fluid cast iron, when contact
with the atmosphere, or the flame of the furnace, is strictly
prevented ; I instituted the following experiment at my Sample
Steel Works, at Birmingham : —

Ten cwts. of Acadian pig metal, and one cwt. of broken glass,
were charged upon the bed of a regenerative gas furnace (usually
employed for melting steel upon the open hearth).

The bed of this furnace was formed of pure silicious sand, and
one object in view was to ascertain whether any reaction takes
place between silica and fluid cast metal, it being generally
supposed that metallic silicon is produced under such circum-
stances by the reducing action of the carbon in the metal upon the
silica or silicates present. The cast metal employed in this
experiment was Acadian pig, containing silicon, 1*5 per cent. ;
carbon, 4*0 per cent.

In the course of an hour, the metal and glass were completely
melted. A sample was taken out, containing silicon, 1-08 per cent. ;

, ( 0'6 per cent, combined carbon,
carbon, 2-90 per cent. | 2.3 ^ ^.^

At the end of the second hour, another sample was taken out
and tested, the result being silicon, '96 per cent. ; carbon, 2*40
per cent., combined.

The physical condition of the metal had now undergone a
decided change. The carbon having wholly combined with the
iron, rendered it extremely hard.

The amount of silicon having steadily diminished, these results
prove that no silicon is taken up ly fluid cast metal in contact with
silica or silicates.

The reduction of the amount of silicon in the metal might be
accounted for by the presence of minute quantities of oxides of

S/A' U'lI.UAM SIEMENS, F.R.S. 241

iron, produced in melting the pig metal, which oxides were now
innvased by the addition of hematite ore in small doses.

At the end of the third hour, another sample was taken,
containing silicon, '76 per cent. ; carbon, 2*4 per cent, combined,
the metal being extremely hard, as before.

Additional doses of red ore were added gradually without
agitating the bath, and the effect upon the fluid metal was
observed from time to time.

At the end of the fifth hour, the samples taken from the fluid
bath assumed a decidedly mild temper ; when the addition of ore
was stopped, and exactly six hours after being charged, the metal
was tapped and run into ingots ; it now contained silicon, 0*046
per cent. ; carbon, '25 per cent., thus both the silicon and the
carbon had been almost entirely removed from the pig metal by
mere contact with metallic oxide under a protecting glass covering.

The quantity of red ore added to the bath amounted to two
hundredweight, and the weight of metal tapped to ten hundred-
weight, five pounds, being slightly in excess of the weight of pig
metal charged.

But the pig metal had contained 1/5 per cent, of silicon, and
4 per cent, of carbon, or a combined total of 5*5 per cent.,
whereas the final metal contained collectively only *296 of silicon
and carbon, showing a gain of metal of 5'5 - '296 = 5*204 per cent.,
or including the five pounds of increased weight, a total gain of
5*7 per cent, of metallic iron.

Supported by these observations, I venture to assert that tJte
removal of the silicon and carbon from the pig iron in the ordinary
inuldlinij or " boiling " process is due entirely to the action of the
fluid ojcide of iron present, and thai an equivalent amount of metallic
iron is reduced and added to the bath, which gain, however, is
generally and unnecessarily lost again in the subsequent stages
of the process. The relative quantity of metal thus produced
from the fluid cinder admits of being accurately determined.

The cinder may be taken to consist of Fe3 0* (this being the
fusible combination of peroxide and protoxide), together with
more or less tribasic silicate (3FeO, SiO3), which may be regarded
as a neutral admixture, not affecting the argument, and silicic
acid or silica is represented by Si O3, from which it follows that
for every four atoms of silicon leaving the metal, nine atoms of
VOL. i. n

242 THE SCIENTIFIC PAPERS OF

metallic iron are set free, and taking the atomic weights of
iron = 28, and of silicon 22'5, it follows that for every 4 x 22-5 =
90'0 grains of silicon abstracted from the metal, 9x28 = 252
grains of metallic iron are liberated from the cinder.

Carbonic oxide, again, being represented by C 0, and the cinder
by Fe3 O4, it follows that for every four atoms of carbon removed
from the metal, three atoms of iron are liberated ; and taking into
account the atomic weights of carbon = G and of iron = 28, it
follows that for every G x 4 = 24 grains of carbon oxidized,
28 x 8 = 84 grains of metallic iron are added to the bath.

Assuming ordinary forge pig, after being remelted in the
puddling furnace, to contain about 3 per cent, of carbon and
2 per cent, of silicon, it follows from the foregoing that in
removing this silicon 'W3 x 2 = 5'6 per cent., and in removing
the carbon ff x 8 = 10' 5 per cent, of metallic iron is added to
the bath, making a total increase of 5'G + 10'5 - 5 = ll'l
per cent., or a charge of 420 Ibs. of forge pig metal, ought to
yield 4G6 Ibs. of wrought metal, whereas from an ordinary
puddling furnace the actual yield would generally amount to only
370 Ibs. (or 12 per cent, less than the charge), showing a difference
of 96 Ibs. between the theoretical and actual yield in each charge.

This difference, amounting to fully 20 per cent., is due to the
enormous waste by oxidation to which the iron is exposed after it
has been " brought to nature " (by the removal of the carbon) ;
when it is in the form of a granular or spongy metallic mass and
during the process of forming it into balls. So great a waste of
metal by oxidation seems at first sight almost incredible, but
considering the extent of surface exposed in the finely divided
puddled mass, it is not at all exceptional, and is in fact almost
unavoidable in a furnace of the ordinary construction, maintained
as a puddling furnace is, at a welding heat. Many attempts have
been made, for example by Chenot, Clay, Renton, and others, to
produce iron directly from the purer ores, by reducing the ore in
the first instance to a metallic sponge, and balling up this sponge,
which is a loose porous mass, somewhat similar to spongy puddled
iron, on the bed of a furnace, but all these attempts have failed,
simply on account of the great waste of iron, a waste amounting
to from 25 to 50 per cent., in balling up the sponge. Indeed, the
loss in an ordinary puddling furnace would probably be greater

S/A' WILLIAM SIEMENS, F.R.S. 243

than -jit per cent, if the metal were not partly protected from the
Hunif by the bath of cinder in which it lies ; for in one instance
in which the cinder accidentally ran out of a puddling furnace
during the balling up of the charge, leaving the iron exposed to
the flame, I found the yield reduced from the average of 413 Ibs.
down to 370 Ibs., showing an increased waste of 43 Ibs., or over
10 per cent., due to the more complete exposure of the metal to
the oxidizing action of the flame.

In order to realize the theoretical result, a sufficient amount of
oxides must have been supplied to effect the oxidation of the
silicon and carbon of the pig iron, and to form a tribasic silicate
of iron (3 Fe 0, Si O3) with the silicic acid produced.

The amount of oxide required may be readily ascertained.

In taking the expression, Fe3 O4, the atomic weight of which is
3 x 28 + 4 x 8 = 116, while that of the three atoms of iron alone is
8 x 28 = 84, it follows that VV x 4G = G3'5 Ibs. of cinder or oxide of
iron are requisite to produce the 4G Ibs. of reduced iron which
were added to the bath. There must, however, remain a sufficient
quantity of fluid cinder in the bath to form, with the silicon
(extracted from the iron), a tribasic silicate of iron, or about
60 Ibs., making in all 1 24 Ibs. of fettling which would have to be
added for each charge, a quantity which is generally exceeded in
practice notwithstanding the inferior results universally obtained.

There remain for our consideration the sulphur and phos-
phorus, which being generally contained in English forge pig
in the proportion of from '2 to 'G per cent, each, can hardly affect
the foregoing quantitative results, although they are of great
importance as affecting the quality of the metal produced.

It has been suggested by Percy that the separation of these
ingredients may be due to liquation. This I understand to mean
that the crystals of metallic iron, which form throughout the
ailing mass when the metal " comes to nature," exclude foreign
substances in the same way that the ice formed upon sea water
excludes the salt, and yields sweet water when re-melted.

According to this view, pig metal of inferior quality will really
yield iron almost chemically pure, to which foreign ingredients
are again added by mechanical admixture with the surrounding
cinder, or semi-reduced metal.

It may be safely inferred that the freedom of the metal from

R 2

244

THE SCIENTIFIC PAPERS OP

impurities thus taken up will mainly depend upon the temperature,
which should be high, in order to ensure the perfect fluidity and
complete separation of the cinder.

The following was the result of an analysis of an inferior English
pig iron before and after being puddled : —

PIG METAL.

Sulphur

Phosphorus . . .
Silicon

Iron and Carbon (by
difference) . . .

•08
1-16
1'97

96-79
100-00

PUDDLED BAB.

Sulphur

Phosphorus .

Silicon

Iron (by difference)

•017

•237

•200

99-546

100-000

showing the extent to which foreign matters are actually removed
by the process of puddling.

These analyses were made a few days since by Mr. A. Willis in
my laboratory at Birmingham.

Led by these chemical considerations, and by practical attention
to the subject, extending over several years, I am brought to the
conclusion that Ihe process of puddling, as practised at present, is
extreme!// "wasteful in iron and fuel, immensely laborious, and
.yielding a metal only imperfectly separated from its impurities.^

How nearly we shall be able to approach the results indicated
by the chemical reasoning here adopted, I am not prepared to say,
but that much can be accomplished by the means actually at our
doors is proved by the result of the working of a puddling fur-
nace erected eighteen months since to my designs by the Bolton
Steel and Iron Company in Lancashire.

This furnace consists of a puddling chamber of very nearly the
ordinary form, which is heated however by means of a regenerative
gas furnace, a system of which the principle is now sufficiently
well established to render a very detailed description here un-
necessary.

The general arrangement of the furnace is shown in the accom-
panying illustrations. It consists of two essential parts :

The gas-producer, in which the coal or other fuel is converted
into a combustible gas ; and

The furnace, with its " regenerators " or chambers for storing

WILLIAM SIEMENS, F.R.S. 245

the waste heat of the flame, and giving it up to the in-coming air
•us.

The gas-producer is shown in Plate 41, Fig. 1 ; it is a rectangular
irk chamber, one side of which, B, is inclined at an angle of
in mi •!."> to GO0, and is provided with a grate, C, at its foot. The
fuel, which may be of any description, such as coal, coke, lignite,
peat, or even sawdust, is filled in through a hopper, A, at the top
of the incline, and falls in a thick bed upon the grate. Air is
admitted at the grate, and in burning, its oxygen unites with the
carbon of the fuel, forming carbonic acid gas, which rises slowly
through the ignited mass, taking up an additional equivalent of
carbon, and thus forming carbonic oxide. The heat thus produced
distils off carburetted hydrogen and other gases and vapours from
the fuel as it descends gradually towards the grate, and the car-
bonic oxide already named diluted by the inert nitrogen of the
air and by any small quantity of unreduced carbonic acid, and
mixed with these gases and vapours distilled from the raw fuel, is
finally led off by the gas flue to the furnace. The ashes and
clinkers that accumulate in the grate are removed at intervals of
one or two days.

E is a pipe for the purpose of supplying a little water to the
ash pit, to be decomposed as it evaporates and comes in contact
with the incandescent fuel, thus forming some hydrogen and
carl ionic oxide, which serve to enrich the gas ; G is a small plug
hole by which the state of the fire may be inspected, and the fuel
moved by a bar if necessary ; and D is a sliding damper by which
the gas-producer may be shut off at any time from the flue.

It is necessary to maintain a slight outward pressure through
the whole length of the gas flue leading to the furnaces, in order
to prevent the burning of the gas in the flue through the indraught
of air at crevices in the brickwork.

Where the furnaces stand much higher than the gas-producers,
the required pressure is at once obtained ; but more frequently
the furnaces and gas-producers are placed nearly on the same level,
and some special arrangement is necessary to maintain the pressure
in the flue. The most simple contrivance for this purpose is the
" elevated cooling tube." The hot gas is carried up by a brick
stack, H, to a height of eight or ten feet above the top of the gas-
producer, and is led through a horizontal sheet-iron cooling tube,

246 THE SCIENTIFIC PAPERS OP

J, (Fig. 1), from which it passes down either directly to the
furnace, or into an underground brick flue.

The gas rising from the producer at a temperature of about
1000° Fahr., is cooled as it passes along the overhead tube, and
the descending column is consequently denser and heavier than
the ascending column of the same length, and continually over-
balances it. The system forms, in fact, a syphon in which the
two limbs are of equal length, but the one is filled with a
heavier gaseous fluid than the other.

In erecting a number of gas-producers and furnaces I generally
prefer to group the producers together, leading the gas from
all into one main flue, from which the several furnaces draw
their supplies.

The puddling furnace, proper, is shown in Plate 41, Fig. 2,
and Plate 42, Figs. 3 and 4.

Fig. 2 is a sectional plan, and Fig. 3 is a front elevation of the
furnace, showing the gas reversing valve and flues in section.

Fig. 4, is a longitudinal section.

The peculiarity of the regenerative gas furnace, as applied either
to puddling, or to any other process in which a high heat is
required, consists in the utilization in the furnace of nearly the
whole of the heat of combustion of the fuel, by heating the
entering gas and air by means of the waste heat of the products
of combustion, after they have left the furnace, and are of no
further use for the operation being carried on. The waste heat is,
so to speak, intercepted on its passage to the chimney, by means
of masses of fire-brick stacked in an open or loose manner in
certain chambers, called " Regenerator chambers," C, E, E1, Cl,
(Fig. 3).

On first lighting the furnace the gas pr.sses in through the gas
regulating valve, B, (Fig. 2), and the gas reversing valve, B1,
and is led into the flue, M, and thence into the bottom of the
regenerator chamber, C, (Fig. 3) ; while the air enters through
a corresponding " air reversing valve," behind the valve B1, and
passes thence through the flue, N, into the regenerator chamber,
E. The currents of gas and air, both quite cold, rise separately
through the regenerator chambers, C and E, and pass up through
the gas and air flues respectively, into the furnace above,
where they meet and are lighted, burning and producing a

.s/A' WILLIAM SIEMENS, F.R.S. 247

moderate heat. The products of combustion pass away through
a similar set of flues at the other end of the furnace into the
regenerator chambers, C1, El, and thence through the flues, M1,
N ' , and through the gas and air reversing valves, into the chimney
flue, 0. The waste heat is thus deposited in the upper courses
of open lire-brick work filling the chambers, C1, E1, so heating
them up, while the lower portion, and the chimney flue, are still
quite cool ; then after about an hour, the reversing valves, B1,
(through which the air and gas are admitted to the furnace) are
reversed, by means of the levers, P, and the air and gas enter
through those regenerator chambers, E1, C1, that have just been
heated by the waste products of combustion, and in passing up
through the open brickwork they become heated, and then on
meeting and entering into combustion in the furnace, D, D, they
produce a very high temperature, probably 500° Fahr. higher than
when admitted cold ; the waste heat from such higher temperature
of combustion heating up the previously cold regenerator chambers,
< . K, to a correspondingly higher heat. After about an hour's
work the reversing valves, B1, are again reversed, and the air and
gas enter the first pair of regenerator chambers, C, E, but which
are now very hot, and therefore the air and gas become very hot,
and enter the furnace in this state, meeting and entering into
combustion, and thus producing a still higher temperature, probably
500° higher still, and again heating the second pair of regenerator
chambers, C1, E1, so much higher, which enables them to again
heat the air and gas to a still higher degree, when the valves, B1,
are again reversed. Thus an accumulation of heat and an acces-
sion of temperature is obtained, step by step, so to speak, until
the furnace is as hot as is required ; for unless cold materials are
put in to be heated, and thus abstract heat, the temperature rises
as long as the furnace holds together, and the supply of gas and
air is continued. The heat is at the same time so thoroughly
abstracted from the products of combustion by the regenerators
that the chimney flue remains always quite cool. The command
of the temperature of the furnace and of the quality of the flame
is rendered complete by means of the gas and air regulating valves
shown at B in Fig. 2, and by the chimney damper. These are
adjusted to any required extent of opening by the notched rods,
Q , R, and S, (Fig. 2,) respectively, so that having the power of

248 THE SCIENTIFIC PAPERS OF

producing as high a temperature as can be desired, there is
also the power of varying it according to the requirements in
each case.

The bed of the furnace, D, D, is of the ordinary construction,
formed of iron plates, and is provided with water bridges at the
ends, as shown, to protect the " fettling," (or oxide of iron used
for lining the furnace) from being melted away. The overflow
from one of the water bridges is led into a sheet-iron tank below
the bed, and then away. The evaporation from this tank keeps
the bottom plates cool and preserves the cinder covering them
from melting off, and the steam is carried away by a draught of
air entering through two holes I, I, (Fig. 3), below the tap hole,
and passing off by small ventilating shafts at the back of the
furnace.

A heating chamber, H, is arranged at each end of the furnace,
in which the charge of pig iron may be heated to redness before
it is introduced into the puddling chamber, D, D.

The advantages of this furnace, for puddling, arc that the heat
can be raised to an almost unlimited degree ; that the flame can
be made at will oxidizing, neutral, or reducing, without interfering
with the temperature ; that in-draughts of air and cutting flames
are avoided ; and that the gas fuel is free from ashes, dust, and
other impurities, which are carried into an ordinary puddling
furnace from the grate. In this last respect the new furnace
presents the same advantages as puddling with wood.

The following tables give the working results which were
obtained from this furnace, as compared with the results obtained
at the same time in an ordinary furnace from the same pig (the
ordinary forge mixture) :

SIR WILLIAM SIEMENS, l-.R.S.

249

TABLE I.
REGENERATIVE GAS FURNACE.

l>:it,-.

Not. of
BMta,

Time
charged.

First Ball
out.

If eta]

charged.

Yield.

1867. FIHST SHIFT.

n.s. iks.

May 27

1

.V2.~,

6-32

410 392

2

8-48

7'BO

433 3'.l<;

3

8-8

'.HI

4:50 410

4

'.)•!.-,

10-7

I:.':.

426

.")

10-20

11-22

126

430

6

11-40

12-46

412 412

SECOND SHIFT.

May 27 1

1-48

2-47

428

410

2

2-:.0

3-47

420

414

3

3-56

(•:,:;

426

418

4

5-0

<;-3

432

417

5

6-5

7-12

425

407

6

7-20

8-15

420 422

THIRD SHIFT.

May 27

1

9-1U

10-15

423 414

2

10-25

11-30

422

412

3

11-3.-)

12-40

420

420

4

12-45

2-0

430

410

5

2-10

3-10

424

418

6

3-16

4-20

420

400

FIEST SHIFT.

May 28

1

5-38

6-45

423

402

2

6-50

8-0

422

400

3

8-6

9-8

430

390

4

9-15

10-28

426

407

5

10-35

11-45

426

420

6

11-68

1-8

430

416

SECOND SHIFT.

May 28

1

2-0

3-1

422

422

2

3-6

4-0

424

415

3

4-5

5-18

423

424

4

B-28

6-27

4i?3

415

5

fi-33

7-46

427

420

6

7-49

8-50

490

406

THIRD SHIFT.

May 28

1 10-0

11-20

420

424

2 11-25

11-33

420

410

3 12-40

1-45

423

412

. 4 1-50

2-58

425

420

5 3-13

4-20

430

418

6 4-30

5-35

422

426

Cwt. qrs. Ibs.

Total charge 130 1 2

„ yield 132 3 7

Being at the rate of . . . 20 2 2

of pig iron per ton of puddled bar.

250

THE SCIENTIFIC PAPERS OF

TABLE II.
ORDINARY FURNACE.

Date.

Time.

Weight of Metal
charged.

Weight of Puddled
Bar produced.

Ibs.

1867,

,C T3

424

May 17

o 2

425

405

«H d,

430

£

430

v £

438

43 If

3

416

-u m

456

§8

oo

410

J3

"*

432

| M .

§

426

pe*« S

M

420

"S °

422

a ^ <M

422

.5 o"1"1

425

« fc1

430

4> 03 0)

450

H

410

1V.D

Mean charge 484

„ yield 426

or 22 cwt. 2 qrs. 20 Ibs. of pig iron per ton of puddled bar.

It will be observed that the ordinary furnace received charges
of 484 Ibs. each, and yielded on an average 426 Ibs., representing
a loss of 12 per cent., whereas the gas furnace received charges
averaging 424 Ibs., and yielded 413 Ibs., representing a loss of less
than 2'6 per cent.

It is important to observe, moreover, that the gas furnace turned
out eighteen heats in three shifts per twenty-four hours, instead of
only twelve heats per twenty-four hours, which was the limit of
production in the ordinary furnace.

This rate of working was attained without the employment of
any arrangement for heating the pig iron before charging it into
the furnace, the heating chambers at the ends not having been
used. The adoption of the plan of heating the metal beforehand,
a system already extensively in use both in this country and on the
Continent, effects a further saving of ten to fifteen minutes in the
time required for working each charge, as well as a considerable
economy in fuel.

The quality of the iron produced from the gas furnace was

A/A' WILLIAM SIEMENS, F.R.S.

251

proved decidedly superior to that from the ordinary furnace, being
what is technically called " best best " in the one, and " best" in
(lit; other case, from the same pig.

The economy of fuel was also greatly in favour of the gas furnace,
but could not be accurately ascertained because some mill furnaces
were worked from the same set of producers. Still, judging from
the experience of several years, in the working of regenerative gas
furnaces as re-heating or mill furnaces, and as glass furnaces, the
saving of fuel in puddling cannot be less than 40 to 50 per cent,
in quantity, while a much cheaper quality may be used.

The consumption of " fettling " was, however, greater in the gas
furnace, and the superior yield was naturally attributed by the forge
managers to that cause, although the writer held a different opinion.

The gas furnace, however, had not been provided with water
bridges, these were subsequently added, and the furnace put to
work again in February last, since which time it has been worked
continuously.

The result of the water bridges has been that the amount of
"fettling" required is reduced to an ordinary proportion, the
average quantity of red ore used being l)2'G Ibs. per charge, be-
sides the usual allowance of bull dog, while the yield per charge of
483*3 Ibs. of grey forge pig has been increased to 485 Ibs. of
puddled bar, as shown by the following return of a series of eighty
consecutive charges in June last :

TABLE 111.
EEGENERATIVE GAS FURNACE.

Date.

No. of
Heats.

Total Charges and Yields. j ^ ^

Ibs. cwt. qrs. Ibs.

Ibs.

June, 1868

80

Pig iron charged .38,668 = 345 1 0

4S.S-3

Puddled bar returned . 38,808 = 346 1 25

186

Red ore for "fettling" 7,406=66 014

92-6

proving that the yield of puddled bar slightly exceeds the charge of
pig metal (representing a saving of fully 12 per cent, over the
ordinary furnace), while the superiority of quality in favour of the
gas furnace is fully maintained.

It is also worthy of remark that these results are obtained regu-

252 THE SCIENTIFIC PAPERS OF

larly by the ordinary puddlers of the works, and that no repairs
have been necessary to the gas puddling furnace since November
last, the roof being reported to be still in excellent condition.

In these investigations I have confined myself to the puddling
of ordinary English forge pig in order to avoid confusion, but it
is self-evident that the same reasoning also applies in a modified
degree to white pig metal or refined metal, the use of which I
should not, however, advocate.

WATER BRIDGES. — Regarding the water bridges, I was desirous to
ascertain the expenditure of heat at which the saving of "fettling"
and greater ease of working was effected. The water passing
through the bridges was accordingly measured by Mr. W. Hackney
(who has also furnished me with the other working data) and found
to amount to 25 Ibs. per minute, heated 40° Fahr. This represents
60,000 units of heat per hour, or a consumption not exceeding 8 Ibs.
to 10 Ibs. of solid fuel per hour, an expenditure very much exceeded
by the advantages obtained where water or cooling cisterns are
available.

The labour of the puddler and of his underhand being very much
shortened and facilitated by means of the furnace, I should strongly
recommend the introduction of three working shifts of 8 hours each
per 24 hours, each shift representing the usual number of heats, by
which arrangement both the employer and the employed would be
materially benefited. The labour of the puddler may be further
reduced with advantage by the introduction of the mechanical
" Rabble " which has already made considerable progress on the
Continent.

By working in this manner, a regenerative gas puddling furnace
of ordinary dimensions would produce an annual yield of about 940
tons of bar iron, of superior quality, from the same weight of grey
pig metal and the ordinary proportion of " fettling."

In conclusion, I may state that a considerable number of these
puddling furnaces have been erected by me abroad, and that, in
this country, they are also being taken up by the Monkbridge
Iron Co., Leeds, and a few other enterprising firms.

The construction of these furnaces has been still further improved
lately by the application of horizontal regenerators, to save deep
excavations, and by other arrangements whereby the first cost is
diminished and the working of the furnace facilitated.

.S/A' WILLIAM SIEMENS, b.R.S.

253

TABLE IV.

K.VTAQE OP CAKUON AND SILICON CONTAINED IN VARIOUS KINDS
or CA-T AND WROUGHT IKON AND STKKL.

Description.

Carlxm.

Silicon.

Authority.

per cent.

per cent.

:i (New Jersey, U.S.) .

6-900

0-100

II«'iiry.

,. (German) . . . .

5-440

0-179

Schafhautl.

„ (MtUen) ....

4-H23

()-!l'.)7

l-'ivMjnius.

i pig iron (Danncmora, Sweden) .

4-809

0-176

Henry.

(iivv pig iron, No. 1 (Tow Law) . .

2-708

4-414

Riley.

Uivy pig iron, No. 1 (Acadian Iron Co.)

3-500

4-840

Tookey.

Grey foundry pig iron, No. 1 (Ne- 1
therton, South Staffordshire) . /

3-07

1-48

( Woolwich
( Arsenal.

Ditto Ditto, No. 2, ditto. . .

3-04

1-27

»

Grey forge pig iron, ditto

3-12

i-i6

»»

Forge pig iron, ditto . . .

3-03

0-83

»

Strong forge pig iron ditto

2-81

0-57

»

Grey pig iron (Dowlais) . . . .

3-14

2-16

Riley.

Mottled ditto, ditto

2-95

1-96

<»

White ditto, ditto ....

2-84

1-21

n

Mottled pig iron (Wellingborough)

2-10

2-11

( Woolwich
\ Arsenal.

White pig iron (Bloenavon) . . .

2-31

Ml

Percy.

Refined iron (Bromford, a. Stafford- \
shire )

3-070

0-630

Dick.

Puddled steel, hard (Kbnigshiitte) . .

1-380

•006

Brauns.

Ditto ditto, mild (South Wales)

•501

•106

Parry.

Cast steel, Wootz

1-34

Henry.

, for flat files.

1-2

A. Willis.

, (Huntsman's) for cutters .

1-0

!

, for chisels ....

•75

. . .

, die steel (welding) . . .

•74

, double shear steel

•7

quarry drills . . . .

•64

masons' tools

•6

• . •

, spades

•32

, railway tyres

•32 to -27

. • .

. rails

•26 to -24

, plates for ships .

•25

Various.

, very mild . . . \
, (melted on open hearth) /

•18

A. Willis.

Hard bar iron (South Wales) . . .

•410

•080

Schafhautl.

„ „ (Kloster, Sweden) .

•386

•252

Henry.

(Russia) . . . .

•340

Trace

»

,, ., },....

•272

•062

>t

Boiler plates (Russell's Hall, South \

•i 'ii i

.111

Staffordshire) . . . . j"

iyu

Jlvv

»

Armour plates (Weardale Iron Co.) 1
too steely J

•170

•110

Percy.

Bar iron (Lb'fsta, Sweden)

087

•115

Henry.

„ (Gysinge, Sweden) .

•087

•066

)>

„ (Osteriiy, Sweden)

•054

•028

ii

Armour plates (Beale & Co.)

•044

•174

Percy.

,. „ (Thames Iron Co.) .

•033

•160

ii

.. „ (Low Moor) .

•016

•122

Tookey.

254 THE SCIENTIFIC PAPERS OF

MR. SIEMENS said he quite agreed with most of the remarks
that had been made.

Mr. Bram well's way of putting the loss of iron was quite correct,
and if ever the puddling process was brought to an absolute state
of perfection the saving would in some cases be 20 per cent, instead
of only 12 per cent, as at present. Undoubtedly what was saved
in weight, was due to the improved chemical action, as named by
Mr. Hawksley and Mr. Mallet, and he was glad to find that his
explanation of the theory of puddling was so readily appreciated,
both by chemists, in Section B, where his paper had also been
read, and by engineers. In working it out he admitted at once
that he had had the benefit of special advantages, by being able
to exclude nearly all free oxygen whilst he was working with a
very high temperature, much higher indeed than usual.

In answer to Mr. Mallet's remark as to the use of the new
furnace for other metals, he might say it was now being used for
zinc, lead, and other ores, with great success.

He was happy to be able to relieve Mr. Jones's mind of the
impression that the furnace cost £3,000 ; the cost of a pair of
puddling furnaces and their gas producers was £450, or £225
per puddling furnace complete.

Caking coal, brown coal, free burning coal, and peat, &c., &c.,
had been in daily use for years in the gas producers with excellent
effect, indeed the arrangement offered a means of utilising for
operations requiring high temperatures, several descriptions of fuel
that had never been used for the purpose before.

He certainly did not heat the boilers in a mill with the waste
heat from the puddling furnaces, simply because he had no " waste
heat " to throw away ; all he made he used, and for the highest
or best paying purposes. With regard to the last objection that
had been raised, he was happy to say he had not found any diffi-
culty with the workmen, they had, in fact, after a few days' work
at the furnace, taken to it very well indeed ; he was sorry to say
that it was with the masters only that he had any trouble, and he
simply mentioned it as the contrary had been stated, but without
the intention of finding fault with ironmasters, who would no
doubt soon see that what was chemically right, and practicable,
must also prove remunerative to themselves.
He would only add one word in explanation of the quality

A/A' WILUAM SIEMENS, F.R.S. 255

obtained, in reference to the remarks made by Mr. Cowper,
Mr. Kolm, and Mr. Hackney ; Mr. Kohn said the silicon might
burn out on the surface of the melted metal, but he thought that
tin- experiments he had produced clearly indicated that both the
silicon and the carbon in the cast metal were oxidized by the
oxygen contained in the fettling. He begged to say that he had
not the slightest idea of suggesting that pig iron containing large
quantities of silicon was preferable ; he preferred the best, as
every ironmaster did, but had given the effects of a furnace on a
bad sample of iron as a severe test of its powers. There was no
doubt that the practical improvement in the quality of the iron
produced in his furnace was due, as Mr. Cowper had stated, partly
to a mechanical cause and partly to a chemical one, for the purer
flame of the furnace did far less injury to the iron, and the higher
temperature at which the metal was balled up without being
burned, caused the cinder to be so much more fluid, as to run out
better from the iron when squeezed, hammered, or rolled, thus
excluding more thoroughly those bad welds, " blacks " and
" greys," so plentiful in ordinary iron.

In the discussion of the Paper, and the Supplementary Paper,

"ON THE FURTHER UTILISATION OF THE WASTE
GAS FROM BLAST FURNACES, AND THE ECONOMY
OF COKE DUE TO INCREASED CAPACITY OF
FURNACE," by MR. CHAELES COCHRANE, of Dudley,

MR. C. W. SIEMENS* said he had listened with great pleasure
to the paper, combining as it did practical experience with theo-
retical considerations ; and as regarded the chemical part of the
question the arguments advanced seemed to be based upon correct

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1869, pp. 33-36, 39, 42-43, 62-68.

256 THE SCIENTIFIC PAPERS OF

chemical proportions, although there was one of the • conclusions
arrived at that he could not agree with.

In reference to the plan of closing in the charging hopper at
the furnace top with the additional covering now described, there
could be no doubt that a saving of gas must be thereby effected ;
and although at first glance it might be thought that the amount
of this saving would not be anything material, as the time that
the furnace top was open on each occasion of charging seemed
only momentary, it appeared from actual observation that this
time of opening really amounted altogether to as much as 1-1 6th
of the whole time of working of the furnace, representing a loss of
l-16th of all the gas evolved from the furnace. At the same time
this loss was attended with the additional objection of interruption
in the flow of gas to the boiler fires and hot-blast stoves where it
was employed for heating purposes ; and it was a decided advan-
tage that these interruptions were now entirely obviated by the
double closing of the furnace top.

With regard to the theoretical minimum consumption of 1\ cwt.
of coke in the blast furnace for the reduction of 1 ton of iron from
the ore, it was certainly true that, with an ironstone containing
40 per cent, of iron, 6| cwt. of carbon would suffice to reduce
1 ton of iron from the ore by combining with the oxygen con-
tained in the ore, and 1 cwt. of carbon in addition would be more
than sufficient for the carbonisation of the iron, making the
1\ cwt. of carbon per ton of iron ; but he did not see that
sufficient heat could then be produced by the combustion of that
reduced quantity of carbon in the furnace. Moreover he did not
consider that it would be possible in practice to attain to this
theoretical minimum consumption of coke by any addition to the
height of the blast furnace. A small additional amount of heat
might indeed be absorbed from the waste gas by raising the
furnaces still higher than at present, but this process could not go
on indefinitely, and there would still be a certain surplus of heat
that would escape ; because the combustion of coke by atmos-
pheric oxygen that was necessary to effect the fusion of a given
quantity of reduced ore at a high temperature must always leave a
certain margin of heat available for reducing and heating the
incoming ore. Moreover as the incoming ore was limited in
quantity and possessed only a limited capacity for heat, it followed

SIR WILLIAM SIEMENS, F.R.S. 257

that even if it were a perfect conductor of heat and presented an
unlimited extent of surface for absorbing the heat from the gas,
thnv would still be a certain amount of heat that would escape at
the furnace top ; and in increasing the height of the furnace, a
neutral point would speedily be reached, beyond which no further
increase of height would produce any further absorption of heat,
for the reason that the quantity of ore charged could not on the
average exceed the quantity of ore smelted during the same
interval of time, and its relative capacity for heat could not be
increased by an augmentation of heat-absorbing surface. The
saving consequent upon increased capacity of the blast furnace
would continue to increase until a certain minimum consumption
of fuel was reached, but this would not be the theoretical minimum
assigned by chemical calculations. The ore could not do more
than take up a certain amount of heat from the gas ; and sup-
posing that its capacity for heat were sufficient to reduce the
temperature of the gas down to 500° or 400°, that would be the
temperature at which the gas must then escape; and no further
saving would be possible. If the relative quantity of ore that
could be charged into the furnace were indefinitely great, then
indeed the escaping gas could be cooled down as low as the
temperature of the external atmosphere ; but inasmuch as the
quantity of ore charged into the furnace top was the same as the
quantity smelted in the lower part of the furnace, this ore could
not absorb more than a certain maximum quantity of heat, and
all the surplus heat in the waste gas must escape. There was also
the further question whether with anything like such a small
proportion of coke a sufficient temperature could be obtained in
the furnace to ensure the perfect fusion of the ore : for this pur-
pose an infinitely high temperature of blast would be requisite,
but with a blast at only 1000° temperature it would certainly be
impossible to attain to anything like the theoretical minimum
consumption, which left no margin at all of extra carbon for pro-
ducing the heat required for smelting the ore ; and it must be
borne in mind that the mere transfer of oxygen from the ore to
the carbon was not attended by any evolution of heat, but by an
absolute absorption of heat.

Mr. C. W. Siemens did not understand how the temperature of
VOL. i. s

258 THE SCIENTIFIC PAPERS OF

the escaping gas could be independent of the rate of driving of
the furnace, unless an excess of heat-absorbing surface had been
provided ; because with a greater rate of driving, a given extent
of surface would then be called upon to absorb a greater quantity
of heat in the same time. When the extent of cooling surface of
the materials was so proportioned as to reduce the temperature of
the escaping gas to a certain degree, short of the ultimate attain-
able point, only a small further quantity of waste heat would be
absorbed by passing a greater quantity of gas over the same
amount of surface in the same time, and the consequence would
be that the gas would escape hotter ; just as, in the case of a
steam boiler, a small extent of boiler surface caused a hot chimney.
This question however was independent of the chemical advantages
that might be obtained by hard driving.

Mr. C. "W. Siemens thought the illustration suggested of the
firebrick regenerator in the hot-blast stoves was not a parallel
case, because the area of cooling surface of the firebrick in the
regenerator was practically unlimited, on account of the means
afforded of reversing the direction of the current through the
regenerator as often as desired, whereby the points of maximum
and minimum heat might virtually be separated by any distance
required. There was no means of taking up a greater quantity of
heat from the escaping gas in the blast furnace, because there was
only a limited quantity of material, which was heated by a fixed
quantity of gas ;• for although in any particular quarter of an
hour it might be possible to fill in more material than corre-
sponded to the average rate of smelting, yet in the course of
a week the total quantity of ore put in was exactly the quantity
smelted ; and for absorbing the extra heat generated in smelt-
ing 1 cwt. of ore at the bottom of the furnace it was not
possible by any means whatever to add more than 1 cwt. of
fresh ore at the furnace top, with the proportionate quantity of
the other materials.

Mr. C. W. Siemens remarked that at the previous meeting,
while appreciating the value of the facts brought forward in the
paper then read, he had been unable to agree with the conclusion

WILLIAM SIEMENS, F.R.S. 259

arrived at as to the quantity of coke that should be sufficient for
smelting a ton of iron in a blast furnace having a capacity carried
to the ultimate limits then indicated. The 7i cwts. of coke per
ton of iron, which had been originally named as the theoretical
minimum consumption, represented no more than the carbon
necessary for combining chemically with the peroxide of iron or
i ah- i ued ore, and for carbonising the reduced metal so as to bring
it into the state of ordinary pig iron ; and the idea that it would
be theoretically possible, by any increase in the capacity of the
blast furnace, to bring down the consumption of coke to that
amount, arose from the error of taking a simple arithmetical
progression to represent the saving of coke consequent upon
successive additions to the size of the furnace. In the supple-
mentary paper now read however, an amended calculation had
been offered, by taking into account the reduced capital, that is
the reduced quantity of coke in the larger furnaces, upon which
the percentage of saving had to be estimated ; so that the same
percentage of saving, out of a total consumption which was
diminished by each successive addition to the capacity of the
furnace, represented a less and less absolute saving as the size of
the furnace was increased. In this way, considering only the
question of abstracting the heat from the escaping gas, the result
now arrived at had been that 17'9 cwts. of coke per ton of iron
was the limit of minimum consumption that could be attained by
simply adding to the height of the furnace, ftb account however
had been taken of the chemical work that had to be constantly
performed within the blast furnace ; and moreover the additional
quantity of heat that could be taken up from the gas by merely
adding more material in the furnace top would be found, he
considered, to be a very small amount. For although he had
shown that, by means of the regenerative system, with a quantity
of cooling material relatively very much smaller than was con-
tained in a blast-furnace top, it was quite practicable to take up
the heat from the outgoing current down to within 20° or 30° of
the incoming draught, yet this was only the case so long as the
two opposite currents were equal in the amount of their total
capacity for heat ; but if the current passing in one direction
through the regenerator had a different total capacity for heat
from that passing in the opposite direction, then whatever extent

s 2

260 THE SCIENTIFIC PAPERS OF

was given to the cooling material or regenerator, equilibrium
would never be established, and it would not be possible to cool
down the escaping current to the temperature of the entering
draught ; or, applying this argument to the case of the blast
furnace, it would not be possible to cool down the escaping gas to
the temperature at which the fresh materials were charged into
the furnace, owing to the inferior total capacity for heat of the
latter.

With regard to the amount of chemical work which was required
to be performed in a blast furnace for smelting a ton of iron, he
had made a calculation, the results of which it might be interest-
ing to compare with those that had been arrived at by adapting
to the Cleveland furnaces the investigations of the foreign writers
referred to, whose researches upon the subject he had not previ-
ously been acquainted with. In the case of Cleveland ironstone
containing 40 per cent, of metallic iron in the calcined ore, the
iron being in the state of peroxide had 17 per cent, of oxygen
combined with it, and the remaining 43 per cent, of the ore would
be silica. To fuse this 43 per cent, of silica, which was equal to
21| cwts. per ton of iron made, 7 cwts. of calcined limestone
would theoretically be required, but in reality he believed 8 or 9
cwts. of limestone was the quantity generally added in the Cleve-
land district ; these would take up about 2| per cent, of iron,
giving a total of about 31 cwts. of cinder per ton of iron made.
The 17 per cent, of oxygen combined with the 40 per cent, of
iron in the ore was equivalent to 8'50 cwts. of oxygen per ton of
iron, and this would require 5 '31 cwts. of carbon to produce the
mixture of f ths carbonic oxide and ^th carbonic acid which he
believed actually resulted from the chemical reactions in the
furnace ; for although it had been calculated in the paper that the
proportions in which these two gases passed off at the furnace top
were about frds carbonic oxide to |rd carbonic acid, he thought
the difference should be allowed for the amount of carbonic acid
evolved from the limestone, and that there would therefore be no
material error in assuming the result of the combustion to be -f-ths
carbonic oxide and ^-th carbonic acid.

The heat necessary to be produced in the blast furnace was
therefore that required for performing the three distinct opera-
tions of melting the iron itself, melting the cinder accompanying

S/A WILLIAM SIEMENS, F.R.S. 26 1

the iron, and generating into a gaseous form the fixed oxygen in
the ore and the fixed carbon in the coke, so as to allow of their
chemical combination taking place. Considering first the heat
n in i red for melting the iron — taking the total heat developed in
funning carbonic oxide as 4000 Fahr. units, and in forming car-
bonic acid 14,000 units — the combustion of 1 Ib. of carbon into
the mixture of -Jths carbonic oxide and |th carbonic acid would
develope 6000 Fahr. units of heat ; and therefore— taking the
specific heat of iron at 0*12 in comparison with water, and
assuming the temperature to which the melted metal was heated
(including its latent heat) to amount to 4000° Fahr. — it would be

„ . . , 20x4000x0-12

found that to melt 1 ton of iron required — 1*60

6000

cwts. of carbon. This estimate was evidently on the safe side, as
no account was taken in it of the further heat necessary for keep-
ing the iron melted in the hearth till casting time, which would
have raised the quantity of carbon nearer to the estimate of
2*70 cwts. given by Mr. Bell. The 1*60 cwts, received a satis-
factory confirmation from actual results which he had obtained in
the working of cupolas, showing that the calculation was sub-
stantially correct. Next with regard to the cinder, taking its
specific heat at 0'20, which he believed was below the mark, and
assuming the total heat for melting it to be the same as that for
iron, namely 4000° Fahr., it would be found that the melting
of 31 cwts. of cinder per ton of iron made would require

31 x 4000 x 0-20 = 4.13 cwtg Qf carbon In reference to the third

6000

supply of heat required in the blast furnace — for transforming
into a gaseous state the fixed oxygen in the ore and the fixed
carbon in the coke — it must not be overlooked that in the combi-
nation of carbon with oxygen, heat was only evolved in conse-
quence of a reduction of volume occurring, and in proportion to
the extent of that reduction of volume. But this could only take
place when both bodies were already in a gaseous state ; and
accordingly by the combination of solid carbon with gaseous
oxygen in the proportion of one equivalent of each, producing
carbonic oxide, the amount of sensible heat evolved (4000 units)
•was less than half of that (10,000 units) which was further
developed when a second equivalent of oxygen entered into com-

262 THE SCIENTIFIC PAPERS OF

bination with the gaseous carbonic oxide so as to form carbonic
acid ; because more than half of the total heat developed in the
formation of the carbonic oxide (6000 out of 10,000 units) was
rendered latent by the solid carbon having first to be volatilised
before it could combine with the oxygen. The absorption of
latent heat was necessarily still greater when the fixed oxygen in
the ore had to be made to enter into combination with the fixed
carbon in the coke, because a supply of heat had first to be
absorbed sufficient for volatilising both the fixed oxygen and the
solid carbon ; and he had found practically that, where a mixture
of ore and carbon was exposed in a close retort to the heat of a
furnace, it was necessary to supply a large amount of heat in
order to bring about a combination of the elements. In the com-
bination of the 8'50 cwts. of oxygen of the ore in the blast furnace
with the 5'31 cwts. of carbon of the fuel, two processes of opposite
effect had to take place : on the one hand, the combustion of the
carbon into the final mixture of |-ths carbonic oxide and ^th car-
bonic acid, developing 6000 Fahr. units of heat per Ib. of carbon,
would give a total of 5'31 x 112 x 6000 = 3,568,320 units of heat
developed ; but on the other hand, the volatilisation and separation
from the iron of the fixed oxygen of the ore required the absorp-
tion as latent heat of, he believed, not less than 6000 Fahr. units
of heat per Ib. of oxygen, and the total absorption of heat thus
occasioned would amount to 8'50 x 112 x 6000 = 5,712,000 units
of heat absorbed. The difference 2,143,680 units, being an excess
of heat absorbed, required therefore to be compensated by an extra

consumption of 2'143'68Q = 357 Ibs. or 3'19 cwts. of carbon.
6000

These three amounts added together gave 1'60 + 4'13 + 3'19 =
8'92 cwts. of carbon, to which had to be added the 7 '43 cwts.
required for chemical combination with the oxygen in the ore and
for carbonising the pig metal, making 16'35 cwts. of pure carbon.
A further addition was required of 10 per cent, or 1'63 cwts. for
ashes and water in the coke actually employed in the blast fur-
naces ; and also another 10 per cent, or T63 cwts. for effecting
the calcination of the limestone and completing that of the ore,
and to allow for vaporising the moisture in the materials and for
loss by radiation. The total amount was thus brought up to
19' 61 cwts. as the theoretical minimum consumption of coke

SfK WILLIAM SIEMENS, F.R.^ 263.

necessary for the smelting of ono ton of iron. The complets con-

Kiiiii|itiini was therefore as follows : —

Cwte.

Melting tin; iron 1'Ou

Melting the cinder 4'18

Meat rendered latent by reduction of the ore 3'19
Carbon for combining with oxygen in ore
and for carbonising pig metal . . 7'43

16-35

Ashes and water in the coke, 10 per cent. . . 1'63
Calcining the limestone, &c., 10 per cent. . . l'G3

Total quantity of coke required . . 19-61 cwts.

In this calculation he had left out of consideration on the one
hand the heat that was carried off from the furnace by the escaping
gas, and on the other hand the heat brought into the furnace by
the hot blast entering at 1000° Fahr. ; but on comparing these
two quantities of heat it would be found that they balanced each
other very nearly indeed. For the 8"9 2 cwts. of carbon made up
by the first three items in the foregoing estimate would require,
in order to produce blast-furnace gas consisting of fths carbonic
oxide and |th carbonic acid, 14' 2 7 cwts. of oxygen, which would
be contained in 62*04 cwts. of atmospheric air entering as hot
blast ; while on the other hand the gas escaping at the furnace
top would contain the whole of these 8'92 cwts. of carbon and
62'04 cwts. of air, together with the 8'50 cwts. of oxygen existing
in the ore and the 5'31 cwts. of carbon supplied into the furnace
to combine with it, making altogether 84'77 cwts. of gas, to which
had to be added about 10 per cent, or 8*47 cwts. of carbonic acid
and vapour of water generated from the limestone and other
materials, thus making up a total weight of 93'24 cwts. of gas
escaping from the furnace top, in the form of carbonic oxide,
carbonic acid, and nitrogen mixed together. Assuming the tem-
perature of this escaping gas to be 600° Fahr. and the specific
heat of the mixture to be the same as that of air or 0*267, the
total quantity of heat thus carried off from the furnace would be
93-24 x 112 x 600 x 0'267 = 1,672,949 units of heat per ton of iron
made ; and at the same time the heat brought into the furnace

264 THE SCIENTIFIC PAPERS OF

by the 62'04 cwts. of hot blast at 1000° Fahr. would be 62'04 x
112x1000x0-267 = 1,855,244 units of heat, leaving a small
excess of 182,295 units of heat brought in by the blast per ton of
iron made, which was probably not more than would cover the
loss of heat by radiation from the furnace.

The actual work done in one of the Cleveland blast furnaces
represented therefore a consumption of between 19 and 20 cwts.
of coke per ton of iron made, supposing the ironstone to contain
only 40 per cent, of iron when calcined, and the blast to be
delivered into the furnace at a temperature not exceeding 1000° ;
and under these conditions he believed no blast furnace could be
constructed to work with less than that quantity of coke,
although with richer ores and higher temperatures of blast there
would of course be a considerable margin for further economy in
consumption of coke. Practically speaking he thought that
22 cwts. of coke per ton of iron made would be the best result
that would be obtained under these circumstances ; and this
would be independent of the capacity of the furnace, which might
be increased indefinitely without improving upon this estimated
minimum consumption.

In the discussion of the Paper

" ON THE DEVELOPMENT OF HEAT, AND ITS APPRO-
PRIATION IN BLAST FURNACES OF DIFFERENT
DIMENSIONS," by I. LOWTHIAN BELL,

ME. C. W. SIEMENS,* who was received with applause, said :
I have listened with great interest to this paper, which, founded
on practical information, deals with blast furnaces as an economi-
cal question. I think, in all operations, we should endeavour
to look at the ultimate results we might possibly obtain, and

* Excerpt Journal of the Iron and Steel Institute, 1869-70. pp. 124-125
and 220-222.

WILLIAM SIEMENS, F.K.S.

265

then try, in practice, to approach these results by the best means
we can think of. The question of the chemical economy of the
blast furnace was discussed, as Mr. Bell will remember, at Bir-
mingham, some time ago ; and, in arguing the point, I disagreed
with some of the speakers on the great results which might be
expected from the mere increase of capacity. I could not see that
this capacity, when it had attained a certain limit, could be bene-
ficial as a mere absorber of heat in order to increase materially the
economy of the furnace. I am glad to see Mr. Bell's elaborate
investigations on the subject have caused him to take the view he
has explained to us. At the discussion in Birmingham I ventured
to state the minimum consumption of coke in a blast furnace in
dealing with ore of a certain chemical composition. Mr. Bell, in
giving the results of his investigations, based partly upon theoreti-
cal grounds, comes to a conclusion — and although the two inves-
tigations are totally distinct, for we view the question from different
points of view — the results we arrive at are the same within a
fraction. I came to the conclusion that the minimum quantity in
smelting Cleveland ore would be 21 cwts., and Mr. Bell's calculation
comes to within 1 cwt. of that figure. This coincidence of result
is, I think, the best proof that can be brought forward of the
correctness of that view of the case. At this meeting we have
had the whole question much more thoroughly brought before us,
and I think the paper will remain a standard paper on the subject.
Mr. Bell made one very valuable suggestion with regard to the
carbonization of iron in the blast furnace. He has discovered
that iron ore absorbs carbon, and the amount of absorption pro-
bably rules the " number " of the pig metal produced. That is a
new view of the question and a very interesting one. I, myself,
have great belief in that view ; it will serve to explain many
apparent anomalies. The next question — that of hot blast — I
am not so well prepared to agree with him. No doubt Mr. Bell
has viewed the question very maturely, and I hesitate to differ
from him with regard to results. But still, in viewing the question
from a different point, I would come to a conclusion different
from that of Mr. Bell, and probably, on comparing the ways, the
result that we will arrive at may, perhaps, in some degree modify
the figures placed before us at this meeting. Mr. Bell shows that
by increasing the blast beyond the ordinary limits of 600°, it

266 THE SCIENTIFIC PAPERS OP

affects the economy in a very slight degree. My view is, that ifc
would affect it very nearly in the same degree as an increase from
60° to 000° would affect it. I come to this conclusion for this
reason. I view the general result of combustion in the blast fur-
nace, and compare it with the general result of combustion in the
heating stove. In the heating stove, you burn fuel into carbonic
acid, and for every pound you burn, 13000 Fahr. units are evolved.
This amount of heat absorbed by the air is carried into the blast
furnace. Now, the action of the blast furnace — the combustion
carried on there — is accomplished in a very different way. The
quantity of heat imported into the blast furnace counts for more
than double that expended in the heating stove. Considering
that the capacity and the specific heat of the air at the higher
and lower temperature is the same within a fractional amount,
for every degree Fahr. by which the temperature of blast is
increased, a proportionate amount of stone ought to be taken
into the blast furnace. As Mr. Bell admits that by increasing
the temperature of the blast he gets a higher proportion of carbonic
acid at the top of his blast furnace, he admits that the effect will
be an increase in ratio. In the measure in which you increase the
temperature of the blast, the amount of carbonic acid at the top
of the blast furnace will augment, and the gas at the top of the
blast furnace will diminish in volume. I am satisfied that an
increase in the temperature of the blast must be highly economical
in the results of the furnace.

In the adjourned discussion of tJie same Paper

MR. C. W. SIEMENS said : With regard to the question raised
by Mr. Cochrane, I think I may be able to throw some light upon
it by putting it in this way. If Mr. Cochrane finds the gases
escape at 700° at the top, and if he was then to throw in a load of
ironstone, he would, no doubt, effect a temporary reduction of that
temperature ; but long before a corresponding amount of ore was
smelted to make room for a further charge, the temperature of
the escaping gases would have risen again to 700° or rather more.
In order to arrive at real temperature of issue, the ore should
be fed, or supposed to be fed, in a continuous manner ; although,
for the purpose of experiment, you can put in a heavy charge

A/A' WILLIAM SIEMENS, F.R.S.

267

suddenly and effect a temporary reduction of temperature in that
manner.

With regard to that point of your paper, sir, on which I took the
liberty of differing from you at the last meeting, I must say that
I still hold that the increase of temperature in the blast would
produce a beneficial result, and the beneficial result would rise
fully in the proportion of the increase of temperature. I think
the chief point of difference between us on that occasion was,
that you, Mr. Chairman, referred the result to a unit quantity
of ore, whereas, I referred it to a unit quantity of blast ; and I
cannot help thinking that the latter is the correct way of taking
it. The quantity of heat supplied depends upon the temperature
to which we heat the blast ; and if I can show that, for every
unit of heat added to the blast, you produce an equal saving of
heat in the blast furnace, then, I think, it is proved that increase
of temperature benefits you to an unlimited extent. But I would
go farther, and say that the benefit goes on in an increasing
ratio ; and I think we can show it from the facts which you
have brought before us in so complete a manner. Your own
figures prove that the higher the temperature of the blast, the
lower is the temperature at which the gases issue from the top
of the furnace. Therefore, this is no longer a question of a certain
bank of heat which you can draw upon to the extent of its
debit and no more ; but your bank of heat is an increasing one,
because, in the first place, the gases escape at a lower degree of
temperature, and, in the second place, the higher the temperature
of the blast, the greater is the proportion of carbonic acid es-
caping from the furnace ; which signifies that more heat is gene-
rated within the blast furnace, inasmuch as every pound of fuel
burnt into carbonic acid produces, roughly speaking, three times
the number of units than is the case when carbonic oxide is pro-
duced. These are two reasons, therefore, why the bulk of heat is
very much increased by raising the temperature of the blast.

Tlie Chairman. You are now referring to the Consett furnaces.

Mr. Siemens. Yes, and to the other examples that have been given
to the same effect. With regard to the very ingenious argument
which you put before us just now, regarding the negative effect of
increased temperature at the boshes, I am disposed to admit it
fully in the abstract, and I am pleased to see such clear proportions

268 THE SCIENTIFIC PAPERS OF

established, regarding the chemical balance of affinities between
carbonic oxide and carbonic acid, in contact with metallic iron at
different temperatures. I am willing to admit these facts, which
are new to me, to the fullest extent, but I would draw from them
different conclusions, inasmuch as I cannot admit that the tem-
perature in the blast furnace increases necessarily with the tem-
perature of the blast. I believe the temperature of the blast
furnace will, on the contrary, be practically the same, whether
higher or lower temperature of blast is employed, and the reason
why I come to that conclusion is as follows : — If blast of very high
temperature is employed, the oxygen it contains takes up carbon
very readily and is converted into carbonic oxide. Whereas, if a
lower temperature of blast is employed, there will be at once a
large proportion of carbonic acid gas produced. The carbonic
oxide gas would not nearly represent such an augmentation of heat
as would be represented by the mixture of the two gases together,
hence the reduced iron would, in the one case, be surrounded by
an atmosphere of carbonic oxide with nitrogen only ; and in the
other case, with a mixture, containing carbonic acid in variable
quantity, but the very chemical balance just pointed out to us
would tend to equalize the temperature in both cases. These con-
ditions would, therefore, not alter the working of the furnace
below, but the result of the higher temperature of the blast would
make itself felt in the upper portion of the furnace where finally
a higher proportion of carbonic acid gas, is realised when blast of
high temperature is used. When these favourable circumstances
are taken into account, we may expect, I submit, a continuance
of the very beneficial result already realized by further raising
the temperature of the blast, notwithstanding the interesting
circumstances that you have brought before us.

SIR WILLIAM SIEMENS, F.R.S. 269

/,•/ tltf (/isriifision of t/w Paper

"ON THE STRENGTH OF IRON AND STEEL, AND

ON THE DESIGN OF PARTS OF STRUCTURES

WHICH CONSIST OF THOSE MATERIALS,"

By GEORGE BERKLEY, M. Inst. C.E.

MR. C. W. SIEMENS* observed that the paper contained a great
deal of valuable practical information ; but he considered enough
had not been said, on the one important question which had been
raised, viz., whether or not there was a definite limit of elasticity
in metals. It was stated in the paper, that a bar of iron 1 inch
square would elongate with a strain of 3 tons to the inch, and
would, on the load being gradually increased, continue to elongate
up to a strain of 9 tons or 10 tons to the inch, beyond which point
it was known that a more rapid rate of elongation commenced.
These results seemed to imply that there was no definite limit
where permanent elongation was produced, but in following up
this argument the evident conclusion was, that the bar of metal of
1 square inch, if exposed for a great length of time to the moderate
strain of 3 tons, must continue to yield, in the same way as though
its elastic limit was really exceeded, and must ultimately yield to
that force. But all experience went directly against such a suppo-
sition. One of the most perfect experiments that could be tried on
that point might be found in the old cathedrals of Italy, where
heavy chandeliers weighing £ a ton, or more, had been suspended
for hundreds of years, by long rods of not more than f of an inch
in diameter. If any permanent injury was produced by such a
load, it would no longer be safe to go into these buildings ; and
the heavy chandelier in the Pisa cathedral, which swung sug-
gestively above the head of Galileo Galilei, must inevitably have
long since fallen to the ground. He thought practically that no
one had any belief in any such action of time, but he was not
disposed, on the other hand, to condemn such experiments as

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXX. Session 1869-70, pp. 268-209.

270 THE SCIENTIFIC PAPERS OF

those of Hodgkinson and Fairbairn. Possibly it might be the
case that, in subjecting a bar for the first time to a strain, elonga-
tion did take place, which might be explained in this way : it was
conceded that in the manufacture of a bar of iron internal strains
must arise, owing to the different action of the rollers upon
different portions of the material, and such " virgin " bar might
be regarded as consisting of a number of aggregate fibres of metal
strained one against the other, the internal metal, which had been
the hottest, being permanently in tension, and the external metal
in compression. In subjecting such a bar to a moderate strain,
only the internal fibres would be permanently elongated, while
the external metal would be simply released from compressive
strain, after which all parts of the bar would bear the strain
equally, and this equalising of the active forces must necessarily
be accompanied by a moderate stretch. Therefore, in order to
ascertain the true limit of elasticity of a material, the first experi-
ment upon a virgin bar should not be taken, but the strain should
be continued till it had reached a certain moderate limit, when it
should be reduced, and then the same series should be gone
through again. In proceeding thus he thought it would be found
that there was no permanent elongation till the true limit of
elasticity was reached. This true limit of elasticity he regarded
as being a very important point. Mr. Phipps had justly remarked
that, in the long series of experiments on breaking strains,
published by various authors, a conclusion was arrived at, more by
guess than by scientific reasoning, to the effect that a load might
be applied to, say, one-fourth of the breaking strain, with tolerable
safety, without any distinction being made regarding the limit of
elasticity of the particular material under examination. It would
be found, however, that two bars of metal, which might break at
nearly the same point, and which might consist, chemically
speaking, of nearly the same materials, would be capable of
bearing with safety a very different load. The one might begin
to elongate with less than one-half the breaking strain, and the
other would not begin to elongate till two-thirds of the breaking
strain had been reached. He had made some experiments with
cast steel, which, being the most refined and uniform metal, was
best adapted for such experiments, proving these discrepancies
according to the mode of its production ; and he had no doubt

SIR WILLIAM SIEMENS, F,R.S. 271

that similar results would be obtained with iron, though it might
not be so easy always to obtain the same degree of regularity with
that material. He thought that the limit of elasticity of a mate-
rial was of much greater practical importance to engineers than
its ultimate breaking strain ; because a structure was safe if no
part of it was loaded past the point of limit of elasticity, but must
ultimately fail if that limit was exceeded.

In the discussion of the Paper

" ON THE TESTING OF RAILS ; WITH A DESCRIP-
TION OF A MACHINE FOR THE PURPOSE,"

By JAMES PRICE, M. Inst. C.E.

MR. C. "W. SIEMENS* said that having lately made some experi-
ments with steel rails and iron rails, he could bear out the author's
remarks, confirmed already by Mr. Bramwell, as to the great sus-
ceptibility of steel to break under the influence of a shock. The
paper which Mr. Bramwell read before the British Association f
pointed out the reasons why it was so. In the case of an iron rail
there were innumerable fibres capable of yielding individually to
take a greater strain when necessary. Now in the case of steel
which was homogeneous, no such independent fibrous action could
take place, and if one part of the line of strength was broken an
indefinite amount of tearing strain was thrown upon the adjoining
part. In trying steel rails under a blow, he found that where an
entire rail would stand the impact of a ton weight falling 20 feet,
in the case of a rail being supported rigidly between 3 feet bear-
ing, a blow from a fall of only 4 feet of the same weight would
break it at the place where the hole was drilled through the web
of the rail. It followed from this experiment that in drilling or

i

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXXII. Session 1870-71, pp. 206-208.

t Vide Report of British Association, 1859, p. 422.

272 THE SCIENTIFIC PAPERS OF

punching a number of holes through a rail the strength of that
rail was diminished to less than one-fourth, as measured by a
falling weight ; and this fact should lead to the preference being
given to a mode of fastening in which no such holes were made.
It was true that in ordinary use no such working strain as the
weight of a ton falling through a number of feet was put upon a
rail ; but engineers liked to err upon the right side in a question
of safety, and were justified, he thought, in giving the preference
to steel rails of mild or ductile quality ; for if the steel was not of
tough quality, there was danger of its being hardened by acci-
dental exposure to cold during manufacture. Steel containing
more than fV Per cent, of carbon would become very brittle if by
accident it was thrown into water when red hot ; and such a rail
might break with a very slight shock. In France it was the habit
to make steel rails containing ^ per cent, of carbon, and as far as
wearing qualities were concerned French engineers were right ;
but there was some danger in using such rails, particularly if the
continuity of the rail was broken by holes. Therefore he did not
agree with the author in his assertion that the test by blows was
altogether inadmissible : he thought it a desirable test to apply
along with others. The particular test which the author proposed
was valuable in so far as it proved the fair wearing quality of the
material ; but he did not consider it sufficient as a test for the
wearing quality of rails under the most trying circumstances.
Mr. Bramwell had pointed out one of the reasons why it was not
sufficient, inasmuch as it did not provide for the propelling action
of the driving-wheels on the rail. But, he might have added, it
did not provide for the sliding action of the wheels. Upon inclines
the wear and tear of rails was much greater than on a level road ;
and that was owing, not to any increase of weight, but to the
different mode in which that weight passed along the surface of
the rails. If the load merely rolled over the surface, the effect
upon the rails was very different to what it was if the same weight
slid over it, burnishing or trying to dislodge every particle from
its neighbour. The machine therefore appeared to him to be
incomplete, unless some means of testing that sliding action upon
the rail were provided in addition to the rolling action. If
stopping the rotation of the wheels would not lead to the expendi-
ture of too much power, it would provide a means of subjecting

WILLIAM SIEAfENS, F.R.S. 273

rails in a short space of time to a test which, in the ordinary
working of a railway, would only be obtained after some months,
or even years, of trial. But if any railway company were public-
minded enough to devote a length of their main line, where there
was a steep gradient or a terminal station for experimental pur-
poses, placing the same at the disposal of engineers to put down
their rails, such a trial would be as valuable, perhaps, as any that
could be desired.

In the discussion of the Paper

« DESCRIPTION OF TWO BLAST FURNACES ERECTED
IN 1870 AT NEWPORT, NEAR MIDDLESBROUGH,"

By BERNHARD SAMUELSON, M.P., M. Inst. C. E.,

MR. C. W. SIEMENS * observed that the information contained
in the paper and elicited by the discussion went far to show that,
by carrying out a chemical operation on a large scale, the utmost
results which theory indicated could be almost realised. Two
years ago, when the question of large furnaces was first mooted,
he made a calculation of the work done in a blast furnace, and of
the heat that was developed in that blast furnace. By taking into
account, on the one hand, the ore to be heated and reduced, the
pig to be melted, the slag to be melted, and the heat and com-
position of the escaping gases at the temperature at which they
were carried away ; and, on the other hand, the temperature of
the blast and the heat due to its combustion with coke, in pro-
ducing the gaseous mixture escaping from the top of the blast
furnace, he found that 20 cwt. of ordinary coke should suffice
theoretically to smelt a ton of pig iron from Cleveland ironstone
containing 40 per cent, of metallic iron, and that 21 cwt. per ton
of iron was a result that might be practically realised. Assuming

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXXII. Session 1870-71, pp. 361-364.

VOL. I. T

274 THE SCIENTIFIC PAPERS OF

the temperature of the blast at 1000°, the best results at that time
were about 26 cwt., but it struck him that Mr. Samuelson had
obtained the calculated minimum result almost completely ; his
consumption being 20| cwt., and the temperature of blast about
1100°. Mr. Bell arrived at nearly the same figures by another
mode of reasoning, and had followed up this subject very minutely,
by examining into the particular stages of the chemical operation
carried on in the blast furnace, deducing very valuable data for
the metallurgist.

Nevertheless, he would not recommend others to follow Mr.
Samuelson under all circumstances in the construction of a blast
furnace. He was at that moment constructing some blast furnaces
of very different dimensions ; they would only have a cubical
capacity of 10,000 feet instead of 30,000 feet ; for he believed it
would be inexpedient in this case to adopt such extreme dimen-
sions. It would require the peculiar conditions of the Cleveland
iron district in order to do so with advantage ; namely, an excel-
lent hard coke and an ironstone rather strong and comparatively
poor. For richer ore and tender coke he believed the size of blast
furnace such as Mr. Samuelson had described would be inappli-
cable.

With regard to the effect of extreme capacity, he agreed Avith
Mr. Bramwell's views, while at the same time he admitted that
Mr. Forbes's remarks were perfectly correct. But Mr. Bramwell
spoke of the mechanical absorption of heat — the regenerative action
within the blast furnace ; — on the other hand Mr. Forbes spoke
of the chemical action between the gases of the blast furnace and
the ores ; and both these actions were subject very much to the
same laws. To get from the fuel a maximum of effect in a steam
boiler plenty of heating surface was given, as in the case of the
Cornish boilers ; and it was reasonable to suppose that more sur-
face would give superior results, simply because the difference of
temperature between the heat-transmitting current and the heat-
receiving surfaces was brought down to a minimum. Where a
current of gas flamed rapidly among a mass of solid matter,
there must be a considerable difference of temperature between
each particle of the fluid and the solid receiving that heat : where-
as, if the mass was greater, and the current relatively slower, there
must be a more perfect adjustment between the two temperatures,

WILLIAM SIEMENS, F.R.S. 275

giving rise to a more economical result. In the same way the
chniiical action would be more perfect if a gas containing a cer-
tain proportion of carbonic oxide (mixed with carbonic acid and
nitrogen) had plenty of time to communicate its carbon to the
pitrc of ore which was ready to take it in yielding up its oxygen.
If the quantity of material present were infinite a perfect adjust-
ment between these conditions would arise, depending upon the
temperature and the proportional amount of carbonic oxide pre-
sent ; but such a condition of equilibrium could never be obtained.
It was therefore necessary in practice to consider how much
would suffice for the purpose, because other considerations entered
into the subject. If a blast furnace was very large in proportion
to the quantity of work that was to be performed in it, a greater
economy would be attained by a relative increase of first cost.
This was the case in Mr. Samuelson's blast furnaces, where 30,000
cubic feet capacity produced only 500 tons of pig metal per week ;
whereas a furnace with a capacity of 10,000 or 11,000 cubic feet
would produce 300 tons. It was therefore clear that these large
furnaces produced approximately only about one-half the amount
of work per cubic foot of capacity that the smaller furnace did.
The question arose how far was it judicious to expend more
capital to obtain additional saving, if other means might be
resorted to of arriving at the same, or even a greater increase of
economy. This point had not been touched upon in the course of
the discussion. Mr. Bramwell had shown that where there was a
greater outgoing current than a heat-absorbing material to meet
it, an unstable condition or waste of heat must ensue ; and Mr.
Forbes has argued in the same way on chemical grounds. In the
blast furnace the object was to waste as little heat as possible, and
to bring out the gases at the top as cool and as thoroughly de-
prived of their chemical or reducing power as was practicable.
Both these objects were obtained by diminishing the relative
quantity of gases. If, for instance, the relative amount of gases
for a given amount of incoming material could be reduced by
one-half, and if the chemical conditions were the same, it was
clear that the smaller quantity would be more thoroughly deprived
of its heat ; but the smaller quantity would have had time to
divest itself more thoroughly of its chemical action also : there
would be a larger relative amount of carbonic acid gas produced,

T 2

276 THE SCIENTIFIC PAPERS OF

and the coke would be consumed in the furnace to greater advan-
tage. This could be accomplished by a blast of high temperature ;
and in this respect his opinion differed very much from the theory
of Mr. Bell, to the effect that there was a theoretical as well as a
practical limit to the beneficial heat of hot blast. He believed, on
the contrary, the blast could not be made too hot for economical
purposes, and that real progress in iron smelting must henceforth
be looked for chiefly in that direction.

In the discussion of the Paper

" ON THE CONDITIONS WHICH FAVOUR, AND THOSE

WHICH LIMIT, THE ECONOMY OF FUEL IN THE

BLAST FURNACE FOR SMELTING IRON,"

By ISAAC LOWTHIAN BELL, Assoc. Inst. C. E.,

MR. SIEMENS* said that Mr. Bell had watched blast furnaces
more carefully than almost any other metallurgist. Some of the
facts brought out in this paper were extremely interesting ;
amongst others the proof that the reduction of iron ore was a
source of heat instead of, as was more generally supposed, the
cause of loss of heat ; and that reduction was effected at so low a
temperature as 420° Fahr. Mr. Bell had endeavoured to show
that in the blast furnace the different operations, constituting the
process of smelting, might be carried on in such a way as virtually
to complete one process before the other commenced ; and from
this the somewhat astounding conclusion was arrived at, that
nearly all the advantages derived from the hot-air blast might be
realised by simply increasing the size of the furnaces. The argu-
ment was not pushed to the practical conclusion, that the capacity
of the furnace should be increased to the point of realising all the

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XXXIV. Session 1871-72, pp. 133-136.

WILLIAM SIEMENS, F.R.S. 277

advantages of the hot blast, but the permissible temperature was
limited to 960° Fahr., at which point it was considered that the
maximum effect could be had with a furnace of limited capacity.
This argument was based upon the supposition that iron ore was
reduced at 420° Fahr. ; and that the second action, which was that
of carbonic acid splitting up in contact with carbon, could only be
accomplished at a higher temperature. If, therefore, the carbonic
acid had all been split up in contact with carbon, and a zone had
now been reached at which no more unburning of carbonic acid
took place, the whole remaining capacity of the furnace would be
reserved for the final action — that of the reduction of the ore to
the metallic condition, and the production of a large proportion of
carbonic acid gas in the gases leaving the furnace. He thought,
however, that iron ores, as a rule, were not reducible at the
extremely low temperature of 420° Fahr., which was really below
that of incandescence. In dealing with calcined Cleveland ore he
had no wish to doubt the fact, but if it was considered what this
ore consisted of, it would be seen why its reduction and carboniza-
tion took place at that low temperature. In calcining Cleveland
ore 25 per cent, of volatilised matter was evolved ; therefore the
calcined ore must be regarded as a sponge, consisting of extremely
divided particles of iron ore. This sponge being brought into
contact with the reducing gas, the carbonic oxide was mechanically
absorbed in large quantities, and the reduction of the ore facili-
tated. The sponge into which Cleveland ore was converted by
calcination might in effect be compared to platinum sponge,
which, when a current of hydrogen gas was directed upon it,
absorbed that gas so violently as to produce ignition. Again,
when a solid bar of iron was exposed to a dry atmosphere, no
appreciable oxidation would take place in one hundred or even in
one thousand years. There were iron bars which, in o.J Roman
buildings, had been exposed to the atmosphere for two thousand
years and were not oxidised ; but the same iron in a pulverulent
condition would take fire instantly in a room, showing that the
same material would behave differently under similar circum-
stances, according to its aggregate conditions.

He had paid considerable attention to the reduction of iron ores,
and he had found that, whereas hydrated ores or spathose ores
were reducible at a low temperature, the peroxides required a

278 -THE SCIENTIFIC PAPERS OF

higher temperature ; whilst there were certain ores which were
not reducible at a temperature below 1500° or 2000° Fahr.
Those were the dense magnetic oxides which had no gases to give
out, but were perfectly dense, like a piece of glass. Such compact
ores, he believed, were not reduced in the upper part of the blast
furnace in the way supposed — viz., by a current of carbonic acid,
but only in a lower part of the furnace after they had become
liquid. Certain ores, such as Ilmenite or Marbella ore, had been
known to pass through the blast furnace and to come out un-
changed with the slag, notwithstanding the heat through which,
they had passed. He thought if such pieces of ore could tell a
tale, they would speak of actions and reactions going on in the
blast furnace which did not altogether tally with the beautiful
stratifications Mr. Bell had shown. The author's conclusion, that
the hot blast could not be beneficial beyond a certain point, was
mainly based upon the argument, that with the temperature of
the blast the heat towards the top of the furnace was also in-
creased. But the temperature of the blast affected the temperature
at the top of the furnace in the contrary way to what Mr. Bell
had in his argument supposed. If the temperature of the blast
were suddenly raised, then indeed a rise would be perceptible
throughout the furnaces, and the gases would escape at a somewhat
higher temperature. In that case the same amount of oxygen
would be introduced, and the same amount of carbon accompanied
the charge ; therefore there must be the same proportion of car-
bonic oxide and of carbonic acid issuing at the top, with an increase
of temperature resulting from the greater importation of free heat
with the blast. But if at the same moment when the temperature
of the blast was raised, the relative quantity of coke in the furnace
could be diminished, to the point of producing not more than the
requisite amount of heat, then a diminished quantity of products
of combustion would have to impart its heat to the same quantity
of ore ; and inasmuch as the relative quantity of matter to be
heated and smelted would be increased, it followed that the
temperature of the escaping gases must be less than before, and
that there must be advantage in raising the temperature of the
blast.

As to the effect of the increased temperature of the blast upon
the temperature of the crucible of the blast furnace, he had

S//? WILLIAM SIEMENS, r.R.S. 279

his views very fully on a former occasion, when the
author of the present paper argued on the other side of the
question.* He might now state that, in advocating high tempe-
rature of blast, he did not look for any great increase of heat in
the zone of fusion. He looked for the advantages of hot blast to
the greater importation of heat into the whole economy of the
blast furnace, resulting, he believed, necessarily, in great saving
of fuel.

In the statements of results which were frequently put forward,
certain elements were generally wanting ; and the most important
of these was the relative quantity of blast used. If the tempera-
ture of the blast was increased without its being stated whether
the same quantity was injected relatively to the amount of carbon
in the charge, no positive conclusion could be arrived at ; but these
discussions would assume a more definite character, if the amount
of blast was in all cases stated, and if the working of the furnace
was only taken after it had been readjusted after a change of
temperature to a minimum percentage of carbon with the charge.
It could be easily conceived that with too much carbon the blast
would produce an excess of carbonic oxide, and this would issue at
the top of the furnace at a very moderate temperature. It might
be argued, in that case, that the hot blast produced unfavourable
results ; but if an increasing quantity of ore had been mixed with
the coke, and the proportion gradually increased with increased
temperature of blast till a positive minimum, to insure fusion of
the mass, had been reached, the result would have been much more
favourable. The general basis which he went upon was, that if
free heat was imported into the furnace with the blast, that heat
was produced economically in burning fuel into carbonic acid ;
whereas the blast furnace could at most produce carbonic oxide,
with a proportion of, say, |th or £th of carbonic acid. The author
had clearly shown that, in producing carbonic oxide, only about
* th the amount of heat was produced from a given quantity of
coal as when carbonic acid was formed. Therefore every unit of
heat applied to the blast outside the furnace represented really 8
units of heat produced by the combustion of coke within.

* Vide The Iroiv and Steel Institute— Transactions, Vol. I. i>. 222 ; see p. 264,
ante.

280 THE SCIENTIFIC PAPERS OP

In the discussion of the Paper

"ON THE PEACTICAL WORKING OF DANKS'S
ROTARY PUDDLING MACHINE,"

By MB. JOHN LESTEB, Wolverhampton,

ME. SIEMENS * said he congratulated the members of the Iron
and Steel Institute on the very able reports •which they had re-
ceived on the subject of Danks's rotary puddling furnace. He
had listened to these reports with great interest, and it seemed as
though the facts brought out conclusively proved the value of the
apparatus that had been brought before them. That was not the first
attempt at a rotating puddling apparatus, as they well knew. He
had, many years ago, seen the apparatus, erected, at Dowlais
Works, by Mr. Menelaus, and he looked with regret at the
difficulty which prevented its practical success. That difficulty
had, to all appearance, been overcome by Mr. Danks, by the in-
troduction of a judicious fettling material. In the report mention
was made of one or two points which called for remark on his
part. In one of them, by Mr. Jones, a paper "On Puddling
Iron " was mentioned, which he (the speaker) read some years
ago before the British Association, setting forth a theory of
puddling which had been much criticised by practical ironmasters,
but which was very- fully confirmed by the results obtained in the
Danks furnace. The reports in question expressed some surprise
at the greater yields realised in puddling grey iron than in puddling
white iron. Now, he would submit that his paper furnished a
complete explanation of that result. He there endeavoured to
prove that puddling was, strictly speaking, a chemical reaction
between fluid cinder and fluid cast-iron, and he showed that for
every pound of carbon in the metal, 3*5 pounds of metallic iron
have necessarily to be reduced and added to the charge. In like
manner, for every pound of silicon in the metal 2*8 pounds of
metallic iron had to be produced. If, therefore, his theory was

* Excerpt Proceedings of the Iron and Steel Institute, 1872, pp. 293-295 and
314-315.

SIR WILLIAM SIEMENS, F.K.S. 281

correct, it followed that a pig metal, rich in carbon and in silicon,
would give greater yields than a metal containing those foreign
substances in smaller proportion, or that grey pig metal would
produce larger yields than white iron. The report also made
mention of the possible application of this furnace to what was
called the " Siemens-Martin " process. He could not fall in with that
view. It would be very difficult to realise such a temperature in
a rotating furnace as was required for carrying out a process in
which they had five or six tons of almost pure iron in a fluid con-
dition— iron containing about one-tenth per cent, of carbon.
Another fatal circumstance would be that the lining of the Danks
furnace was composed of oxides of metal, and in the presence of
oxides of metal they could not have fluid mild steel. The steel
would immediately part with its small percentage of carbon, and
become wrought metal. Therefore, it could not be applied, he felt
certain, to carrying out that process. But he might remark that
some years ago his attention had been directed towards a
rotating apparatus, not for puddling, but for accomplishing just the
reverse operation — that of reducing oxides into the metallic con-
dition. He had steadily followed out those experiments, and before
long he might have the pleasure of bringing them before the
Institution. Before sitting down he wished to express that he
was entirely satisfied with the able, and evidently strictly impartial,
reports which they had received.

Mr. I. Lotvthian Bell said he should like to put one question to
Mr. Siemens, who mentioned that the richer the pig iron was in
silicon, the better the yields. That, he supposed, he ascribed
to the action of silicon upon the oxide of iron. But he (the
speaker) believed a very general impression prevailed amongst
puddlers that that could only be true to a moderate extent ;
because the silicon, by taking the oxygen, was changed into
silica, and that, not being able to exist in a puddling furnace in
its uncombined state, combined with oxide of iron, and was thereby
a source of waste.

Mr. Siemens in reply stated that with regard to the carbon, the
result had been proved in the most conclusive manner, by ex-
periments which he himself had made in the regenerative gas-
puddling furnace. The richer the pig iron was in carbon, the
greater was the yield of wrought metal produced. The same

282 THE SCIENTIFIC PAPERS OF

chemical reason and the same practical result applied to silicon,
although he would quite admit certain drawbacks, which Mr. Bell
had alluded to, to the presence of much silicon in the pig metal.
The silicon in the pig metal had to be combined with oxide of
iron to form a tribasic slag, and this oxide of iron had to be
supplied by the fettling. Therefore, if they puddled an iron con-
taining much silicon they would require an extra amount of
fettling to dissolve the silicon afterwards in the slag, and in the
absence of this extra supply of fettling the operation would not
progress favourably ; but the chemical reasoning still held good.
They could obtain from pig metal rich in silicon a larger yield
than from the same pig metal containing no silicon, the simple
reason being that the atomic weight of silicon was considerably
less than the atomic weight of iron. In his paper read before the
British Association he had shown that in puddling ordinary forge
pig iron the theoretical increase of weight was about 8 per cent.,
and that he had already obtained weight for weight in practical
puddling — taking the average result of six months' working —
without using more fettling ore than was necessary under all
circumstances to accomplish the operation.

ME. SIEMENS said he had certainly misunderstood the remarks
made by Mr. Jones on the previous day, perhaps it was because
he had not explained them fully. He had understood that
Mr. Jones thought it possible that the Siemens-Martin process of
making steel could be carried out in the Danks furnace. If the
question was one of puddling Cleveland pig so as to make it
suitable for this process, he would not say a word against it, except
that he must see the analysis of the puddled material before he
made up his mind as to its merits. If that analysis were shown to
him he would be able to say at once whether the metal was suit-
able or not for the Siemens-Martin process, because he knew
exactly how much phosphorus and how much sulphur he could do
with ; and if in the first instance Mr. Samuelson had put clearly
to him the question whether he would be able to make steel from
Cleveland pig, he (the speaker) would most decidedly have told
him that he did not believe he could. He thought it could not
be done unless they could show him that they could remove
phosphorus to a much greater extent than the ordinary puddling

S/X WILLIAM SIEMENS, F.R.S. 283

furnace had yet accomplished. With regard to the yield he agreed
entirely with the views that had fallen from Mr. Snelus, and
which were in conformity with the chemical reasoning the speaker
had advanced in his paper on puddling iron. He had showed on
that occasion that he could gain 8 per cent, in puddling Cleveland
pig in a theoretically perfect manner, and he felt glad to see that
results so near perfection had been realised with the Danks furnace.
He accorded with Mr. Danks's view that it was not at the expense
of the fettling ore that they obtained the increase. In the ordinary
furnace it happens that although they had much silicon in the
pig metal, a portion of that silicon was burnt off by the direct
action of the flame, but in a furnace where there was but little
oxidation by flame, very little of the silicon was burnt, and the
great bulk of it was oxidised by chemical re-action with the oxides
of iron present ; but they must have the oxides present, no matter
whether they intended to make iron of them or not, and the
dillerence between economical and wasteful puddling was due to
whether they retained the iron from these oxides, or whether they
wasted it after getting it. In Mr. Danks's furnace, and also in a
regenerative gas furnace, if properly conducted and worked, they
need not waste the iron, and they could then at least get " weight
for weight."

ON SMELTING IRON AND STEEL.
By C. WILLIAM SIEMENS, D.C.L. (Oxon.), F.R.S.*

[A Lecture delivered before the Chemical Society, March 20th, 1873.]

ON the 7th of May, 1868, I had the honour of addressing you
on the subject of " the regenerative gas furnace f as applied to the
manufacture of cast steel," my object being at that time to point
to the important part which that furnace was likely to play in
such metallurgical processes where intense heat is required. At
that time I described a method of producing cast steel upon the
open hearth of a regenerative gas furnace by the dissolution either

* Excerpt Journal of the Chemical Society, 1873, pp. 661-678.
t The joint invention of C. William and F. Siemens.

284 THE SCIENTIFIC PAPERS OF

of scrap iron or of ores in a more or less reduced state in a bath of
intensely heated pig metal. These processes have now received
very considerable practical development at the works of the Lan-
dore Siemens-Steel Company, of Messrs. Vickers and Company of
Sheffield, and at several other works. Two processes are employed
at these works, the Siemens-Martin process, which consists in dis-
solving scrap metal or steel in a bath of pig metal to which spiegel-
eisen is finally added, and the ore-reducing process in which pig
metal and ore in a more or less reduced condition are employed.

The process chiefly employed at the Landore works consists of
introducing on the bed of an intensely heated regenerative gas
furnace, as shown in Plates 43 and 44, about six tons of pig metal,
which may be No. 3 or 4 hematite pig. When a fluid bath has
been formed, oxide of iron (which should by preference have been
melted beforehand with such proportions of lime or other fluxing
materials as to form with the silica in the ore and in the pig metal
a fusible slag) is added, or natural ores may be used in their raw
condition if they contain lime and manganese, as, for example, the
African Mokta ore. When about 30 cwt. of this ore have been
dissolved (with ebullition) in the metallic bath, it is found that a
sample taken from it contains only about O'l per cent, of carbon ;
a point which can easily be detected by the eye of the workmen
owing to a peculiar bright appearance of the sample when chilled
in water and broken by the hammer.

In order not to pass beyond this point, samples are taken out
from time to time during the latter part of the operation, and
such a series of samples, which have been kindly supplied to me
by Messrs. Vickers and Co., together with samples of steel made
by the process, are now placed before you. The requisite point of
decarburisation being reached, the supply of ore must be stopped,
and from 8 to 10 per cent, of ferro-manganese or spiegeleisen
added to the bath. When this has been well incorporated by
stirring, the metal is ready to be tapped into a ladle mounted
upon wheels, which is afterwards propelled into the foundry, and
discharged either into ingot moulds, to be hammered and rolled,
or into dried clay moulds for the production of steel castings.

Considerable difficulty was experienced to find a material to
resist the excessive heats necessary for carrying out this process ;
ordinary Dinas bricks, which are considered the most refractory

SIX WILLIAM SIEMENS, F.R.S. 285

material in general use, would be rapidly melted, but a brick,
specially prepared by crushing pure quartz rock, and mixing it
with not more than '2 per cent, of quicklime to give cohesion,
answers well. The hearth of the furnace is made of white sand
with u small admixture of more fusible fine sand, which mixture
sets exceedingly hard at a steel-melting heat, and possesses the
advantage of combining into a solid mass with fresh materials
introduced between the charges to make up for wear and tear.
The hearth and the furnace-roof, if of the materials just specified,
are very little attacked when the Siemens-Martin process is used,
although the heat must be sufficient to maintain wrought iron
containing only a trace of carbon, in a perfectly fluid condition.
If pig metal and ore (previously fused with the necessary amount
of flux) are used, the furnace also stands well, but the use of raw
ore entails the disadvantage of a more rapid destruction of the
furnace ; even magnetic oxide of the purest description necessitates
the addition of raw lime for the formation of a fusible slag, and
the dust arising from the lime and through the decrepitation of
the ore causes the silica bricks to melt away rapidly, so that after
perhaps two months' usage, the 9-inch arch of the furnace is re-
duced to the thickness of from 1 to 2 inches. It is evident that
silica is chemically speaking an objectionable material to be used
in the construction of these furnaces, because it prevents the for-
mation of basic slags, and that a furnace bed constructed of pure
alumina or lime would be preferable. My friend M. Le Chatelier,
Inspecteur-General des Mines, whose valuable labours for the ad-
vancement of iron metallurgy are well known, suggested to me
years ago the use of Bauxite (from Baux in France, where it was
first discovered), a mineral consisting chiefly of alumina, for
making the furnace-bed, but I was not able to succeed with this
owing to the great contraction of the mass when intensely heated,
and non-cohesion with the same material introduced for the pur-
pose of repair. In attempting to construct the sides and roof of
the furnace of Bauxite bricks, these were not found to be equal in
heat-resisting power to silica bricks, which latter are indeed unob-
jectionable, except when raw ore and limestone are used.

If good pig metal, such as is used in the Bessemer process, is
employed, a metal of high quality is the result, equalling in most
respects the steel produced by melting Swedish bars in pots by the
old Sheffield process.

286

THE SCIENTIFIC PAPERS OF

There is manifestly an analogy between this and the Bessemer
process, both of them being processes of decarburisation of pig
metal ; yet there are very important differences, both as regards
the nature of the chemical reactions and of the metal produced.
In the Bessemer process the silicon and carbon are fully oxidised
by the action of the blast, whereas sulphur and phosphorus are
known to remain unoxidised. Manganese is oxidised in the Bes-
semer process only to a certain extent, and therefore it is not
necessary to add spiegeleisen at the end of the operation, when
the pig metal employed contains a moderate amount of metallic
manganese, as is the case in Sweden and in Styria! Mr. W.
Hackney, the manager, and Mr. A. Willis, the resident chemist
at the Landore Steel Works, have analysed Bessemer metals made
without the final addition of spiegel, and found them to contain
no less than 0*3 per cent, of manganese. Notwithstanding the
non-oxidation of manganese, from 8 to 10 per cent, of iron is
oxidised in the Bessemer process, although iron has naturally less
affinity to oxygen than manganese. The oxidation of this amount
of iron in the Bessemer converter is, in so far, a fortunate circum-
stance, as without it the amount of heat necessary to liquefy the
resulting malleable iron could not be produced, and the metal
would necessarily set in the converter. In the ore-reducing pro-
cess above described a totally different result ensues, as is shown
by the following experiment.

A manganiferous pig was melted, and ore charged in the usual
manner. After the pig was completely melted, it contained the
ingredients stated in the first line of the following table, from
which it will be seen that silicon and manganese were eliminated
first, and then very little change took place in the carbon until the
two other constituents were completely removed : —

•

Contained.

Samples taken.

Carbon.

Silicon.

Manganese.

When pig was melted .

1-90

•57

1-14

1 hour later

1-8

•233

•576

2 hours later .

1-7

•183

•2

3 , „ ...

1-65

•05

•08

4 , „ . ' -

1-6

none

none

o , „ ...

1-1

•6

f Manganese and silicon

7 , „. % .

•2

( completely gone.

.s/A' WILLIAM SIE.MEXS, I'.R.S. 287

Tin- sulphur and phosphorus are oxidised to a considerable
extent; a circumstance which enables us to use pig metal con-
taining a certain small percentage of these admixtures, which are
not absolutely objectionable unless they occur in excessive propor-
tions, the quantity admissible of both phosphorus and sulphur
bt ing about '08 per cent.

The decarburising agent employed being ferric oxide, no cast-
iron can possibly be oxidised, and a ton of pig metal with its
quota of spiegel, containing from 9 to 10 per cent, of foreign
substances, yields fully 21 cwt. of steel ingots. In this process
the addition of spiegeleisen at the end of the operation is an abso-
lute necessity, however much manganese may have been contained
in the pig metal employed, because the manganese is rapidly oxi-
dised, as can be seen from the foregoing table.

The reason of the different reaction in the Bessemer process is,
in my opinion, that the silica liberated from the beginning of the
operation requires a base for its saturation, and that its great
affinity to protoxide of iron determines the oxidation of this metal
(free oxygen from the blast being present) in preference to man-
ganese.

A characteristic difference between the two metals, in their fluid
condition, is that Bessemer metal on cooling in the moulds sets up
a violent ebullition, probably resulting from a reaction between
occluded oxygen and carbon, which is counteracted by stoppering
the moulds. The ore process metal, if it has been made with care,
on the contrary, contains no occluded oxygen, and sinks in the
moulds on cooling in the same way as " dead melted " steel does,
produced by the old Sheffield process.

It has been an open question between metallurgical chemists
whether manganese is really required to be present in malleable
steel, some maintaining that its beneficial action is confined to the
elimination of sulphur during the process of production. Such,
however, is not the case, as may be seen from the following extract
from Messrs. Hackney and Willis's report : —

" Our observation, based on five years' experience, is that no
chemical reaction takes place between the manganese of the
spiegeleisen and the other elements of the steel, but that it acts
simply as an alloy. With regard to manganese not removing any
sulphur in a bath of steel, it is a curious fact that the reverse

288 THE SCIENTIFIC PAPERS OF

takes place in a blast furnace. As the percentage of manganese
in the pig increases, so does the sulphur decrease ; for instance, if
to ore and coke, which in the ordinary way will produce a pig
containing from 2 per cent, to 3 per cent, of sulphur, mangani-
ferous ore is added so as to put 2 per cent, of manganese into the
pig, the sulphur will be reduced to "05 or '08 per cent. ; but when
3 per cent, of manganese is found in the pig it never contains
more than a slight trace of sulphur. "We have had occasion lately
to analyse several hundreds of samples of pig iron made under
these conditions, and can vouch for the accuracy of this state-
ment, and we think it a fact well worth the attention of iron-
masters who are troubled with an excess of sulphur either in their
ores or fuel.

" Every one must have noticed the absence of sulphur in all
analyses of spiegeleisen, although much of it is made from coke,
and most of the ores used contain a considerable quantity of pyrites.

" That the manganese acts by its actual presence, and not by any
chemical reaction, is. we think, proved by the fact that, if through
any accident or carelessness of the workmen, the charge is left in
the furnace for more than twenty minutes after the spiegeleisen is
fairly melted the steel is invariably bad, owing to the oxidation of
the manganese, and therefore if such delay occurs, a small addition
of spiegeleisen is made, and no bad results follow."

About the impurities of steel the same gentlemen report : —

" The presence of sulphur, it is well known, makes steel red-
short, and the only question is, how much may it contain without
injury ? Our experience is, that up to 0'08 per cent, will do no
harm, provided there is also present 0'3 per cent, of manganese,
and we have even found '112 per cent, without having any com-
plaint from the hammermen.

"Phosphorus does not appear to affect either the rolling or
hammering of steel, even if 0'2 per cent, is present, but if it ex-
ceeds 0*08 per cent., the rails become so cold-short that they will
not stand the severe test to which they are subjected.

" The presence of arsenic and phosphorus increases the hardness
of steel at the expense of its toughness.

" The effect of copper is not yet understood ; according to
Percy it renders steel red-short to a greater extent even than the
same amount of sulphur.

.S7A' \\-ll.I.L\M SIEMENS, FJLS. 289

- I-'rom the nature of the process, as a rule, very little if any
silicon is found in Siemens steel. The carbon and silicon appear
to be eliminated equally when manganese is not present, and, as
the carbon is generally in excess of the silicon, none remains when
the carbon is reduced to 0'2 per cent. The small amount found
in the steel is derived from the spiegeleisen."

At the Landore Works upwards of 1,000 tons of cast steel are
produced weekly by these processes, and other works, such as
Vickers and Company, Krupp, of Essen, etc., are using the same
for the production of steel of high quality.

Both in the ore-reducing and in the Siemens-Martin or scrap
process, pig metal forms the principal basis, being used either as
such, or as puddled iron or Bessemer scrap-metal, resulting from
the conversion of pig metal by a previous process.

In my former lecture I expressed my belief that the direct con-
version of ores into iron or steel would ultimately be accomplished,
and, having since been actively engaged on the problem, it is now
my chief object to lay before you the results I have up to this time'
attained.

I am aware that, on the one hand, the direct conversion of iron
ores into wrought iron or steel is no novelty, inasmuch as the
ancient Indians and Romans produced their iron by a direct
process from the ore ; and that, on the other hand, the blast
furnace offers immense facilities for the wholesale extraction of
metal from the ore, by the side of which the puny efforts of the
ancients producing half a cwt. of metal intermixed with half-fused
cinder by a day's toil, and with the expenditure of large quantities
of rich ore and charcoal, sink into utter insignificance.

Mr. Riley says, in his able lecture " On the Manufacture of
Steel and Iron," read before this Society on May 22, 1872 : —

" The great improvement lately introduced in the manufacture
of pig iron, the enormously increased production at a diminished
cost of what may be considered to be the raw material from which
we start in the manufacture of iron and steel, are, I think,
sufficient to convince not only practical men, but scientific men
also, that in any improvement in iron manufacture we must
commence with the pig, and consider that our starting point."

Mr. Riley here gives expression to the prevailing opinion
amongst metallurgists ; but still I do not despair of being able to
VOL. i. v

290 THE SCIENTIFIC PAPERS OF

prove to you that upon theoretical grounds the blast furnace is
open to very grave objections, inasmuch as it produces iron com-
bined with nearly all the objectionable substances contained in
the materials used, that only expensive fuel, such as coke, can be
employed, and that its combustion is necessarily imperfect.

Plate 45 shows the distribution of temperature in the blast
furnace, as given by Mr. Lowthian Bell in a paper read by him
before the Institution of Civil Engineers, in 1872.

It shows that the reduction of the metallic oxides to spongy
iron is accomplished within the first 20 feet of their descent
in the furnace, and at a comparatively low temperature. This
upper zone is followed by one where the limestone is decom-
posed and the carburisation of the spongy metal is commenced.
Between this second zone and the zone of fusion in the boshes of
the furnace a zone of great magnitude intervenes, where apparently
no other change is effected than an increase of temperature of the
spongy metal, but where in reality a very powerful reducing action
is accomplished of substances which had much better not be joined
to the iron. It is well known that almost all the phosphorus con-
tained in the iron-stone, the lime-stone, and the coke is here
incorporated with the spongy iron. The silica is reduced to its
metallic condition, and, together with sulphur, arsenic, and other
bases which may be present, combines with the iron. The final
action in the blast furnace consists only in fusing those reduced
substances and forming the slags which envelop and protect the
fused metal.

As regards the fuel question, it will be observed that the result
of the combustion in the blast furnace is for the most part car-
bonic oxide, and that the heat developed in this combustion
amounts to only 2,400 heat units per pound of pure coke con-
sumed, whereas a perfect combustion (to carbonic acid) of the
same coke would be attended by the development of 8,000 heat
units. It is practically impossible to obtain more than one-fifth
of carbonic acid to the carbonic oxide issuing from the top of the
blast furnace, and taking also into account that the gases leave
the top of the blast furnace at 350° C., it may fairly be asserted
that only one-third of the heat-producing power residing in the
coke is utilised. A portion of the heat thus left undeveloped may
.be used, it is true, in burning the gases for the generation of

WILLIAM SIEMENS, F.R.S. 2QI

steam or for heating the air blown into the blast furnaces ; but
fliis utilization represents only a small proportion of the value of
the coke or charcoal charged with the ore into the furnace ;
whereas a much cheaper material might be employed for the pur-
poses just named. The introduction of hot blast was unquestion-
ably a very great improvement in blast furnace economy, because
the heat thus introduced is obtained by means of the perfect com-
bustion of fuel, and, in reducing the combustion necessary within
the furnace, the quantity of products of combustion as compared
to a unit quantity of ore is greatly reduced, the effect being that
the incoming ores have a relatively greater capacity for absorbing
the sensible heat from the products of combustion, which latter
must therefore issue from the top of the furnace at a lower
temperature.

Relying upon this argument, I am inclined to believe that the
consumption of coke in a blast furnace must materially diminish
with increased temperature of blast without limitation of degree,
and in this respect I venture to differ from my friend Mr.
Lowthian Bell, who maintains that mere increase of capacity of
furnace up to a certain limit produces the same effect as increase of
temperature of blast, and that no beneficial effect can be obtained
by increasing the temperature of blast beyond 515° C. In taking
however, the best examples of blast furnaces, the products of com-
bustion escaping from the top, four parts out of five, as CO with-
out counting the N, and at not less than 350° C. of sensible heat,
carry with them fully two-thirds of the heat they would be capable
of producing, if they were burnt to carbonic acid.

In my former lecture I described a plan of reducing iron oxides
by feeding them, mixed with carbonaceous materials, into a rever-
beratory furnace through inverted hoppers of fire-clay, the inten-
tion being to effect the reduction of the iron ores into spongy
metal during their descent, and the fusion of the spongy metal so
produced on the open hearth of the furnace, pig metal being used
to facilitate the fusion. It was found, however, that the quantity
of heat that had to be transmitted through the sides of the fire-
clay hopper was so great that the process of reduction proceeded
very slowly, and the hoppers themselves were rapidly destroyed by
the intense heat of the furnace. The reduction of the metallic iron
in closed chambers of this or any other form which I have tried

u 2

2Q2 THE SCIENTIFIC PAPERS OP

cannot be effected in less than about 36 hours, and that at a great
expenditure of fuel for heating the chamber externally.

This unsatisfactory result directed my thoughts to another
method of producing spongy iron by means of a rotative furnace.
This furnace consisted of a long cylindrical tube of iron about 8
feet diameter, mounted upon antifriction rollers ; the brick lining
of it was provided with longitudinal passages for heating currents
of air and gas prior to their combustion at the one extremity of
the rotating chamber. The flame produced passed thence to the
opposite or chimney end, where a mixture of crushed ore and car-
bonaceous material was introduced. By the slow rotation of this
furnace the mixture advanced continually to the hotter end of the
chamber, and was gradually reduced to spongy iron. This dropped
through a passage constructed of refractory material on to the
hearth of a steel-melting furnace, where a bath of fluid pig metal
had been provided. The supply of reduced ore was continued till
the carbon in the mixture was reduced to the minimum point
before indicated. The rotation was then arrested to prevent
further descent of reduced ore ; spiegel was added ; and the con-
tents of the melting furnace tapped into a ladle and thence into
ingots, as before described.

This rotary furnace was erected by me at the Landore Works in
18G9, and it was so far successful, that the reduction of the ore
was accomplished in a comparatively short time. A difficulty,
however, presented itself, which led to its immediate abandonment ;
it was found that the spongy metal produced, absorbed sulphur from
the heating gases, and was rendered unfit for the production of
steel ; the spongy iron moreover, upon its introduction into the
steel-melting furnace, floated upon the metallic bath without being
readily absorbed into it, and was in great part reoxidised and con-
verted into slag by the action of the flame in the furnace.

These experiments convinced me that the successful application
of reduced ores could not be accomplished through their conversion
into spongy metal, and fully explained to me the want of success
which has attended the previous efforts of Clay, Chenot, Yates,
and others, to produce iron directly from the ore. On the other
hand, I had observed that in melting iron ores no sulphur was
absorbed from the flame ; and it occurred to me that by melting
ores mixed with fluxing materials in a furnace so arranged as to

\\-JLIJAM SIEMENS, F.R.S.

293

accomplish its fusion in a continuous manner, and on a large scale,
the fused ore might be acted upon by solid carbonaceous matter,
so as to separate the metallic iron in a more compact form, while
the earthy constituents of the ore would form a fusible slag with
the fluxing material. Experiments proved that this reduction by
precipitation of the iron could be accomplished only at an intense
heat, exceeding the welding heat of iron, but that the iron so pro-
duced was almost chemically pure, although the ores and the fuel
used might contain a very considerable percentage of sulphur and
phosphorus. The specimens of iron which I exhibit were made in
this manner from ores of various descriptions, and the following
tables give the analysis of the ores and the iron, as also the yield
of iron obtained from a ton of ore : —

RESULTS OP PRECIPITATION PROCESS AT LANDORE.

Orea.

Charge.

Yield,

Total.

Coal charged
in furnace.

Mokta ore . .12 cwt. )
Scale . . . 8 „ /

20 cwt.

6 cwt.

14 cwt.

Mokta ore . . 10 „ 1
Scale . . . . 8 ., /

18 „

4 „

14 cwt. 2 qrs.

Mokta ore . . 12 „ \
Scale . . . 8 „ /

20 „

6 „

11 „ 2 „

Mokta ore . . 6 „ \
Scale . . . 4 „ /

10 „

2 „

12 „

Mokta ore . . 10 „ \
Scale . . . 10 „• /

20 „

6 „

15 „ 2 „

Mokta ore . . 14 „ 1
Scale . . . 12 „ /

26 „

6 „

18 „

Total ....

114 cwt.

80 „

85 „ 2 „

RESULTS OF ROTATOR.

Ore.

Lime.

Coal.

Iron.

Vickers, Sheffield . .

984

48
40

216

492

.. Birmingham

7S4

176

MM

294 THE SCIENTIFIC PAPERS OF

RESULTS OP PRECIPITATION PROCESS AT BLOCHAIRN.

Charge.

Yield
of iron.

Analysis of iron.
Impurities.

No.

~{
*{

Ores.

Lbs.

Total.

Lime.

Coal.

Purple . .
Red

4501
150 /

600

56

356

137 j

S =-055 per cent.
P=-114 „
Si = -13 .,

Purple . .
Red

168 \
56 /

224

12

170

165

»!

Melted . .
Coke (powder)

3.10 \
210 /

560

160

4i

Purple . .
Blackband

300)

600 /

900

90

390

210

•I

Blackband
Clayband
Purple . .

600 )
600 V
300 )

1500

150

420

210

•i

Blackband
Clayband . .
Purple .

300 )

300 \
300 )

900

90

145

202

•I

Cleveland . .
Purple .

900 \
300 /

1200

120

290

268 ]

S =» -041 per cent.
P=-150 „
Si = -173 „

WORKING RESULTS OF THE CASCADE FURNACE AT LANDORE.

Charge in Cwts.

Yield.

Mokta ore.

Roll scale.

Total ore.

CoaL

Puddled bar.

cwt. qrs.

8

8

16

4

7 2

10

8

18

8

15

6

6

12

3

8

4

8

12

4

11

10

8

18

6

13

6

4

10

2

12

12

8

20

6

14

10

8

18

4

14 2

12

8

20

6

11 2

10

10

20

(,

15 2

14

12

26

6

18

102

88

190 cwt.

140

WILLIAM SIEMENS, F.R.S. 295

The ore employed was Moktaore, of the following description : —

\ Fe 60-8

FcO

Mn304

CaO

SiOa .

Loss on ignition

82 charges, 14,338 Ibs. ore
„ 5,952 „ coal

„ 7G8 ,, lime

7:i-7 I
6-48
52-92
•52
•25
4-75
5-11

99-72

Yield in slabs.

> 8,115 Ibs. = 56-6 per cent.

Other specimens of steel shown are produced by fusion of the
iron thus obtained on the open hearth of a regenerative gas
furnace, which are undoubtedly of superior quality.

The furnace used for carrying out this process of fusion and
precipitation consists of a reverberatory gas furnace having two
beds formed by the ore itself ; on the upper bed a lake of fused
ore is formed which can be let off into the lower bed by piercing
the intervening bank of unfused ore ; the lower bed is divided
into two compartments, used alternately, each provided with a
working door. The dense carbonaceous material, such as an-
thracite or hard coke, to be used for the precipitation of the iron
in the lower bed, is reduced to a state of powder and mixed with
about an equal weight of pulverulent ore. This mixture is spread
over the bottom surface of the working bed, and the fluid ore is let
in upon it. By stirring with a rabble it is transformed into a
pasty and foaming mass, which in the course of from 40 to 50
minutes is shaped into a metallic ball in a bath of fluid cinder, which
may be shingled in the usual manner and formed into bar iron or
transferred to the pig-iron bath of a steel-melting furnace, where
it readily dissolves. The accomplishment of this process involves,
however, a certain degree of manual labour and skill, as, if it be
carelessly conducted, the yield of iron will be unsatisfactory ; the
analysis of the slags shows a variable percentage of iron amounting
rarely to less than 15 per cent., but reaching occasionally up to 40
per cent.

296 THE SCIENTIFIC PAPERS OF

It was evident that if iron and steel were to be produced largely
by direct process, that process must be a self-acting or mechanical
one, and here my attention again reverted to the rotating furnace
above-mentioned. If I could succeed in furnishing such a rotating
furnace with a lining capable of resisting the high degree of heat
requisite for the precipitation of the iron, and at the same time
capable of resisting the chemical action, I felt confident that this
mode of conducting the process must succeed. My attention here re-
verted to M. Le Chatelier's former suggestion of the use of Bauxite,
which possesses the requisite qualities, if only it can be put into a
compact form and rendered sufficiently infusible.

A series of experiments to form solid lumps by using different
binding materials have shown that 3 per cent, of argillaceous clay
suffice to bind the Bauxite powder previously calcined. To this
mixture about 6 per cent, of plumbago powder is added, which
renders the mass practically infusible, because it reduces the per-
oxide of iron contained in the Bauxite to the metallic state. Instead
of plastic clay as the binding agent, waterglass or silicate of soda may
be used, which has the advantage of setting into a hard mass at
such a comparatively low temperature as not to consume the plum-
bago in the act of burning the .brick. When the lining is completed,
the interior surface of the bricks is preserved against corrosion
by fluid cinder, added to bind them together, which also prevents
contact with the flame. A Bauxite lining of this description
resists both heat and fluid cinder in a very remarkable degree, as
I have proved by lining a rotative furnace at my Sample Steel
"Works at Birmingham, partly with Bauxite and partly with
carefully-selected plumbago bricks. After a fortnight's working
the brick lining was reduced from 6 inches to less than half an
inch ; whereas the Bauxite lining was still 5 inches thick and
perfectly compact. It is also important to observe that Bauxite
when exposed to intense heat is converted into a solid mass of
emery of such extreme hardness, that it can hardly be touched
by steel tools, and is capable of resisting mechanical as well as
the calorific and chemical actions to which it is exposed. The
Bauxite used for this lining was of the following composition : —

Alumina .... 53'62 per cent.
Peroxide of irou . . . 42'2G „
Silica . . 4-12

S/K WILLIAM SIEMENS, F.R.S.

297

Other Bauxites which I have had occasion to analyse were
composed as follows : —

ANALYSES OF VARIOUS BAUXITES.

FRENCH BAUXITES. — First Group.

SiOa.

Al.O,.

Fea03.

1-7

66-2

19*7

4-2

54-2.

28-7

37

58-9

25-7

3-2

55-9

28-1

8-2

60-4

25'1

3-9

56-4

22-1

8-2

61-2

23-1

5-9

58-8

23-1

2'8

60-7

25-8

| Raw

. 3-5

59-2

24-5

'° \ Calcined

. 4-01

67-89

28-09

Second

Group.

SiOa.

A1.0..

Fea03.

2-4

52-5

29-0

2-6

41-8

43-7

1-7

25-2

53-5

2*1

38-7

50-2

1-4

38-7

48-5

1-1

39-2

49-3

ro

38-7

48-0

1-9

41-2

41-7

( Raw

. 1-75

39-5

45-5

0 \ Calcined

. 2-02

45-53

52-45

AUSTRIAN BAUXITE.—

- Wocheinite

(Carniola.}

Raw.

Calcined.

A1203 .

.

64-24

85-88

FeaOs

> . .

2-40

3-20

SiOa .

. .

6-29

8-40

CaO .

> . .

•85

1-13

MgO .

. .

•38

•50

S03 .

•20

•26

P04 .

^

•46

•51

H.O .

. .

25-74

HaO.
12'7
18-0
12-4
12-2
11-5
13-5
11-3
11-8
11-8

1-2-1

HaO.
15-9
13-0
10-05
11-8
12-0
9-4
12-5
15-9

12-57

298 THE SCIENTIFIC PAPERS Ofr

IEISH BAUXITE.

Raw. Calcined.

A1203 . . . 35-0 44-58

Si02 . . .3-5 4-45

Ti02 . . . . 2-0 2-54

Fe203 . . . 38-0 48-40
H20 .... 21-5

The complete rotary furnace, such as is now ill use at Messrs.
Vickers and Co.'s, at Sheffield, and at my Sample Steel Works at
Birmingham, consists of a set of four regenerators of the usual
construction with reversing valves and gas-producers. The
rotative chamber is constructed of iron, and rests upon four anti-
friction rollers. Wheel-gearing is applied by which either a very
slow rotative velocity of from four to five revolutions per hour can
be imparted to the chamber, or a more rapid velocity of about
60 to 80 revolutions per hour. The chamber is about 7' 6"
in diameter and 9' 0" long, and is provided with a Bauxite
lining about 7" thick. A tap-hole is on the working side for dis-
charging the slag into the cave below, where it is received in
vessels mounted on wheels. At the two extremities of the cylin-
drical rotative chamber with its truncated ends, are large orifices,
one of which, on the side of the regenerators, serves for the in-
troduction of the heated gas and air as well as for the exit of
the products of combustion, and the other facing the work-
ing platform is closed by a stationary door hung before it in the
usual manner. Although the passage for the introduction of the
gases in combustion is separated only by a vertical partition wall
from the passage through which the products of combustion are led
away, the chamber is heated very perfectly, care only being taken
that the gases enter the chamber with a certain velocity, which
sends them forward towards the door and makes them reach the
exit passage only after having traversed the rotative chamber to
and fro.

This rotative furnace is worked as follows : —

The ore to be smelted is broken up into fragments not exceed-
ing the size of peas or beans ; to it is added lime or other fluxing
material in such a proportion that the gangue contained in the ore
and flux combines with only a little protoxide of iron into basic
and fluid slag. If the ore is hematite, or contains silica, I prefer

.S7A' WILLIAM SIEMENS, F.R.S. 299

to add alumina in the shape of aluminous iron ore ; manganiferous
ir-in ore may also be added with advantage. A charge of say 20
<-\vi. of ore is put into the furnace when fully heated, while it is
slouly ivvuhin;.r. In about forty minutes this charge of ore and
fluxing material will have been heated to bright redness, and at
this time from f> cwt. to 6 cwt. of small coal of uniform size (not
larger than nuts) are added to the charge, whilst the rotative
velocity is increased for a short time in order to accelerate the
mixture of coal and ore. A rapid reaction is the result : the
peroxide of iron being reduced to magnetic oxide begins to fuse,
and at the same time metallic iron is precipitated by each piece of
carbon, while the fluxing materials form a fluid slag with the
siliceous gangue of the ore. The slow rotative action is again
resorted to, whereby the mass is turned over and over, presenting
continually new surfaces to the heated lining and to the flame
within the rotator.

During the time of this reaction, carbonic oxide gas is evolved
from the mixture of ore and carbon, and heated air only is intro-
duced from the regenerator, to effect its combustion within the
rotating chamber. The gas from the gas-producers is entirely, or
almost entirely, shut off during this portion of the process. When
the reduction of the iron ore is thus nearly completed, the rotator
is stopped in the proper position for tapping off the fluid cinder ;
after this the quick speed is imparted to the rotator, whereby the
loose masses of iron contained in it are rapidly collected into two
or three metallic bails. These are taken out and shingled in the
usual way of consolidating puddled balls ; the furnace is tapped
again and is ready to receive another charge of ore. The time
occupied in working one charge rarely exceeds two hours ; and
supposing that 10 cwt. of metallic iron is got out per charge, the
apparatus is capable of turning out at least o tons of puddled bar
per 24 hours. If anthracite or hard coke is available for effecting
the reduction of the ore, it should be crushed much finer than
when coal or brown coal is used, the idea being that each particle
of the reducing agent should be fully consumed during the period
of chemical reaction. If wood is used, it has to be charged for
the same reason in still larger pieces.

If it is not intended to make iron, but cast steel, the balls may
be transferred from the rotator to the bath of a steel-melting

300 THE SCIENTIFIC PAPERS OF

furnace in their heated condition, and without subjecting them to
previous consolidation under a hammer or shingling machine.

It is feasible, however, to push the operation within the rotator
to the point of obtaining cast steel. If this is intended, the relative
amount of carbonaceous matter is somewhat increased in the first
instance, so that the ball, if shingled, would be of the nature of
puddled steel, or contain even some carbon mechanically enclosed.

If now, after removing the cinder by tapping, from 10 to lo
per cent, of ferro-manganese or spiegeleisen is thrown in, and the
heat within the rotator is rapidly raised by urging the influx of
heated gas and air from the regenerator, the metallic balls will
soon be seen to diminish, and presently a metallic bath only will
be found in the furnace, which may be tapped into moulds and
hammered and rolled into steel blooms or bars in the usual
manner. Experience alone can determine which mode of working
will ultimately prove the best ; but it is probable that for the
production of cast steel on a large scale it will always be more
profitable to transfer the metallic balls to a separate melting
furnace, a series of rotating furnaces working in concert with a
series of steel-melting furnaces, so as to produce charges of 5 or 0
tons of fluid steel.

In comparing upon theoretical grounds this method of producing
metallic iron with the operation of the blast furnace, it will be at
once perceived that, whereas in the blast furnace the products of
combustion consist chiefly of carbonic oxide, and issue from the
top of the furnace at a temperature exceeding 350° C., the result
of combustion in the rotative furnace is carbonic acid, which
issues from the regenerative furnace into the chimney at a tempe-
rature rarely exceeding 175° C. This proves at once a great
possible saving of fuel in favour of the proposed method, and to
this saving has to be added the fuel required for converting pig
metal into wrought iron by the puddling process.

It may however be asked, why the rotating surface should admit
of the complete combustion of carbon, whereas in the blast furnace
such complete combustion is, as is well known, not possible,
because each atom of carbonic acid formed would immediately
split up into two atoms of carbonic oxide, by taking up another
equivalent of carbon from the coke present. The following
explanation will serve to elucidate this point :

.SYA1 WILLIAM .S7A.1//-W.V, l-.R.S. 30!

In the rotative furnace streams of carbonic oxide are set up
within the mass under reaction ; and this carbonic oxide on
reaching the surface meets the current of intensely heated air
proceeding from the regenerators, and completes with it perfect
ci.iiilnistion within the free space of the chamber. The carbonic
acid thus generated comes in no further contact with carbon or
metal, consequently it cannot split up, but is drawn away un-
changed into the chimney, while the evolved heat is taken up by
the sides of the chamber and transmitted by reverberation and
conduction to the mixture of ore, fluxes, and coal.

In this process we have, therefore, to accomplish two things,
viz., the deoxidation of the ore, and the fusion of the earthy
matter mixed with it. If we take (say) hematite ore, consisting
of peroxide of iron with 10 per cent, of silica, we shall determine
the quantity of carbon necessary for its reduction from the
formula Fea03 + 3C, which gives 2Fe + 3CO ; and according to
which the consumption of carbon (taking its atomic weight at 12

Q x 1 9

and that of iron at 56) amounts to ' = *32 Ibs. per Ib. of iron

2x56

reduced.

The heat absorbed in this reaction amounts, according to Dr.
Debus, to 892 units * per Ib. of iron produced, but on the other
hand the further combustion of '32 Ib. of carbon from the condi-
tion of carbonic oxide to carbonic acid (CO to C0a), by means of
the free oxygen introduced into the rotative chamber from the
regenerator yields "32 x 5600= 1792 units of heat, leaving 1792 -
892 = 900 units available for heating the materials and for melting
the slag.

The quantity of materials to be heated per Ib. of iron produced
would amount to ore, 1'59 ; lime or other fluxing materials, '16 ;
making a total of 1'75 ; and taking the specific heat of Fe203 at
•154 as determined by Herman Kopp, and the temperature to
which the materials have to be raised at 1500° C., the heat
required for this purpose would not exceed 1'75 x '154 x 1500
= 404'25 units. To this consumption would have to be added
the latent heat absorbed in liquefying the slag. The slag would
amount to '16 Ib. silica + -16 Ib. lime = '32 Ib. per pound of iron

* Dr. A. W. Williamson gives 885 '3 units as the result of his calculation, which
two figures agree sufficiently for my present purpose.

302 THE SCIENTIFIC PAPERS OF

produced, and, although we have no precise data from which we
could ascertain the latent heat absorbed in liquefaction, we can
hardly estimate it at more than 150 units per lb., or at -32 x 150
= 48 units, which, with the above 404*25, makes 452'25 units,
whereas 900 heat-units are available, as resulting from the cal-
culation above given, proving that '32 lb. of pure carbon would,
theoretically speaking, amply suffice to produce 1 lb. of puddled
bar from ordinary hematite ore, without counting, however,
losses of heat by radiation and from other causes.

In the production of cast steel, three operations are essentially
involved, viz., the deoxidation of the iron, the fusion of the slags,
and the fusion of the metal itself with such proportion of carbon
and manganese as is necessary to constitute steel of the temper
required.

The theoretical quantity of fuel required to accomplish these
operations would exceed that of making wrought iron by the
fusion of heated metal, which may be estimated at, say, 1000

units, or at = '125 lb. of carbon per lb. of steel produced.

8000

which have to be added to the '32 lb. used in reduction.
- In fine, a ton of iron ought to be producible from hematite ore
with 6 "4 cwt. of carbonaceous matter, or say 8 cwt. of common
coal, and a ton of cast steel with 8'90 cwt. of carbon, or say 11
cwt. of coal. In giving these figures, I do not wish to imply that
they will ever be completely realised, but I maintain that, in all
our operations, we should fix our eyes upon the ultimate result
which theory indicates, which, owing to the imperfect means at
our command, we shall never completely reach, but should con-
stantly endeavour to approach.

In taking incidental losses by radiation through imperfect com-
bustion and through imperfect absorption of heat into account,
we find that the actual consumption exceeds the theoretical
limits about three times, or that a ton of iron can practically be
produced with a consumption of 25 cwt. of coal, and a ton of cast
steel with 40 cwt. of coal, which consumption represents a great
reduction as compared with other methods of production.

WILLIAM SIEMENS, FJl.S. 303

In tlif discussion of his Paper

"ON THE MANUFACTURE OF IRON AND STEEL
BY A DIRECT PROCESS."

[NOTE ON MR. SIEMKNS'S PAPER.* — The Paper f "On the Manufacture
of Iron and Steel by a Direct Process," was prepared under severe
pressure of business, with the view of being amplified and corrected
after delivery. Unfortunately, however, the rough paper, with marginal
remarks having reference only to the reading, was handed to the
printer without my having had an opportunity of revising and ampli-
fying it, as I should have wished. — C. W. S.]

DR. SIEMENS \ said, amongst the observations which had been
made, he would notice only those that required an answer from
him. Some of them had been answered fully or partly by other
speakers, and need not therefore be referred to by him again,
because their time was naturally valuable. The first speaker was
Mr. Snelus, and he had spoken very much in favour of reducing
iron from the fluid condition, and had referred to a method, which
he had informed them was in course of being tried at Middles-
brough, of forcing, he supposed, carbonic oxide gas through fluid
ores. Now, he had tried that method of proceeding and he cer-
tainly had not succeeded. Carbonic oxide, or any other poor
description of gas, was so near the point of saturation with carbon
that, if it took up any oxygen from the ore which it contained, it
soon arrived at the point where it took no more, and there the
reaction would stop. The result would be that a vast amount of
gas had to be driven through the fluid ore in order to produce a
given amount of reducing work, but that large amount of gas had
the disadvantage upon the fluid ore that it cooled it, and thereby
checked the reaction that was going on. It was upon those re-
sults becoming apparent, which, however, he (Dr. Siemens) partly
expected, that that mode of proceeding was given up and solid

* Excerpt Journal of the Iron and Steel Institute, Vol. I. 1873, p. 253.

f As both the paper "On Smelting Iron and Steel," see p. 283, ante, and this
one cover the same ground, the former has been reprinted, with the above dis-
cussion on the latter.

J Excerpt Journal of the Iron and Steel Institute, Vol. I. 1873, pp. 84-90.

304 THE SCIENTIFIC PAPERS OF

fuel resorted to in all cases. With regard to the elimination of
sulphur and phosphorus, he had to draw a particular line of dis-
tinction between the ore-reducing process carried on at Landore
and the process which he now had only partially substituted in
aid of it. In the furnace, such as was worked at Landore, and at
several other places, it was possible that the elimination of phos-
phorus and sulphur proceeded only at the same ratio in which it
would proceed in the Bessemer converter, but there was this differ-
ence that the operation took eight hours, whereas the Bessemer
process was completed in a quarter of an hour ; therefore, it was
probable that a larger proportion of sulphur and phosphorus was
evolved in the ore reducing process, and that, he believed, was the
case in actual practice. In drawing a distinction betAveen the two
processes, he did so without the least intention of depreciating the
Bessemer process, which, in many ways, he admired excessively,
but as drawing attention to an essential difference between the two
processes. "With regard to the price of bauxite, he could hardly
speak with confidence as yet. The mineral was found in large
quantities in the south of France, in Austria, and also in Ireland ;
and as there was no present use for it, the probability was that it
would be obtained cheaply, and the process of converting it into
material for the lining was certainly not an expensive one — not more
expensive than that of fire-brick making. He had not been aware
that Mr. Snelus had turned his attention to lime for linings ; in
the paper, allusion had been made to lime also, and they had,
especially in the ore-reducing process, used lime bottoms to some
extent. Hitherto he had not been able to obtain lime in such a
form as to resist permanently the effects of scoriae, but it was
quite possible ; and he might say that he was now engaged upon
a course of experiments at Birmingham, with a view to substitute
lime, in some cases at any rate, for bauxite in forming a lining.
Mr. Snelus had also alluded to a point in the furnace which cer-
tainly required explanation — whether the parts of the furnace
throat, where it entered out into the rotating chamber, were not
very rapidly attacked ? He (Mr. S.) had some little misgivings
on that, but he found that experience proved decidedly more
favourable than anticipation ; those parts (pointing to some on
drawing) had now stood for five or six weeks of working, and they
certainly were none the worse for it. It had to be borne in mind

.s/A' \VIIJ JAM SIEMENS, FJt.S. 305

that tin- bauxite lining did not actually touch the silica brickwork,
but was pt 'i-hups an inch apart from it, and that the flame in
making a turn through the furnace did not carry with it so much
dust outward as might be the case if it swept across a furnace. He
t'< 'inid also in the puddling furnace a very great difference in favour
of the return current ; it seemed as though the dust were swept out
of the flame and little ashes were deposited in the passages. They
could get any amount of heat, even coming up to full steel melting
heat, inside the rotating furnace without artificial blast. With
regard to manganese and its necessary presence in steel, he thought
that Mr. Riley was perfectly right in saying that tool steel was
found without more than a trace of manganese, but they must
remember that such steel contained one per cent, of carbon, and
generally was very free from sulphur ; but it had been established
in a very absolute manner at Landore, that mild steel containing
sulphur must necessarily also contain manganese in a given propor-
tion to the amouut of sulphur present, and that this proportion
increased with the mildness or the relative absence of carbon in the
metal. Thus, if the metal contained '08 per cent, of sulphur, they
considered it not safe to have less than three-tenths per cent, of
manganese in the metal, and whenever, through accidental causes,
the manganese in the metal was less, the steel did not work so
well, nor did it do so well when cold. He (Mr. Siemens) was much
impressed by the kind observations made by Mr. Bessemer. Mr.
Bessemer had been an early labourer in that field, and the great
successes which he had attained were vividly before them. It was
no small compliment to him (Mr. Siemens), he considered, that
Mr. Bessemer had some years since been thinking of extending his
labours in the direction which he (Mr. Siemens) had taken, and if
circumstances had not prevented him (Mr. Bessemer), he doubted
not would ultimately have arrived at the same, if not superior
results, but he (Mr. Siemens) had given attention to this question,
not a few years, but many years ; and if results of a practical
kind had been obtained, Mr. Bessemer, he was glad to think,
would be the last man in the world to dispute or discredit them.
Mr. Parks referred to the analysis of a bar of iron (which had
been produced at Birmingham) containing one-third per cent, of
copper. Now it was an undoubted fact that the process under
discussion fetched all the copper out of the ore, and, therefore, if

VOL. I. X

306 THE SCIENTIFIC PAPERS OP

the ore contained copper, the iron would contain it no doubt also ;
but in the particular instance referred to, the presence of copper
was accounted for by a previous charge which was made with ore
containing about a half per cent, of copper, and some of the
charge having remained attached to the sides of the furnace, it
worked into the succeeding charge. Some other speakers had
spoken with regard to the effect of copper on iron, and he could
fully confirm the observations made by Mr. Fothergill, that a very
small percentage of copper entirely prevented the iron from
coming together ; instead of forming balls, it remained separate,
and would not hold together. Dr. "Wright was the first speaker
to touch upon the fuel question, which question had been very
fully taken up by the President, and carried somewhat against the
conclusions arrived at by himself. Now, with regard to that
question, he wished to explain his views concisely, but more fully
than he had done in the paper. He must say that the furnace in
Birmingham was no criterion with regard to the economy of the
process, because that furnace was of comparatively small dimen-
sions, and was erected simply for the sake of experiments, the
object being to see the effect produced on the lining by different
treatment, and on the quality of the iron produced by various
mixtures and fluxes. Lately, he put in a lining which had not
been baked beforehand, and the result was that the balls would
not hang well together ; the bauxite came off in small filaments
from the lining, and acted, as he might say, like soap between the
nodules of iron, preventing their cohesion. That difficulty was
removed after exposing the lining to two days' intense heat, and
now the results were better, as might be seen by the two balls sent
up, which were then in the yard at the back of the building, and
had been produced in that little furnace. They were balls of
nearly 600 Ibs. each from charges of a thousand pounds ; one of
blackband and the other of Acadian ore. The normal way of
working the rotary furnace was as follows : — The furnace was
charged with say a ton of pulverized ore, and the slow rotation
went on until the mass was heated through and through to full
redness. Then the carbon was introduced and mixed with the
charge in giving a rapid rotation to the apparatus. The conse-
quence was that reaction and fusion set in immediately, and it
was very important that at that point the action should be very

S/A WILLIAM SIEMENS, F.R.S. 307

en. -r^etic in order to weld the particles of iron t <>•.;•' •( her, while at
the same time the earthy material with the ore was fused and
!'"r;i u.'d a slag with the fluxing material employed. If that opera-
tion was normally carried out, the iron formed collected in what
he might call a grisly condition — small masses hanging together
and slightly rolling over each other — whereas a lake of perfectly
liquid cinder formed at the bottom, and after that liquid cinder
had been tapped off, more rapid rotation was given, and the mass
formed readily into one or several balls of metallic iron. In that
process the two actions of the formation of the metallic iron and
of the fusion of constituents of the ore were carried on simultane-
ously, and that was an essential condition to the making of good
and clean balls, and in this the process differed essentially from
the processes of Chenot and others, in which spongy iron envelop-
ing earthy matter was first formed which absorbed sulphur, and
could not, therefore, yield pure metal. Now, with regard to the
fuel question, he had certainly appeared to be very bold to
prognosticate such results as he had done, but the action in the
furnace had hardly been fully understood, or his results could
hardly have been questioned. They had there the ore and the
carbon mixed together in a state of reaction ; that reaction pro-
duced currents of carbonic oxide which rose to the surface, where
complete combustion took place, and the whole of the carbonic
oxide was consumed under the most favourable circumstances
possible, because it was consumed at the moment of its generation
without being first taken through a cooling tube or exposed to
loss of any kind, and it was burnt by a current of intensely hot
air by oxygen heated to at least 2,000 degrees Fahrenheit, perhaps
3,000 degrees Fahrenheit. Therefore, the carbonic oxide thus
formed by chemical reaction which was lost in the blast furnace
was consumed under the most favourable conditions of the furnace,
and the heat thus developed accounted fully for all the heat that
was necessary for bringing up the material to the state of heating
required, and for melting the slag, the regenerators acting as fly-
wheels for taking up an excess of heat at one part of the process,
and yielding it at another. Theoretically speaking, '32 cwt. of
carbon would be necessary for the production of a hundredweight
of metallic iron, but. the result which he had to put forward as
being practically attainable and virtually attained by himself,

x 2

308 THE SCIENTIFIC PAPERS OF

was one and a-quarter ton of fuel per ton of iron. Now, in
dividing the one with the other, it would be found that he
claimed only 25 per cent, of the theoretical result. Mr. Krantz,
in his investigation of his regenerative gas furnace, gave him
credit for utilizing from 15 to 20 per cent, of the theoretical heat
residing in fuel in heating iron, but he (Mr. Siemens) claimed
25 per cent, in the rotative furnace, because fully one-half of the
work was done by reaction of the heated oxygen upon the car-
bonic oxide produced there and then in the furnace, and upon
that portion there was no sensible loss by refrigeration in cooling
tubes or through leakage. Mr. Bell, when he visited his (Mr.
Siernens's) little works at Birmingham, had seen the reaction
within the rotator going on for a considerable time without any
gas being admitted from the gas-producers, showing that the
carbonic oxide generated within the mass sufficed to produce the
requisite amount of heat to support the process. He (Mr. S.)
therefore believed that he should be borne out by practical experi-
ence when he asserted that one ton and a-quarter of fuel would
suffice for producing a ton of iron. With regard to the quality of
that iron, he had already stated in the paper, that it was excellent
material for being transferred to the melting furnaces in order to
produce steel, being almost entirely free from sulphur and phos-
phorus, and although the material produced had not yet been
practically proved for iron-making, he felt pretty confident of
ultimate success in this direction also. The reason why sulphur
and phosphorous did not combine with the iron was that they
were in an oxidised condition in the ore and combined with other
substances, and that the reducing action of the apparatus was
insufficient to move them. If the reaction was insufficient for
their reduction, then there was no reason why they should be
present at all in the iron, and in that respect the direct process
claimed an advantage over the blast furnace, where the reducing
action was so energetic that everything reducible was brought
into combination with the iron. He thought he had now touched
upon all the points referred to in the discussion, and had only to
thank them for the patient hearing they had given him.

WILLIAM SIEMENS, F.R.S. 309

In the discussion of

THE PRESIDENT'S (MR. I. LOWTHIAN BELL)
ADDRESS,

DR. SIEMENS * said he had not thought of making any remarks
of his own, but the President in delivering his address had, as
usual, given them a good deal to think about, and he was rather
sorry that the discussion had turned only upon one of his remarks,
though indeed to discuss all the points raised would fill up the
whole time of their meeting. With regard to one question that
had been touched upon, Sir John Alleyne had already mentioned
that he had tried, with tolerably successful results, the furnace for
puddling that he (Dr. Siemens) had brought before the Institution
last year. The chief features of that application were, that the
regenerative gas furnace was used, and that the products of com-
bustion left at the same side of the heating chamber as that at
which the flame entered. Puddling furnaces, similar in arrange-
ment, but not rotative, were now largely used. He mentioned in
his paper last year the works of Messrs. Nettlefold and Chamber-
lain, who had then 10 furnaces at work upon that plan, and they
had now 20 ; these were working very successfully, and producing
excellent iron and good weight. Of course the question of fettling
was the important one as regards the rotative furnace, and what-
every they might do, it would be a difficult and delicate question
to make good fettling, unless they had a large supply of very pure
oxides. At Butterley, they were particularly badly off in that
respect, as their ores were of a poor description, though in the
end they made good iron. Still, however, they were very poor
ores, and their quality was not nearly sufficient to supply pure
oxide, or even pure enough to use for fettling their furnaces. This
difficulty was less formidable in working ores in the manner he
had proposed to do last year. That process, although not yet
carried out on a commercial scale, was progressing towards practical
results. His sample works at Birmingham were steadily working

* Excerpt Journal of the Iron and Steel Institute, 1874, pp. 49-50.

310 THE SCIENTIFIC PAPERS OP

experimentally at the details of the process, and several applica-
tions had been made of it at other works, but they were hardly
sufficiently advanced in practice to give any working results.
With regard to the President's observation concerning his ore
process, he must admit that it was a legitimate criticism, namely,
whether in the rotary furnace, carbonic acid could be kept away
from the metal as it was forming. He (Dr. Siemens) still main-
tained the argument he used last year, that the carbonic acid
remained separate from the ore, and that in this respect he had an
advantage over the blast furnace. Of course, there was no abso-
lute separation of the carbonic acid formed from the ore, and
practice alone could show whether the reduced metal was effectu-
ally protected against re-oxidation under these circumstances.
The results he had obtained showed that they could get about 80
per cent, of the iron from the ore, which did not seem a good
result as contrasted with the blast furnace practice, but if they
compared it with the working of a blast furnace and the subse-
quent puddling furnace combined, it was by no means a bad
result, and he (Dr. Siemens) claimed to be judged by that standard.
His advantage consisted, however, chiefly in economy of fuel. He
was glad that the President had alluded to a project which had
found vent in the public press, of using chalk instead of coal.
He thought that, on the whole, the President had dealt too
leniently with that project. They should remember that chalk
was the result of a double combustion — a combustion of calcium
and a combustion of carbon — both combustions being perfect in
themselves, leading to a further combustion of the two combined
together, and, therefore, if the originator of that scheme had
looked all the world over, he could not have found a more
thorough cinder.

.SYA' \V1LLIAM SIEMENS, fi.R.S. 3! I

In t/ie discussion of the Paper

" ON THE MANUFACTURE AND USE OF SPIEGELEISEN,"
By MB. GEO. J. SNELUS,

DR. SIEMENS* said the Landore Works being producers of
spiegeleisen, he could, from his experience there, bear out nearly
all that had been said by the author of the paper. He agreed
with him that there was no great difficulty in producing spiegel,
as, indeed, all things were simple if they knew how to deal with
them, though up to that point they were difficult. At first there
had been a difficulty in getting the slag right, and according to
their experience the slag must be as basic matter as possible. They
had succeeded in making spiegel containing as much as from 18
to 20 per cent, of manganese, and they could vary it so as to make
it from 10 to 12 per cent., which was what was preferred for
common use, the richer spiegel being reserved for producing extra
mild steel. There was one point upon which he would like to ask
Mr. Riley to give a little more information, that was, why did
spiegeleisen never contain even a trace of sulphur ? As he under-
stood the paper, it was owing to the manganiferous slag having a
greater capacity for sulphur, but he (Dr. Siemens) thought that
that explanation was hardly sufficient for the extraordinary fact
that they did not find even a trace of sulphur in spiegeleisen.
As regarded phosphorus, it had been contended that it acted as a
substitute for carbon, and was rather a desirable constituent in
spiegeleisen, but he would much rather be without it, and not risk
the beneficial effect of phosphorus on the steel. Another interest-
ing point with respect to the use of spiegeleisen, was the nature of
its effect upon the steel ; was it that the manganese simply
deoxidised the iron, or was it that the manganese must be present
in a metallic state, in order to produce malleable steel ? He
inclined to the latter hypothesis, because his observations went to
prove that manganese was necessary in malleable steel, especially
if sulphur was present. Manganese seemed, in fact, to have a

• Excerpt Journal of the Iron and Steel Institute, 1874, pp. 84-86.

313 THE SCIENTIFIC PAPERS OF

specific action in neutralizing the effect of sulphur, and no chemi-
cal explanation, that he was aware of, had ever been even given of
that fact. On the other hand, he (Dr. Siemens) did not think
that the deoxidising action of the manganese was the essential
effect of that substance in the process with which his name had
been connected, because they could regulate the oxidising action
on the bath to such a degree, that the absorption of oxygen by the
bath had not commenced before the spiegel was added, and yet
the addition of spiegel was necessary. If the operation of decar-
burisation was carried on to the point at which oxygen was
absorbed from the atmosphere, that is, until the carbon was
reduced to below 1 per cent., it was very difficult to get rid of the
oxygen. After spiegeleisen had been added, and the bath had
become perfectly quiescent, they would find the steel, when poured
into the moulds, set up a violent ebullition. Now, that ebullition
proved that oxygen existed in the fluid metal and probably in
combination with carbon, which was liberated suddenly after the
steel had been lowered in temperature to near the point of solidifi-
cation, but which the manganese had not been capable of removing.
He was aware that, in the Bessemer process, when spiegeleisen
was added, a powerful ebullition was set up ; but that was not
the case in his process, in which ore was used as the decar-
burising agent, and the steel when poured into the ingot moulds,
or moulds for casting purposes, showed no sign of ebullition, if
carefully prepared ; on the contrary, the steel contracted, or what
was technically called piped, but if the operation had been carried
too far, it seemed to acquire an extraordinary affinity for oxygen,
and that oxygen at very high temperatures did not seem to go out
again, notwithstanding the addition of manganese, but remained
associated with the metal until it had cooled down near to the
point of solidification. He would like to have that important
point cleared up in this discussion.

WILLIAM SIEMENS, F.R.S. 313

In ihe discussion of the Paper

"ON IRON AND STEEL FOR SHIP-BUILDING,"
By MR. N. BARNABY,

DR. SIEMENS * said, there can be no doubt that steel is a metal
which is worthy of the highest consideration of ship-builders,
engineers, boiler-makers, and all persons, in fact, engaged in
construction of that kind. One thing is necessary, namely, that
those who use that material should know what it is. We hear of
comparative results of steel and puddled iron as produced in one
furnace and another furnace ; and one would naturally come to
the conclusion that steel was a definite compound varying only as
regards quality. What is that compound may I ask ? Steel in
the form of a needle, an edge tool, or of a punch, is of a hardness
approaching that of the diamond ; steel in the form of a spring is
of an elasticity unequalled by any other metal, or any other sub-
stance in nature. Then, again, steel in the form of a milled plate
is, with few exceptions, the toughest material in existence, tougher
than copper or wrought iron. It can be moulded into almost any
shape in a cold condition. Therefore, we ought first to under-
stand what we mean by steel before we consider its merits or
demerits for structural purposes. You may say that the hardness
depends upon the proportion of carbon in the metal. Does it ?
I have lately experimented with steel containing four-tenths per
cent, of carbon. That steel, if treated in a certain way, would
make an excellent punching or cutting tool. If treated in another
way and annealed carefully, it is so tough that you could work it
into the form of your hat. Then, again, drawn into wire and
annealed in another way — that is to say, heated in a certain way
and then put into oil — it assumes a tensile strength of nearly
100 tons per square inch ; whereas, in the other form just before
mentioned, when annealed carefully its tensile strength would not
exceed 85 or 36 tons. Now, if the mere fact of after-treatment

* Excerpt Transactions of the Institute of Naval Architects, Vol. XVI. 1875,
p. 140.

314 THE SCIENTIFIC PAPERS OF

can produce such enormous changes in the behaviour of steel, how
important it must be to give to steel the nature which we desire
for structural purposes. Now, if you take a plate and throw it
down in a wet court-yard, and then punch it and use it for the
manufacture of a boiler, there can be very little satisfaction in the
result obtained from the material. Yery likely in the same plate,
if you could cut out pieces, you would find very different qualities,
and some of these qualities would be valuable in one place and
would be destructive to the plate in another. It may be laid
down as a general rule, that a higher material and a higher pro-
cess require higher intelligence to deal with them. If you ride a
common cart-horse you may go to sleep on him, but if you ride a
fine spirited horse all your wits must be about you, or else you
may be landed in a ditch. So with steel, you have to understand
first of all what quality of steel you have to produce, how much
carbon, and how much manganese should be mixed with the
material, and how little sulphur and phosphorus you can put up
with. Until you have done that, you do not know what you are
speaking of. It is a material belonging to a group varying
between the hardness of the diamond and the toughness of copper ;
and it is also of the highest importance that the manufacture
throughout, and the construction throughout, should be carried
on with superior intelligence. Now, should we shrink from using
a material, because intelligence is required in working and using
it ? Surely that would be a very poor compliment to this age of
progress. We should have no difficulty in finding what are the
conditions necessary to produce steel of such and such a quality,
and should see to it that we obtain this quality and obtain it
always. With regard to the question of obtaining it always, I
maintain that we have a greater power in our hands than with
regard to iron ; iron, as we know, is produced in small quantities ;
the puddler produces sometimes a ball which is rather young, and
at other times produces it overheated ; and this material is piled
again, put through the rollers, and in the end we get a sort of
average between the qualities of the different balls. In making
steel we formerly dealt with it in small quantities also by melting
it in pots, but Mr. Bessemer has shown us how to deal with it in
large quantities in his converter. I have had considerable ex-
perience in dealing with it in large quantities in the open hearth

S/X WILLIAM SIEMENS, F.R.S. 315

furnace. There I know that we can produce six, eight or ten tons
of steel of perfectly uniform quality. We can take out samples
before pouring that steel to assure ourselves of having the quality
(1. This metal is thoroughly mixed — it is a perfectly fluid
mass — and, therefore, there can be no reason why there should be
a difference in the behaviour of one part of this metal from the
behaviour of another part. Now, I have lately seen steel of very
mild quality produced which is eminently suitable for structural
purposes. This steel contains hardly any carbon at all, perhaps
one-tenth per cent, only ; but it contains manganese in a larger
proportion than has been given to it hitherto. It is possessed of
a toughness which is unapproached by any other kind of metal ;
and before it breaks it yields even to 50 per cent. Now, if such a
material can be produced, and if such a material will resist say
80 tons, which is quite enough for all purposes, I think it is the
very best material for structural requirements.

In tlie discussion of the Papers

"ON THE MANUFACTURE OF STEEL," by WILLIAM
HACKNEY, B.Sc., Assoc. Inst. C.E., and

"ON BESSEMER STEEL RAILS," by JOSIAH TIMMIS SMITH,
M. Inst. C.E.,

DR. SIEMENS * said that he regarded Mr. Hackney's Paper as so
comprehensive that it offered room for observations on many
different branches of the subject. His remarks would be confined
to the first two portions into which the subject was divided,
namely, the chemical and the manufacturing. The author had,
perhaps, dwelt too much upon the question of name, and hail
proposed to change the prevailing denominations of mild steel,

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
XL1I. Session 1874-1875, pp. 91-95.

316 THE SCIENTIFIC PAPERS OF

hard steel, spring steel, and so forth, and to call "steel" all
malleable iron which had passed through the fluid condition.
This definition would exclude, however, natural, blister, and
puddled steel, and was therefore in his opinion inadmissible. On
the other hand, Dr. Siemens considered it would be difficult, if
not impossible, to define steel by its mechanical qualities. Steel
was almost the hardest substance in nature, if treated in a certain
way ; treated in another way it was the most elastic of metals, if
not the most elastic substance in nature ; and treated in another
way it was nearly the most ductile of metals. It was decidedly
the strongest substance in nature. Dr. Percy had said he con-
sidered steel a metal in which iron was the chief constituent, and
which would harden. If that definition were adopted, a limit
would be arrived at where it would be uncertain whether a sub-
stance was steel or iron. One rail delivered by a manufacturer
would come under the definition of steel, and another rail delivered
in the same contract would come under the definition of iron,
simply because one had rather less carbon than the other ; and
after all, the mildest steel, and iron itself, would admit of
tempering. Dr. Percy had said that molten iron, according to
Mr. Hackney's definition, ought to be called steel ; but it had
been already remarked that such iron could not be treated
mechanically ; it would not stand rolling and hammering, and
therefore it was not steel. He thought, however, a definition
might be found which would answer all requirements, namely,
that steel was a compound of iron with any other substance which
tended to give it superior strength. This definition would embrace
the different kinds of steel, from the hardest tool steel down to the
mildest, and would embrace those compounds in which manganese,
tungsten, chromium, phosphorus, or sulphur, replaced the carbon
of ordinary steel. It had been objected that steel was uncertain
in its nature ; that it sometimes was very ductile and very strong,
and at other times would break short and was not be trusted.
Mr. Walker had said that, on that account, he preferred iron to
steel for structural purposes. Steel was a material of a much
higher nature than iron. It was much stronger, and could be
made to possess nearly any degree of strength, hardness, and
ductility, between wide limits, that it was desired to give it ; but
it was only natural that such a material should require greater

51//? WILLIAM SIEMENS, F.R.S. 317

care than iron. If extreme toughness was required, care must be
taken that the material was not put partially through a process of
lur.li'iiing or chilling, and that its chemical constitution w;is
uniform. The higher material required a higher intelligence to
<l«-il with it, and it was not sufficient that the chemist, or the
manufacturer, should possess a knowledge of its characters under
dill'rrent treatments. The engineer, also, had to be thoroughly
conversant with its properties, and in giving his orders for a supply
of the material should make out a proper specification, defining its
chemical constitution and its temper. Sir Joseph Whitworth had
proposed a method of defining steel according to its two principal
qualities, toughness and strength ; he had one number signifying
strength, and another toughness, and combining these two expressed
the quality of the steel. That was a rule which furnished a fair
standard of comparison, and might be worked out with great
advantage. He had repeatedly heard it stated, that mild steel,
such as was used for engineering purposes, ought to be rejected if
it had an absolute strength exceeding 27 or 30 tons. Such a
limit was, he thought, objectionable because it stopped the road to
improvement. He considered it quite possible that steel might be
extremely tough, and at the same time possess very great tensile
strength. As an instance of this he would mention wire, which
had undergone the process of annealing in oil at a certain tem-
perature. He had lately had occasion to experiment on some steel
containing 0'4 per cent, of carbon, which, if properly annealed,
was extremely strong and yet sufficiently ductile ; if hardened, it
would take such a temper that it could be put into the form of
a punch ; and if drawn into wire, and annealed by passing it at
a certain temperature through oil, its strength rose from 35 tons
per square inch to about 100 tons per square inch. That was
an instance of one material presenting itself under very different
aspects, in consequence of slightly different modes of treatment.
Et was necessary that the engineer, in dealing with steel, should
inquire into these different conditions, and should insist upon the
utmost care being taken in order to bring the material really into
the condition which he required. Cast steel was produced in the
Bessemer converter, or in the open hearth of a regenerative gas
furnace, in masses of from 5 to 10 tons ; and absolute uniformity
could be obtained if all conditions were properly carried out.

318 THE SCIENTIFIC PAPERS OF

Mr. Hackney, after paying the high compliment which was due
to Mr. Bessemer for the introduction of his process, had referred to
one process with which Dr. Siemens's name had been associated.
That process might perhaps be appropriately called the chemical
process, or the cooking process, to use a more vulgar expression ;
because, given a hot receptacle, such materials were put into that
receptacle as would, when melted together, produce steel of the
quality required. The beginning of the operation must be a fluid
bath, and the material natural for such purpose was pig metal. To
this pig metal, when heated to a temperature exceeding 2,000°
centigrade, either scrap iron, scrap steel, puddled blooms, or iron
sponge might be added, and thus a bath might be obtained, which
would gradually come to contain less and less carbon, until it was
reduced to the condition of fluid iron. It then received such
additions as would give it the necessary manganese, carbon, or
other substances, to make steel of the required quality. This
process had the advantage that samples could be taken out, from
time to time, in order to make sure that the operation was going
on properly, and that the chemical and physical qualities of the
material were such as might be desired. The essential condition
for carrying out such a process was a very high temperature ,• it
was perhaps the highest temperature used in the arts, excepting
only the fusion of platinum. Considerable difficulty had existed,
at first, in obtaining a vessel to withstand such heat, continuously,
and the only substance capable of doing so was silica, containing
only from 1 to 2 per cent, of binding material. He noticed
diagrams of the furnaces in which the processes were carried out ;
and one diagram of a modified form of the furnace, by M. Pernot,
which was certainly ingenious. The bath was circular, and was
rotated on an inclined axis ; and the pig metal flowing round and
round in that vessel would wash over any solid substance put into
it. He rjowever doubted very much whether such a vessel would
resist the high temperature which was necessary for the production
of mild steel by that process. He understood it was used at
St. Etienne, in France, where engineers were content to have steel
rails containing 0'8 per cent, of carbon, which melted at a much
lower temperature than the mild steel required in this country.
Nor would it probably work out altogether satisfactorily for the
process of producing steel from pig metal and ore ; because, in that

.S7A' \\-lI.UAM SIEMENS, F.R.S. 319

process, considerable ebullition ensued, and he should be afraid
that the )>« tiling mass would go between the rotating lip of the
vessel and the standing part of the furnace, and cause it to stop.
Mr. Hackney had alluded to experiments carried on in America
I iv Mr. Blair, who appeared to have found that, if the ore and
rarhon were heated to a sufficient temperature, and left to them-
selves, the heat retained in the mass would be sufficient to
complete the reduction of the ore. Dr. Siemens maintained that
that was an entire fallacy. Although the ore might be hot enough
for the reaction to commence, a very large influx of heat would be
required, during the operation, to make the whole of the oxygen
contained in it combine with carbon. Mr. Lowthian Bell had
already proved the fallacy of the assumption that ore, under such
circumstances, would combine with carbon and form itself into
spongy metal. He had himself paid considerable attention to the
reduction of ore, in order to obtain material for the bath, and he
had experienced great difficulty in reducing ore into spongy metal.
The iron, in that spongy condition, took up sulphur very readily,
and sulphur vitiated the steel produced from it. These failures and
difficulties induced him to resort to another mode of obtaining
metallic iron from the ore, by passing the ore, in a heated condition,
through a rotating chamber, and there treating it with reducing
material, not at a temperature to produce sponge, but at a melting
temperature at which cohesive wrought metal was at once produced.
That part of the process was as yet in an experimental stage, and
he could not at present reply to Mr. Williams's inquiries regarding
it ; but before long it would be at work on a complete scale, and
he would then be in a better position to speak definitely regarding
it. The chemical process of making steel might be varied almost
ad libitum, and that variety was undoubtedly the best which,
from impure ore or impure pig metal, would produce material of
comparatively high quality. Hitherto phosphorus and sulphur
had been regarded as the enemies of steel, but recent experience
had shown that they might be innocent constituents under certain
conditions. If phosphorus was contained in the metal used for
steel-making, it was necessary that the metal should contain very
little carbon ; in fact, the carbon should be reduced to a mere
trace ; and manganese should be introduced in order to produce a
malleable metal. This was being carried out very successfully in

320 THE SCIENTIFIC PAPERS OF

France and in this country, and the results were very satisfactory.
Years ago the use of rich ferro-manganese had been found to
afford great assistance, but manufacturers had not succeeded in
producing a soft and tough metal containing so large a proportion
of phosphorus as had been produced latterly, more particularly at
Terrenoire, in France, in using the Siemens-Martin process. He
had very few observations to offer with regard to the working of
steel. In solidifying from the melted state it had a natural
tendency to become honeycombed, either through the generation
of gases, or through mere contraction of the semi-fluid mass, and
it was necessary to close these cavities as thoroughly as possible,
because steel, unlike iron, did not weld ; and if a hollow space
existed in an ingot, that hollow space would remain a break of
continuity, no matter through what process it was put afterwards.
Probably hammering was the most efficacious of those contrivances
which had hitherto been adopted for closing those bubbles ; but
Sir Joseph Whitworth had elaborated a plan for remedying the
evil at the proper moment. He compressed the steel while it was
still in a fluid or semi-fluid condition ; Dr. Siemens expected
great results from that system. He was not prepared to say
whether it would ever supersede hammering and rolling, for
ordinary rail-making, but for such purposes as gun-making, he
considered that Sir Joseph Whitworth's plan commended itself as
one of great importance.

In the discussion of the Paper

"ON FLUID COMPRESSED STEEL AND GUNS,"
By SIR JOSEPH WHITWOETH, BART., D.C.L., F.E.S.,

MR. C. W. SIEMENS * observed that the paper now read was
remarkable on account of the very striking results which had

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1875, pp. 285-288.

WILLIAM SIEMENS, F.R.S. 321

been arrived at, and which the members would have an oppor-
tunity of witnessing at the author's works. But perhaps the most
ivmarkalilr l< 'uture of the subject was the systematic manner in
which those results had been arrived at. The question of gunnery

h-iil 1 n taken up by Sir Joseph Whitworth in 1854 in connection

with the rifle, and in 1868 in connection with the construction of
stei-1 guns ; and on the occasion of his reading a paper before the
British Association in 18G9 on the penetration of armour plates by
long slid Is tired (»l»lii|iiely, the remark had been made by himself
• Mr. Siemens), as President of the mechanical section, that "Mr.
Whitworth having taken up the question of gunnery has viewed
it from a mechanical point of view, and has arrived at purely
mechanical results in this problem. His solution commends itself
to all who have gone into the question ; and it is natural there
should be some diversity of opinion under such circumstances."
The solution at that time arrived at by Sir Joseph Whitworth in
connection with the question of constructing guns had reference
to the method of rifling that he proposed ; and it appeared to be
conclusive. Nothing was more natural than that his attention
should next be directed to the production of the material from
which the gun was to be constructed ; and having persevered in
thoroughly investigating this portion of the subject, he had suc-
ceeded in arriving at the remarkable results brought forward in
the present paper.

In the plan now carried out, the steel after it had been produced
was dealt with under a method entirely different from those tefore
adopted, being here compressed while in a fluid or semi-fluid state.
He had at first felt considerable doubt as to the effect of the new
method. It was said that in applying hydraulic pressure upon
fluid steel the gases contained in the fluid metal would be driven
out ; but he could not see how that was to be done by mere pres-
sure. For in applying pressure to a fluid, the pressure acted in all
directions equally ; and why a particle of gas held in suspension
in the fluid should go in one direction rather than another, and
should get away from the pressure to which it was subjected, it
seemed difficult to conceive. The facts however spoke for them-
selves ; and these being ascertained, it was more easy to find an
explanation of what took place. The result he suggested might
be accounted for by the circumstance that the fluid steel congeal-

VOL. I. Y

322 THE SCIENTIFIC PAPERS OF

ing first on the outside of the mould offered more resistance there
to the motion of the plunger, and the outside thus became com-
paratively speaking porous, while the fluid portion in the centre
received a larger amount of compression than the outside which
had more power of resisting the pressure. The particles of gas
entangled within the fluid mass would therefore encounter rather
less resistance towards the outside than towards the inside, the full
hydraulic pressure being transmitted to the centre of the fluid
mass. In that way the expulsion of the gases from the fluid
metal might perhaps be accounted for ; and he should be glad if
Sir Joseph Whitworth would give his explanation of the matter.
The fact being admitted, it was clear that the steel produced by
that mode of treatment must possess many great advantages
over metal treated in the ordinary way by hammering ; for it
was hardly to be supposed that hammering would be capable of
driving out the gases.

With regard to the mode in Avhich these gases entered the
metal, he did not think they were merely entrapped mechanically
at the time of pouring out the metal into the mould ; because in
working melted steel in the open hearth of a regenerative furnace
he had found that the metal could be made at any moment to
evolve gases in great quantities by simply plunging a cold bar of
iron to the bottom of the fluid mass. The fluid metal evidently
absorbed carbonic oxide to a very great extent ; and it was due to
the partial congelation of the metal that the gases were suddenly
set free. Similarly in the Bessemer process a great ebullition took
place on pouring the fluid metal out of the converting vessel into
the iron moulds, and the top of the moulds had to be closed by a
stopper to prevent the metal being thrown out by the ebullition.
It was clear therefore that the metal contained a large quantity of
gas occluded within itself ; and if this was retained in the metal
it became a source of weakness. However small the bubbles of
gas might be, their presence would have the same effect as the
presence of particles of foreign matter between the particles of
metal, and must necessarily weaken it.

In reference to the proposal to designate as steel any metal
bearing a tensile strain of 28 tons per square inch, he thought it
would be wise on the whole to fix a limit of strength, but some
further limitation seemed also to be needed. For instance, a

WILLIAM SIEMENS, F.R.S. 323

in- tul produced in the puddling furnace, with or without being
converted into steel by the ordinary cementation process of making
blister steel and shear steel, would have the required amount of
tcinile strength, and would therefore pass as a steel. But he
(..nsidered a broad distinction should always be made between
steel which had passed through the fluid condition and that pro-
duced by other processes, because the latter was deficient in one
essential quality which was always sought for, namely, uniformity
of strength. He would therefore willingly accept the suggested
definition of steel and iron according to tensile strength and
ductility, if it were confined to metal that had passed through the
fluid condition.

It would appear from the paper that it was immaterial by what
process the steel was produced, provided it had been produced
from pure materials and would bear the required tensile strain.
But however true it might be that the quality of the steel was due
to its chemical nature, yet this chemical nature could only be
obtained in a uniform and regular manner if the process of manu-
facture was in conformity with the conditions necessary to produce
the desired result. Such steel, he believed, it would never be
possible to produce by the old processes of puddling aud cementa-
tion ; and this led him to suggest that the demand for the quali-
ties generally associated with steel should be confined to what was
known as cast steel. The results of the experiments in testing
small cylinders with gunpowder, as shown in the table accom-
panying the paper, were most instructive, and appeared to leave
very little room for doubt that mild steel must be the proper
material for the production of guns. Whether guns made of one
block of steel like the old cast-iron guns, or those having an inner
core hooped with steel rings, were the best, was a question for
gunmakers to solve, and must depend mainly upon the dimensions
of the structure. But the results now arrived at sufficiently
showed what extraordinary strength and ductility combined might
be obtained in constructing guns of a solid block of steel ; and
he hoped the Admiralty and the War Office would soon sec their
way to affording this system a thorough trial.

Y 2

324 THE SCIENTIFIC PAPERS OF

In the discussion of the Paper

" ON PRICE'S PATENT RETORT FURNACE,"
By ME. I. LOWTHIAN BELL, M.P., F.R.S., Middlesborough,

DE. SIEMENS * said, that when he first saw the paper on that
furnace, he could hardly think it was Mr. Isaac Lowthian Bell
who had written it, because they all knew that Mr. Bell was
thoroughly conversant with the phenomena of heat and the
chemical actions that occurred in the combustion of fuel. Yet
he found in that paper, prominently put forward, a comparison of
Price's furnace with his regenerative gas furnace, in which it was
stated that there was an absolute loss of 30 per cent., arising from
the action of the gas-producer in dealing with the carbonaceous
matter while reducing it to the state of carbonic oxide, before it
passed into the furnace. Seeing this statement in Engineering^
he took the opportunity of calling attention to the fact that the
gas-producer had several duties to perform. The carbonaceous
matter contained in the coal was burnt to carbonic oxide, but in
addition to that, all the hydrocarbons were developed from the
coal, and besides a quantity of vapour of water being decomposed
into its elementary substances, and thus a considerable amount of
the heat that would otherwise be lost, was simply transferred from
the state of sensible heat to that of latent heat, in order to be
re-developed in the furnace. Considered from this point of view,
it would be found that the loss in the gas-producer was not 30 per
cent, as stated, but about 12 ; it was a subject that could be
thoroughly cleared up by theoretical argument, and by measuring
the temperature of the gases as they came from the gas-producer.
He was glad to find Mr. Bell admitting that he was wrong in his
calculation, but still he did not seem disposed to accord to the
regenerative gas furnace the reputation for economy which it has
acquired. The regenerative gas furnace was sufficiently well
known, and its economy was proved in various applications, and
he would not take up the time of the meeting by reasserting its

* Excerpt Journal of the Iron and Steel Institute, 1875, pp. 496, 497.

WILLIAM SIEMENS, F.R.S. 325

merit. In liis letter, he did not make comparisons between the
t\v»i furnaces, he merely endeavoured to correct an error which
Mr. I'.ell li;ul incidentally fallen into. Rc^anlin^ Price's retort
furnace, IK- ini^ht observe, that at one time he had retorts in con-
nection with his own gas-producers when he found that he could
not get them gas-tight, and also that if, as in Price's furnace, the
coal passed through the retort, surrounded by the spent gases on
tin ir way to the chimney, the spent gas would speedily take up
all the gas that leaked through the joints, and make away with it
Dzioonsnmed ; moreover, it was very difficult to construct a vertical
retort in such a way, and to heat it to that nicety that the coal
would part with its hydrocarbons, and would not adhere to the
sides. The probability was that if they gave that furnace into
the hands of an ordinary fireman, the retort would soon be choked,
the bricks get overheated, and the whole thing come to grief.
How such accidents were prevented in Price's furnace did not
appear ; probably at Woolwich, where extreme care was bestowed
upon it under the author's own eyes, and where any failure could
readily be put right, without counting much the cost of doing it,
it might answer for a certain time, but judging by his own
experience, he should be apprehensive of a considerable amount of
difficulty if such a furnace were put into ordinary hands. The
furnace, while working, would no doubt give a fair result, and
develop a considerable amount of temperature.

In the discussion of the Paper

"NOTE ON THE MANUFACTURE OF ANTHRACITE
COKE IN SOUTH WALES,"

By ME. W. HACKNEY,

DR. SIEMENS* said, he could fully substantiate the results
which Mr. Hackney had stated in his paper as having been

* Excerpt Journal of the Iron and Steel Institute, 1875, pp. 530-532.

326 THE SCIENTIFIC PAPERS OF

obtained at Landore ; but it must be remembered that the coke
which had been used hitherto there was very weak coke, and it
would be unfair, perhaps, to expect the same difference of results
if that anthracite coke were tried, for instance, in the Durham
district where coke of a much harder kind was obtained by direct
distillation of coal. It would be admitted by the author that the
process of making artificial coke from anthracite and binding
material was, of course, not new. He (Dr. Siemens), had first
seen it in operation many years before at the Creusot Works, and
he had then formed a very favourable opinion of the coke produced ;
and when about seven years ago he had been requested by the
Landore Siemens Steel Company to design two blast furnaces, he
designed them with the idea that such coke would be used. The
question which had then most occupied his mind was what height
he ought to give to those furnaces. It was at a time when the
general opinion of iron-masters was very much in favour of high
furnaces. Some had suggested 70 or 80 feet furnaces. Others
had said that GO feet was the least height that should be employed.
His idea had been to make the furnace only 52 feet high, as being
the height most suitable to the fuel at his command in the locality ;
but he had added another 2 feet, and the first furnace was made
54 feet high. Nothing had been changed since in the proportions
of the furnaces. The blowing engines worked splendidly, and the
Cowper stoves had given no trouble, and yet the result had not
been such as he had anticipated. That was due entirely to the
weak coke that had been employed, and though before they started
those furnaces all the machinery had been ordered for making
anthracite coke, their erection had been postponed, as coke of that
description was at the time not well known in this country. It
had been a very fortunate circumstance, therefore, that Messrs.
Penrose and Richards had taken up the question of making an-
thracite coke. They seemed to have come to the conclusion that
to make hard anthracite coke, besides binding coal, a high binding
material, such as pitch, was required ; and he (Dr. Siemens),
willingly admitted that pitch was necessary, in addition to binding
coal, if it was their intention to make coke very rich in anthra-
cite ; but he was not quite so certain, notwithstanding the results
which they had recorded, that it was desirable in all ca*ses to have
coke containing quite so much anthracite as 60 per cent. He

.S7A' WILLIAM SIEMENS, F.R.S. 327

looked upon tin- l>last furnace as requiring a fuel which would live
down to the crucible, and there perish. If they were to charge
graphite, they would get it all into the crucible, and have to tap
it out with the slag, unconsumed ; and what Professor Williamson
had just said showed the great resistance of hard coke to combus-
tion, and its great advantages in certain respects ; but they might
ha\v too much hardness and so make the coke imperishable in the
furnace. He had therefore proposed that before deciding upon
the proportion of anthracite, comparative trials should be made,
using coke containing a very large per ceutage of anthracite,
formed with pitch, and weaker coke containing only from 25 to 30
per cent, of anthracite, made without pitch. He had with him a
piece of coke containing 25 per cent, of anthracite and no pitch,
which appeared to him probably hard enough for their height of
furnace and their blast ; but he might be mistaken. It might be
that the harder coke gave still better results ; but he must say
that if the hardness exceeded a certain limit, he expected that a,
large amount of carbonaceous matter would come out with the
slag, and in fact, their experiments already showed that evil to
exist to some extent, but they had hardly been carried on for
a sufficient length of time to say definitely to what extent.
There could be no doubt whatever that the production of coke
of such hardness that it would resist the action of hot car-
bonic acid in the upper portion of the furnace and furnish
carbonizing material down to the very bottom of the furnace,
was a matter of very great importance. One result which had
been observed and which Mr. Hackney had not stated, as far
as he could gather from his paper, was this, that while they
were troubled with sulphur in using their ordinary coke, which
contained a good deal of sulphur, the sulphur disappeared as by
magic the moment the anthracite coke came down to the bottom
of the furnace, producing, in fact, No. 1 pig metal with only a
trace of sulphur, instead of No. 4 pig, containing a very large
proportion of that objectionable ingredient. The coke now
brought before them had been produced in the horizontal coke
oven, commonly used in South Wales, but he thought that u
still denser quality would be produced in vertical retorts of the
Appold type, and in making his designs for the Landore Com-
pany, this had been the form of coke-oven he , had specified ;

328 THE SCIENTIFIC PAPERS OF

his reason having been that the powdery material of which an-
thracite coke was constituted, required a certain degree of com-
pression to bring the particles into contact, and that the absence
of this pressure caused the coke to be more or less spongy near
the upper surface of the mass.

In the discussion of the Paper

"ON A METHOD OF PRODUCING PURE CHARCOAL
STEEL DIRECTLY FROM THE ORE,"

By HENRY LAEKIN,*

Dr. SIEMENS, F.R.S.,* said the old Sheffield method of making
cast steel might be called a method by error. They took the ore
and smelted it in a blast furnace, turning out a product which
consisted of iron mixed with slag, and a great deal of carbon, all
the carbon that could be gathered from the ingredients put into
the furnace. All that foreign matter had then to be removed, and
so it was puddled. The result was iron mixed with slag, but with-
out any trace of carbon, or at any rate in no definite combination
therewith. This was not what was wanted, which was iron com-
bined with carbon, or steel, and so it was again subjected to a very
different process, termed cementation, whereby it was kept heated
for a fortnight in a furnace, packed with charcoal. This gave a
metal which was certainly steel, but it was mixed with slag and
full of blisters, so that it was not yet fit to be used. After some
years they adopted the method of fusing it in pots, and so pro-
duced cast steel such as was used in Sheffield to the present day.
Any means of producing this material by a short process must be
of great importance to the country, since steel was becoming more
essential every day. Several methods of producing steel direct from
the ore had been mentioned, but others had been omitted. He him-

* Excerpt Journal of the Society of Arts, Vol. XXIV. 1875-1876, p. 93.

5/7? WILU 'AM SIEMENS, F.R.S. 329

self introduced the system of reduction by a rotator in 1868, and
illicit state the results he obtained. The cheapest and most
direct method of reducing iron ore was in a rotary furnace ; the
iron ore and the carbonaceous material were introduced at one
end of a large rotating cylinder, and brought out in the condition
of spongy iron continuously at the other end ; nothing could be
simpler or cheaper, but still he saw reason to abandon it. First,
because the spongy iron thus obtained absorbed in the very process
a good deal of sulphur, and this could not be remedied unless you
could prevent its contact with the sulphurous acid produced in the
combustion of the coal. Then, again, it was not easily dissolved
in the bath of metal, even if that were sufficient, because it floated
on the top, and rapidly oxidised under the influence of the flame
of the furnace. Moreover, unless you used such materials as could
not practically be employed on a large scale, the spongy iron
became enclosed in the gangue of the ore, and would not melt in
the metallic bath, the gangue forming a screen which prevented
intimate contact between the fluid and the solid sponge. He
therefore modified the process without abandoning the principle of
the rotary furnace, and now raised the heat to the point of
melting, not the ore but the gangue contained in it, and thus
effected the reduction. The result was a metallic ball enveloped
in scoria, which did not absorb sulphur, and which could be
hammered out into bars, or transferred to the open hearth furnace
and melted down into steel. He believed that would be found the
cheapest and most direct method. He alluded chiefly to the pro-
duction of mild steel in large quantities ; with regard to tool steel
he was not sure if the presence of a little foreign matter, such as
silicon and sulphur, should be entirely avoided. The famous hard
steel of India contained a considerable proportion of these
elements, and generally compounds ol' iron were harder than the
pure metal.

330 THE SCIENTIFIC PAPERS OF

In the. discussion of the Paper

"OX SOME RECENT METALLURGICAL PROCESSES,"
By J. A. PHILLIPS,

THE Chair was taken by Mr. C. W. SIEMENS,* who in closing
said : The process which Mr. Phillips had brought before them was
one of many which distinguished the metallurgy of the present
day. Ancient methods were always found to be very wasteful,
both in fuel and materials, because it never occurred to the ancient
metallurgist to separate the silver or gold from the copper, or to
make use of the iron as a secondary process. He selected his ore
for the one purpose he had in view, whether it was copper, silver,
or iron, and used his fuel very wastefully in extracting it. This
process was conducted at the expense of the iodide, which had
only lately come into use mostly through photography, and at
first sight he had been struck by the apparently great expense
attaching to it. The discussion, however, had shown that this
was really of very little importance, because, though iodine was
relatively expensive, the process was so perfect that it was used
over and over again. This reminded him of another process
lately come into use for the production of soda, where another
expensive material, ammonia, was used over and over again to
produce carbonate of soda directly from common salt, without the
intermediate stages of soda ash, soda cake, and so forth. The
question had been asked whether the extraction of silver dete-
riorated the value of the copper, and the answer was that it did
not, but rather enhanced its value, and he could fully believe that.
Lord Palmerston once defined dirt as "matter in the wrong
place," and certainly silver in the copper must be reckoned as
dirt, and those who dealt with copper as a metallic conductor of
electricity regarded it as dirt of a very objectionable character, for
a very slight admixture of silver or any other foreign matter in
the copper would reduce its conductivity to perhaps one-fourth of
the proper amount. There could be no question, therefore, that

* Excerpt Journal of the Society of Arts, Vol. XXIV. 1875-1876, p. 291.

.S7A' WILLIAM SIEMENS, F.R.S. 331

it was advantageous to produce the metals separately and in their
utmost purity. With regard to the use of the residue in the blast
furnace, with all respect to Mr. Ililey's opinion, he still thought
copi><T was an objectionable element in wrought iron ; for if any
considerable amount of copper, say ^ per cent., got into a puddled
bar, it did not behave well under the hammer, and therefore the
copper must be classed in that case with Lord Palmerston's dirt.
He had had these residues examined repeatedly for sulphur and
for copper, and found they yielded on the average '6 per cent.,
which was equal to 1 per cent, of the metallic iron contained in
the ore. Therefore, although this material might be extremely
useful in the blast furnace to assist in producing quantity, it
ought to be used with great caution, in order not to increase
unduly the proportion of copper or sulphur. With regard to the
extraction of silver, he only hoped, for the benefit of this process,
that the late discoveries in California, which it is said would
reduce the price of silver to an almost nominal amount, would
not interfere with its financial development. He concluded by
proposing a vote of thanks to Mr. Phillips, which was carried
unanimously, and the proceedings terminated.

In t/ie discussion of tlie subject

"THE USE OF MOLTEN IRON DIRECT FROM THE
BLAST-FURNACE FOR BESSEMER PURPOSES,"

Introduced by Mr. J. T. SMITH,

DR. SIEMENS * said he hoped the two processes would live side
by side and help each other, instead of being considered antagonistic.
He should not have troubled them with any remarks of his own,
except for the remarks which had fallen from Mr. Hackney. When
the works at Landore were contemplated, he (Dr. Siemens) was

* Excerpt Journal of the Iron and Steel Institute, 1876, pp. 33, 34.

-THE SCIENTIFIC PAPERS OF

asked to prepare plans. One arrangement proposed was to run the
metal from the blast furnaces that were intended to be erected,
directly into an open hearth furnace. Those furnaces were to have
been worked with anthracite coke — coke similar to what was used
at Creusot — and he had no doubt that if coke of that description
had been employed at the time, instead of that made from the coal
available at the works, pig of sufficient purity to be run at once
into the steel-melting furnaces would have been produced. He
believed that with weak impure coke it would be impossible to
obtain regularity and such freedom from sulphur in the pig as
would give satisfactory results, and it appeared to him that the
success of that — invention he would not call it, because it was
contemplated by Mr. Bessemer, and it must have occurred to most
of them as a possibility — that the success of that addition or im-
provement in the Bessemer process, or the process with which his
own name was connected, must depend upon the working of the
blast furnace, and if they could get such results as had been pro-
duced at Seraing, at Creusot, and as no doubt were produced now
in this country, there was a clear margin of benefit ; but in order
to attain success, it would be necessary to work the blast-furnaces
with greater care than was usually exercised. At the time when
he contemplated using pig metal from the blast-furnace, he thought
it preferable to work with an open top ; in fact the blast-furnace
specified was almost identical with what Mr. Whitwell had just
stated was now used at Seraing. The open top was preferred,
because it gave them greater power of regulating the charge, and
he would throw out the suggestion to those eminent iron and steel
makers who were there present, whether it would not be worth
while to make a comparative trial between the open top and the
close top furnaces. No one would be able to give them sounder
advice on this subject than his friend Mr. Bell. A good metal
could no doubt be produced with a close as well as by an open top,
but it appeared to him (Dr. Siemens) there was a greater margin
of safety in working with the open top. With regard to the ex-
periment which Mr. Hackney had tried in using pig metal re-
melted with impure coke in a cupola, he did not wonder at the
want of success that had attended it for two reasons : first, the
molten metal would contain a great deal more sulphur than there
was in the pig charged, especially if No. 3 or 4 pig metal was

S/A WILLIAM SIEMENS, F.R.S. 333

used, as was the case at Landore, and secondly, it would be pro-
(l:i<-i'<l at a comparatively low temperature. Much also would
depend upon the condition of the steel melting furnace at the time
when the fluid metal was run into it ; if the furnace was in a
rumpurativcly oool condition (for instance, if the bottom had just
made up with cold material, and the comparatively cold pig-
metal from the cupola was run in), then considerable time must
be lost before both the furnace and the fluid pig metal were
brought up to the temperature at which the raw ore would act
upon the latter, and this time might be quite equal to that required
for fusing pig iron in the melting-furnace ; but if the furnace was
heated to full steel melting heat before the molten pig iron was
run in, and if the metal was taken direct from the blast-furnace
at the highest temperature attainable, he had no doubt that an
advantage would be realised analogous to that which had been
obtained by the Bessemer process.

In the discussion of the Paper

"ON THE CASSON-DOKMOY PUDDLING-FURNACE,"
By MR. E. FISHER SMITH,

DR. SIEMENS * said Mr. Scattergood had given them a somewhat
amusing account of the great precautions he used in making his
experiments. He trusted to nobody but himself ; but it struck
him (Dr. Siemens), in listening to his observations, that in one or
two particulars perhaps he had been over-cautious. He was not
satisfied with puddling the Earl of Dudley's iron, but took iron of
his own — pig metal of a very inferior kind — with a view of putting
the furnace upon its trial ; but it appeared to him (Dr. Siemens)
not improbable that that very circumstance might have contributed
to the superior results which he reported, viz., the shortness of
time which it took to bring his iron to nature, and to the com-

* Excerpt Journal of the Iron and Steel Institute, 1876, pp. 134-135.

334 THE SCIENTIFIC PAPERS OF

paratively small consumption of coal. Inferior irons were often
puddled much more rapidly.

Mr. Scatiergood said it was not white iron, but grey common
iron.

Dr. Siemens, continuing, said that even then it depended very
much upon the constitution of that iron, upon the silicon it
contained, and the way in which carbon was combined with
iron ; but as regarded the time necessary to bring pig-metal to
nature, another observation which the same speaker made struck
him as being contrary to generally accepted views. Mr. Scattergood
said that less fettling was required than in the ordinary furnace,
because there was less surface, and he went on to say that if the
furnace could be constructed without any sides, no fettling would
be required. That perhaps was a slip only, because they all knew
without fettling they could not carry out the puddling process ;
they must have oxide of iron in order to oxidise the carbon and
silicon in the pig and bring it down to the condition of wrought
iron. He wished to make a few observations on the paper itself.
Mr. Smith very fairly stated that the furnace contained no
striking element of novelty. Of course the dandy was a well-
known thing ; it had often been tried in this country, and it had
never found favour with the men. On the Continent it was used
almost universally, and it produced no doubt very considerable
economy. Blast under the grate was used, moreover, to a very
large extent, and the points of novelty in the furnace before them
seemed to reduce themselves to the' round form of the puddling
chamber and the mode in which that chamber was cooled. Now,
with regard to the cooling of the chamber, he might observe that
he had once tried a similar arrangement without success. He had
placed under the puddling chamber an open vessel containing
water very similar to what had been shown, but he had had to
give it up very soon, because it was found that the vapour rising
from the water found its way through the joints into the puddling
chamber and affected the iron prejudicially. He should like to hear
Mr. Smith's observations on that point, and whether or not he
found any inconvenience on that score ; also with regard to the
inclined side of the puddling chamber, whether it had been found
that the side held the fettling as well as the vertical side, his
(Dr. Siemens's) own observation being that on the sloping side the

S7K WILLIAM SIEMENS, F.R.S. 335

IHtling — unless it was of a very strong nature, that is to say of
it very pure oxide — slipped down and did not stand up so well to
i:s \vork as on the vertical side.

In Vie discussion of the Paper

"ON IMPROVED CASTING ARRANGEMENTS FOR

THE SIEMENS-MARTIN PROCESS,"

By ME. MICHAEL SCOTT,

DR. SIEMENS * said the paper referred to a subject which was
one of considerable practical interest. At present they usually
tapped the metal into moulds, as they were aware, which were
stoppered at the top, and there was a good deal of uncertainty
as to the amount of metal which they filled into these moulds.
The metal, if it was at all frothy, boiled and gave only uncertain
indications of its height in the moulds, which made it difficult for
the workmen to stop the flow of fluid steel at the right moment.
There was another effect resulting from the mode of filling at the
top, viz., atmospheric air was carried in and mixed with the steel,
to be liberated again at the moment of solidification, and thus
giving rise to spongy metal. It was well known in ordinary practice
that gases were set free at the moment of congelation in the moulds ;
the steel seemed to take up air, which was, he thought, combined
with iron, forming a sub-oxide, which oxide of iron appeared to be
dissolved throughout the mass, without combining with the carbon
contained in the steel at the high temperature. The temperature
of steel being in excess of the point at which carbon and oxygen
combined, such metal might be kept for any length of time in a
fluid condition, at that temperature, without getting rid of the free
oxygen which was evolved at a lower temperature in the moulds.
It had been the endeavour of several gentlemen, both in this

* Excerpt Journal of the Iron and Steel Institute, 1876, pp. 161-163, 165-
168.

336 THE SCIENTIFIC PAPERS Of

country and abroad, to overcome these difficulties by filling the
mould from below, amongst which he would only mention Pinks's
process, which had received a considerable amount of public
attention, and realised to some extent the advantages of not
exposing the metal to the oxygen of the atmosphere, and of getting
the ingots of definite size and shape. There had been several
difficulties met with, however, in the attempts made to realise those
advantages, and Mr. Scott, who had had a considerable amount of
experience in the matter, had devised a mode of overcoming those
difficulties by running his metal through refractory material, which
could be easily heated up beforehand, and which was surrounded
by a non-conducting substance, namely, sand, and he had thus
overcome one of the difficulties, that of the metal setting in the gits,
and of not filling moulds right up to the top. Another part of the
arrangement which Mr. Scott brought forward was to have the
moulds divided longitudinally instead of casting them solid. That
really was going back to the old Sheffield plan of making the
moulds, and a very good plan it was. The difficulty connected
with it, in using such large moulds, was to fit the two parts well
together, and that Mr. Scott proposed to do by an elastic clamp,
which would bring the surfaces home, one against the other with
a great degree of certainty. He (Dr. Siemens) did not know that
the plan in its entirety had yet been tried, but he learnt, and in
fact knew, that it had been partially tried, and with, he believed,
very encouraging results. In fact, there could be no doubt that
the arrangement would work, and it was only by continued
practical experience that the exact amount of saving and con-
venience, as compared with present modes of working, could be
ascertained, but, as far as he could see, these advantages would be
largely in excess of any corresponding drawback arising out of the
method of arranging the ingots as Mr. Scott had shown.

Dr. Siemens asked Mr. Snelus whether he considered that dead
melting depended upon manganese, as if so, it was certainly quite
a new principle to him. He had looked for the cause of dead
melting in quite another direction, and might mention, in support
of his view, that the Sheffield pot steel contained comparatively
little manganese, at all events much less than Bessemer metal,
which he mentioned, not with a view to disparage it, but as the
metal with which they were most familiar. His view regarding

WILLIAM SIEMENS, F.R.S. 337

dead melting was, that it was chiefly a question of temperature,
and tin- ivsuli of maintaining the steel for a considerable time in
a fluid condition at a comparatively low temperature, at which the
which was chemically combined with iron would leave the
and go to the carbon, and escape as carbonic oxide, gradually
re the metal was poured. If Mr. Snelus had any other decided
view on that subject, it would be important, he thought, that it
should be brought forward.

Mr. Snelux finished his reply with the words : — If they put
plenty of manganese into the steel, he believed the temperature at
which the metal could occlude the gas was raised, and therefore
the whole of the gas came off. The alloy with manganese appeared
to have a less power of holding those gases in solution than the
alloy without it. That was his idea upon the subject.

Dr. Siemens said he should like to be allowed to express his
views on that point. They all knew Mr. Snelus had great ability
as a chemist, and his view on such a subject was entitled to the
utmost consideration ; at the same time, on a purely scientific
point, each man was entitled to his own opinion until he saw
sufficient reason to change it, and his view was in effect that the
amount of gas held by liquids in solution decreased generally with
the increase of temperature — for instance, they drove carbonic acid
out of soda water when they heated it, and their champagne got
flat when it had not been kept cool before it was poured out. In
the same way he should argue that if it was simply a question of
occlusion of carbonic oxide gas in the fluid steel, that the gas would
be driven out in a greater measure at the high temperature in the
Bessemer converter than it was at a much lower temperature just
before the metal was going to congeal in the mould ; and a further
point which he would mention in favour of his view was, that in
Sheffield it was the practice to lower the temperature of the pot
furnace during the latter portion of the operation of melting, and
that that lowering of the temperature produced the effect of dead
melting. After all, the effect would be similar if Mr. Snelns's
view could be proved by absolute chemical test. It would simply
show that occlusion in steel took place to a greater extent at a
higher temperature than at the lower ; whereas, according to his
(Dr. Siemens's) view, the oxygen at the high temperature was
combined with iron, and its combination with carbon was due to

VOL. r. z

338 THE SCIENTIFIC PAPERS OF

a somewhat lower temperature. Now, Mr. Snelus said, in opposi-
tion to that view, that they could never finish a Bessemer blow
under those circumstances. He did not quite agree with him.
there. In finishing a Bessemer blow, the carbon first went out,
because the temperature was comparatively low, and towards the
end of the blow he expected very little carbon combined with
oxygen, when iron was chiefly acted upon, producing intense heat,
a fusible silicate which in its turn would react upon the carbon to
a certain extent. They all knew that carbon, which might exist
in a proportion, say, of t^ths per cent, in pure metal at that tem-
perature, would not be able to exist in the same percentage in the
presence of oxide of iron, and it was all a question of how much
oxide of iron was necessary to push out the carbon. According to
this view, the amount of carbon remaining in the Bessemer metal
after the blow would depend upon the amount of oxidation of the
iron which had taken place, but he should be much surprised to
learn that the Bessemer metal at the end of the blow was entirely
without carbon.

In the discussion of the Paper

"ON THE PERMANENT WAY OF RAILWAYS,"
By R. PEICE WILLIAMS, M. Inst. C.E.,

DR. SIEMENS * said there seemed to be a great divergence of
opinion with regard to the quality of steel composed of different
percentages of carbon. Mr. Riley, who had given considerable
attention to these questions, no doubt could have supplemented
the information by further chemical details, which perhaps he did
not like to do because there was at present doubt as to the specific
effect produced by other materials, such as phosphorus, sulphur,
silicon, &c. Dr. Siemens might perhaps be able to supply some

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers,
Vol. XLVI. Session 1875-1876, pp. 203-205.

S/X WILLIAM SIEMENS, /.A'..s. 339

!v._rar«ling steels in which the proportions of different mate-
rials miclu-.l tlu-ir limits on the one side or on the other. It had
been said that steel containing a considerable percentage of carbon,
i>- 1 per rent., was very soft. That might be partly owing to the
tempering, for if the steel had been slowly cooled it would be soft,
but it might also be owing to the manner in which the carbon was
combined. Carbon was not always chemically combined even in
steel. In the process of mixing spiegeleisen with the blown metal
at the last moment in the Bessemer process there was hardly time
to form a chemical combination, and hence metal might be pro-
duced which contained a considerable proportion of carbon, and
yi't was essentially soft metal because the carbon was not chemi-
cally combined. He had observed, in boring a large cylinder of
steel cast from metal of that description, that at one point the
boring tool went deeply into the metal, and at another point there
was a resistance as though the metal were hard. In such a case
no effect of temperature, of sudden cooling, was involved, because
of the particular form of the metal. The phenomena showed
clearly that steel was not always homogeneous unless special care
had been used to make the mixture perfect. Mr. Cowper had
attributed to manganese the quality of making the metal tough,
whereas carbon made it hard. That was not universally the case.
In the case of steel containing as much as 0'4 or 0*5 per cent, of
carbon, little manganese was desirable or necessary to make the
metal forge properly ; but in the case of very soft metal, and metal
containing phosphorus, the manganese was an essential condition.
For instance, 0*2 per cent, of phosphorus, which was about the
maximum amount admissible, necessitated about 0*4 per cent, of
manganese to make the metal at all workable and to prevent its
extreme brittleness when hard ; but in metal containing hardly
any carbon, less than 01 per cent, manganese was of the utmost
importance, and without it the metal could not be got to work or
stand against the grinding axle. The manganese seemed to bind
the particles of metal together, and to make it more homogeneous.
Mr. Riley had mentioned the difficulty of determining such a slight
amount of carbon by the ordinary colour test. No doubt the diffi-
culty was great, but Mr. Willis, of the Landore "Works, had intro-
duced a test for determining a very slight proportion of carbon,
sufficiently accurate for all practical purposes, by dissolving the

Z 2

340 THE SCIENTIFIC PAPERS OF

metal in nitric acid and observing the effect of colour on the solu-
tion. The nitric acid did not act upon the carbon, and it produced
a brown tinge of sufficient intensity to be distinctly observable,
even though the proportion of carbon should be considerably below
O'l per cent. This test would enable manufacturers to produce
steel of extreme mildness, such as was now used largely for engi-
neering purposes. It was impossible to estimate so slight a differ-
ence in carbon as 0'05 or O'l per cent, by any mechanical test ;
but by a ready chemical test of that kind the manufacturer was
enabled to bring the metal down to an exact point considerably
below O'l per cent. No doubt these questions were at present
only partially understood, and much work was still required in
order to arrive at certain conclusions regarding the conditions
necessary to produce steels of definite qualities. If steel was pro-
duced from pure iron, such as was found in Sweden, a metal could
be made containing hardly any manganese, and only a trace of
carbon, and no appreciable percentage of phosphorus or silicon ;
and metal of that description, which was as nearly as possible of
pure iron, was perhaps the toughest that could be made, exceeding
in toughness copper or even silver. A further advantage of such
metal was that it was less liable to rust. Experiments lately made
in France showed a percentage of rust, as compared with wrought
and cast iron, in the proportion of 0'4 to 1'4, being much in favour
of a pure metal such as a very mild steel — a quality of great value
for shipbuilding and purposes of that kind.

SOME FURTHER REMARKS REGARDING THE PRODUC-
TION OF IRON AND STEEL BY DIRECT PROCESS.
BY MR. C. WILLIAM SIEMENS, D.C.L., F.R.S.*

IN mixing comparatively rich iron ore in powder, with about 25
per cent, of its weight of pounded coal, and in exposing this mix-
ture for some hours to the heat of a common stove or of a smith's

* Excerpt Journal of the Iron and Steel Institute, 1877, pp. 345-359.

SJX WILLIAM SIEMENS, l-.R.S. 341

fire, metallic iron is formed, which, on being heated to the welding
p lint, on the same smith's hearth, may be forged into a horseshoe
i.t' cxirllcnt quality. The admixture with the ore of some fluxing
materials, such as lime or clay, will, in most cases, be of advantage
to rid the iron of adherent slag.

The simplicity of this process is such that it naturally preceded
the elaborate processes now in use for the production of iron and
suvl upon a gigantic scale, nor can it surprise us to find that
attempts have been made from time to time, down to the present
day, to revert to the ancient and more simple method. It can be
shown that iron produced by direct process is almost chemically
j.uiv, although the ores and reducing agent employed may have
contained a considerable percentage of phosphorus, sulphur, and
silicon, and that, if freed from its adherent slag, it furnishes a
material superior in quality and commercial value to the ordinary
iron of commerce.

The practical objections to the direct process, as practised in
former days, and as still used, to a limited extent, in the United
States of America and in some European countries, are that —

1. Very rich ores only are applicable, of which about one-half is
converted into iron, the remainder being lost in forming slag.

2. The fuel used is charcoal, of which between three and four
bons are used in producing one ton of hammered blooms.

3. Expenditure of labour is great, being at the rate of 33 men,
working twelve hours, in producing one ton of metal (see Percy).
Iron produced by direct process in the Catalan forge is therefore
expensive iron, and could not compete with iron produced by
modern processes except for special purposes, such as furnishing
melting material for the tool steel maker.

But, it may be asked, could not the advantages of the direct
process be combined with those of modern appliances for the pro-
duction of pure and intense heats, and for dealing with materials
in large masses, without expenditure o.f manual labour, and cannot
chemistry help us to obtain larger yields and facilitate the using
of comparatively poor and impure ores ?

A careful consideration of these questions led me to the con-
clusion, some years ago, that here was a promising field for the
experimental metallurgist, and that I possessed some advantage
over others in the use of the regenerative gas furnace as a means

342 THE SCIENTIFIC PAPERS OF

of producing the requisite quality of heat without the use of
charcoal and blowing apparatus. I engaged, accordingly, upon a
series of experimental researches at my sample steel works, at
Birmingham, and, in 1873, I had the honour of submitting the
first fruits of these enquiries to the Iron and Steel Institute, in a
paper entitled " On the Manufacture of Iron and Steel by a Direct
Process." Encouraged by the results I had then obtained, I ven-
tured with others upon some larger applications, the principal one
of which has been made at Towcester, in Northamptonshire.

If viewed by the light of present experience, it would have been
wiser to have fixed upon another locality with fuel, skilled labour,
and better ores within easy reach ; but in extenuation of the error
committed, it may be urged that the site was fixed by force of cir-
cumstances rather than by selection, the chief temptation being
an ample supply of small Northamptonshire ore at a very low cost.
It was, however, soon discovered that this ore, although capable
of producing iron of good quality, was too poor and irregular in
quality to yield commercial results unless it was mixed with an
equal weight of rich ore, such as pottery mine, Spanish ore, or
Roll-scale, all of which, as well as the fuel, are expensive at Tow-
cester, owing to high rates of carriage. It is in consequence of
these untoward circumstances that the works at Towcester have not
been completed by the addition of rolling mills, the intention being
to transfer the special machinery ultimately to existing ironworks
when the process has been sufficiently matured for that purpose.

The Towcester Works were visited, in the autumn of last year,
by two eminent metallurgists, Professors Von Tunner, of Leoben,
and Akermann, of Sweden, who have published the results of their
observations in separate reports.* The results noted down by Mr.
Von Tunner are referred to by our past-president, Mr. I. Lowthian
Bell, in his paper on the " Separation of Carbon, &c.," which was
read in March last, and will be discussed at the Newcastle meet-
ing. The criticisms contained in these publications are conceived
in the fairest possible spirit, and form indeed a most valuable
record of the progress achieved up to that time, but they furnish
me with an inducement to break silence sooner than I had in-
tended, regarding the further progress which has been effected,

* Das Eisenhiittenwesen von L. Ritter von Tunner, Wien, 1876.

SIR WILLIAM SIEMENS, J-.K.S. 343

and the conclusions I am disposed to draw, from past experience,
regarding the direct process of the future.

The leading idea which guided me in these researches was to
operate upon such mixtures of ores, fluxes, and reducing agents as
would, under the influence of intense heat, resolve themselves
forthwith into metallic iron and a fluid cinder, differing essentially
f roiii the methods pursued by Chenot, Gurlt, Blair, and others,
who prepared spongy metal in the first place, by a slow process,
which is condensed into malleable iron or steel by after processes,
but assimilating to some extent to the process first proposed by
.Mr. \Vrn. Clay. In my paper of 1873, I described two modes of
effecting my purpose, the one by means of a stationary, and the
other by means of a rotative furnace chamber, the former being
applicable chiefly where comparatively rich ores are available, and
the latter for such poorer ores as occur near Towcester.

At the Towcester works, three rotative furnaces have beeu
erected, two of them with working drums 7 ft. in diameter, and
9 ft. in length, and the third of smaller dimensions. The gas
flame both enters and passes away from the back end of the fur-
nace, leaving the front end available for the furnace door, which
is stationary. The ends of the furnace chamber are lined with
Bauxite bricks, and the circumference with iron oxides, resulting
from a mixture of furnace cinder enriched with roll-scale, or
calcined blackband in lumps. About 30 cwts. of ore mixed with
about 9 cwts. of small coal having been charged into the furnace,
it is made to rotate slowly for about 2£ hours, by which time the
reduction of the metal should be completed, and a fluxed slag be
formed containing the earthy constituents and a considerable per-
centage of ferrous oxide. The slag having been tapped, the heat
of the furnace and the speed of rotation are increased to facilitate
the formation of balls, which are, in due course, taken and treated
in the manner to be presently described.

These balls contain on an average 70 per cent, metallic iron and
30 per cent, of cinder, and upon careful analysis it is found that
the particles of iron, if entirely separated from the slag, are pure
metal, although the slag may contain as much as G per cent, and
more of phosphoric acid, and from 1 to 2 per cent, of sulphur. In
shingling those balls in the usual manner, the bulk of the cinder
is removed, but a sufficient residue remains to impart to the

fractured surface a dark appearance without sign of crystallisation.
Upon being worked, the metal shows what appears to be red-
shortness, but what should be termed slag-shortness. In re-piling
and re-heating this iron several times this defective appearance is
gradually removed, and crystalline iron of great purity and
toughness is produced. But a more ready mode of treatment was
suggested by Mr. Samuel Lloyd, one of my co-Directors in the
Towcester Company, in reverting to the ancient refinery or char-
coal hearth. The balls as they come from the rotator, are placed
under the shingling hammer and beaten out into flat cakes not
exceeding an inch in thickness. These are cut by shears into
pieces of suitable size and formed into blooms of about 2 cwts.
each, which are consolidated under a shingling hammer and
rolled into bars.

The bars have been sold in Staffordshire and Sheffield at prices
varying from £7 to £9 per ton, being deemed equal to Swedish
bar as regards toughness and purity.

It may therefore be asserted, as a matter of fact, that iron and
steel of very high quality can be produced from ores not superior
to Cleveland ironstone by direct process, but the question remains
— at what cost can this conversion be effected ? The experi-
mental works at Towcester are unfortunately not sufficiently com-
plete to furnish more than the elements upon which the question
of cost may be determined, the principal reasons being that the
one re-heating furnace and a 30 cwt. hammer at the works are not
sufficient to deal with the iron produced by the three rotators, that
the iron has to be finished at a rolling mill elsewhere, and that
transport weighs heavily upon the cost of production. The prin-
cipal factor in the calculation of cost is unquestionably the
rotator, and Table I, furnishes the working result of eighteen
consecutive charges as taken from the charge book. The mixture
of ores consisted for each charge of 12 cwts. of Towcester ore
(containing about 38 per cent, metallic iran), 8 cwts. of calcined
Great Fenton ore, 1 cwt. of tap cinder, 1 cwt. of limestone, and
6 1 cwts. of small coal. The time occupied for each charge was
about 4j hours, and the yield of hammered blooms was, on an
average, 6 cwts. 2 qr. 13 lb., whereas the metal contained in each
charge, amounted (by estimate) to 8 "2 cwts., showing a loss of
19'3 per cent. This loss is, however, partially recovered in using

Sffi WILLIAM SIEMENS, F.R.S. 345

a portion of the cinder again in succeeding charges, but the pro-
l»onion of cinder that may be so used with impunity, depends
upou the amount of phosphorus, sulphur, and alumina contained
in the ore.* The coal used in the producers amounted to 2 tons
per ton of hammered blooms produced, and in pricing the mate-
rials used and labour engaged upon the work, the table — prepared
by the manager at the Works— gives £3 8s. as the cost per ton
of hammered blooms. To this amount must be added repairs
and general expenses, and the cost of rolling the hammered blooms
into bars, which in the case of Towcester practice are very heavy,
but of which an experienced ironmaster would form his own esti-
mate. The cost of working the metal in the hollow fires is also
not included, and this may be taken to add from 25s. to 80s. to
the ton. The refined iron so produced will, therefore, cost from
£5 5s. to £5 10s. per ton.

Tables II. and III. give the analyses of irons produced from
various descriptions of ores, and Tables IV. and V. give Kirkaldy's
tests of the mechanical properties of the iron ; but it should be
under-stood that these tests were taken with a view to ascertain
the influence of different methods of manipulation rather than to
show high results. Only a small proportion of the samples had
been subjected to the refinery process, and the variable percentage
of phosphorus may be taken really as indicative of the extent to
which the cinder had been removed from the metal.

Table VI. gives the analyses of slags produced in the process.
These are no doubt rich in iron, but it must be remembered that
in the case of comparatively pure ore they can be used almost
entirely in succeeding charges, and that in the case of ores con-
taining much sulphur and phosphorus they are the recipients of
those impurities — in the same way as the puddling cinder carries
off the same impurities in the puddling furnace — and thus serve a
useful end.

If rich ores, such as hematites, are available, it is more advan-
tageous to use a stationary furnace, and to modify the process as
follows.

* Since this paper was prepared, further experience has shown the possibility
of obtaining greater yields of iron from the ores employed, and of making, at the
same time, a greater number of charges per 24 hours ; some of these results are
recorded in Tables VII. and VIII., which may be considered as appendices to this
paper.

346 THE SCIENTIFIC PAPERS OF

A mixture is prepared of pulverulent ores combined with fluxing
materials and reducing agents in suitable proportions, and of this
4 to 5 tons are charged from a platform into the heated chamber
to the depth of some 12 or 15 inches. But, before charging the
mixture, some coke dust or anthracite powder is spread ov<jr the
bottom and sides of the chamber for the protection of its silica
lining. The heat of the furnace is thereupon raised to a full
welding heat, care being taken that the flame is as little oxidising
as possible. The result is a powerful superficial action upon the
mixture or batch, causing simultaneous reduction of the ore and
fusion of the earthy constituents. In the course of two hours a
thick skin of malleable iron is formed all over the surface of the
mixture, which, on being withdrawn by means of hooks, is consoli-
dated and cleared of cinder under a hammer, and rolled out in the
same heat into rough sheets or bars, to be cut up and finished in
the refinery furnace or charcoal hearth. One skin being removed,
the furnace is closed again, and in the course of 1 \ hours another
skin is formed, which, in its turn, is removed and shingled, and so
on, until after three or four removals, the furnace charge is nearly
exhausted. A fresh charge is then added, and the same operation
repeated. Once every 12 hours the furnace should, however, be
cleared entirely, and the furnace lining be repaired all round.

The shingled metal, so produced, forms an excellent melting
material for the open-hearth or Siemens-Martin process ; but if
ores, both rich and free from sulphur and phosphorus are used,
together with roll and hammer scale, which form an admirable
admixture, I simplify the process still further in causing the fusion
to take place in the reducing furnace.

The furnace having been charged with say five tons of batch,
the heat is allowed to play on it for four or five hours, when about
two tons of hematite pig iron are charged upon the surface, by
preference in a heated condition. The pig metal on melting con-
stitutes a bath on the surface of the thick metallic skin previously
formed, and gradually dissolves it on the surface while it is form-
ing afresh below, and in the course of from three to four hours
the whole of the materials charged are rendered fluid, consisting
of a metallic bath, with a small percentage of carbon, covered with
a glassy slag containing about 15 per cent, only of metallic iron.
The carbon of the bath is thereupon brought down to the desired

.S7A' WILLIAM SIEMENS, F.R.S.

347

point, of only about 1 per cent, of carbon, and spiegeleisen or ferro-
manganese is added, and the metal tapped in the usual manner.
By these means the direct process of making cast steel is earned
to a further limit than I have been able to accomplish before, and
no difficulty has presented itself in carrying it into effect. The
steel so produced is equal in quality to that produced by the open
hearth process as now practised If light scrap, such as iron and
steel turnings or shearings are available, these may be mixed with
advantage, with the batch to increase the yield of metal.

These are, in short, the more recent improvements in the direct
process of producing iron and steel which I have been able to
effect, and which I should have been glad to lay before the Iron
and Steel Institute in a more complete form than I am able to do
at the present time.

TABLE I.

CHARUES IN "A" FURNACE AT TOWCESTER.

No.

4H4
435
430
437

438
439
440
441
442
443

444

445
446

447

448
449

450
451

Description of Charge.

In Charge.

Time.

Yield.

Remarks.

Iron.

Coal.

Blooms.

Loss

(12 cwt. Raw, Towcrst, r

Cwt.
8-2

Cwt.
6*

H. M.

3 50
3 15
3 50
4 0

5 0

4 0
3 50

C. Q. L.

530
500
7 0 14
6 0 14

910

700
430
7 0 16
630
7 0 14

700

520
030

S 0 0

800
520

ti 0 14
6 1 0

°/o

30
39
14
26

15
42
14
18
14

15

33

18

3

3
33

26
24

f Great Fenton Coal.
\ Gain 12 %
•i »» it

The average loss is 19 -3 %
( The average production

\ C. Q. L.

( per charge is ti 2 13

< The average time per
( charge is -Hi. 21m.

r Amount of coal used in
(producers (regular
continual working),
2 tons per ton of
blooms produced.

1 ., Tap Cinder
1 ,, Limestone . .

»

u

;: ' . :

»

»

3 50
5 45
5 35

4 25

4 0
3 45

5 5

5 0
4 0

4 20
4 50

,. . •

,,

„

. .

'•

»

„ . .

ii

»

ESTIMATE FOR THE PRODUCTION or 1 TON BLOOMS. £ » </

Ores, 3 tons (including fluxes), at the average price of 10». per ton . 1 10 0

Coals, 1 ton for reducing at 8s 080

2 tons for heating at 4* 080

Labour 0 15 0

Hammering 070

Total cost

£380

348

THE SCIENTIFIC PAPERS OF

TABLE II.
ANALYSES OP IRON.

1

2

3

4

5

6

7

8

9 ! 10

Iron . .

98-30

98-73

99-45

98-97

98-909

99-907

99-128

98-728

Silicon

•745

•243

•643

•565

•582

•640

•717

•932

0. Carbon .

•235

•225

•15

•100

trace

traces

Phosphorus
Sulphur
Manganese .

•080
•065
•144

•032
•071
•101

•030
trace
traces

•019
•085
•126

•020
•106
•158

•090

•128
•035
traces

•125

•03
•228

traces

traces
•34

•024 & 172

traces

traces

of CaO

of CaO

and

and

A1..O.,

A12OS

99-334

99-177

100-123

100-000

100-000

100-000

1. Iron hammered from Charge 60 A.— Aug. 9th, 1875.

2. Iron made from Canadian ore.— Aug. 18th, 1875.

3. „ „ „ Towcester ore.— Aug. 25th, 1875.

4. Towcester iron, rolled fairly at Darlaston —Sept. 15th, 1875.

5. „ „ „ badly ,, ,, ,,

6. ,, „ hammered badly, contained slag.

7. Charge 94 A, Chatterley.— Oct. 13th, 1875.

8. Towcester iron, rolled at Darlaston. Charge 94 A to 102 A. '717 Si corresponds to
1-540/oSiO,,.

9. Indian iron. Charge 132 A.— Nov., 1875. '932 Si corresponds to 2% Si Oa

10. Phosphorus in Towcester iron.

TABLE III.
No. 3.— ANALYSES OF IRON.

1

2

3

4

5

6

7

8

Iron .

99-711

99-278

99-1988

Silicon . . .

•065

•316

•400

•316

•155

•230

•026

•027

C. Carbon .

•12

•12

•150

trace

trace

trace

.10

•10

Phosphorus . .
Sulphur
Manganese . .

•077
•027
trace

•073
trace
•213

•0502
trace
•201

•019

trace
trace

•046
trace
trace

•083
•060
trace

•090
•021
trace

•093

100-000

100-000

100-0000

1. Homogeneous iron, made from Towcester iron, at Darlaston, June, 1876.

2. Towcester iron, rolled at Darlaston, August, 1876.

3. Towcester iron, rolled at Northampton, August, 1876. Charges 95 A and 98 A.

4. Towcester iron bloom, February, 1877, contained 317 per cent. slag.

5. Broken iron bloom 25 C. fracture crystalline. Slag— 1-58 per cent.

6. „ „ „ „ fibrous and dirty. Slag— 2 461 per cent.
7 and 8. Iron from blowing fires.

Iron from
Blowing Fires. (C.)

Iron ; . . 99'6 99'716

C. Carbon ........ '19 '120

Sulphur nil '012

Silicon '028 '021

Manganese traces traces

Phosphorus '048 -031

99-800

100-000

(C.)— Iron made from mixture of Crewe-Scale, and Soniorrostro, and afterwards passed
through the blowing fires.

Si/! WILLIAHf SIEMENS, F.R.S.

349

2 E
•

« X

» «

P9 a

O ^

H «

§ §

H X *

C3 » f

a hi

03 '

QQ to

tH S

S 5
W <

is
II

3!

111

e

J

t. J

h- 0 00 •* «0 <0

c

I

i£|

N ft ft ft I ft

a

1

S 1 S S * *

., J3

™ c£ 3 2 cS S ?2

J~-ld

1 4 <• M • • • a

tl co eo eo Sj eo
II II II II II II

•

M §• fa

OQ

rf I S 3 1 i g 3

,a J* 00

g Sa

00 CO rt CO -H PH

g ^S

1 g

o « e» o i- CT

1

5 |

P ? 5 ? ? ?

1

1

5 ! 1 I I g

»^ O ift ^* SO CO

o o -^ 10 -5 m

N

X X X X X X

OQ

CO O O <M -^» O
•fi ip 10 -p CO 0

C-l Ol (N (?1 <M O»

•s •

.

M O OJ i-H lO »> r-<

£s

1

oa

2 *

M

,2 IM e>« S « 55 S

M II II II II II II

R-

« o

55

1

h S

j JH 2 °^ 2T *"" ^*

41 O

.2 °

rf

5

3

O O O O O O

» —i ^ c. oo e>« i-

JJ

C •

6-1

2.S .

|

!!!.!!!

i|

c

-** ^h O rH O O

lit

5

1

X X X X X X
g c» eo o e» g

eo eo eo eo co eo

^"2=0

S 5-^

- = 1

o £5

.

.

|»1

33

m at «K at ai «

S^3!

0

•X.'X *x 'X "X 'X

- ': :-:

'I

'.= '.£ '=' •£ -.5 •-.

,CO .CO .M .30 .CO .«

0 H

u

0

-» $ - «

s

1

« § 00^ \ 0? 00_ S

350

THE SCIENTIFIC PAPERS OF

tt 55

21

HI

Appear-
ance of
Fracture.

i -si - -1

JL* * S = r5£ - = |

g «£ M

a

0

o

«> ip h- <p co p eq oo

00 C<l O O* O C* r-» rH

O* rH •** O t-l •"* O '.O O '.O CO 00 CO <M 00 GO

£

Jj i> AH co co oo IN n I-H £- co o ooi'j'oo

5

X

H

3

<N .-1 •* O i-"»i5O«O O-£>COGC CO S-) GO CO
OimScO CO 0> H IN rHt-COO O >N -* GO

t-H

i
i

i

o

Square Inch
of Fractured
Area.

„; IN <O OJ <N <N SS

^»l O 00 OS O-. •* K 00
^HOO r-ll^ CiO OO
^05H OJ<M CJIM «<M

(MCO ^t*«C) rHCi Ot^
UJO5O M*Ci OOOi (M1^
^DCCI- (M-* .-Hi— 1 O500

ij cT <>r F-T cT t-T cT t-T t-T

1- 1C 1- O O 0 <0 0

|

8 o

CO»O COO5 OOO COQO
C-* I— ' OlrH WrH CN f— I
Oi-COO OCOiOS Of'OO rtO«CO

1 S

OOt-C> •M'J'CiCi iOC^*»iO Ot-OO

1

5 -

§'^) oo o i~ c; *-O *o c-3 « ri oo c^ ^ o 3i
-O O '-O O 1 - 'C *-O OO t - O -i< Cs X O •*

I

•<

3

OOC<ICOO »-»OS»OtC ^CiOr-* *tSO-*fO
(M ?O t^ c5 C1 -^ *O '-O CO CO '-O !- C-l ^1 O 1—
INC^IMIM Ol'MM'M C^IMMOI IMUJMcM

•<)

>Om!NOO M«OO ifJtOOOOS Md'J'OO
Or-i(Mr-i O^'MO) OO"^ OOrt^-1

s
a

XXXX XXXX XXXX XXXX

OOOO^'-O iCp-<f-lrH C<5«OCO OOC^COO
.— IC^C^tM •— t (M <M C-1 (Nd(MC^ »-l<MC^<N

S^N<NIN (MINININ (N«N(M 'M^l^'M

nate Stress.

r Square Inch of
Original Area.

2»at^ h-eo ?£><>i r-CT
Swc^i C3C$ e^w ^HI^-I

^«<M 'NIN C^IM Nffq

oot~ -<<-t< or- ^o

Jl- ^O -^ CO Ci !M . 1— rH
X5 O O O_ >I^ CO_ -O 0__

of i-T i-T cT cT oT oo~ r-T

OO 4O*O *O-^ ••*•**<

t-ocoi— i o^-»c; o -^to^ci cotc^oo
. _^.ij.i_ c-. o-f^i o«r-i- or7OrH

S O5Cq»r3tf3 CO'XH-CO <O>OCO!N CirfOO

2

&

T (N(MOr- O>-<CiO OOOCJi OOOOCltO

11 o >o >o o -o •« •* «n ic tn >o •* >*<•*•*•*

•3

i- -o o o o "0 i- « o i~ o o >o ea c^ o
«: tococot— «»TiOto i— IOI-M »orH^oo

_2 COi-IJ-O OOiTftO Ol-Ot- •*r^--OO

I

J 1- t- •£, 1— to '•£> CD tD r-D »O O O >O iC »O •*

i

00 00 r-l i— 1 COOOrHt-t tOMCiCi (OlMOICi
O3C1WCO C^OICOCO f^^Hl— I^H ^Hf-lrHrH

j

•<

a

•a

ssi^ 8883 s?sss -s?s;s5^

o

1

8

XXXX XXXX XXXX ,XXXX

COXOJOO C/)OOCOGO COOCQOOO 'OOQOGCQO

"N N ffl Cl (NIM"MC^ (MC^(NC^ CIC^C^C^

s

0

i

1 1 ++ 1 1 ++ 1 1 ++ 1 1 ++

~0 . .g-6

ri-||c ^ ^"3^' 3

^^o __."« . . . -S 2 ® -.-"5 ......

1

R

Srtv C w3v C

C — rt _H S

w«S -H <"=£ •<!
&x °x

-* -*

ox '^x

1- i— ( Oi CO OO(NO-* lOOit— rH <OOOO<M
OOOO ClC>OO OOO^ SS^S

^jT^T-^ -^ •*"•*' -t **" -** ••*'•'* •* •^•<t -^T-^

•4 ;
Q o

,S.S>

jgO3

r

s

Si

WILLIAM S/KMENS, F.K.S.

351

TABLK VI.
\v u.YSEs OF SLAGS FROM ROTARY FURNACE.

1

2

8

4

5

6

'•• of iron . . . .

l->-63

traces

19-50

7-05

• If. of iron .

.•Ir of lllilllpllll-M- .

Sulphuric uc i>l
Phosphoric acid

Alumina ....
Slli.M

I.imi'

40-20
•50
•458
4-94
15-56
29-00
2-68

17*00

'•99

5-40
18-90
•82

55-83

i-io

3-05
" -869

10158

1-65
•518
257
10-20
14-50
5-60

I.;--,
•489
1 -032

10-50
28-10
2-09

!••_•:

•768

20-40
18-80
trace

Total

SHl'338

99-02

100-009

100-008

100 381

99-753

Metallic iron

35-93

.13-64

43-42

44-39

30-51

43-25

No. 5 first tapping.

No. 6 second tapping.

TABLE VII.
CHARGES MADE IN "C" FURNACE AT TOWCESTEU

Description of Charge.

In Charge.

Time.

Yield.

Iron.

Coal.

Blooms. Gaiii %

Loss0/,

Lbs.

Lbs. H. M.

Lbs.

Scale . 1,008 Ibs. . 1

1,250

T-'s 3 55

l,l-»8

8-16

Cindi-r. . . 784 Ibs. .

M

4 50

1,152

7-84

Soinorrostro . 224 Ibs. . )

,,

5 0

728

4-17

Towivster . . 224 Ibs. .

,,

3 30

1,180

5-60

Limestone . 112 Ibs. . J

M

3 35

1,232

1-44

» »

,,

3 SO

1,15(5

,.

7-52

Same charge as preceding ones,'!
with the addition of 108 Ibs. of 1
Fentonorelumpsusedforfettling f

1,330

4 10

1,232

7-86

4 15

1,200

8-00

4 30

1,456

9-47

4 10

1,038

23-15

3 50

1,883

7-36

4 0

1,174

11-73

5 10

972

26-91

3 50

l,4tt

7-67

5 15

1,488

11-88

3 10

l,-24(>

e'-ai

4 15

1,092

17-89

4 5

l,17(i

11-57

4 0

1,386

4:21

4 30

1,900

5-26

5 5

1,280

8-76

3 55

1,.V-'0

14-28

4 25

1,168

ms

3 45

1,106

16-84

3 50

1,218

8-42

3 25

1,162

12-63

2 52

1,156

13-08

Averages .

131'."2 728

4 8

1.232

6-07 °L low.

352

THE SCIENTIFIC PAPERS OF

TABLE VIII.
CHAKGES MADE IN " C " FURNACE AT TOWCESTEE.

No.

Description of
Charge.

In Charge.

Time.

Yield.

Remarks.

Iron.

Coal.

Blooms.

Gain %

Loss %

Lbs.

Lbs.

H. M. LbS.

1

fSame as charges -
Nos. 2 to 13 follow- 1
j ing, but without I
•\ Fenton lumps, 1
and with ' only
1 784 Ibs. scale J

.1,100

728

4 0

966

12-18 (

Small charge
to clear fur-
nace of ac-
cumulations
on lining.

2

Scale. . l,0081bs.^

1,310

728

4 10

1,248

4-65

3

Cinder . 7841bs.

4 0

1,153

11-90

4

Somorrostro 2241bs.

3 35

1,309

o-oo

5

Towcester . 2241bs. )

3 50

1,360

3-89

6

Fenton lumps 1121bs.

n

,

3 20

1,195

870

7

Limestone. 561bs.

,

3 30

1,288

1-68

8

Coal . . 728 Ibs. J

§

3 50

1,090

17-55

9

n

j

3 40

1,153

11-91

10

n

j

3 45

1,097

16-18

11

yj

t

s

3 25

1,283

1-98

12

,,

M

,

3 40

1,092

16-64

13

,,

,

3 40

1,103

15-57

(Same as charges

14

Nos. 2 to 13, with

(

(

3 5

1,029

21-45

15

the limestone in- }

.

3 25

1,032

21-22

16

creased from i

3 35

1,005

23-20

561bs. to 1121bs. J

17

,,

n

3 0

1,059

20-68

18

,,

j,

3 5

1,069

18-31

19

,,

,

3 5

1,122

14-27

'20

»

i

3 0

1,019

22-14

21

„

?|

. j

3 15

1,044

20-22

(Same as charges^

22

Nos. 14 to 21, Fen- 1

1,250

^

3 15

963

22-88

23

ton lumps being f

t

3 15

951

23-84

omitted . J

24

j

3 0

1,123

10-88

25

3 0

1,156

7-52

26

j

3 20

1,106

11-36

27

3 20

1,004

1800

28

j

2 50

1,021

18-24

29

3 10

1,091

12-48

30

j

3 10

1,003

19-68

31

)

2 55

1,126

9-84

32

3 15

900

28-00

33

3 0

1,128

9-68

34

2 40

1,121

10-80

35

2 50

1,178

5-60

36

j

(

1,205

4-32

37

!

S"d

1,223

2-08

38

2 45

934

25-20

39

3 0

1,159

7.04

40

3 0

1,031

17-22

41

2 55

1,100

11-92

42

M

!

3 0

1,310

4'"80

43

ii

..

.

2 45

1,313

5-28

\ ^__

/

Averages . .

1,274-4

728

3 12

1-113

12-6 % loss.

S7/? lVILL!A.\f SIEMENS, f.R.S. 353

In the discussion of the Paper

•ON THE SEPARATION" OF CARBON, SILICON, SUL-
PHUR, AND PHOSPHORUS in the REFINING and
PUDDLING FURNACES and in the, BESSEMER CONVERTER ;
with some Remarks on the MANUFACTURE AND DURABILITY OF
RAILWAY BARS," by Mr. I. LOWTHIAN BELL, M.P., F.R.S.,

and some further remarks regarding

"THE PRODUCTION OF IRON AND STEEL BY DIRECT
PROCESS," by Mr. C. WILLIAM SIEMENS, D.C.L., F.R.S.

THE PRESIDENT * (Mp,. C. "W. SIEMENS) said that perhaps he
might be permitted to explain a point to which Professor
Williamson had referred, as it might save time hereafter.

Professor Williamson wished to know how it was that in
a rotative furnace the phosphorus did not go back to the
metal, seeing that the temperature must be as high as in a
puddling furnace. This question would give him the opportunity
of explaining the action as it took place in a rotative furnace.
Thirty hundredweight of material was introduced into the fur-
nace ; and heat was applied — intense heat, if they liked — to work
upon this mixture. At first the temperature of the mixture did
not rise, because the work that had to be done was carried on, as
it were, at its own temperature — that of the reducing point of iron
ore. But as the deoxidation of the iron proceeded, the tempera-
ture would rise, and a point would be attained where metal would
become malleable ; but at this point the fluxing materials, which
had been mixed with the ore, would have commenced to be fluid,
and would form fusible slag which, immediately upon fusion, would
away from the mass, sinking to the bottom of the rotative
whereas the solid material formed on one side, and rolled
about until it went into balls. Hence, although the flame that

* Excerpt Journal of the Iron and Steel Institute, 1877, pp. 389-395.

VOL. I. A A

354 'TH-E SCIENTIFIC PAPERS OF

was brought into the furnace would produce a full welding heat if
it was continued, yet the essential work done was accomplished at
its own temperature, just as water would be evaporated at its own
temperature in a boiler, whatever might be the intensity of the
heat applied outside. Probably to this circumstance was attribut-
able the fact that the iron when taken out of the furnace was
really free from phosphoric acid. But he attributed in large part
this freedom from phosphoric acid in the iron to the circumstance
that the phosphorus was combined in the ore — very often combined
— with oxygen and lime as phosphate of lime, and to the fact that
it would take a greater heat to disturb this compound which
already existed than would suffice to retain the phosphorus in com-
bination with metallic iron after having been combined with that
metal in the blast furnace. This argument would lose its force
probably if the time allowed for these reactions was unlimited, and
it must further be borne in mind that the final chemical reaction
in both the puddling and direct process did not probably exceed a
quarter of an hour, during which dephosphorization would go on
in the one and phosphorization in the other process to a limited
extent. No one would be better able to judge than Prof.
Williamson to what extent this question of time might be expected
to influence the final result in the two processes.

Before proposing a vote of thanks to Mr. Bell for his valuable
communication, he would reply very shortly to some of the
remarks that had been made with reference to the paper he
had had the honour to bring before the Institution. Mr. Ridley
wished him to explain how it was that in dealing with an ore
containing much phosphorus he should be able to get an iron
free from phosphorus, whereas in the puddling furnace a cinder
containing 5 or 6 per cent, of phosphoric acid would readily com-
municate that phosphorus, or a portion of that phosphorus, to the
iron. In dealing with chemical processes they had, of course, to
take into account all the circumstances of temperature, and the
relative proportion in which the substances liable to act upon one
another existed. Now, he explained yesterday that in treating
poor ores, and poor ores containing a great deal of phosphoric acid,
in a rotatory furnace, such as the one erected at Towcester, care
was taken that a very fluid cinder was the immediate result of the
heat acting upon the mixture. The first action of the heat upon

SfK WILLIAM SIEMENS, F.R.S. 355

lixture was no doubt the reduction of the iron, but as the
n •dilution got more and more completed the fusion of the earthy
iluents immediately set in, and it set in practically before the
n'.Iiu'iiuii was completed. The moment fusion took place, which
was necessarily at the minimum temperature, the cinder drained
a \viiy from the iron, accumulated at the bottom of the rotating
1, whereas the iron separated more and more from it, and as
soon as the cinder had accumulated in any considerable quantity
it was tapped from the vessel. The result was that the first cinder
coming from the vessel contained a considerable quantity (he had
seen and tested it even up to 8 per cent.) of phosphoric acid. At
that time the temperature of the vessel was not sufficient to re-act
upon the phosphate of lime or other chemical combination in
which the phosphorus was held in the slag so as to bring it back
to the metal. If the temperature was now pushed forward in the
rotative vessel, that action, which had been so ably described by
Mr. Hell in reference to the puddling process, no doubt would set
in there also. The iron would gradually but slowly take up the
phosphorus again, but, in order to break up the combination in
rhich the phosphorus was by that time comfortably settled, time
iras required ; and in that process, although the total duration of
3h operation was between three and four hours, yet the time
luring which the metal was in contact with the fluid cinder was
rery short indeed, in fact, the ball was brought out as soon as the
ictal would hang together, and it was purposely brought out at as
>w a temperature as possible, in dealing with impure ore, and
cinder squeezed out as he had described in the paper, not by
jueezing it or hammering it in large masses, but by either
lering it flat out or rolling it into thin sheets, squeezing the
ider out, and dealing then with the bar metal either in the
linary way in the re-heating furnace, or, what was better, in a
lollow fire, where the cinder was more thoroughly washed out from
3 metal, without increasing the temperature very rapidly, to such
extent as to make the phosphorus go back to the iron. There-
)re, he thought there was nothing contradictory in those results
' all those circumstances were taken into account, and at any rate
analysis showed that iron was produced from very impure ores
ctically free from phosphorus, and what was very curious,
)metimes charges from ore containing a great deal of phosphorus

A A 2

356 THE SCIENTIFIC PAPERS OF

were almost entirely free from that ingredient, whereas other
charges made of richer and better ore, contained phosphorus in
somewhat more tangible proportion, owing no doubt to the
presence of cinder which had not been sufficiently removed before
the iron was consolidated. Mr. Bell, in criticising the process
which he (the President) had submitted — and criticising it very
fairly, and contrasting it with the method of treating iron rich in
phosphorus which he (Mr. Bell) proposed — observed that, in
dealing with Towcester iron, he (Dr. Siemens) found it necessary
to mix richer oxides with it, and that that would put the process
altogether out of consideration in dealing with similar ores in the
blast furnace, but it must be remembered that Towcester ore could
not be smelted in the blast furnace by itself either. It was a very
valuable ore as an admixture with other ores, on account of the
large proportion of alumina which it contained, but all attempts to
smelt it by itself in a blast furnace had hitherto practically failed,
and, therefore, it was but reasonable that they should find it
necessary to mix other ores with the Towcester ore. Then his
friend, Mr. Bell, said that he (Dr. Siemens) required forty-two
rotators to produce the metal yielded by one of his great blast
furnaces.- That was very true, and if the two apparatus produced
the same results he would say at once the blast furnace had the
advantage ; but it must be remembered that the blast furnace pro-
duced pig metal, whereas the rotator produced wrought metal, and
that probably forty-two puddling furnaces were required, in
addition to the blast furnace, in order to produce the same result ;
therefore he, at any rate, had the advantage of one.

Mr. I. L. Bell, M.P,, remarked that he had been talking of
steel.

The President said he would come to that presently. There-
fore, as far as wrought iron was concerned, the number of apparatus
required was certainly not greater, and it must be remembered
that the results which he had placed before the meeting were by
no means final results (that they were tentative results, and that
they hoped to do a great deal better), whereas with the blast
furnace, after the powerful mind of his friend Mr. Bell had
been given to it for so long a time, he supposed they had very
nearly reached finality. Then, his friend found fault with his
consumption of coal, three tons being required to produce a ton of

S7X WILLIAM SIEMENS, F.R.S. 357

malleable iron. Well, that was higher than he wished to see it,
but Mr. Bell admitted it was no higher than his consumption in
smelting nud puddling iron.

Mr. I. L. /Icll, M.P. : No ; smelting and making into steel.

The President : I am now speaking of iron.

Mr. I. L. Bell, M.P. : Pardon me, Mr. President, for interrupt-
ing you. 1 was contrasting the conversion of iron from the ore
into steel by the two methods, considering the fact that we have
steel to look forward to.

The President : Mr. Bell says he wishes to compare my iron
with his steel, but in the description which he gives of his process
I do not see a trace of steel. He describes a method of purifying
the pig metal of its phosphorus and then transferring it — remelting
it, it appears, before he can puddle it, and then he says this puddled
iron may be converted into steel in an open hearth furnace ; but
he fails entirely to bring forward any proof that by his method he
produces steel in one operation.

Mr. L L. Bell, M.P. : You must let me explain, please. I am
quite willing, sir, to admit that if you can make bar iron by your
process from a poorer ore containing phosphorus, it is obvious
that there are certain advantages in doing so ; but if you inform us
that you require richer ores free from phosphorus, this is a totally
different matter. I did not intend to compare your process with
the one I had previously described, and I hope nothing I said
induced the meeting to imagine that such was my intention. All
I wished to draw attention to was that portion of your charge
which consisted of hematite, and the question I raised was whether
such ore was most economically treated by being made into malle-
able iron, and then converted into steel, than by being smelted in
the blast furnace, and the pig iron then treated in the Bessemer
converter or in the open hearth for the purpose of obtaining steel.

The President said he wished, first of all, to reply to the obser-
vations with regard to the production of iron, considering iron as
the goal to which they tended. If he had rich ores, and wanted
to convert these rich ores into steel, he modified the process
entirely ; and, indeed, he described the modus operamii which he
adopted in that case. If he had rich ores and wanted to make
steel, he put five or six tons of a mixture of that rich ore with
fluxing material and carbonaceous matter on the bed of an open

358 THE SCIENTIFIC PAPERS OF

hearth furnace — a regenerative gas furnace — and he applied heat
to it for several hours, until a thick crust of metallic iron was
formed all over the surface. He then added, on the surface of
that, pig iron, perhaps in the proportion of about 25 to 30 per
cent, to the amount of iron contained in the ore, and that pig
metal as it melted acted upon the crust of metallic iron formed in
the furnace, melting it gradually, and constituting with it a bath
of steel, which, after being adjusted and mixed with the requisite
amount of manganiferous material, was tapped into ingots and
worked in the usual manner. In that case he did not touch the
ore from the time he charged it into the furnace to the time of
tapping it from the furnace as liquid steel, and he believed a more
direct process could not well be conceived. But that process of
converting rich ore into steel was not applicable to poor ores and
to ores containing a great deal of such materials as phosphorus
and sulphur, which would undoubtedly go into the ingot metal if
they were left in contact with it at a high temperature. There-
fore, they must draw a broad line of distinction between the
process of making iron and that of making steel. In dealing
with poor and impure ores they could still make steel from
them, but, in that case, they must eliminate the earthy matter,
the sulphur, and the phosphorus, at a low temperature, and then
re-melt the material at a high temperature. That was an extra
operation, which ought to be paid for by the difference in the cost
of the ore. In that case, if they dealt with those impure ores,
they had to go so far on the road of making wrought iron as to
take the metal out of the furnace and squeeze it or roll it in order
to get rid of the cinder. Then it was in a proper condition to
transfer it, either to the re-heating furnace to be worked up into
wrought iron, or into the bath or open hearth furnace to be
converted into steel. With regard to his friend's process (and he
supposed he would allow him now, on his side, to make some
objections), his objection to it was that it was an additional
process, which necessarily would cost money to accomplish, and he
doubted very much whether the iron resulting from his refinery
furnace would work in a Bessemer converter.

Mr. I. L. Bell, M.P. : I said this was doubtful.

The President: Mr. Bell admits it; therefore, it had to be
treated and used as iron, or that iron would be melting material

WILLIAM SIEMENS, F.K.S. 359

for the open hearth process, but sti.l in that case they would have
to compare his results with those obtained by simple puddling,
i he President) had lately made some experiments at the
request of Messrs. Hopkins, Gilkes, and Co., which proved in a
satisfactory manner that Cleveland pig metal, when puddled
r.nvfully in a rotative furnace, and then melted on the open hearth
furnace, produced a very good steel. He believed Mr. Hopkins
had exhibited some of the steel upstairs which was so produced,
and Mr. Cramptou had given them proofs long before that time of
the suitability of Cleveland pig to produce steel if treated in his
(Mr. Crampton's) puddling furnace, and subsequently in his (Dr.
Siemens's) open hearth furnace ; therefore, in that case they would
have to compare the results obtained by the refinery process
proposed by Mr. Bell with those of ordinary puddling done in the
most careful manner. In other respects he was ready to compli-
ment Mr. Bell upon his very valuable communication. He had
placed before them experimental facts which he (Dr. Siemens)
believed were entirely reliable, and which in themselves consti-
tuted most valuable information. Whether or not the particular
process which he had lately followed would lead to practical
results or not, was a matter which perhaps did not interest even
Mr. Hell very much, because he was searching for information at
present, and in those researches he was powerfully assisted by the
North Eastern Railway Company, who — acting very differently to
railway companies in general — encouraged research instead of
rather throwing impediments in the way of those who proposed
any new method of manufacture. His (the President's) experi-
ence had been that even when they had got a result it was very
difficult to make people take the trouble to test that result, and it
was a very refreshing sign that here a great railway company not
only encouraged research, but found large means for carrying that
research into effect. One of the implements proposed to be used
in carrying out those researches was a very large hydraulic
squeezer. That machine had been constructed by Sir William
Armstrong & Co., and they would have an opportunity of seeing
it to-morrow.

A vote of thanks having been passed to Mr. Bell for his
valuable paper,

Mr. Menelaus said the President could not very well include

360 THE SCIENTIFIC PAPERS OF

his own name in that vote of thanks, but he (Mr. M.) would ask
them to pass a hearty vote of thanks to their friend, Dr. Siemens,
for the paper he read yesterday, and on which, with Mr. Bell's
they had had such an interesting discussion, although the two
were opposed to each other on certain points. He was quite sure
that if his (Mr. B.'s) process was to be successful, he must go
hand-in-hand with Dr. Siemens, and use his open hearth for
converting the material which he was modest enough to say he
was purifying for the purpose of using it in the shape of iron.
He was quite sure that his (Mr. Bell's) meaning was to make steel,
and, if he was going to make steel, he must go hand-in-hand with
Dr. Siemens and use his open-hearth furnace.

A vote of thanks was accorded to the President by acclamation.

On the presentation of

THE BESSEMER MEDAL
To PROFESSOR VON TUNNER,

THE PRESIDENT * (MR. C. "W. SIEMENS) observed that, thanks
to the liberal provision of our past President, Mr. Bessemer, it
was the pleasurable duty of the President of the Iron and Steel
Institute, at the ' beginning of each year of office, to present
this medal to any gentleman who, in the opinion of the Council,
had highly distinguished himself in that branch of science which
the Institute specially cultivated.

The choice of the Council had this year fallen on a gentle-
man who, all would agree, was fully entitled to the distinction.
Members of the Institution all knew Professor Tunner by repute,
and those who did not know him personally would also be pleased
to learn that this compliment was paid to a gentleman who,
although drawing towards the close of the average period of active
life, yet retained all the energy and all the active interest in his
work that made it still productive of useful results — a gentleman

* Excerpt Journal of the Iron and Steel Institute, 1878, pp. 8-10.

S//? WILLIAM SIEMENS, F.R.S. 361

who, by his amiable, modest manner, impressed every one very
much in his favour. It was, however, not his personal qualities
wliivh the Institute had to look to. Professor Tunner was, it
iniirlit be said, at the head of a very ancient iron district. Styria
hud produced for centuries iron of very high quality. The cutler
of Damascus had for centuries forged his blades of Styrian iron —
at any rate, Styrian iron had entered largely into the composition
of the material of which those blades were made ; and he remem-
bered that when, in 1851, the International Jury made certain
itnents upon the scythes sent from Styria, they were perfectly
surprised to find that the keen edge of those implements would
cut off the head of a nail without being injured by the operation.
Nature had been very kind to the ironmaster of Styria in providing
him with a spathose ore of extraordinary purity, and with a fuel in
abundance of the highest quality in the form of enormous and
beautiful forests. Still, with all those advantages, results such as
we heard of could not be realized without skill and the application
of scientific knowledge. In this country, we were very proud of
our achievements in blast furnaces, but the Styrian furnace,
although a different thing from the blast furnace of this country,
was one of peculiar merits, and the economical results which had
been obtained from it were very surprising. He learnt that a ton
of iron was smelted with something like thirteen cwt. of charcoal.
That result was a remarkable one, although, of course, it could not
be fairly compared with the results obtained from coke and the
inferior ironstones of this country. Professor Tunner might be
said to be the leading spirit in that centre of industry. It was
unnecessary to enlarge upon his literary works, which were well
known to members, and which were largely quoted in our excellent
"Metallurgy" by Percy ; but Professor Tunner had for many years
been active in educating young metallurgists who had done good
work in Austria, and who had also given us indirectly some
benefit. We were now receiving from the district of Styria a
large supply of ferro-manganese, a material which we could not
produce from substances of our own. Professor Tunner had been
active on many commissions, in which his practical sense, and his
thorough mode of investigation, going to the very foundation of
every matter that came before him, had left a remarkable impress
on inquiries into metallurgical subjects. I therefore think (the

362 THE SCIENTIFIC PAPERS OF

President concluded), that members will approve the action of the
Council in giving our annual award of the Bessemer Medal to
Professor Tunner. I have great pleasure, Mr. Bell, in presenting
this medal to you, as representing Professor Tunner on this
occasion, and I hope you will convey to him the high appreciation
in which he and his works are held amongst us.

In the discussion of the Paper

"ON THE SEPARATION OF PHOSPHORUS
FROM PIG IRON,"

By J. LOWTHIAN BELL, M.P., F.R.S., &c.,

THE PRESIDENT* (MR. C. W. SIEMENS) remarked that Mr.
Bell had contributed some valuable facts on a subject which
interested many members.

Mr. Bell had found that he could, by treating Cleveland pig
metal in a rotating furnace (such as was to be seen at the Clarence
Works), reduce the phosphorus to the extent of from T3oths to
-rijths per cent. ; but other results had been given, showing that
the phosphorus had been reduced below TVth per cent., while the
slag produced gave- 6 and a fraction per cent, of phosphoric acid.
These figures relating to phosphorus in the slag and in the metal,
agreed well with the results which he himself had found in treat-
ing the subject in another way. At Towcester, in the rotary
furnace, according to figures given by him at the Newcastle meet-
ing, the slag was generally found to contain from 6 to 7, and
sometimes 8 per cent, of phosphoric acid ; but the metal produced
in that process had been brought to a very low percentage indeed.
So far, therefore, we might consider it an accomplished fact that,
whether they dealt with pig iron or with ores containing a large
percentage of phosphorus, that phosphorus could be eliminated to

* Excerpt Journal of the Iron and Sttel Institute, 1878., pp. 41-44.

SIR WILLIAM SIEMENS, F.R.S. 363

such :IM extent as to make the metal perfectly applicable to the
production of steel. The question as to the relative merit of the
two modes of doing it was one of expense, and on that subject
they were still without sufficient information. Mr. Bell had given
rertjiin data which commenced with Cleveland pig metal. He
subjected the material to a refining process, and in that operation
he used 10 cwt. of oxide of iron to the ton of metal. Now with-
out desiring in any way to discourage Mr. Bell and his friends,
he thought they would have to modify that process before they
could tempt steel manufacturers to use it practically. The
employment of 10 cwt. of rich oxide of iron, which could nob be
valued at less probably than lUs., would form an item of expense
altogether out of proportion to what was now practically admis-
sible. Adding to that, the cost of the operation itself, and the
fuel used, he thought there would be no margin in favour of using
the cheaper Cleveland as compared with the dearer Cumberland
metal. But, as Professor Williamson had said, some bases might
be found to take the place of oxide of iron ; and Mr. Snelus had
brought forward a scheme, which he had been working out, to
employ lime for lining. For his own part, he could not look with
much favour on that proposal. At his experimental works at
Birmingham, he had made an experiment with a lining of that
description. The difficulty he found was that the lining was not
sufficiently hard to resist the action of the charges. He thought
that as much heat was applied as the material could stand, but
the limestone employed contained a percentage of magnesia, which
rendered it fusible to some extent, and probably prevented satis-
factory results from being obtained. With regard to the extent
of phosphorus allowable in the steel, he might speak with some
confidence, having made a great many experiments, and having
seen the results on a large scale in a process of making steel from
materials containing a considerable percentage of phosphorus.
There was no difficulty, for instance, in converting old iron rails
— which might be to a great extent Cleveland rails — into steel
rails of such quality as would satisfy all the mechanical tests that
engineers insisted on. The limit he would assign to phosphorus
in such a material was j per cent. In increasing the amount of
phosphorus in steel beyond £ per cent., there was great danger of
making the steel brittle — that was to say, of making it liable to

364 -THE SCIENTIFIC PAPERS OF

fail under the test of a falling weight. That difficulty increased
very much with the reduction of temperature. On a hot summer's
day such a rail might resist the test, but on a cold winter's day it
would not. Another very essential condition was that the steel
rail containing \ per cent, of phosphorus should contain certainly
less than -^ths per cent, of carbon, and considerably less in order
to withstand the falling weight test, and it must contain not less
than T7oths per cent, of manganese. Such metal would stand
almost any ordinary mechanical test that could be applied to it.
He could quite believe the result which had been given by Mr.
Price, that he had produced a metal containing T%ths per cent, of
phosphorus, and -^ths per cent, of carbon, which would stand the
Admiralty test, but he could not go with Mr. Price further than
that. It might be said that manganese was to the steel maker
what charity was to religion : it covered a multitude of sins, and
these were of a lucifer kind — phosphorus and sulphur. If you
gave him the worst class of iron, and wished him to make steel of
it that would stand the Admiralty test, he would undertake to do
it ; but such steel, nevertheless, should not be used in such appli-
cations as were now contemplated of mild steel — no boiler should
be made of it, no forging should be made of it, and he thought
also no ship should be made of it. Steel of that description
depended for its mechanical qualities on the larger dose of
manganese which it contained. Now every practical steel maker
knew that manganese was a somewhat treacherous element in
steel. Its distribution was never so uniform as to make a homo-
geneous compound. Then, in the working of a plate containing
a large amount of manganese, the manganese was apt to be burnt
out, and in attempting to flange it, although sometimes it would
stand the test, at other times it would break. That was a question
to which the users of steel should direct particular attention in
order to prevent failure in the application of a comparatively new
material. Good mild steel, such as could be safely recommended
to the boiler maker, should contain only mere traces of man-
ganese ; TVth per cent, was about the limit which he considered
such steel could contain with entire safety, when the material
had to undergo certain preparations in order to put it into its
definite shape. Then, again, we heard a good deal about the
liability of mild steel to corrode, and some results which had lately

S/K WILLIAM SIEMENS, f.R.S. 365

to light seemed to throw great doubt upon the desirability
of nsiii^ mild steel, especially in the Navy. He believed that a
tendency to corrode was also the result of excessive manganese,
and some experiments which he had himself made went certainly
in confirmation of that view — that as the manganese in mild steel
was increased, so its liability to corrosion increased. He quite
agreed with the suggestion of Mr. Fox that the subject of the
(|ii;ility and treatment of mild steel should be carefully inquired
into. The Committee of Lloyd's Register were collecting experi-
mental data regarding it, and the Board of Trade had also turned
their attention to it recently ; but whether individually or col-
lectively, the Iron and Steel Institute should not be indifferent to
a subject so important.

In tfw discussion of the Paper

"ON SOME EECENT IMPROVEMENTS IN THE
MANUFACTURE OF IRON SPONGE BY THE
BLAIR PROCESS," by MR. J. IRELAND,

THE PRESIDENT * (MR. C. "VV. SIEMENS) said the converting of
ore into spongy iron was not new. It was first proposed by Mr. Clay,
many years ago. "We had also heard much of the Chenot process,
and after that there had been many other attempts to produce
spongy iron. He himself gave very considerable attention to the
subject at one time, and he had produced spongy iron by methods
somewhat analogous to that which had been brought before the
meeting. He did not think, therefore, that any particular name
could be attached to the process of making iron sponge for the
purpose of utilising that spongy metal in the after-process for the
production of iron and steel. The difficulties which made him
desist from his endeavours to arrive at a practical solution of the

* Excerpt Journal of the Iron and Steel Institute, 1878, pp. 57-58.

366 THE SCIENTIFIC PAPERS OF

problem in that way were several. The iron sponge locked up in
it all the earthy materials of the ore, and if you threw it in the
open hearth furnace these materials prevented liquefaction of that
sponge in the metal. Another great difficulty was that the iron
sponge absorbed sulphur at a most extraordinary rate from flame.
In fact, he would deprecate any process of making the sponge or
using the sponge afterwards, which brought it into contact with
coke gas of any kind, as even the purest coke gas might give so
much sulphur to the sponge as practically to make it unfit for
steel. Mr. Ireland proposed a novelty which consisted in melting
the sponge in a cupola furnace, and then puddling or otherwise
treating the liquid pig metal so produced. He altogether failed
to see the advantage of these processes. "We had an apparatus
which made spongy iron better than any other, and that was the
blast furnace. Take a blast furnace at a certain heat, and you
had your spongy iron produced at the lowest possible cost and in
a heated condition. Why cool it, and then melt it in a cupola ?
There it was, perfect sponge, in the lower part of the blast furnace,
heated and ready to be melted. He quite failed to see what
advantage you could gain by dividing the process into two.
Altogether, the difficulties connected with spongy iron were very
great indeed, and were such as had induced him, at any rate, to
abandon that mode of proceeding entirely, and to endeavour
rather to treat the iron ore in such a way, that while the iron was
reduced to the metallic condition, the earthy matter contained in
the ore was at the same time fused and combined with the fluxing
material. The product had been obtained by a method which
involved but one instead of two intermediate processes, and which
produced iron at once in a perfect metallic condition. He was
sorry to be obliged to speak thus discouragingly of a paper brought
before the meeting, but, as Mr. Bell had said, it was much kinder
to point out plainly and clearly the objections one felt because it
could not be to the advantage of the inventor or patentee to be
encouraged in a direction which was open to objection. If Mr.
Ireland could show that these objections were inapplicable to the
process he brought forward, he should certainly be most happy to
congratulate Mr. Ireland on having overcome difficulties which
had made himself desist.

SfX WILLIAM SIEMENS, F.R.S. 367

/// t/M discussion of th* Paper
" ON STEEL FOR SHIPBUILDING," by B. MARTELL,

DR. CHARLES W. SIEMENS * said, with reference to the observa-
ti<>n that fell from Mr. Reed, as to the difficulty of rolliug steel
plates to dimensions that would present a reduction of weight
n|;i:il to 20 per cent., T would mention that when the material
for the Iris and the Mercury were ordered from the Landore
Company, a wish was expressed by the authorities that the weights
of the plates should in no case exceed those specified, but that it
would be desirable to make those plates perhaps 2 per cent, less in
weight than would follow from the calculation. I believe the
Landore Company acted strictly up to that, and that the total
weight of the material supplied was just 2 per cent, below that
calculated, showing that there really is no difficulty in working by
weight instead of dimensions. I would suggest that it would be a
far more rational thing to order the plates by weight per square
foot, rather than by vulgar fractions of an inch thickness. With
regard to riveting, I, for one, would strongly advocate steel rivets
in preference to iron. It is true that at Pembroke the experiments
tend rather to show that no material advantage would be gained
by the employment of steel rivets, but at that time the question
was hardly sufficiently understood. Now steel has been produced
which I think answers its purpose perfectly, and reduces the risk
of rivet heads falling off to a minimum. I may mention that at
another Institution, Mr. Boyd, the engineer of the Slipway Com-
pany, will read a paper to-day giving the result of the construction
of several steel boilers in which steel rivets were employed, and he
states in that paper that not a single rivet gave way or gave any
trouble whatever in the manipulation, showing that with a little
care difficulties that have been complained of, such as overheating,
may be overcome. The advantage of steel riveting over iron is
not only in the rivet itself, but it is in a great measure in the
plate, which has to be cut away to a much less extent, and thus

* Excerpt Transactions of the Institute of Naval Architects, Vol. XIX. 1878,
p. 29.

368 THE SCIENTIFIC PAPERS OP

leaves a larger nett area of metal to resist the strain. Then again,
if the rivets are put the same distance apart as they would be put
in an iron plate 20 per cent, thicker, it is natural that these rivets
stand too far apart for this reduced thickness of the plate, and
thus throws a buckling strain on the plate. I would suggest that
in riveting steel plates together, the distance between rivet and
rivet should be the same as that which would be adopted if a thin
iron plate were used — in fact, the reduction should be propor-
tionate. The moment you use one material for the plate and
another material for stitching the plate together, there is a dis-
proportion between the two elements introduced which naturally
must tend to weaken the whole. Perhaps I may be allowed to say
one word before I sit down with regard to corrosion. Certain
laboratory experiments which have been tried seem to show that
steel corrodes faster than iron, whereas most of the working results
that have come to my knowledge seem to show the contrary, and
that steel wears longer. We have heard to-day of steel ships from
Mr. Martell himself, that have stood wear and tear exceedingly
well. But I have no doubt that some steel may be introduced
which is liable to corrode at a greater rate than iron would ; and
although I perhaps have no right to speak absolutely and definitely
on the subject, I believe it is the result of an excess of manganese
in that steel, and it would perhaps be well for steel users to turn
their attention to that subject. Manganese is an exceedingly con-
venient agent in order to give to steel, or homogeneous iron, the
high degree of ductility which is desirable ; but it is by no means
a necessary element, because at least the same ductility, and the
same strength, may be obtained without an appreciable percentage
of manganese in the material ; and my experience, which is some-
what limited as to the wearing quality and resistance to corrosion
of the two materials, goes distinctly to prove that with an increase
of manganese beyond a very narrow limit the steel becomes more
corrosive, does not weld with the same facility, and is not so abso-
lutely uniform as when a little manganese is used.

/•:!//•:. v\. I--.K.S. 369

fn I fit- discussion of the Paper

«ON THE USE OF STEEL FOR MARINE BOILERS

AND SOME RECENT IMPROVEMENTS IN THEIR

CONSTRUCTION," by W. PARKER,

I )u. Si KMKXS* slid, I think this paper of Mr. Parker's has brought
• us some very important matter, and the questions arising
upon it have been so well discussed already that I feel it is almost
unnecessary for me to say anything.

The Chairman. There is a question upon which you can give
us some important information, namely, whether the thing called
steel and the thing called iron exist any more as separate
existences ?

Dr. Siemens. Chemically the only difference is, that what we
call mild steel is pure iron — it is more essentially iron than what
is sold as such ; but as regards the physical properties of the
material, it differs in many points from iron and ought not to be
confounded with it, because if we were to call it iron and confound
it with iron, we should think it right to treat it in the same
manner, whereas it requires in many respects different treatment.
In fact, Mr. Parker has given you one or two instances. I had
occasion to allude at another place to one point which was brought
before my notice very forcibly with regard to the steel boilers
which Mr. Boyd has constructed. After testing samples of the
plates that were supplied by the manufacturers the rumour was
spread that the material was wholly unreliable — that instead of
breaking through the riveted seam, through a narrow strip of
metal, it broke through a line of 50 per cent, greater extension
rather than through the narrow place ; and as iron would not
have given way in that manner, and as the experience of engineers
has been derived from iron, it was at first concluded that the new
material was not reliable ; but it was soon discovered upon further
investigation that the method of riveting employed had been such

* Excerpt Transactions of the Institute of Naval Architects, Vol XIX. 1878,
pp. 188-190.

VOL. I. B B

370 THE SCIENTIFIC PAPERS OF

as to be unsafe for this mild steel. Two large rivets had been put
in a forward position and other rivets to make up the holding
strength between the piece under test, and the shackles yielding
were put back. In dealing with a material like this mild steel,
the strain comes entirely upon the forward rivets, whereas the
backward rivets being, as it were, the elongation of a much larger
piece of metal, do not feel the strain at all. As it appears to me
these two forward rivets were near the edge of the metal, and they
set up a tearing action. Now, it was remarked that although the
sample was expected to elongate 25 per cent, before rupture took
place, it did not elongate at all, but broke in the manner already
alluded to. In further testing this metal it was found to be per-
fectly homogeneous and capable of an amount of elongation of
25 or 28 per cent, before breaking.

If you take a strip of india-rubber you may elongate it to several
times its length, but if you make the least nick in the edge of the
india-rubber, and try to elongate it, it will tear readily. In the
same way, if you give this mild steel a chance of tearing it will tear.
In that case no elongation will take place, because in tearing the
strain is confined to an infinitesimal amount of area, and its
course in fracture will be the result not of the total amount of
resisting area, but of the direction the tearing action may choose
to take. With regard to riveting, Mr. Parker states, and perfectly
correctly I need hardly say, that in punching a plate such as is
used in the construction of large boilers, the metal in the act of
punching is reduced 30 per cent, in strength. Mr. Kirk has
already offered an explanation which is similar to the explanation
I should give ; but with your permission I will draw attention,
perhaps a little more, to the precise action that I conceive takes
place in punching a thick plate of this material. Suppose the
plate is being punched, the punch being above and the supporting
die below it. As the punch penetrates into the metal, the latter
does not immediately sever ; it will first yield by compression,
and it will, as it were, be forced laterally into the mass of metal
surrounding the punch, where it will create a zone of highly com-
pressed metal all round the hole that is being punched. The metal
beyond the zone of compression will be to a certain extent under
tension. If the piece punched out is afterwards examined it will be
found to be about 10 per cent, thinner than the original thickness of

.SYA1 ll'lUJAM >•//; .J//:. V.S", /^.S1. 371

the plate, and that diminution in thickness is accounted for by these
forces of compression and extension set up within the metal. If
the metal is now subjected to a tearing strain, it has to resist besides
this, the abnormal states of strain into which it has been thrown by
the punching action, and is therefore destroyed by a force con-
siderably less than that, it can resist in the original condition, as
the paper shows. Now, if this plate after the punching has taken
plan- is subjected to annealing, then this strain between metal and
metal will be relieved and the plate will regain its full strength ;
or if the strained metal be rimed away, and the hole be en-
larged, the plate will have its full strength. But I have the
results of some experiments which were made by Mr. Riley at the
Landore Works, near Swansea, proving that punching does not
••ii the plate, but increases the strength of it. His experi-
ments were not confined to one or two cases, but comprised a
whole series that, strange as it may appear, went to prove that the
punching absolutely increased the bearing strain of the punched
plates, without annealing, to something beyond the strength of
the plate per square inch before it was punched. I should first
say he used the precaution to make the die larger, to perhaps a
greater extent than is usually done, and the result was, to my
mind, that the mass of metal under compression of which I have
spoken was not formed, and that the plate was thus relieved from
the abnormal strains generally produced in punching. Under these
circumstances I can understand that a slight increase of strength
should take place, for the same reason that when you subject a
bar of metal of any form to strain, its strength will be increased.
It is a subject, perhaps, not sufficiently understood, but I have
had an opportunity of observing very extraordinary results in that
way. Take a bar of iron or steel and subject it to a strain exceed-
ing the elastic limit, and allow this strain to be active for some
time, in testing the same bar again you will find not only the
elastic limit has been raised, but that the ultimate breaking strain
is considerably increased. In that way whatever strain you put
upon a piece of metal you increase its strength, unless the strain
is applied in such a way as to set up contending forces within the
metal. Mr. Parker has alluded to two forms of construction with
regard to marine boilers. The one is the admirable furnace of
corrugated form which is represented by the drawings, and which

B B 2

372 THE SCIENTIFIC PAPERS OF

I quite believe is a step in the right direction. The other is an
experiment of construction which I had occasion to make, and
which, by Mr. Parker's request, I have put before the meeting. I
was asked to design a vessel that could resist the pressure of at
least 1,000 Ibs. per square inch, and with a capacity of not less
than 100 cubic feet. The mode in which such high pressure air-
vessels have been hitherto constructed has been in making them
multitubular, but in resorting to that construction you get many
joints and you have also a voluminous structure. It occurred to
me that a better result would be gained by rolling cylindrical
bands of metal from solid steel ingots — cylinders of 1 foot in
depth and 44 inches in diameter. In rolling metal in this fashion
you get the whole strength of the metal in the right form for re-
sisting a bursting strain. The difficulty was to make joints between
the cylinders, and this was accomplished by turning in the face
of each a V groove of about \ inch deep, rings of copper wire of
f inch thickness are thereupon placed between groove and groove,
and the whole set of cylinders are screwed up and bolted together
by sixteen or twenty long bolts made of steel of great elasticity
containing about \ per cent, carbon, with a resisting power of
about 50 tons to the square inch on the ends of the vessel.
After screwing down these bolts hydraulic pressure was applied,
and the vessel stood 1,300 Ibs. to the square inch without a leak,
but at that point a great many joints began to weep. The bolts
were then drawn up |th of a turn, when the vessel was found to
be perfectly tight at the pressure of 1,300 Ibs., but the joints
began to weep' again at 1,400 pounds. We might have continued
the operation of drawing up and have reached a higher limit of
pressure, and seen whether all the parts of the vessel were not
strong enough to resist 2,000 Ibs. to the inch, but inasmuch as
the vessel in use was only to resist a pressure of 1,000 Ibs. to the
square inch, it was thought sufficient to tighten these bolts to the
limit of 1,400 Ibs., because we felt that if ever the pressure should
reach that point the bolts would yield owing to their elastic
property, and allow the fluid within to escape, thus avoiding the
risk of an explosion taking place. In that way vessels I believe
can be constructed which are perfectly safe and capable of resist-
ing a very high pressure. I may mention that in this form steel
will bear a pressure of 45 tons to the inch with perfect ease and

.s/A' WILLIAM SIEMENS, F.R.S. 373

v. It occurred to me, therefore, that marine boilers might
be constructed on the same principle, by rolling a number of rings
in this fashion and placing them one behind the other, with these
copper rings inserted, and lx>lting the whole down. In that way
we should have at any rate a boiler-shell capable of resisting
•jon (.r ;;oo Ibs. pressure without the least risk of an explosion,
wlu-reas in riveting very heavy boiler plates of an inch in thick-
there is always some difficulty in treating the material.

In the discussion of the Papers

"ON THE MECHANICAL AND OTHER PROPERTIES
OF IRON AND MILD STEEL," by DANIEL ADAMSON,
C.E., Manchester ; and

"ON CERTAIN MATTERS AFFECTING THE USE OF
STEEL," by M. ERNEST MARCHE, Paris,

The PRESIDENT * (ME. C. W. SIEMENS) said he would proceed
to make a few observations upon the papers before them, com-
mencing with Mr. Adamson's, which he thought was one of great
practical merit.

He agreed with Mr. Adamson both in his objection to short speci-
mens and in his reason for such objection, but regretted that he
had adopted a 10-inch specimen instead of the standard length of
8 inches ; for the 8-inch standard, or the length of 200 millimetres,
had found its way into very large practice. It had been adopted
first of all by the French Admiralty, and afterwards by the English
Admiralty and by Lloyds, and it was of very great importance to
have uniformity. There was, of course, no a priori objection to
Mr. Adamson's standard, but it would render comparison of his
results with those obtained by the Admiralty standard less easy,

* Excerpt Journal of the Iron and Steel Institute, 1878, pp. 414-427 and
457-458.

374 THE SCIENTIFIC PAPERS OF

and this was the more to be regretted as, in his opinion, ^Ir.
Adamson's paper would be largely referred to hereafter regarding
the mechanical properties of steel.

"With regard to weldings, he would observe that all the speci-
mens which Mr. Adamson had brought before them were very
beautiful indeed, and he had succeeded in welding up an angle-
iron ring in a very perfect manner ; but he (the President)
objected altogether to welded rings. Why not roll them at once
as they would roll a tire ? This had been done already to some
extent in England, and he thought it was a very excellent prac-
tice. The results which Mr. Adamson had placed before them
regarding corrosion were very interesting. This question of cor-
rosion, as they were aware, had occupied the attention of the
English Admiralty, and had been discussed on several occasions.
The results which Mr. Adamson now brought before the members
of the Institute made steel come out much better than they could
have hoped after the results that had been brought before them
by the Admiralty Committee ; and he must say he could not see
why a really pure metal like melted iron should corrode more than
an impure metal where different substances were at war one against
another.

The reason why some specimens of mild steel had given such
unfavourable results was in his opinion (and he simply repeated
the opinion now) due to an excessive quantity of manganese.
Manganese was a highly oxidisable metal, and if eaten out of the
mixture, there remained only a sponge, which was easily de-
stroyed. Altogether he fully maintained the view he expressed
that manganese was only a cloak to hide imperfections in steel,
and when they wanted a really high-class metal, they should have
it as pure as possible, containing a little carbon in order to give
strength, but as little of other materials as they could possibly do
with. Manganese, no doubt, was a good metal for the maker. It
enabled him to push stuff through the rolls, whatever it might be
made of ; but now that they spoke of really high-class metal, he
thought manganese was one of its worst impurities. Mr. Adam-
son had also alluded to steel higher in carbon than the mild steel
which was now used for engineering purposes. He thoroughly
agreed with Mr. Adamson that it was absurd, in a bridge whore
they wanted any moderate extensibility and great strength, to

A7A1 \ni.UAM S//-.1//-.VS, F.K.S.

375

employ a nirt.il that \\oiild hear without breaking a tensile strain
•..us, when th.'V mi-Ill have for tin' same money and
with; (•;!<'• a iiii'lal I hat . \v<ml.l resist 40 OF 45 tOUH. The

practical ditlictllty to tin- udupt.iMii of thi- higher metal \v«inli| lie
found in ollicial circles. The I '.«. an I of Trade 1 1 ad, after a great
deal i.f argument and a -ival deal of endeavour on the part of
rii'jiii'ii- and sled-makers tu induce them to recognise steel at
all, consented in allow steel to !«• used in engineering construc-
tions. Inn they would only credit it with ('..I tons per square ineh,
mid so long as that rule was maintained, the engineer was power-
leu to adopt material of a higher class; but the hindrance, no
iloiiht, must be removed by the force of fact, and .Mr. Adamson's
would do valuable service in furnishing further data for
them to consider and to act upon.

Wiih regard to Professor Akerman's paper, it was a valuable
commiinieatii.il on the exhibits of steel shown at the Exhibition,
but they could expect only a very high-class paper from one of
I'nife.s-'or A kerman 'swell-known attainments and position. There
however, some obsenaiinns as to which he must beg to
dill'er from that gentleman. Professor Akerman said — "No
|'f"l' :' dill'Tene.- I.ei \\een Bessemer and Siemens-Martin plates
n.iild he disi-oxeivd in tli'- course of these exjicriiiiciits, which

< iprehend both complete analyses and tension tests," &c. No\\

h«- (Dr. Siemens) had always avoided comparisons between the
op> n-iieaii h and the Bessemer metal. Ho had been always a greai
admirer of the Bessemer process and a great friend of .Mr. l'«
s 'iiier, but he thought Mr. Bessemer himself would probably take
r\e.-pt jon to that sentence. Was it probable, he would ask, that
if they commenced with the same pig metal, put a portion of it
into a Bessemer converter, and another portion into an open-
In ail h furnace, treating the one by blowing air through it,
diminishing its quantity by 15 per cent., and treating the other
with rich ore such as they used for the reduction of the carbon in
pig metal on the open hearth, that the results would be the same ?
II could not sec how the results could possibly be the same. It
might be better in the Bessemer process than in the open-hearth
for au^ht he knew, but it could not be the same. If absolutely
pure pig iron were used, such as is found in Sweden, then the
analysis might show no difference in the amount of phosphorus

376 THE SCIENTIFIC PAPERS OF

and sulphur in the final result, because taking the half of nothing,
it was again nothing. But if they had an impure pig metal to
deal with, the simple diminution in weight must, as it was well
known it did, increase the percentage of phosphorus and sulphur
in the Bessemer metal, whereas the ore in acting upon the bath
not only substituted for every pound of carbon and silicon it took
away a pound of pure iron, but the slag took away at least a small
portion of the impurities from the metal, especially if lime were
used in the process. One of the essential differences between the
two metals was this (and it was well known in practice at those
works where the open-hearth furnace was carried on), that if they
wanted a certain ductility, they must put TVth per cent, more
carbon into the open-hearth metal than they would require in the
Bessemer metal. That was common practice. It was alluded to
in Mr. Akerman's paper, but with a doubt thrown upon it. He
could inform them that it was a positive fact, and it was a signifi-
cant fact also ; because if for the attainment of a certain degree
of ductility they could put in TVth per cent, more carbon, it meant
that they had the extra strength due to the -j^th per cent, of
carbon in their metal, and for that reason he believed that the
open -hearth metal would have its own special application, in
which it would take the first position, whereas the Bessemer metal
was very excellent indeed for perhaps the majority of purposes.
He thought it was but right that he should call attention to this
difference without wishing in any way to detract from the value
of the paper. Mr. Barnaby, the Chief Constructor of the British
Xavy, was, he believed, in the room, and as he was a very large
user of steel and iron, they would be glad to hear his views.

In closing the Discussion,

The PRESIDENT said he only wished to refer to two points in
the discussion. One was an observation by Mr. Kitson, to the
effect that steel did not stand heating and cooling as well as iron.
Now, he had received from Messrs. Easton and Anderson, only the
previous day, a statement of some very interesting experiments
which they had made with steel boilers. The cylindrical bodies
were heated to redness, then plunged into cold water, and again
reheated and flanged. In other experiments they were flanged

-S7A' WILLIAM .s'/AM/A'A'.S1, F.R.S.

377

while in a state of redness, and then plunged into water. With-
out going into the details of those experiments he would, with the
jxTiuission of the Meeting, hand them to the Secretary, in order
that they might appear in their "Transactions." He had the
misfortune, though he very highly esteemed Mr. Bell, and put
store upon his views, of often differing from him, but in this
instance he believed Mr. Bell had been perhaps too severe upon
him. He certainly took the view that if they wanted to make a
high-class steel, such as Mr. Adamson had brought before them,
they must exclude, from that material foreign substances, not only
phosphorus and sulphur, but also manganese. He knew from his
own experimental facts, and from long experience, that they could
not produce a very high-class material unless they made it exceed-
ingly pure ; and with all deference to Mr. Bell, he thought there
was a broad line of difference between carbon and manganese.
Carbon, they knew, associated itself with the iron, and for every
loth per cent, of carbon they got a definite increase of tensile
strength, therefore when they wanted tensile strength they put in
carbon, but that was not the case with manganese. They did not
put l£ per cent, of manganese into their rails in order that they
might resist the action of corrosion better, or in order that they
might stand the test better to which they were subjected, but it
was well known that they used that amount of manganese in order
that the stuff might roll better, and, as regards rails, he had not
much to say against it ; on the contrary, he looked upon the steel
which had latterly been brought forward so much by the Terre-
Noire Company as a feather in the cap of the open-hearth process,
and therefore he had no reason to find fault with it ; but he saw
great danger in the indiscriminate use of manganese for the pro-
duction of the higher classes of steel. When they heard such
extraordinary results as Mr. Kitson had alluded to, when they
heard from an authority like M. Fremy such results announced to
them as that, under certain circumstances, steel utterly failed, he
believed that if the case were to be examined chemically it would
be found that the steel was, as it were, forced into a condition of
mechanical aptitude for its purpose by the addition of silicon,
which was another of those substances which, he thought, under
the circumstances, ought to be strictly excluded from high-class
steel. Mr. Adamson had called iron a " concrete " of " slag " and

378 THE SCIENTIFIC PAPERS OF

" iron." He (the President) thought the same appellation applied
to a material containing If per cent, of manganese, besides an
ample allowance of phosphorus and sulphur, and, however perfect
the material might be for such a purpose as rail-making, he was
quite certain it would not stand the higher tests required of a
material such as Mr. Adamson had brought before them, and
when they wanted to melt it and heat it rapidly and cool it to
stand the tests alluded to by Mr. Kitson, or when they wanted to
expose it for a length of time to corrosive action, they would find
out the difference.

ON THE PRODUCTION OF STEEL, AND ITS APPLI-
CATION TO MILITARY PURPOSES.

BY C. W. SIEMENS, Esq.,* D.C.L., LL.D., F.R.S., &c.

THE Chairman, Lieut.-General Sir Henry Lefroy, K.C.M.G.,
R.A., F.R.S., said, It is scarcely necessary for me to remind you
that you have before you one of the most distinguished physical
philosophers of the present day. "We have to do to-night, not
with the distinguished electrician, with the constructor of the
Faraday, with the inventor of the bathometer, with the fertile in-
ventor whose range has gone over subjects as various as the setting
of type, and the measure of the depth of the ocean, but with one
of the most scientific metallurgists of England, the inventor of
many remarkable processes in that art, and particularly of one
\vhich I daresay he will allude to for the direct extraction of steel
and iron from the ore, and of whom it may be very safely said that
he has touched no subject which he has not adorned.

Dr. Siemens. The subject-matter regarding which I propose to
engage your attention this evening is not new. Steel was known
to the ancients, and is still produced by semi-barbarians in a
similar manner to what we find described in ancient records.

* Excerpt Journal of the Royal United Service Institution, Vol. XXIII., 1879,
pp. 536-553.

.svA- WILLIAM >•//•;.!//•:. v.v, I-.R.S.

379

Rich ferruginous ores were placed upon ignited charcoal in a
cavity formed in the side of a hill, and as the result of a day's
hard labour in activating goat-skin bellows, a lump of metal mixed
with charcoal and slag was produced, which after being subse-
quently forged, would prove sometimes of a comparatively soft
nature, when it was called iron, and at other times harder, when it
was denominated steel. This shows that the two metals iron and
steel are substantially the same, and that they are distinguishable
only by difference of physical qualities which are the result of very
small chemical admixtures.

The steel produced by the ancients was of very high quality,
remarkable for its great hardness, and for its, power to resist
abrasion. Who has not heard of the blades of Damascus, and at a
somewhat later period of those of Toledo, and of the remarkable
swords that were made by the Norsemen ? So much value indeed
did the Norsemen place upon the production of cutting edges of
great hardness coupled with tenacity to resist chipping, that those
who could produce a blade of unequalled edge were highly esteemed
and in one or two instances even rewarded with the purple of
royalty in being elected sea-kings.

Notwithstanding the antiquity of the metal called steel, its pro-
duction and its application have taken a new and very remarkable
stride within our recollection. It was, however, as early as the
year 1722, that Reaumur, the distinguished French philosopher,
proposed to produce steel on a large scale by fusing cast or pig
metal with wrought metal or scrap. He put these ingredients into
a crucible, and melting them together produced a metal partaking
of the nature of steel. The difficulty he encountered, however,
was insufficiency of heat. As far as we can make out, Reaumur's
suggestion amounted to little more than a proposal, and it was not
until 1820 that steel melting was introduced into commerce in a
successful manner by Huntsman of Sheffield, using coke in a
furnace actuated by intense draught, such as we know at present
as an air furnace. Huntsman succeeded in melting steel in con-
siderable quantity in pots, and from that date steel has become of
great value in commerce for various applications.

Steel has variable properties depending upon very slight
differences in chemical composition. Thus one steel when hardened
will be next to diamond in its power to resist abrasion, and in its

380 THE SCIENTIFIC PAPERS OF

suitability to cut other metals, or when drawn out it will show a
permanent elasticity not equalled by any other substance ; it is
susceptible of retaining magnetism, and will become what is called
a permanent magnet, which property is shared with it to an
inferior degree by only two other metals — nickel and cobalt. It is
now produced in another condition known as mild steel, in which
it manifests a quality of a totally different kind, a ductility not
equal only but superior to that of copper and silver.

I hold in my hand a vase that has been wrought from a bar of
this mild steel by an ordinary blacksmith, or I should rather say
by an extraordinary blacksmith, because upon examination the
workmanship displayed in the production of this vase is found to
be really astonishing. The vase is hollow throughout, and not
more than J^ of an inch in thickness and perfectly sound, without
weld or soldering. For this work of art I am indebted to my
esteemed friend Mr. Henri Schneider, of the celebrated Creusot
works in France, who has forwarded it to me as a sample of what
he had produced by the open hearth process of steel making, of
which process I shall have occasion to speak further on.

It was not, however, until the year 1856, that a means was pro-
posed of producing steel at a cheap rate. This was the year when
Mr. Henry Bessemer read his famous paper at the meeting of the
British Association at Cheltenham. His paper was entitled " The
manufacture of malleable iron and steel without fuel," and it
naturally created the greatest possible interest throughout the
country. Mr. Bessemer, however, did not immediately succeed in
producing by his process such steel as could be used. It was not
until the year 1862, the time of the Universal Exhibition in Lon-
don, that Bessemer steel attained a decided position in commerce.
But, in speaking of this very important process, I feel in justice
bound to make reference to the name of Mushet, who, when he
heard of the Bessemer process, thought that it would require an
addition — analogous to what Heath made in the Sheffield pot-
melting process — that of manganese. Mushet proposed and
patented a mode of adding spiegeleisen (or pig metal containing a
considerable percentage of manganese) to the Bessemer metal
whilst it is still in the liquid state, thereby separating from it the
oxygen held in suspension in the metal in consequence of the
blowing, and as we now know, adding to it some manganese which

5/A' U'l I.I.I AM SII'.MKNS, F.R S. 381

is essential in order to make the metal thoroughly malleable. We
h:i\< here then a process that has done more than any other in-
vent inn in modern times to revolutionize, I may say, the most
important industries of the land. At present not only railway
machinery, but the very rails upon which we travel, are made not
of iron, but of steel, and if I say that steel rails have shown a power
of endurance five or six times greater than those of iron, I may
add, with some regret (and I now speak not as a consumer, but as
one connected with the production of steel as a manufacture) that
they are unfortunately produced at a price almost cheaper than
iron or any other metal that could be named. Perhaps, however,
the manufacturer will learn to produce them at those prices and
yet clear some profit.

Almost at the same time that Mr. Bessemer made his remarkable
invention experiments were instituted, at a distance of not 100
yards from this place, which have led to another process of pro-
ducing steel upon a large scale. I, in conjunction with my brother,
Frederick Siemens (who had previously been my pupil), erected
an experimental furnace at Scotland Yard, by which we proposed
to attain very high degrees of heat, and it was almost from the
first that I looked upon that furnace as capable of accomplishing
what Reaumur, and after him, Heath, had proposed to do, namely
to produce steel in large quantities upon the open hearth.

At first our attention was confined to melting steel in crucibles,
to melting glass, and to other applications of this mode of pro-
ducing intense heat ; the difficulties encountered were very great,
and it was not until the year 1861 or 1862 that the prejudices in
the way of the practical application of the furnace were sufficiently
overcome, and that the furnace itself had assumed such a shape as
to enable us to show that it could be applied with commercial
advantage. And it is a curious coincidence that it took us just as
long to mature this furnace as it took Mr. Bessemer to mature his
process. In the year 1861, a large furnace was erected at the
glass works of Messrs. Lloyd and Summerfield, near Birmingham,
which has been at work up to the present time, and has realized
those results that we, up to that time, had only hoped to attain.
The success then achieved encouraged me to commence a series of
experiments in the direction of producing steel on the open hearth,
but, in order not to weary you, I will proceed to describe the

382 THE SCIENTIFIC PAPERS OF

regenerative gas furnace in the form in which it is now applied
for the production of steel on the hearth, in quantities of from
5 to 10 tons at a time.

The regenerative gas furnace is so essential a part of the process
of open hearth steel-making that it is indispensable to describe its
principle and construction to some extent. It consists of two
distinct parts — the furnace proper, with its reversing valves,
regenerators, and melting chamber, and the gas-producer, in which
the raw fuel (mostly small coal) is converted into gaseous fuel,
which is thus separated from all the drossy and dirty constituents
in the coal. It would appear, at first sight, a roundabout opera-
tion to convert fuel from the solid into the gaseous condition and
then to take and burn this gas in a furnace elsewhere ; the gas, as
it passes from the producer, is in a heated state, and in its transit
to the furnace a great deal of that heat must necessarily be lost ;
therefore, it might well be asked, why make this conversion of
solid into gaseous fuel ? Surely, in burning the gas less heat is
obtained than if the fuel were burnt in the heating chamber of
the furnace, and produced its effect there. That would be per-
fectly sound argument, if it was not for the regenerators of the
furnace. These regenerators are by far the most important part
of the whole arrangement, and, in order to understand the general
principle of the furnace. I will first describe their action.

Plate 46, represents the furnace in longitudinal section ; the
gas from the gas-producer passes in through a reversing valve,
by means of which it is directed into the bottom part of the re-
generator chamber-. The gas flowing up through the mass of
brickwork the chamber contains, and which is placed so as to form a
large aggregate surface, with intricate zigzag passages, will become
heated, providing any heat has been accumulated therein. In the
first place, there will be no heat, and the gas will pass unheated
through this chamber and thence to the combustion chamber of
the furnace. At the same time, a current of air is admitted
through the air-reversing valve into the air regenerator chamber,
which is larger than the gas chamber. The air passing up through
the chequer work will reach the same point as the gas does at the
entrance into the combustion chamber of the furnace. Now,
since both the air and gas are cold, and as they meet for the first
time at the entrance into the furnace, they will, if there ignited,

WILLIAM .sv/-;.i//-;.\'v, r.R.s. 383

a heat not certainly superior to what would be produced
it -olid fuel had been burned there instead ; on the contrary, gas
of the description we are dealing with is a poorer fuel than solid
fir -I. and the heat produced in the furnace will, therefore, be very
m 'derate indeed. But the flame, after passing over the bed of the
furnace, does not go to the chimney direct, but has to pass through
two iTg'-Merative chambers, similar to those already described ; the
larger proportion of the heated products of combustion will pass
through the air regenerator chamber, simply because it is the
largest channel, and another portion will pass through the gas
regenerator. The products of combustion pass from these
chambers through the reversing valves, and are by them directed
into the passage leading to the chimney.

The operation, therefore, is simply this, that the air and com-
bustible gas pass up into the furnace through the one pair of
chambers, and pass away, after combustion, towards the chimney
through the other pair. But in passing through the second pair,
the heat of the products of combustion is given up to the brick-
work. The upper portions of this brickwork take up the first,
and, therefore, the highest degree of heat, and, as the burnt gases
are passed downwards through the regenerators, they are, by
degrees, very completely deprived of their heat, and reach the
bottom of the chambers and the chimney comparatively cold.
After this action has been going on, say, for an hour, the reversing
valves are turned over. They are simple flaps, acting like a four-
way cock, and, by throwing over the levers which work them, the
direction of the currents is reversed. The gas and air will enter
now through the second pair of chambers, and the air passing up
one regenerator and the gas passing up the other, will take up heat
from the bricks previously heated by the descending current. The
gases so heated, say, to 1000° Fahr., will enter into combustion,
and if the heat produced at the former operation was 1000°, it
ought this time to be 2000°, because the initial point of temperature
is 1000° higher. The products of combustion will also escape at
2000°, and passing through the chequer work of the first pair of
regenerators, its uppermost ranges will be heated to very nearly
2000°. The temperature will diminish by degrees in descending
till the gaseous currents have again reached the bottom nearly
cold. Again reversing the process, after another hour or half-hour,

384 THE SCIENTIFIC PAPERS OF

as the case may be, the gas will take up the heat to the extent of
nearly 2000°, and since another 1000° is again produced in com-
bustion, the temperature of the furnace will this time attain 3000°,
and in this way it might be argued that, unless work is done in
the furnace, the heat developed in combustion will, step by step,
increase the temperature of the furnace 1000°, or something less,
each time the reversal of the valves takes place, till we arrive at
the practical limit imposed by the melting point of the most
refractory substance we can find (pure silica, in the form of Dinas
brick), of which the melting chamber is usually formed. This
high temperature is obtained by a gradual process of accumulation,
and without any such current as would be likely to destroy, by
oxidation, the metal in the bath, or cut away the sides and roof of
the melting chamber.

There is, however, a theoretical as well as a practical limit to
the degree of heat obtainable in combustion, which was first
pointed out by M. H. St. Claire Deville, namely, the point of
dissociation at which carbonic acid would be converted back into
its constituents, carbon and oxygen. If carbonic oxide or any
other combustible gas and air enter the furnace at a temperature
very nearly equal to the point of dissociation, it is evident that
association or combustion cannot take place, and thus nature
fortunately steps in to restrict the increase of heat by accumula-
tion, within comparatively safe limits. In a furnace fully heated
up to the melting point of iron, this action of dissociation can be
very clearly observed. At first, when the gas and air are com-
paratively cold, combustion takes place sluggishly, the gases will
flow through the furnace and produce only a dark-red flame ; the
next time the valves are reversed a whitish flame is produced ; the
next time a short white flame ; and after having reached a full
white heat, exceeding the welding point of iron, the flame will
again become a long one, but this time not red, and of little
apparent power, but bluish white, and flowing in clouds. This
indicates the near attainment of the point of dissociation ; com-
bustion can no longer take place, except in the measure of the
heat being dispersed to surrounding objects, or to the metal in
the furnace, and that is about the degree of heat required for the
process of making steel on the open hearth.

Before I leave the question of the furnace, I must refer back to

S/X WILLIAM SIEMENS, F.R.S. 385

t'l-- apparatus in which the solid is converted into gaseous fuel.
This is a very simple apparatus, consisting of a cubical brick
chamber of about 8 feet side, one side of which is cut off in a
slanting direction. Fuel descends on this inclined plane, to the
grate at the bottom where combustion takes place. The result of
this combustion is carbonic acid at a high degree of temperature,
ami if this product of combustion was allowed to pass up the gas-
collecting channel, and through the overhead tube to the furnace,
there would be nothing to burn ; and the only result we should
probably observe would be that the iron tube would very soon
become red hot, and be melted down. But the carbonic acid as
it is formed near the grate, encounters a further layer of fuel
descending from above, which is also incandescent, but which
cannot be consumed on the same terms, because there is no longer
any free oxygen present The first result of combustion being
carbonic acid, a compound of one atom of carbon and two of
oxygen, this carbonic acid in passing through the subsequent
layers of incandescent fuel is broken up, and a second molecule of
carbon is added to the first, thereby producing carbonic oxide,
which is a combustible gas. But coal is not simply carbon, it
consists also of volatile matters, hydro-carbons, water, and the
constituents of ammonia, and the hot carbonic oxide in passing
through a further thickness of the fuel which contains these
gaseous constituents, acts upon them in the same manner as heat
does upon the coal in a gas retort. This action absorbs a portion
of the free heat in the carbonic oxide, and the result is a gas
consisting of carbonic oxide, hydrogen, hydro- carbons, aqueous
vapours, and nitrogen, which latter, being a constituent of atmo-
spheric air, necessarily passes with it through the fuel, and dilutes
the combustible gas produced to the extent of about 50 per cent,
of the total volume. This combined gas leaves the producer not
at 3,000°, the temperature of direct combustion, but at about
700° F. only. This remaining heat is thrown away and purposely
so, and many criticisms have been made in consequence of this
apparent waste of heat in the regenerative gas furnace, but I
think I can prove that although there is loss of heat, no waste is
incurred.

The gases passing from the gas-producer could be forced to the
furnace by mechanical means, but this would be very troublesome
VOL. i. c c

386 THE SCIENTIFIC PAPERS OP

and costly, and the duty performed by the 700° of heat is to give
them onward motion in the direction of the furnace. The hot
gases rising in the uptake represent a column of heated gas at a
temperature of 700°, at which its density will be about half the
density at ordinary temperature. From the uptake they pass
through a long tube of sheet iron or steel, and on their journey
through this horizontal tube they part with most of their heat,
so that when they reach the downtake their temperature has
probably fallen from 700° to 200°, having parted with 500° of
heat by radiation from the tube. The consequence is that the
descending column will be of about twice the specific weight of
the ascending column, and therefore a continual flow of the gas
will take place, ascending on the one side, and descending on the
other in forcing its way towards the gas furnace ; by this means
the heat apparently lost in the gas is utilized to produce useful
mechanical effect.

But suppose that the gas passed from the producer to the
furnace without being allowed to cool, what would be the result ?
The gas would enter the regenerative chamber at a temperature
not of 200° but of 700° ; it would, in ascending, take up more
heat and enter the heating chamber at the temperature previously
imparted to the upper ranges of the chequer work by the
descending current or product of combustion. The same tem-
perature would be attained by the gas if it entered at 200*, the
only difference being that the regenerator in the case of the cooled
gas would work through a greater range of temperature by 500".
But a regenerator will work with the same economy through a
greater range as through a less range ; therefore, this heat, if it
could be saved, would be of no benefit whatever to the gas
furnace. The only difference in result would be that the gases
would get less cooled in descending on their way towards the
chimney, and that we should have a hot chimney instead of a
comparatively cool one. Therefore, no loss to the furnace is in-
curred in cooling the gas on its way to the regenerative chambers,
and the temperature of the gas is utilized to produce the very
essential mechanical effect of urging the gas from the producer to
the furnace.

The economical action of the furnace depends upon the circum-
stance that the products of combustion reach the chimney, not at

S/K WILLIAtf SIEMENS, F.A\S. 387

:"mperature of the heating chamber, as is the case when
ordinary f urnaces are employ t-<l, lnit at a temperature not exceeding
T l<>0° F., thus rendering nearly all the heat produced in
actual combustion available for accomplishing useful work. It
will I) • readily perceived that the economy of this system must be
greatest in melting steel or in accomplishing operations of melting
or heating at very high temperatures, whereas for the attainment
of low temperatures, such as the heating of boilers, the economy
would be comparatively small. Its practical economical result for
ligh temperatures is well illustrated by the fact that in melting
steel in pots in the ordinary air furnace at Sheffield, 3 tons of
Din-ham coke are required to melt a ton of steel, whereas a ton of
small coal suffices to melt a ton of steel in the same pots when the
regenerative gas furnace is employed. In melting steel in bulk
upon the open hearth, the consumption of fuel is further reduced,
and does not exceed 12 cwt. of coal for the production of a ton of
steel. In re-heating iron, the practical economy effected in the
regenerative gas furnace over the ordinary furnace amounts to
from 40 to 50 per cent., owing to the inferior degree of tempera-
ture required. When applying the system to inferior tempera-
tures, there is advantage in suppressing the cooling tube and gas
regenerator, and in approaching the gas-producers to the furnace,
to consume the gas at its initial temperature.

At large works such as are now erected for carrying out the
open-hearth steel process, a cluster of producers are put up outside
the works ; and the fuel is delivered from the railway at an
elevated point in order to be put into the producers in which it is
gradually consumed, and flows as a gas through the large overhead
tubes into the works, where a number of furnaces are supplied for
the production of steel. I think these observations may suffice to
describe the furnace which plays a most important part in the
process to which I shall presently refer.

I have already stated that one of the chief objects I had in view
in maturing this furnace was the production of steel on the open
hearth, but, as usual, in introducing a new process, great difficulty
was encountered in first attempting to carry that idea. The
question arose whether steel could be melted and maintained as
steel upon the open hearth of a furnace at a temperature exceeding
the melting point of most fire-bricks. The general opinion of

c c 2

388 THE SCIENTIFIC PAPERS OF

practical men was entirely opposed to the idea of accomplishing
the object, and it is, therefore, perhaps natural that its realization
was a question of time. The first attempt to make steel on the
open hearth of a regenerative furnace was made by Mr. Charles
Atwood, of Tow Law, who, in 1862, agreed to erect such a furnace
— a small one, it is true — to my design ; but although he was
partially successful, he abandoned the attempt because he was
afraid that the steel so produced would not be of the proper
quality. In the following year, another attempt was made in
France. A large furnace was erected at the Monthi£on Works,
and my colleague in the experiment was a very celebrated French
metallurgist, the late M. le Chatellier, Inspecteur-General des
Mines. The experimental results were on the whole satisfactory.
We obtained some charges of metal that was decidedly steel, but
unfortunately, the roof of the furnace soon melted down, and the
company who had undertaken the erection of this furnace were so
much disheartened, that they, for the time at least, abandoned the
idea of following up the trials. After two or three very similar
disappointments, I decided to erect experimental works at Bir-
mingham, where the processes of producing steel on the open
hearth have been gradually matured, until they were sufficiently
advanced to entrust them into the hands of others. But another
French manufacturer, MM. Martin of Sereuil, undertook to erect
a regenerative gas furnace that could be used for making steel on
the open hearth, but which, in the first place, was to be used as a
furnace for heating wrought iron. While I was engaged at
Birmingham with experiments to produce steel of good quality by
my process, MM. Martin also succeeded in obtaining results with
the furnace I had designed for them. At the time of the French
Exhibition, in 18G7, MM. Martin brought forward their excel-
lent exhibits, for which they soon got a considerable name. I
also sent samples of steel produced by me at Birmingham, differ-
ing from those sent by MM. Martin, as regards the material used
in the process ; they had turned their attention to the production
of steel by dissolving wrought-iron in a bath of cast iron, whereas
my efforts were directed, from the first, to the use of cast iron and
ore for the production of open-hearth steel.

In the process as it is now carried on at the Landore and other
works, both scrap metal and ore are employed, in conjunction

S/ff WILLIAM SIEMENS, F.R.S. 389

with pig metal and such other ingredients as serve finally to
adjust the quality of the steel. The process may be described as
follows : — The furnace having been heated up to the steel-melting
point, or say, 3,500° F., the first duty of the steel-melter is to see
that the silica bottom and tapping hole are in the proper condition
for work. If, in consequence of wear caused by previous charges,
the surface-bed should be pitted, white sand, previously calcined,
is introduced in such quantities as to fill up the inequalities, and
heat is allowed to act for eight or ten minutes with the furnace
doors closed, by the end of which time the silica or white sand
introduced will be partially melted and consolidated with the
older portion of the furnace-bed. The tapping hole is filled up
with white sand mixed with powdered anthracite or coke, which
serves to prevent its entire consolidation, and thus facilitates the
tapping of the furnace at the end of the operation.

These preliminary operations completed, the furnace is charged
with say six tons of pig metal, mixed with two tons of such iron
or steel scrap, as gits, spillings of previous operations, old iron or
steel rails, that may be available. The furnace doors are there-
upon closed, and heat is allowed to act upon the charge for two
hours and a half, when it will be found to have fused, and analysis
would prove the metal to be an intermediate condition between pig
iron and steel, its percentage of both carbon and silicon being greatly
reduced. The subsequent work of oxidation of these ingredients
consists in the introduction, at intervals of about half-an-hour, of
rich ores or oxides of iron, in charges of about 5 cwt. each ; the
immediate effect of the introduction of each charge is an active
ebullition, through the reaction of the oxide of the ore upon the
carbon of the metal, producing carbonic oxide. This gas escapes
to the surface, whereas the iron contained in the ore or oxide
becomes metallic, and is added to the bath. "When about 25 cwt.
of ore have been thus added, a sample is taken from the bath, by
means of a small iron ladle, and subjected to a simple mechanical
test, whereby the percentage of carbon remaining in the metal is
readily, though somewhat roughly, ascertained. If it appears
from the test that the carburization is nearly completed, no more
ore is added, but 3 or 4 cwt. of limestone are thrown into the
lurnace, which has the effect of combining with the silicon con-
tained in the slag, and of liberating ferrous oxide from the same,

390 THE SCIENTIFIC PAPERS OF

which latter, being thus set free from its combination with silicon,
continues the action of decarburization of the metallic bath.
Samples are again taken, until the steel-melter finds that, upon
breaking the sample, the peculiar silky fibre is obtained which is
indicative of a reduction of the carbon in the metal to O'l per
cent. The metal is now ready for final adjustment, according to
the strength or temper of the steel required.

If it is intended to produce ordinary rail metal, from 7 to 8 cwt.
of spiegeleisen, containing 20 per cent, of manganese, previously
heated to redness, are charged in, the bath is stirred by means of
a rabble, and, after being allowed to rest for a few minutes, is
tapped either into a ladle or directly into ingot moulds, arranged
in groups, whilst the slag that followed the metal through the tap-
hole is collected in a pit or mould prepared for its reception. The
amount of slag produced depends chiefly upon the percentage of
silicon in the pig metal used, and also upon the degree of purity
of the ore employed in effecting the reduction, and amounts gene-
rally to 2 tons in an 8-ton charge. The yield of metal should be
within 1 or 2 per cent, of the total amount of pig metal and scrap
(if of a solid description) charged into the furnace, because,
although the pig metal would contain some 7 or 8 per cent, of
carbon and silicon, which have to be expelled, this loss of weight
is made up by the metallic iron given up by the ore. The lime
added near the end of the operation is useful in taking up some
of the other impurities, such as sulphur and phosphorus, from the
metal, although the amount so taken up is only small.

After tapping, the steel-melter again inspects his furnace-bed,
repairing any slight defects that may have arisen, fills up the
tapping-hole, and introduces the next charge.

The time each charge occupies is from seven to nine hours,
according to its character and to the heat of the furnace. When
pig metal and scrap alone are used, a charge can be worked in
from six to seven hours, and the proportion of pig metal employed
can be reduced to from 10 to 15 per cent, of the total charge,
whereas when no scrap at all is used, the amount of pig metal
charged must be equal to the total amount of steel to be produced,
and the reaction between the ore and pig metal extends the time
of each operation by about three hours. In this respect, then, the
scrap or Siemens-Martin process has an advantage over the ore

SIR WILLIAM SIEMENS, F.R.S. 391

process, which is compensated for, however, by the correspond! ug
advantage in favour of the ore process, that it is not dependent
up' in the irregularities appertaining to scrap metal, and upon the
purifying action produced upon the fluid metal first by the oxide
and thereafter by the lime. It has been found generally that for
small applications the scrap or Siemens-Martin process is the more
advantageous, while for large applications the ore process has the
advantage.

For the production of steel of special quality, such as is em-
ployed for boiler and ship plates and castings, the process is
different only towards the eud of the operation from that already
described. The reduction is carried to a still lower degree than
O'l per cent, of carbon, and in order to make sure that the right
degree of carburization is attained, chemical analysis of a sample
is resorted to. Instead of spiegeleisen, a rich ferro-inanganese is
employed, together with a small proportion of silica iron (a pig
metal containing about 10 per cent, of silicon), which latter metal
has the effect of taking up oxygen from the fluid iron, and thus
preventing blow-holes in the casting. Another method of con-
solidating steel is that which has been introduced and so success-
fully carried out by Sir Joseph Whitworth. The steel upon which
he operates is made upon the open hearth of the furnace in the
manner I have described ; but steel when it is poured from the
ladle into the moulds shrinks, and during shrinkage little air-
spaces or hollows are formed, which break the continuity of the
steel, although the cavity may afterwards be closed. Sir Joseph
Whitworth, by applying great pressure through hydraulic agency
to the steel while in the fused condition, closes up these cavities,
and steel is thus produced perfectly continuous in its nature, and
of such great hardness and tenacity combined, that when put, for
instance, into the form of shells, some of these shells have gone
three or four times through thick iron armour, and have been
quite fit to go into the gun again.

The rich ferro-manganese now introduced into the market
affords the steel-maker great facility for producing sound metal,
notwithstanding its admixture with a considerable amount of
impurities, notably of sulphur and phosphorus. By its means
such inferior irons as the Cleveland can be rendered suitable for
the production of steel rails, and large contracts have been carried

392 THE SCIENTIFIC PAPERS OF

into effect by some of my licensees for converting old iron rails of
mixed or doubtful parentage into steel rails.

The facility thus offered to steel manufacturers introduces,
however, a danger upon which the steel user will have to fix his
attention. A steel rail containing from 1 to 1| per cent, of
manganese may look well and resist the tests for toughness and
strength which are usually applied, and may yet contain more
than \ per cent, of phosphorus or sulphur, or both these sub-
stances. It is only at temperatures below the freezing-point that
the presence of phosphorus will make itself felt by symptoms of
cold shortness, and it would therefore certainly not be advisable to
put down rails of this description in cold climates.

For the construction of ordnance and other high class purposes,
more than a trace of manganese in the metal is, in my opinion,
decidedly objectionable. Manganese, though very efficacious in
hiding impurities in the steel, is in itself an impurity inconsistent
with high quality of the material produced. Its admixture with
the metal is purely mechanical, and upon analysis of different
portions of the same ingot it is found that its distribution is very
irregular. Being more oxidizable than iron, a metal containing a
considerable percentage of manganese cannot be re-heated without
deterioration ; is pitted by exposure to sea-water ; and its strength
and toughness are also found to be below those of really pure
metal when subjected to crucial tests. It is important, therefore,
that steel for war purposes where high temper and great tensile
strength is required, should be practically free from manganese,
as well as from all other admixtures, with the sole exception of
carbon. Extra mild steel, which is so remarkable for its extreme
ductility, should contain in 100 parts 99*75 parts of metallic iron,
and only 0'25 per cent, of all foreign substances put together.

It is for the production of these special qualities of steel that
the open^hearth process has come into extensive use, being
employed, either wholly or partially, by many of the leading works
both in this country and abroad. The total production of open-
hearth steel (both the Siemens-Martin and Siemens variety)
amounted in 1877 to 275,000 tons, since which time its produc-
tion has gone on increasing, notwithstanding the extreme depres-
sion which continues to prevail in the iron and steel trades.

The French Admiralty were the first to take up the subject of

WILLIAM SIEMENS, F.R.S.

393

constructing ships of very mild steel, and the British Admiralty
n«'\v use it largely in naval construction, with such results as I
k'lu-vc will shortly be placed before another Institution by their
chit f constructor, Mr. Barnaby.

Although I have described the open-hearth process, as dealing
with pig metal and ore — or, when it can be had, pig and scrap metal —
my attention has been directed for many years to the accomplishment
of a process, in which the ore used is put through a preparatory pro-
cess of reduction and precipitation, and only a minimum quantity
of pig metal is employed to impart fluidity to the mass in the melting-
furnace. In my early experiments in this direction I followed the
lead of Chenot and others in producing what is called spongy iron,
or iron deprived of its oxygen by heating it to redness, in com-
Mi.ation with carbonaceous material. I soon convinced myself,
however, that no practical results could be obtained by this means,
inasmuch as the spongy iron contains, bound up with it, the gangue
of the ore, which can with difficulty be separated from the metal,
and afterwards encumbers the melting-furnace with excessive slag.
All the hurtful impurities contained in the ore, such as sulphur,
phosphorus, arsenic, &c., remain, moreover, in the spongy iron ;
and, as regards sulphur, its quantity is much increased on account
of a powerful absorbing action, exercised by the spongy iron upon
the sulphurous acid contained in the flame of the furnace. It was
necessary, therefore, to devise some plan by which the metallic
iron could be simultaneously separated both from the ore and its
impurities. This object I have succeeded in accomplishing by
means of a rotating furnace, which has, however, hitherto received
only a limited application. The furnace, which is represented
in Plate 47, consists of a gas-producer, the air-regenerator, a
reversing- valve, and the revolving drum. No gas regenerators
are employed in this furnace, but the gas passes from the pro-
ducers through an oblong channel continuously into the revolving
chamber, where it is brought into contact with the heated current
of air passing in from one or the other of the air regenerators.
The flame thus produced rushes forward into the heating chamber,
and after heating the material therein, passes back again towards
the inlet side, whence the products of combustion pass through the
second air regenerator and the reversing valve into the chimney
stack. By this arrangement the front of the rotatory furnace is

394 THE SCIENTIFIC PAPERS OF

left free for access, and is provided with a charging and discharging
door placed eccentrically, for the convenience of withdrawing
masses or balls of iron from the furnace on a level with the lining
when the furnace is stopped with the aperture in its lowest
position. The lining of the furnace is made of highly aluminous
or bauxite bricks covered on the inside with a certain thickness of
iron oxide produced by melting hammer scale and rich ores in the
furnace, which set while the furnace is kept slowly rotating. The
rotation of the furnace is effected by means of a small Brother-
hood or other engine, and suitable gearing.

The modus operandi is as follows : — The furnace being already
lined and heated, a batch of ore mixed with fluxing and reducing
materials, in a proportion depending upon the chemical con-
stitution of the ore employed, is charged from an elevated platform
in front, the rotator being stopped for this purpose with the
charging orifice in its upper position. From :30 to 40 cwt. of
batch is thus charged, upon which the door is closed, and the
furnace chamber is made to rotate at the very slow rate of six or
eight revolutions per hour. A high temperature is produced
within the chamber by the combination of the gas with the highly
heated air from the generator, causing the mass rapidly to become
hot, whilst slowly rotating, so as to present continually new
surfaces to the heat. No chemical action takes place under these
circumstances until the temperature of the mass is raised to a full
red heat, when reaction between the carbonaceous matter and the
ore will take place, giving rise to the development of carbonic
oxide, which, meeting the heated air proceeding from the re-
generators, is burned, and thus adds to the heating action of the
flame. When this reaction has fully set in, the supply of producer
gas may be almost entirely stopped, and thus no sulphurous gas
is admitted into the furnace during this critical interval. The
heat now rises rapidly, and fusion of the earthy constituents of the
ore occurs simultaneously with a continuance of the reducing
action. In the course of an hour and a half after starting, the
charge consists of metallic iron in a more or less agglomerated
condition, found on analysis to be almost chemically pure, and of
a liquid mass of cinder containing the earthy constituents of the
ore and other foreign matter. The rotation of the furnace is now
stopped with the tapping-hole in its lowest position, and the bulk of

WILLIAM SIEMENS, F.R.S.

395

tlir finder is discharged ; the tapping-hole is thereupon closed
again, and the furnace made to rotate somewhat more rapidly with
a \ ifw of facilitating the agglomeration of the metallic iron ; by
tin- timely introduction of a rabble, the agglomeration of the mass
can be so regulated as to induce the formation of two or three
balls of convenient size for handling. The balls being formed,
the furnace is stopped with the large door in its lowest* position,
which, upon being removed, admits of the charge being withdrawn.
This is effected by the introduction of tongs supported by pulleys
running upon overhead rails for transferring the balls in rapid
succession from the furnace to a squeezer (which has for its purpose
to expel adhering cinder), and from the squeezer to the bath of
such a steel melting-furnace as already described.

The great purity of the metal thus reduced from the ore, and
the rapid and comparatively inexpensive nature of the reducing
process, are conditions highly favourable to the production of steel
of high quality by this method at reasonable cost ; and it is my
intention gradually to complete the open-hearth process of pro-
ducing steel by combining with it the mode of preparing the
material to be melted just described.

What I also wish on this occasion to call attention to are certain
properties of steel which are of importance in considering its
applicability to engineering and military construction. We know
that steel varies between very large limits in its hardness and in its
ductility, but it is not so generally known that steel up to a
certain point is of the same strength to resist strain when it is mild
as when it is hard, and I have prepared a table of results which
shows very clearly the nature of both mild steel and hard steel if
compared under different strains.

Upon reference to the table it will be seen that loads from G to
15 tons per square inch affected all the bars equally, whether hard
or soft, annealed or unannealed, and with the exception of one bar
only this uniformity holds good up to 18 tons ; up to this limit,
the elastic elongation of all the samples was equal ; but with the
strain of 18| tons, the mild steel has become permanently elongated,
whereas the harder steel shows a normal increase of elastic elonga-
tion. With the hard bars experiments were continued, and 21
tons to the square inch was applied, producing an elastic elonga-
tion '108, which is entirely normal, and no permanent set is pro-

396

THE SCIENTIFIC PAPERS OF

P !.'

^ §

S K)

fe ^

<K ^

i

H

(rt

fc
O

03
H
4
O

U5 COINOOOO

>O CC(MOC(NCO

1J

IO tO <M OO O >O IO 15 «ff (M CO <M CO IS (M

10 ecwoooo

'

S5 <M lO GO i

S/K WILLIAM SIEMENS, /-.A'..s.

397

but \\lien 24 tons were applied, then a permanent set
occurred also in the hard bars ; therefore the elastic limit of the
hard steel was 24 tons, whereas that of the milder steel was only
18 tons.

r.nt there is a very peculiar circumstance connected with these
elongations. If a bar of mild steel is taken from the rolls and
subjected at once to a test of say 18 tons, it will very likely be
found to show a permanent elongation, but if the same bar is first
subjected to a strain of say 17 tons, for several hours, it will then
be capable of resisting perhaps 19 or 20 tons before showing any
permanent elongation. One might also say the bar of steel can be
taught to resist a higher strain without yielding permanently. Sir
"William Thomson has lately made some elaborate researches on
this point, and perhaps he will favour the Meeting with some
account of them.

The question of using steel for the purposes of engineering or
military construction depends a great deal upon the particular
application. Mild steel has the peculiar quality of yielding to an
enormous extent to strain before giving signs of rupture. Bars
are tested generally to 28 tons, at which mild steel, such as is used
in naval construction, generally breaks after showing an elongation
of 25 per cent. The steel used in the construction of boilers,
which is made still milder, will stand only a total strain of 24 tons,
but will show a still greater elongation before breaking. From
this we can go upwards, and produce steel that will bear a break-
ing strain of 50 tons, with an elongation of perhaps 12. per cent.
Still harder steel shows a strength of GO tons, and an elongation of
only 7 or 8 per cent, before breaking, whilst Sir Joseph Whit-
worth has shown that the absolute strength of steel in the form of
bars, of the proper temper, may be brought up to 90 tons per
square inch, by subjecting it to a process of oil hardening ; such a
tensile strength is hardly exceeded by carefully tempered steel
wire. Therefore we have a range of strength which we can apply
under different circumstances with great advantage. It must be
borne in mind that the harder steel is apt to become brittle when
suddenly cooled, and therein consists the great safety of using the
milder description of steel for engineering and military purposes.
With regard to this last matter I would say something before con-
cluding, with reference to the construction of ordnance.

39^ THE SCIENTIFIC PAPERS OF

Years ago a great advance was made in gun manufacture by Sir
William Armstrong in producing his well-known mode of con-
struction. At that time wrought iron was the strongest material
practically available for the gun-maker's use, and this was put into
the strongest possible form by the construction of coiled rings,
which method places the fibre of the metal in the direction of the
strain. The Woolwich or Fraser system of gun construction being
a modification of the Armstrong system comprises the same mode
of putting iron into the condition of greatest strength ; but it is
time I think to enquire whether, after the recent advances made
in the production of mild steel, which is a material of superior
strength, tenacity, and uniformity to iron, the mode of construct-
ing ordnance should not be modified to suit these altered
circumstances.

It is important then to appreciate wherein consists essentially
the difference between iron and mild steel.

Mild steel is a metal consisting of 99*75 per cent, of the elemen-
tary substance, iron ; and only a quarter per cent, of manganese,
carbon, and such impurity as phosphorus and sulphur in the
smallest possible quantity ; whereas wrought iron of commerce
generally consists of 96 to 97 per cent of metallic iron and be-
tween 3 and 4 per cent, of other material, for the most part slag.
Now, it seems not a very great matter that in 100 Ib. of iron there
should be 3 Ib. of slag, but if we represent this proportion to our
eyes we see that it is not such a very inconsiderable quantity,
considering that slag is both voluminous and entirely devoid of
tensile strength. I hold here two cubes, one of 4^" side, and the
other of H" side, representing as nearly as may be the one the
metallic iron, and the other the slag which when mixed together
form wrought iron. If the slag was mixed up amongst the mass
of metallic iron in an irregular way, the strength of the iron would
probably be very little more than is due to that of the glassy slag,
because filaments of slag might go right across ; but in drawing
out the iron, again and again, the little original globules of metal
become elongated into strings or fibres of iron, held together by
filaments of slag, and thus we get in iron a great apparent increase
of strength by drawing it. But even if we draw it out to the
utmost, we lose strength to the extent of the sectional area taken
up by the slag, and thus get less resisting power than if the pure

SIR WILLIAM SIEMENS, F.R.S. 399

metal is separated from the slag in subjecting it to the melting
jinn-ess, when we get the maximum strength of which the metal is
r;i|i:il)l''in all directions, and we have, in fact, metal of the greatest
rth for moderate strains that can be obtained, as we have
sivn, that even the hardest steel elongates as much as the mildest
metal when subjected to moderate strains.

At another Institution Mr. Longridge has severely criticised the
coil system of construction, which is still followed at Woolwich,
upon the ground that the stresses are not properly distributed ;
but whilst not agreeing with him, chiefly as regards the limits to
which shrinkage can be advantageously resorted to, I cannot, on
the other hand, but think it is wrong in principle to use the
harder and more resisting material in the inside, and the weaker
material outside the gun, as is still practised at Woolwich.

Mr. Longridge says that the inner tube of the gun should be
under compression, and therefore one or several layers of rings
should be shrunk on with such increasing force as to bring the
metal into considerable tension, in order that when the powder
gas acts expansively upon the inside of the gun, the compression
of the inner ring may be such as to resist the first part of the
impact of the powder, and then after having come to its condition
of neutrality take up its proper portion of the tensile strain. But,
practically, I believe the shrinkage is not carried to any such
extent, and it appears to me reasonable that it should not be, not-
withstanding Mr. Longridge's argument, because if the large mass
of iron he suggests to use is put under compression to the extent
indicated by theory, it would inevitably crush or permanently
deform the inner tube. But if the inner tube is of steel and the
outer portion of iron, a metal of less elastic range than steel, it
follows that by repeated expansive actions, the external metal will
be strained beyond the limit of elasticity before the metal of
greater elastic range in the interior.

Again in firing the gun the inner metal will be expanded by the
heat, which will increase the pressure exerted by the inner ring
against the outer rings. This action will result in excessive strain
on the outer rings tending to enlarge and loosen them, or in com-
pressive action upon the inner tube, tending to produce the same
result through crushing.

It appeara, therefore, to me to be evident that, in constructing a

400 777^ SCIENTIFIC PAPERS OF

gun, steel only should be used, in which the strains should, if
possible, be so distributed that when the powder pressure acts,
each portion should offer the same resistance to the strain. This,
I think, might be effected in a very thorough manner by a process
analogous to that employed by Admiral Eodman in the construc-
tion of cast-iron guns, only that cast-iron is perhaps the material
least adapted for the purpose. If a steel gun or a ring forming
part of the same was put into a furnace, heated up to a tempera-
ture of say 600° C., and the inside was subjected to cooling action,
while the outside was maintained at the temperature of the fur-
nace, a distribution of stress would result which would be highly
advantageous to the strength of the gun. The chilling of the
inside surface of the ring or gun would cool the metal towards the
inside circumference. This metal could not shrink, nor would the
inner diameter of the gun diminish, because the diameter would
be determined by the mass of metal still in the heated condition.
The vacancies produced in the cooled metal would be filled up by
the inflow of metal from the heated mass outside, resulting in
equilibrium of the metal at the diameter originally due to the
heated mass, and the temperature will gradually vary from say
100° inside to 600° outside. If the gun was afterwards taken out
and allowed gradually to cool, but without stopping the cooling
action from within, the whole mass will cool down to the minimum
temperature. If we imagine the ring to consist of a succession of
concentric cylinders, each cylinder would acquire a tensile strain
due to its previous temperature, which being a minimum on the
inside and a maximum on the periphery, there would result a
distribution of tension throughout the mass, being negative or
compressive in the interior and more and more tensile towards
the exterior. Then when the full pressure of powder gas was
active, the strain upon each portion would be equal, and the
resulting strain would be opposed by the whole elastic strength of
the metal. The internal portion of the tube would, with such a
mode of construction, have no other function to perform than to
resist the abrasion that is necessarily going on in the gun. This
question is just now very much before the scientific public, and
therefore I thought it well to bring before you my own view of
the matter.

Before quite concluding, I would call attention to a machine

S/K WILLIAM SIEMENS, F.R.S. 401

which lias lately been sent me from America for testing pieces of
steel or iron, and after the meeting is closed I will break a test bar
of mild steel, that the members may see the amount of elongation
of which such a bar is capable before breaking.

Dit. SIKMEVS. With regard to an observation made by General
Younghusband, I should ask permission to say one word. I do
not wish to suggest for a moment that a gun, such as is now
turned out at Woolwich, is not a very excellent mechanical pro-
duction ; and at tha time the materials that were to be used were
de3ided upon, thos3 were ths very bast materials that could be
obtained. And with regard to iron, I admit the value of putting
the strain in the direction of the fibre, but at the present time we
have made a very considerable step forward, which I think should
be taken advantage of in the construction of guns. If the coil
is limited in its length, it gives, a? General Younghusband says,
no longitudinal strength, and the inner tube becomes a matter of
much greater importance than it would necessarily be if the out-
side of the gun, instead of being multifarious, was solid. But
there seems to be no reason why, if the outside of the gun was of
steel, and the elasticity of that mass of steel was properly distri-
buted, it should be divided, and not be of a single piece. If that
was the case, the inner tube would have only one function to
fulfil, that of taking the rifling ; but it is most essential, I con-
sider, that it should be as mild as is consistent with its power to
resist abrasion, inasmuch as it cannot be true in principle, although
it may be perfectly workable, to have the material of which the
lining is composed of greater resisting power than the material
surrounding this inner tube, upon which the strength of the gun
depends. The lining must necessarily be subjected to very con-
siderable expansion through the firing itself. One fact I should
like to mention, viz., that a shrinkage of 1 in 1,000 produces \\\
to 12 tons of strain to the square inch, and an elongation of 1 in
1,000 is produced by heating the material to 130° Fahr. ; there-
fore, if a gun is built up by shrinkage, and the inner tube after-
wards heated up to 180° Fahr., the amount of fight between the
inside metal and the outside must be just double. If the heating
should exceed that limit, as it appears to me it would be likely to
do, the strain between the inside and the outside must increase in

VOL. I. D D

402 THE SCIENTIFIC PAPERS OF

the same ratio ; and this action between the two principal portions
composing the gun could only be obviated if the inner tube was
of such a material as to yield absolutely to the outside strain.
Therefore, it was not with a view of criticising, but rather with a
view of suggesting, that I ventured to make the observations I did.
I thank you very much for the kind attention you have given me
this evening.

Tn llw, discussion of the Papers

"ON IKON AS A MATERIAL FOR ARCHITECTURAL
CONSTRUCTION," by J. ALLANSON PICTON, F.S.A., and

" ON MILD STEEL," by A. W. B. KENNEDY, M. Inst. C.E.,

DR. C. "W. SIEMENS, F.R.S., Hon. Associate,* said : I am sure
I cannot do less than congratulate you, Mr. President, and the
Institute of Architects upon the interesting and valuable papers
to which we have listened. Mr. Picton, in surveying the whole
question of the application of iron and steel to architecture,
appears to have taken such a broad and enlightened view of the
subject, that I hardly know what to add. Professor Kennedy
has dealt more particularly with steel, the latest development of
iron as it may be called, and with its preferential merits to iron ;
and as my name has been connected with one of the processes
now largely used for producing this material, I may be expected
to offer a few remarks chiefly with reference to the properties
of steel and its applicabilities for architectural purposes. Steel
is a material, which, as Professor Kennedy very correctly said,
has not been well defined. We call steel the material of which
watch-springs and cutting tools, needles, &c., are made, and we
all know it is very remarkable for its hardness, its great elasticity
or rather, I should say, its high limit of elasticity, and also for
its brittleness the moment that limit is exceeded. We call steel

* Excerpt Transactions of the Boyal Institute of British Architects, 1879-
1880, pp. 173-176.

SIX WILLIAM SIEMENS, F.R.S. 403

the material now used largely in railway structures, both in pro-

ducing wheel tyres and the rails upon which they run. This

ial is remarkable for great strength coupled with a sufficient

i •!' toughness to resist very heavy blows, and we now call

: In- material which exceeds even copper in ductility, stretching
as it does 25 per cent, before breaking. Yet I am in favour
«.f continuing the appellation of steel in speaking of all materials
consisting mainly of iron, which are malleable and have been
proilucctl by a process of complete fusion. It is to this cir-
cumstance of complete fusion that the superior uniformity and
the certainty of character of the steel of the present day are due.
If we follow the production of wrought iron through all its

s we almost wonder how even such uniformity of strength
and character is produced as we find it to possess. In puddling
iron the workman produces about 1 cwt. of metal that has been
massed together from a semi-fluid condition, and the nature of
which depends upon his skill, the temperature of the furnace,
and other variable conditions. The consequence is that one ball
is never quite similar to another. One may be what is called
young iron, and another is more matured when it leaves the
furnace. The young iron will, after consolidation by hammering
and rolling, show a crystalline fracture, and the other more en-
tirely decarburized material will possess a fibrous texture. In
order to harmonize the differences in the character of the iron
in successive balls, the process of re-mixing or packeting is re-
sorted to, which consists in welding together bars produced from
different charges, and rolling them anew. Thus by very careful
treatment the inequalities naturally appertaining to puddled iron
may be reduced, and iron of high quality, such as Yorkshire
plate, may be produced ; but such care is not always taken, and
hence arises the uncertain character of the iron of commerce.
Even the best commercial wrought iron is not pure iron, but
an agglomeration of that metal with cinder (as may easily be
seen by viewing a section of a bar through an ordinary magni-
fy ing glass), and not unfrequently with more hurtful substances
such as sulphur and phosphorus. In producing steel, however,
and particularly in producing the steel now used in engineering
and architecture, a mass of some ten or twelve tons may be seen
in the furnace in a state of perfect fluidity, in which state it

D D 2

404 THE SCIENTIFIC PAPERS OF

is tested by means of samples both as to its chemical and
mechanical condition. It is thereupon run into a ladle and from
that ladle into ingot moulds, and all we have to do afterwards
is to give it the particular form or shape required for our pur-
pose. It is therefore nothing more than natural that a material
so produced should be much more uniform and trustworthy than
the iron of former days. There is, however, still a prejudice —
I cannot call it by any other name — in the minds of many of the
users of these materials, because the steel of former days was
of an uncertain nature, and occasionally gave way when least
expected. Professor Kennedy has sufficiently alluded to the
causes of this, and I have no doubt but that very shortly this
idea will be got rid of entirely, and steel will be used in lieu
of iron for a great many purposes to which it has as yet been
unapplied, and that it will be trusted to a higher limit of its
strength than has hitherto been allowed either by the Board of
Trade or by the insurance companies. The advantages of steel
as a material to be used in architecture, I would consider to be
the following : — For purposes where boldness and grandeur of
outline are essential no material can rival steel ; when we want
to bridge a third of a mile in span, or to construct a roof or
dome of enormous size, there is no material that can serve our
purpose like steel ; if the object is simply to get tensile strength,
without subjecting the material to cross strain, as is the case
in the chains of a suspension bridge, the use of steel wire enables
us to attain a limit of strength exceeding 100 tons per square
inch, or as much as five times the tensile strength of wrought
iron. The chains supporting the great American suspension
bridges across the Falls of Niagara and across the Hudson River
at New York are constructed of this material, and in the latter
case 120 tons per square inch is the breaking strength to which
the wires are tested before approval. If, instead of wire, links
or bars are used in the construction of bridges subject to con-
tinuous strain, the breaking strength (which is really optional,
being dependent upon the percentage of carbon admitted into
the material) may be conveniently fixed at from 40 to 50 tons
per square inch, or say twice the strength of wrought iron.
This same material lends itself best for rolling girders to span
large openings, and it will be readily conceived that, although

SIX WILLIAM SIEMENS, l-.R.S.

405

1 1 •solute strength of such steel is only twice as great as that
of iron, the girders themselves, if of considerable length, may be

'•(1 in section in a much greater proportion as compared
with iron, because a considerable portion of the duty of a

in»n girder consists in its having to carry its own weight.

•Blueing that weight to a half in taking advantage of the
extra strength of steel, a greater proportion of the total strength
Incomes available for carrying load, which latter being a fixed
i|iiantity renders a further diminution of scantling possible. It
is in consequence of this compound effect, resulting from the
use of high-class material, that its great advantage in large
structures arises, and it is difficult to assign the limit of size
which skilful treatment may not accomplish by its use. These,
however, would be exceptional instances, and the more important
question after all is : How does steel compare with its old rivals,
iron and wood, in ordinary construction ? I would venture to
say that the use of steel in certain portions of ordinary build-
ings is not only better and safer, but absolutely cheaper than
the use of such a material as wood. In using steel girders instead
of wooden beams in the construction, for instance, of a ceiling,
what is the comparison as regards efficiency and cost ? The
flooring, supported on steel girders, would be more permanent,
being both fire-proof and not subject to gradual decay ; it
would, moreover, if made as rigid as a wooden flooring, be four
to six tunes as strong when subjected to exceptional loads. As
regards price, granting the material would be dearer, we should
have to take into our calculation an element which is not at
once apparent. Whereas the depth of wooden beaming of suffi-
cient strength would be, say twelve inches, if constructed of steel
a depth of seven or eight inches would amply suffice ; and if we
were to repeat this, say five times in a house, a total saving in
height of two feet would be the result. Now comparing the cost
of the cubical contents in the structure built in the two ways, we
should find that in a house of an area of thirty feet by sixty, in
taking the price of construction of one cubic foot of capacity at
Is. 2d., without taking into consideration fittings and ornamenta-
tion, we should save £210, to be placed to the credit of the steel
girders. If allowance were made for this or even half of this it
would be found that the steel girder would probably be the cheaper

406 THE SCIENTIFIC PAPERS OF

construction of the two. I may mention here a case with which
I had to deal in a small way a year or two ago. I have at my
house in the country a terrace, and beneath that ten-ace a billiard
room ; I wanted to lower the terrace six inches and to raise the
ceiling of my billiard-room at least a foot. The billiard-room
had been constructed according to ordinary practice. I managed,
by putting steel girders over this billiard-room (which was about
seventeen feet span), by filling in between girder and girder with
cement and covering it with the same sheet of lead that had
formerly been used, to save the eighteen inches I required. The
result was a perfectly dry room of increased loftiness, whereas
before I had considerable difficulty in keeping the water out.
This simply shows how, by the use of the stronger material,
structural advantages, besides saving in cost, may be obtained.
But it has also been reproached against iron that it is not fire-
proof. I have heard Captain Shaw say he liked houses with
wooden beams and columns, because he knew exactly when they
would give way, whilst this was not the case with iron. The
remark is no doubt the result of very extensive experience, but I
expect that it has arisen through the use of cast iron. ISTow cast
iron, I should think, is a very unsafe material to be used in archi-
tectural structures. You can never know with certainty whether
the iron is sound throughout, when put in the shape of a column
or girder, or whether there are not unperceived cavities Avithin the
shell that make it unsafe. Another danger is that in case of fire
cast iron, when heated partially, is apt to crack, and the fire
brigade would receive no warning as to when the building would
give way. But in using steel a very different condition of things
obtains. Steel is produced at the highest temperature attained in
the arts, and it sets or solidifies at a temperature far exceeding
that reached in a burning house. Therefore, although the struc-
ture contains this material in parts where strength and solidity
are required, it would be improbable that any portion of the fabric
would be heated, in case of conflagration, to a temperature suffi-
cient to make it yield on account of that temperature. The
temperature would depend upon the amount of combustible
material accumulated at any one place, and by the employment of
steel instead of wood in the structure the quantity of combustible
material would be necessarily much diminished ; it would be

S/A WILLIAM SIEMENS, F.R.S.

407

in fuct to furniture and stores, and it is probable that a
conflagration would remain confined within very narrow limits.
If liy concentration of flame the heat should nevertheless reach a
hiirh point in any portion of the building, the steel when heated
dness would certainly deflect to an enormous extent before
giving w;iy, ;uid the same indication which the fire brigade men
receive in the case of wood they would have in probably a still
greater measure in dealing with steel. But an almost complete
protection of the steel employed in structures would be obtained
by constructing the floors of girders filled in between with plaster
or cement, a method which is more largely resorted to in France
and other portions of the Continent than in this country. In
employing steel in the construction of columns the metal may be
conveniently rolled in the section of a cross and be enclosed in a
casing of plaster or cement, which latter will in that case fulfil
the double purpose of protecting the steel against accession of
heat and of lending itself to such structural effect as the architect
may desire to produce. In making large girders of steel, riveting
has to be resorted to, and the quality of steel that should be used
for such purposes differs widely from the harder and stronger
description of material applicable for tie rods, rolled rafters or
columns. These harder descriptions of steel would lose strength
in an extraordinary degree through break of continuity such as is
necessarily produced by punching and even drilling for the intro-
duction of rivets. Although steel is immensely stronger than iron,
it is more apt to tear from a point of discontinuity, and this
liability to tearing action in steel increases with its strength. By
careful annealing, risk from this source may be reduced ; but an-
nealing itself is a delicate and therefore risky operation, and no
work would be safe that depended upon its due accomplishment.
The steel-maker of the present day is not at a loss, however, to
supply a material combining extraordinary toughness with a
strength still greatly exceeding that of iron. This " mild steel,"
which has recently nearly expelled iron from naval construction,
has an absolute strength of about thirty tons per square inch, but
its toughness is such that if a bar of eight inches in length is
subjected to increasing strains it assumes a length of ten inches
before giving way. Nor is this toughness dependent upon previous
annealing ; on the contrary this mild steel may be heated to red-

408

THE SCIENTIFIC PAPERS OF

ness and cooled suddenly by plunging it in water, without losing
its yielding property to any great extent. Although the absolute
strength of this material does not much exceed that of best iron,
its superiority consists in its uniformity and power of yielding
before breaking, which makes it a safer material to be used when
weighted even to a point nearer its elastic limit than it would be
safe to go to in dealing with iron. The Board of Trade have fixed
upon six tons and a half as the weight allowable per square inch
of steel in bridge work, instead of five tons, the weight allowed
for iron, but this rule, which makes no distinction between steel
and steel, will in all probability be yet considerably modified in
favour of the new material. It has been mentioned in Mr. Pic-
ton's paper that iron, when applied to architecture, has very often
been hidden in order to obtain its strength without acknowledg-
ing its use. That it could be worked into elegant shapes he
proved by the productions of mediaeval times, but it appears that
this art of working in iron has been — if not abandoned — very
much less used, and one of the reasons is probably to be found in
the fact that, after the introduction of the blast furnace, wrought
iron was no longer produced of its former excellent quality, and
that there was a difficulty in working inferior wrought iron into
those forms which are admired in old churches, gates, &c. But
in mild steel we have a material which is above all others capable
of being put into perfect form. We have in this material not
only the greatest uniformity combined with the greatest strength,
but we have the utmost security against accident and breakage,
either when loaded in excess or in case of fire, and it is also
a material capable of being wrought into the highest artistic
forms.

S/Jt WILLIAM SIEMENS, F.R,S.

409

In the (fiaritssion of ttie Papers

• "X THE MECHANICAL PROPERTIES OF IRON AND
M 1 U ) STEEL," by D. ADAMSON,

••ox THE USE OP STEEL IN THE CONSTRUCTION
OF BRIDGES," by A. N. MAYNARD; and

"ON THE USE OF STEEL IN NAVAL CONSTRUC-
TION," by NATHANIEL BAKNABY, C.B.,

DR. SIEMENS * said he thought he should not be entirely silent
when a discussion took place on steel such as they were now
engaged in. They had before them three papers dealing with the
use of steel : one by Mr. Adamson. another by Mr. Mayuard, and
a third by Mr. Barnaby. With regard to Mr. Adamson's paper he
would like to say a few words. They had had a very admirable
paper from Mr. Adamson at their Paris meeting, and he had added
some very valuable information that day. He was glad to see that
Mr. Adamson had arrived at a conclusion to which he (Dr.
Siemens) had sometimes given expression, namely, that in rivet-
ing steel plates together steel rivets should be used, and that those
rivets should be put as close together as they could be conveniently
placed. He was glad to see that Mr. Adamson had by his in-
vestigation brought this matter so clearly to light; but it was a thing
which had often come before him in dealing with those very subjects
that had been alluded to by Mr. Parker and others. Mysterious
failures occasionally had occurred. It so happened that the one
failure which Mr. Parker mentioned had been brought to his
(Dr. Siemens's) notice at the time ; and the very nature of
that failure proved, he thought, that the cause was only hidden
below the surface, and was not beyond their ken to unravel. It'
a cylindrical tube was intended to fit into a flanged neck, and the
tube was too small, and if, after being punched all round or
drilled, it was forced out mechanically to fit the larger diameter, it
stood to reason that the metal extending from the rivet holes to the
edge had to be stretched considerably. This extension of the outer

* Excerpt Journal of the Iron and Steel Institute, 1879, pp. 82 and 107.

4IO -THE SCIENTIFIC PAPERS OF

edge would throw a corresponding tensile strain upon the metal inside
the line of rivet holes. The result was, that if in any of these
holes there was the slightest defect or the slightest discontinuity of
the even surface a tear would begin. It was said that metal which
ought to stretch 25 per cent, before breaking should not have torn
under these circumstances, because they could find that it had not
stretched at all. But, in arguing this point, gentlemen forget the
fact that in tearing a material they did not extend it. A tear
meant applying the whole of the force available upon absolutely
one point of the metal. At that one point the metal might
be capable of stretching even 50 per cent., but this would
not save it, inasmuch as the force was confined to one
point. At that point it would tear, and afterwards at another
point, and so on. All the mysterious failures reported might in
nearly all cases be attributed to tearing action, which was gene-
rally set up at rivet holes if the rivets at those holes took more than
their due proportion of the strain. He knew of another instance
where failure took place in a testing bar. It was not intended to
test the strength of a riveted joint. In fastening the piece it was
held by two large rivets on each side, followed up by smaller rivets.
With a metal of high elasticity, such as mild steel, a great strain
would be thrown upon the part most forward. The supports,
which were backward, could not come into full bearing condition
till the forward part had taken the proper strain. Wherever that
strain exceeded certain limits, tearing action would take place
from any part where the continuity was broken. One of the chief
advantages of drilling over punching was that the drilling made a
smoother hole, and there was less susceptibility to tearing; but,
as stated by Mr. Barnaby, mild steel should not require either the
particular attention or expense due to drilling, nor should it
require annealing. The failures spoken of were generally at-
tributable to some fault in the treatment. He would rather not
touch upon the subject of a comparison between Bessemer metal
and the metal with which his name was connected. He had more
than once expressed his great admiration of the Bessemer process,
and of the excellence of the metal which could be produced by it.
At the same time it was not probable that he should have gone to
the great labour and expense of developing another process if he
had not thought it possessed some particular merits. It would

WILLIAM SIEMENS, F.R.S.

411

have been entirely waste labour, and he might ba excused, perhaps,
if he dropped in a word in favour of the latter process without in
any way wishing to detract from the Bessemer process. Mr.
Bessemer had brought before them very interesting specimens,
showing that at the very early stages of the process excellent
material was produced thereby. Therefore he might ask why was
it that this material did not come for a number of years into such
use as the mild steel was now put to ? The answer, he thought,
had been given indirectly by Mr. Bessemer himself. Mr. Bessemer
had said frankly that he did not always produce such excellent
quality, and that there were irregularities which sometimes gave
rise to failure. It might be added, as remarked by Mr. Sneius,
that there was another cause of difficulty, and that was the
chemical composition. If the Bessemer process was carried out
properly, and good Swedish metal or high-class hematite was used,
he did not doubt that excellent results would be produced by it ;
but he thought that in a process such as the open hearth, which
allowed plenty of time for the taking of samples and for adjusting
the composition of the metal at the end of the operation, if they
started on equal ground, with equal care, there must be some ad-
vantage in favour of the uniform production of such metal as they
were speaking of. Another reason he might allege was that of the
chemical composition. Starting with pig metal of the same per-
centage of phosphorus, silicon, and sulphur, in the case of the
open-hearth process, where they treated the metal with ore and
limestone, they did not diminish the weight of metal ; that was to
say, for a ton of pig metal put in they got out within 1 per cent,
of a ton of steel. That in itself meant no reduction in the per-
centage of phosphorus and sulphur. Whereas in the Bessemer
process there was a certain loss of weight, meaning a certain con-
centration within a less weight of those materials. It might be
very little ; it might be nothing if absolutely pure metal was used ;
but if impure metal was used it would be something. On the
following day they would have a process brought before them by
which Bessemer metal could be produced from Cleveland iron. He
did not doubt the chemical result ; but if there was merit in it,
that merit might be claimed in favour of the open-hearth process
as carried on hitherto, inasmuch as in the Siemens process they
had always used ore and limestone as necessary additions to the

.THE SCIENTIFIC PAPERS OF

bath. In 1863 and 1865 he had tried, moreover, many ex-
periments with basic linings, including lime lining, in connection
with the open-hearth process, but he never met with such an
amount of practical success in his endeavours to render such
linings permanent as to justify him in recommending their
adoption to his licensees. He believed the Bessemer metal was
perfectly good for nearly all the purposes to which mild steel was
put, but he was equally satisfied that the open-hearth process was
best suited for the production of such special qualities of steel as
were now under discussion.

DE. SIEMENS remarked that with regard to the mysterious
breakages which had just been alluded to by Mr. John, perhaps
he (Dr. Siemens) had not expressed himself as clearly as he should
have wished on the previous day, if he had said that it was owing
to its great ductility that steel tore more readily than iron. What
he meant to say was, that steel tore more readily because it
was the more uniform and more homogeneous metal. In iron
each fibre was a place at which a tear would stop. In steel, if a
tear was once started, it would run along just as it would, for
instance, with india-rubber, the most elastic of all materials. If
they took a strip of india-rubber and elongated it ten times, it
would return to its original length upon removal of the strain upon
it. If, therefore, the edge of the india-rubber was slightly nipped with
a pair of scissors, and it was again stretched to the same extent as
before, the india-rubber would tear right across from the point of
discontinuity produced by the incision. The same kind of tearing
action was set up in mild steel, when it was treated in a way that
was not suitable to it. In working mild steel drawn out in a
half cold state and allowed to set, great tension arose in the mass ;
and unless the metal was subsequently annealed, a tear might
commence at a rivet hole, and that tear would run along more
readily than in iron. Care should be taken to prevent mischief
from this cause in dealing with mild steel.

WILLIAM SIEMENS, F.R.S.

413

/-/ the discussion of the Paper

"ON THE ELIMINATION OF PHOSPHORUS," by
MR. SIDXKV 0. THOMAS and MR. PERCY G. GILCHRIST,

DR. Si KM HNS* said they should distinguish two things in the
discussion — the proposition to use a basic lining, and the chemical
results which had been brought before them. As to the first of
these, he had no experience and no knowledge with regard to the
Bessemer converter, but as regarded the open-hearth furnace, he
had to say that the question of the advantages which would result
from a basic lining had occupied his serious attention as early
as sixteen years ago. In 1863, he, in conjunction with M. le
Chatelier, conceived and patented a lining for the open-hearth
furnace, with which he was then experimenting for the purpose of
producing steel. They had arrived, by inductive reasoning, at the
conclusion that in order to carry out the process properly, a basic
material ought to be adopted for the lining ; and M. le Chatelier
suggested bauxite (alumina with a certain proportion of peroxide
of iron). They commenced experiments in France, at the Four-
chambault Works, with a lining composed of bauxite, crushed
and rammed into the furnace. But they did not find that they
could get that lining to stand when the fluid charge came upon it.
He afterwards made bricks of bauxite (at the present time he had,
perhaps, 50 or 100 tons of those bricks in Birmingham), and by
the use of these they had succeeded in forming a lining which
would stand for a series of charges. But although the result with
the use of this lining had been good, he confessed that he had
no such marvellous effects to put before them as were stated in
Messrs. Thomas and Gilchrist's paper. He also tried lime mixed
variously with clay and other binding substances, and he found
that the difficulty of getting the binding to stand increased very
much in using that material. He had also used a lining made
of magnesia formed into bricks, and burned at a very high
temperature ; a magnificent lining was thus produced, but it was
expensive, and on that account not commendable. The reason

* Excerpt Journal of the Iron and Steel Institute, 1879, pp. 153-156.

414 THE SCIENTIFIC PAPERS OF

why he did not persevere with the basic lining was, that although
for a few charges it seemed to stand exceedingly well, the slag
containing silica and oxide of iron acted upon it more rapidly than
was desirable ; and it was subject, moreover, to the inconvenience
that, unlike a silica lining, it could not be repaired. In dealing
with silica they had only to put in a few shovelfuls of sand where
the lining was defective, and after the application of heat for a few
minutes, the new material was found to have perfectly combined
with the old ; whereas with a basic lining, the lining had to be
taken out when defective and entirely renewed. These results
discouraged him from continuing the application of these linings
to the open-hearth furnace. He still continued, however, the use
of bauxite brick lining in his rotative furnace, in the manner
described in his paper before the Iron and Steel Institute in 1873.
Reference to the same subject would also be found in his printed
lecture, delivered to the Fellows of the Chemical Society in 1868.
With regard to the subject now brought before them, he found
that there was something wanting in the papers communicated.
Mr. Thomas, in his paper, stated that the amount of basic material
added usually exceeded 2 cwt. per ton of pig iron. He wished
Mr. Thomas had given them, not what it exceeded, but the exact
amount. They knew that basic material would absorb phosphorus,
but the President in his able address had already called attention
to the circle through which they seemed always to wander in
developing the different processes for the manufacture of iron and
steel. The puddling furnace had originally a silica lining, and
only after a basic lining had been adopted were good results
obtained. It was probable, therefore, that in making steel they
were only now passing from the first stage to the second. The
difficulty they had to contend with in this case was, however, a
more serious one, inasmuch as the temperature was very much
higher than in the puddling furnace, in which the oxide lining
could be as easily maintained as the silica lining that was
previously employed ; but it was to be apprehended that the use
of a basic lining in the Bessemer converter and in the open-hearth
furnace would entail a good deal of difficulty and expense in the
renewal of the lining — on which subject, however, he should
be glad to elicit more definite information from the authors of the
paper.

SIX WILLIAM SIEMENS, F.X.S. 415

Tli«? lining itself did nothing towards the removal of phos-
phorus; but «t was the added basic material that effected that
ivsuit, and the question was how much of that basic material was
requisite to remove a large amount of phosphorus from the iron.
.Mr. Bell hud brought forward last year a system of eliminating
phosphorus from iron. He (Dr. Siemens) at the time questioned
whether the amount of rich oxide proposed to be employed by
Mr. Bell, would not be too great to pay for the difference in cost
of hematite and Cleveland pig metals. He would only repeat the
same question to Messrs. Thomas and Gilchrist. Was not the
quantity of basic material necessary to absorb from .the Cleveland

g the amount of phosphorus it contained, such a serious item
as to make it an open question whether they had not better go on
using "West Country iron for steel ? They must also bear in mind
the materials of which the basic linings were composed. Lime-
stone itself was certainly cheap enough, but the question was,
would limestone, unaided by more expensive reagents, carry away
the phosphorus which had to be eliminated ? The paper admitted
that ferrous oxide was employed in the process, and he should wish
to ask how much of that material was necessary to accomplish the
effect contemplated, and how much of it was produced by oxidation
in an overblown charge. In overblowing the charge they had to
oxidise some of the pig metal, and that might be a convenient way
of providing the ferrous oxide, but he should like to know how
much of the iron under treatment was destroyed. He had no
reason whatever to doubt the chemical results brought before
them by Messrs. Thomas and Gilchrist in their paper — results
which were certainly most interesting, and which he would not
have been prepared to expect. The question, therefore, seemed to
him to narrow itself into one of cost, which could only be fully
solved by experience.

41 6 THE SCIENTIFIC PAPERS OF

In the discussion of the Paper

"OX THE DEPHOSPHORISATION OF IRON" AND
STEEL," by M. A. POURCEL, Terrenoire, France,

DR. SIEMENS* said he wished he had had an opportunity of
reading M. Pourcel's paper previous to the meeting, because there
were so many points of interest attaching to it, and involving
chemical complexities, that it was almost impossible to discuss
such a paper without having previously perused it. He would,
therefore, confine himself to one or two matters only, in which
M. Pourcel made reference to processes and experiments he
(Dr. Siemens) had carried out. He was disposed to agree with
Mr. Snelus in his criticisms of the three principal factors upon
which M. Pourcel said the elimination of phosphorus depended,
these factors being liquid basic cinder, high temperature, and, if
he understood right, the absence of oxygen. He was disposed to
think, however, that the presence of oxygen was an advantage in
the elimination of phosphorus, for it stood to reason that as the
phosphorus was removed by oxidation, oxygen could not interfere
with the process. He further could not agree with the author
when he said, that by the ore-reducing process only rail metal was
producible, and that in order to obtain very mild metal the scrap
process had to be resorted to. It was, indeed, quite the contrary ;
for in order to make a very mild metal, such as was now generally
used in naval construction, the ore process was almost exclusively
used, and scrap was only added for two objects, — the one being to
get rid of the cuttings and other scrap produced in the mills, and
the other to quicken the output. In working with ore and pig
only, the time taken up was naturally greater than in adding at
once a proportion of perhaps 20 or 30 per cent, of scrap metal, which
would bring down the carbon of the bath by one-third to begin
with ; and there was this further advantage in. it, that in dealing
with a silicious pig a very considerable amount of slag was pro-
duced in using ore and pig only, which was diminished relatively

* Excerpt Journal of the Iron and Steel Institute, 1879, pp. 367-370.

417

in adding ivlinrd iiiaal. M. IVmrcel wont on to say that an
aihantauv might l»e obtained by tapping off the slag at a certain
. and, after dephosphorization had taken place, adding the
frn-o-manganese ; and then he said, "All this is very complicated.
Dr. Siemens himself will acknowledge that this is no longer his
simple method. It is a mixture of the Krupp and the forno-
convcrtisseur processes" (p. 22). Now he (Dr. Siemens) should
be sorry to interfere with the carrying out of Hen* Krupp's pro-
cess, but if it consisted in tapping off the slag, the misfortune was
that he had done that already ten years ago, and he must, there-
f«.iv. pardon him for interfering with that part of the process. A
very great advantage could be gained through tapping off the
silicious slag at a certain stage, and then adding the final quanti-
ties of ferro-manganese and silicon iron, if the latter was used for
bringing the bath into its final and proper condition. But it was
difficult in actual practice to insist upon such nice operations.
The tapping off of the slag took time, and his experience was that,
after having used it anywhere, and finding that it worked very
nicely, the next time he came to the place he found it had been
discontinued. Except for the additional trouble it involved, the
tapping off of the silicious slag, in order to produce towards the
end of the operation a basic slag by adding lime, was a very
advantageous mode of working, and one that he had for many
years insisted on.

Mr. Richards had made some interesting remarks, giving the
negative results of experiments he had made at Middlesborough,
with a view of saving time in deoxidising the metal, by means of
silicon and manganese combined, after the blow in the Bessemer
converter. He (Dr. Siemens) could confirm Mr. Richards from
his own experience, it being the established practice in some of
the works where the open-hearth process was carried on, to add
silicon iron and spiegeleisen about twenty minutes before the final
addition of ferro-manganese, this interval of time being necessary
in order to effect the formation and separation of the particles of
slag, formed when silicon, manganese, and the oxygen absorbed in
the metal combined. This modus operandi, he believed, was first
suggested by M. Gautier of the Terrenoire Works.

M. Pourcel alluded to the rotary process as carried on at
Towcester, and said that there could be little advantage in work-

VOL. I. E E

41 8 THE SCIENTIFIC PAPERS OF

ing that process, because the removal of phosphorus must depend
on finding oxide of iron in the slag. It was quite true that it did
do so, but at the same time Dr. Siemens thought that the rotary
process was a most effectual and thorough one for the removal of
phosphorus, or rather for the production of a metal free from
phosphorus. He was much surprised at that portion of M. Pour-
eel's remarks, because on the preceding page he said, "In this
operation only an imperfect reduction of the ore is obtained, and
if the phosphate of iron can be partially reduced at a low tempera-
ture, nearly all the phosphoric acid remains in the slag, and as
phosphorus was generally present in ores only in the state of
phosphate of lime, it was not reduced by the ordinary reducing
action of the furnace" (p. 15). He (Dr. Siemens) found, in fact,
that in operating upon ore containing phosphorus, the phosphorus
never went to the metal ; it remained in the oxidised condition,
and he had been enabled to produce metal almost entirely free
from phosphorus from an ore containing a very considerable per-
centage of phosphoric acid in combination. It was necessary, in
order to cany out the process properly, to tap off the first slag
before the metal was brought into the form of balls. This first
slag contained certainly not an excessive quantity of oxide of iron,
and as much as 3 or 4 per cent, of phosphorus. If done
systematically, there was no great difficulty in obtaining metal
free from phosphorus without incurring great loss. The cinder
must be a rich cinder in order to produce the welding quality of
the ball. He could bear his testimony to the great value of
M. Pourcel's paper, which opened out a very important subject for
discussion.

The, President asked Dr. Siemens whether he had anything to
say about the last paper — that of Mr. Bull.

Dr. Siemens said, that with regard to the last paper read, they
had in it the revival of a very old friend. He had heard of the
blowing in of steam into molten metal even before the days of
the Bessemer process, when their friend Mr. James Nasmyth pro-
posed and patented such a plan. Sir Henry Bessemer himself had
also proposed to blow steam through the metal. There was,
therefore, nothing new in the conception. At the same time, he
did not doubt the chemical reaction, but, as had already been
pointed out, he doubted very much its practicability, because when

S/A WILLIAM SIEMENS, F.R.S. 419

he had tried experiments of that sort he had always found a very
rapid ivduction of temperature. Steam acted like a wet blanket
on tho metal, making it pasty almost immediately, and he expected
that this would be the difficulty which the author of the process
would experience.

In tlw discussion of t/te Paper

"OX THE PROGRESS OF IROX AND STEEL AS
CONSTRUCTIVE MATERIALS,"

By J. A. PICTON, F.S.A., Liverpool,

DR. SIEMENS* said he rose, as one representing the steel-
makers more than the steel-users, to make a few remarks, and in
answer to Mr. West he should say, " Do not diminish your tests,
but rather increase their severity." The President, in his remarks
on the previous day, had said that iron, which was not perhaps
first-rate, had answered well for ship-building, and why should
they be so very particular with regard to steel ? Now, iron and
steel differed in more than one important particular. Iron always
held together. It was apt to break when ifc got a blow, but it
held to a considerable extent although it might be of poor quality ;
whereas steel of high quality would bend like leather, would
extend, and was almost unbreakable. But the moment they left
that high pedestal of perfection, they got a very treacherous
material — a material of which they had heard it stated that it was
riveted in the evening, and on the following morning it was found
rent from end to end. They must not relax the tests with regard
to steel. Then, with regard to the degree of carburisation — the
mildness or the hardness of steel — Mr. Adamson had no doubt
had great experience of steel, and everything coming from him
should be listened to with great attention. But in science, pure
and applied, there was no infallibility, and he must differ from

* Excerpt Proceedings of the Iron and Steel Institute, 1879, pp. 408-411.

E E 2

420 THE SCIENTIFIC PAPERS OF

that gentleman when he said they should have a harder steel for
such purposes as ship-building, and that they should do away with
the bending test.

Mr. Adamson : No, no.

Dr. Siemens, continuing, said he should strongly advocate the
maintenance of that test. Mr. Adamson had alluded to the
material for the Frith of Forth Bridge, and he (Dr. Siemens)
agreed with him, that where the bars of such a structure as a
bridge had to be subjected to a permanent strain, a comparatively
hard steel was the most suitable to be employed. The question of
the links for that bridge had indeed occupied his attention for
some time, and when Sir Thomas Bouch, in the summer of 1878,
came to consult him as to their construction, he had an idea of
riveting mild plates one-eighth of an inch thick together, but he
(Dr. Siemens) strongly advised him to abandon that and to use
solid links, making them of a material which would stand a
breaking strain of 50 tons, and which they could load with perfect
safety to the working strain of 10 tons. While it was perfectly
safe to make such a structure as a chain bridge of this material, it
would not be safe, and there would be no advantage, he main-
tained, in making a ship of it. It might be supposed that the
harder material was stifler, and that they might therefore make
the plates lighter ; but if they tested a bar of mild steel of a
square inch sectional area, and by the side of it a bar of hard steel
which would break with a strain of 50 tons to the square inch,
they would find that both elongated exactly to the same extent
per ton of weight put upon them, up to, say 15 tons ; therefore.
it would be wrong to say that the hard steel was, within that
limit, more rigid than the mild steel, and this was a fact which he
thought was not sufficiently known or borne in mind by engineers.
If they continued their tests, increasing the weight beyond 15 tons
per square inch, then only a difference became perceptible. The
mild steel, when it was strained beyond that point, began to take
a permanent set, whereas the hard steel, having a much greater
elastic range, would return to its original length until the elastic
limit of, say, 25 tons per square inch was reached.

In selecting a material for ship-building, they should put the
problem to themselves in this way : Do we intend to strain this
material in any direction beyond 15 tons per square inch ? He

WILLIAM SIEMENS, F.R.S.

421

" No," as the limit for steel according to the Board of Trade
Jea was at present (>£ tons compared with ."> tons for iron ; and

w nut likely that this limit would be materially extended in
favour of steel for riveted work, because a certain thickness of
plate was necessary to prevent buckling and rapid deterioration
through oxidation. There was, then, at 15 tons with mild steel
an enormous margin of safety within the elastic limit, and a
vessel constructed of mild steel would be just as rigid as one
constructed of hard steel, using the same scantling in both cases.
But occasions would arise upon which a ship was bumped, and
then it would be strained far beyond 15 tons per square inch ;
and what would be the result ? If they had a ship of mild steel,
it would simply yield to that extreme strain ; if they had a hard
steel, it might break. It was, therefore, a question for the ship-
builder to consider. " Is my ship as a whole structure, taken as a
girder, to be strained beyond 15 tons per square inch, and must I
therefore make the plates strong enough to resist that strain, or
shall I limit the extreme strain of the structure to within 15 tons
per square inch, and have a material which, if subjected to an
excessive strain or blow, will yield and not break ? " He should
say that it was far safer to have a material that would, under
such circumstances, yield. He would, therefore, maintain com-
paratively low tests for absolute strength, which he would limit to
perhaps 30 tons per square inch breaking strain, whilst insisting
upon an elongation of at least 20 per cent, of an 8-inch bar.
There were other advantages connected with mild steel which
would be lost in using a hard material. A mild steel could not be
injured by sudden cooling, and it need not be annealed, but a
hard steel had to be annealed. If a comparatively hard steel
plate was punched, its strength would be diminished to a consider-
able extent, whereas it was fully proved that a very mild steel
might be punched without injuring its strength any more than by
drilling ; and at some of the Government yards, after careful
consideration of the whole subject, they had given up the practice
of drilling plates whenever it was found more convenient to resort
to punching. These arguments were all in favour of a veiy mild
steel. Then with regard to what fell from his friend Mr. Adamson
as to the bending test not being a test of elongation but a kind of
mixed test of elongation and solid flow, he thoroughly agreed

with this observation, but he would still maintain the bending
tests, after sudden cooling, as most important. If they had a
material that would elongate to a certain extent, and that did not
possess at the same time the quality of flowing under great
compression, that material would break in the case of a sudden
shock such as a ship might receive in knocking against a rock or
against another ship : if it had the power of flowing, it was a
much safer material. That quality of flowing came chiefly to the
fore in making material for the tin-plate trade. Instead of using
charcoal bar, tin-plate makers now used to a great extent a very
mild steel for making tin plates, and their test was entirely a
bending test. They took the tin plate, doubled it, doubled it
again, and doubled it again ; put it on the anvil and hammered
it quite flat with an iron hammer ; they then bent it back, and if
it stood that three times, it was considered a pretty good material.
Now that test would not be stood by metal which would bear
almost any amount of elongation ; the quality required of such a
metal was a quality which he would call the power of solid flow,
that was possessed only by very pure iron indeed. He would,
therefore, as one representing more the manufacturer than the
user of steel, say to them, do not drop your tests, but rather
increase their severity.

In the discussion of the Paper

"ON IRON AND STEEL AT LOW TEMPERATURES,"
By JOHN JAMES WEBSTEE, Assoc. M. Inst. C.E.,

DE. C. "W. SIEMENS * observed that the impression produced on
his mind by the paper was that it dealt with a limited question
purely for a practical purpose, and that within those limits the

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
LX. Session 1879-1880, pp. 204-207.

WILLIAM SIEAfENS, P.R.S.

423

results \\vre much && might have been expected. The author had
dealt with various metals at the limits of 5° and AO" Fahr. The
limit was a narrow one, and it could not be expected that the re-
sults of the experiments could be such as would lead to a general
ri inclusion regarding the effect of temperature within wider limits
(in materials of that description. There could be no doubt that
toU-rably pun; material must be stronger at the lower temperature
than at the higher, because the last particles of the material were
closer in contact at the lower than at the higher temperature. For
a similar reason the material would resist a blow rather less at the
lower temperature than at the higher. But he agreed with the
observations of previous speakers that no influence was produced
by those degrees of temperature on the absolute structure of the
metal. It was quite out of the question to expect to change the
metal from a crystalline to a fibrous condition, and back again
from a fibrous to a crystalline condition, by lowering or raising
the temperature a few degrees. The paper, though valuable of its
kind, was suggestive chiefly of what had been omitted to be taken
into consideration ; for instance, the author had not in the first
instance given a chemical analysis of any of the materials he had
employed, though that omission had, he believed, since been reme-
died ; but at any rate he had drawn no conclusion from such
analysis. Nor had he considered the elastic limit of the materials,
although to an engineer the elastic limit and the condition of the
material up to the elastic limit were more important than the
breaking strain. Engineers did not want to break down with
their materials ; but they wanted to see to what extent it was safe
to use them. It would have been most interesting if the author
had given the elongation due to the rate of elastic extension at
different temperatures. It was Mr. Barlow, the President, who
first established by experiments that steel of various tempers was
equally strong up to a certain limit. In other words, that if a bar
of very mild steel having a sectional area of 1 square inch pure
iron, as Mr. Adamson had correctly described it, was weighted
with a ton weight, the same absolute extension would ensue as in
weighting a bar of very hard steel of 1-inch sectional area and of
the same length, and so on, each ton additional weight producing
the same amount of extension in both bars until the elastic limit
of the mild metal was reached. These results, which he must

424 THE SCIENTIFIC PAPERS OF

confess at first appeared hardly credible to him, had since been
fully confirmed by experiments of his own. It was a remarkable
fact for engineers to consider that in constructing a bridge it came
to the same thing whether they took mild steel that would break
with 28 or 30 tons strain to the square inch and would come to its
elastic limit at about 1 5 tons, or whether they took a material that
would find its elastic limit only at 25 or 30 tons, and would break
at 50 or GO tons ; therefore the advantage that would be derived
from the hard and strong material would only be met with after
exceeding the elastic limit of the milder steel. He was not aware
of any experiments giving the influence of heat upon the coeffi-
cient of elastic extension of those materials. If it should so happen
that a bar of iron or steel at the temperature of 5° would follow a
less rate of elastic extension than a warmer bar, a steel bridge that
was ordinarily loaded to a strain of 8 tons per square inch, when
it cooled down, would contract rather more than in the ratio of
simple thermal contraction ; and afterwards, if the bridge changed
from the one temperature to the other, it would alter its length in
all its members due to the two ratios of absolute contraction and
a contraction due to a different rate of extensibility, It would be
interesting to know how the modulus of elasticity was influenced
by temperature. It was often thought that metals were fatigued
by strain ; and in the case of inferior classes of iron, it might be
true that, if a bar were strained a little beyond its elastic limit, its
fibres would be ruptured to some extent, and the material would
receive an injury ; but in the case of high-class material, such as
mild steel, there was an absolute advantage in straining it beyond
the limit of elasticity. It was possible, by careful manipulation,
to raise the breaking strain of a bar of a given sectional area to a
remarkable extent by gradually accustoming it to the strain. By
taking a bar of mild steel of 1 square inch sectional area and load-
ing with a weight of, say, 15 tons, and leaving the weight on
twenty-four hours, it would be found that both the elastic limit
and the breaking strength of the bar were materially increased.
That was a quality of high-class material which was, perhaps, not
yet thoroughly understood, and which certainly gave a great factor
of safety in favour of its employment. The chemical analysis,
which had not been dealt with in the paper, was of great import-
ance in reference to the object the author had in view. The

WILLIAM SIEMENS, F.R.S.

425

mithor had spoken of certain brands of material— of certain
s, for instance by a firm like Krupp's. Now, he happened to
know that Kriipji made steel by at least three different processes,
and it was difficult to say by what process the particular bar was
produced that had been referred to in the paper, nor, if the pro-
cess were known, could it be affirmed what were the materials used
in its manufacture. If it was a cheap material, probably Mr. Krupp
would have put in some cheap iron, and the result might be a bar
that would give a very fair absolute strength, but would give way
to a bending and breaking test before its time ; whereas if he
knew that the metal would be tested very severely, he would pro-
duce something that would be capable of resisting very high im-
pacts. Some years ago Dr. Siemens made a series of experiments
on the power of materials of particular composition to resist blows
at different temperatures. His object was to see what effect phos-
phorus had on mild steel with regard to that quality. From those
experiments he had arrived at the conclusion that the power to
resist blows in mild steel, at any rate, was dependent almost
entirely on the presence or absence of phosphorus. It was well
known that phosphorus might be counteracted in the steel, to a
great extent, by the addition of manganese ; but let the steel be
cooled down to freezing point, or below, and the metal would
break short : whereas true mild steel, containing only traces of
phosphorus and of manganese, would be affected to a very slight
ex cent by any depression of temperature to which it might be
subjected. He believed that pure metal was influenced very little
by changes of temperature, whereas metal containing foreign
matters, and particularly phosphorus, would be influenced to a
great extent. It would, therefore, be erroneous to draw general
conclusions from such experiments as were given in the paper.
The author seemed to have attempted to find out results without
the chemist; but the chemist now-a-days would put his finger
into everything. If they called a thing iron, he would call it a
mixture of iron, phosphorus, sulphur, silicon, carbon, and other
earthy materials, and it was useless to try to do without him.
They must work hand in hand with him in order to arrive at such
results as would guide them safely in their practice.

426

THE SCIENTIFIC PAPERS OF

In the discussion of the Papers

"ON STEEL IN THE SHIPBUILDING YARD," by
Mr. TV. DENNY, and

« ON STEEL FOR SHIPBUILDING," by Mr. WEST,

DB. C. TV. SIEMENS * said : My Lord and Gentlemen, we have
listened to two, in my opinion, very valuable papers, one written
by an eminently practical shipbuilder, and the other by a gentle-
man who has had great experience in the testing of a material now
largely used in construction. I agree with many of the opinions
advanced in both these papers. Both seem to advocate advance
in the same direction ; but there are certain points on which I
must, with all respect to the authors, differ from them in opinion.
The all-important question which has been put forward in these
papers, is the question of the absolute maximum strength to be
required or accepted in mild steel. If you take a bar of compara-
tively hard steel — steel that will stand 50 tons to the inch — it will
at the time of its breakage have elongated perhaps only 6 or 7 per
cent. But take a material of 30 tons to the inch and it will have
elongated 20 per cent. Now, to begin with, it is not strictly
correct to say that this latter material broke under a tensile strain
of 30 tons to the inch, because whatever might have been the
sectional area of the original bar at the time it broke, it had
bodily elongated 20 per cent. ; therefore, the strain was no longer
30 tons to the inch, but 36 tons to the inch, and I claim for the
mild material at any rate this advance of strength. What, how-
ever, is the condition of things before we have reached this
ultimate limit ? Take one bar with a breaking strain according
to the usual test of 30 tons to the inch, and another of 50 ; weight
both these bars with 5 tons to the inch and you will find that these
bars have elongated to exactly the same extent. Weight both
these bars to the limit of 10 tons to the inch, and again it will be
found that the elongation is the same. Therefore, up to that

* Excerpt Transactions of the Institute of Naval Architects, Vol. XXI. 1880,
pp. 216, 217.

SfK WILLIAM SIEMENS, F.K.S.

427

point at any rate, I claim as great a strength for the mild steel as
the hard. Increase that strain even to 15 tons to the inch, and
< ioiiLMtion will still be the same. Take the load of both bars and
they will both come back to the original lengths. Go to 20 tons
and you will find a difference. The hard steel — the steel that will
bn-ak at 50 tons to the inch — will again return to its original
length when the load is taken off, whereas the material breaking
with 80 tons to the inch of the original section will show a
permanent elongation. This, you will say, is a sign of weakness
and condemns that bar ; but I would venture to maintain that
t li is is a sign of the strength of the latter bar. Up to the point of
15 tons to the inch the two materials are precisely alike, and if
the naval constructor takes care that no portion of his ship,
viewed as a girder, should receive a greater strain than 15 tons,
both materials are able to bear the strain, and they will both
deflect to precisely the same extent. But it may be said, materials
may be accidentally subjected to a higher strain, and in that case
greater strength will be needed. Taking the strain of bumping a
ship against a rock, or the ground, or against another ship, the
hard steel will no doubt stand a heavier blow, but there is just the
possibility that when the blow comes it will fracture, whereas the
soft steel we know will lead almost to any extent to sudden impact.
Therefore, as far as the safety of the structure against an action of
that description is concerned, the soft metal is decidedly the
stronger. Then again it has been put very forcibly by Mr.
Denny, that very mild steel requires no annealing ; that it is very
nearly as strong and as ductile annealed as unannealed ; whereas
if you have a harder material, a material containing more carbon,
annealing becomes necessary after bending ; after punching
certainly, and even after drilling. It becomes a necessity. But
by annealing, you take away some 20 per cent, of the strength, and
therefore I say it is much better to make at once a material that
requires no annealing, that does not assume a strained condition
when put into a definite shape, and which maintains its strength.
The mild material without annealing, is as strong, or at any rate
is as ductile practically, as after annealing ; and, I believe, at some
yards punching without annealing has been resorted to without
any practical drawback. These are powerful arguments, I think,
in favour of the very soft material. There is one more argument

428

THE SCIENTIFIC PAPERS OF

which I wish to adduce, that is, if the material of the low elastic
limit happens to be chilled accidentally after manufacture, this
chilling does not interfere with its strength to resist blows or
sudden strains ; whereas with a hard material you will even come
to the point of contemplating with great concern the safety of the
structure when made, I believe it is much safer to adhere to a
material which stands an extreme bending test, and to assign to it
the same practical limit of load which you would to a harder
material. In dealing with wrought iron the practical limit
generally allowed is one quarter breaking strain, whereas in dealing
with cast iron we allow one-sixth breaking strain as the safe
working limit. What I maintain is, that in dealing with mild steel
the safe working limit may safely, for similar reasons, be extended
to one-third the breaking strain ; and if this were agreed to, a
ship constructed of mild steel need not be made any thicker in its
scantlings than one having a breaking strength of 35 tons, such as
I am sorry to find is now being advocated by ship-builders. Re-
garding the remark made by my friend Mr. Bramwell, I am ready
to agree with him in principle to this extent : that if a steel could
be produced of increased ultimate strength without in any way
trespassing upon the quality of extra mild steel, of maintaining
its ductility after tempering, and after punching or bending, that
material would deserve the preference ; but so long as the strength
has to be preserved at the expense of ductility, I must maintain
my objection to an advance, or rather a retrogression, in that
direction in ship-building. I may add one word of entire approval
of the opinion put forward by Mr. Denny, that all plates ought to
be tested at the works of the maker, in the most careful and
rigorous manner possible, and not at their destination, and also
that for riveting steel plates, steel rivets should be used.

S/A' WILLIAM S/K. \fENS,

429

/// the discussion of Hie Paper

"ON THE STEEL-COMPRESSING ARRANGEMENTS
AT THE BARROW WORKS," by MB. ALFRED DAVIS,

I>it. 0. W. SIEMENS* said the advantages to be derived by com-
ing steel in a fluid condition had been proved by Sir Joseph
Whit worth, who for a number of years had produced steel of a
very high quality, made in the open-hearth furnace, and subjected,
while in a state of fluidity, to very high pressure by hydraulic
pumps. Other steel makers had tried to arrive at the same
result by similar means. This was perhaps the first paper that
had been brought before them giving a distinct account of an
attempt to obtain the same advantage by pressure exerted upon
fluid steel, without resorting to the very thorough but expensive
plan adopted by Sir Joseph Whitworth. At various times he had
himself tried to bring pressure, resulting from the spontaneous
generation of gases within the closed ingot mould, to bear upon
steel ; but he had not obtained altogether satisfactory results.
Accordingly, when last year the subject of steam compression was
brought before the Iron and Steel Institute, he immediately
offered to make an experiment ; but he had found that for mild
steel, such as he operated upon, a very high pressure was certainly
necessary. Sir Joseph Whitworth had found that 2 tews per
square inch was the pressure necessary to produce solid metal.
They now heard of 100 Ibs. or 150 Ibs. per square inch being
sufficient to produce success ; and from the photographs exhibited
by Mr. Richards it was evident that a certain degree of success
was obtained, although the holes in the metal were not entirely
got rid of. In the case of very mild steel they were also troubled
with holes near the surface, which they called honey-combs ; and
it was to get rid of these that the heavy pressure seemed to be
necessary. The large cavities formed in the centre of the ingot
would no doubt be, if not removed, very much reduced in size by
such moderate pressure as had been mentioned ; but he was quite

* Excerpt Minutes of Proceedings of the Institution of Mechanical Engineers,
1880, pp. 403-407.

430 THE SCIENTIFIC PAPERS OF

certain that, in the case of mild steel, to get rid of the honey-
combs they would have to resort to a pressure of 1 ton, if not of
2 tons, per square inch.

With regard to the interesting question raised by Mr. Richards,
as to how the pressure acted upon the gases which had been
occluded, he confessed that he was in the same difficulty as the
President. He could not conceive how, by applying pressure to a
fluid mass, they could induce one ingredient out of several to go
from the centre to the outside, or from the outside to the centre.
The pressure was the same throughout over the whole surface ;
and all he could conceive was that by that pressure the volume of
say one cubic inch of gas would be reduced, if the pressure were
high enough, to say one-tenth of a cubic inch ; or it might be
that the gases would be reabsorbed in consequence of the pressure,
and thus return to their former combination with the steel. Mr.
Richards had drawn a comparison between steel and soda-water.
Now when the cork of the soda-water bottle was lifted, the whole
contents became a froth, and might be called spongy soda-water.
If the cork were pressed down again, the frothing would immedi-
ately cease, the gases being again absorbed in the liquid. There-
fore it Avas quite conceivable that by the application of a sufficient
pressure to fluid steel, although the gases were not expelled, they
might remain occluded in the steel. He imagined that this would
be the real solution of the problem. But there could be no doubt
that, if the cavities could be prevented in the ingots, a great gain
would be secured ; and he should be glad to see that, by the appli-
cation of so moderate a pressure as was mentioned in the paper,
this result could be obtained.

While discussing the question of steel, he should be glad, with
the permission of the President, to make a few observations upon
the more general question of mild steel. Within the last week or
two a great deal had been said about this steel not being reliable,
and they had heard of boilers giving way mysteriously under a
very moderate pressure. It so happened that he had been made
cognisant of some of the circumstances regarding the failures
which had been prominently alluded to ; and he might say that,
although the first boiler had failed under pressure, the second was
found to be rent in precisely the same manner without any
pressure having been applied : clearly showing that it was not a

.s/A- \V1I.UAM SIEMENS, F.R.S. 43 1

case of weakness of the metal, under so very moderate a pressure
0 Ibs. per square inch ; but that from one cause or another
tin- metal had been cracked and broken previous to testing. He
rniild not speak as to the cause of the metal being in that condi-
tion ; and he would only say at present that mild steel, properly
made and properly put together, would not burst under any
ire whatever. If a boiler made of mild steel were subjected
to an increasing pressure, it would be found impossible to burst it.
And it was natural that this should be so. A material that would
stretch 30 per cent, before rupture, would naturally give way first
at the weakest sections, namely those through the rivet-holes ; the
round rivet-holes would become oblong, until sufficient water or
steam leaked out to balance the amount of water pumped in or
steam generated. That was not a mere hypothesis of his. He
had witnessed some experiments made by Mr. Dean, the locomo-
tive superintendent of the Great Western Railway, in the presence
«>f Mr. Parker, the chief surveyor of Lloyd's, and described in a
paper by Mr. Parker before the Institution of Naval Architects *
in 1878. The leakage began at 560 Ibs., and the highest pressure
reached was 800 Ibs. Messrs. Easton and Anderson had also tried
a boiler made of mild steel ; and, in order to increase the severity
of the test, between each trial they put the boiler into a furnace,
made it red hot, and then took it out again, caulked it where it
appeared necessary to close the seams, and then again subjected it
to a pressure of several hundred pounds per square inch. He
believed that Mr. Greig had also made similar experiments. It
might therefore almost be taken as an axiom that a steel boiler
could not be burst ; it might be made leaky, but that was all.

With regard, then, to the great question as to whether mild
steel was a reliable material or not, he would answer most un-
hesitatingly that it was the most reliable material they knew of ;
they might punch, shear, bend it, or do what they liked with it,
but they would not in any way destroy its tenacity. By way of
practical proof he might mention one other fact. The steamers
"Iris" and "Mercury" had been constructed of mild steel about
three years ago for the Admiralty : not only the shell-plates of the
vessels, but the angles and the whole of the boilers were all con-

* See Transactions of the Institution of Naval Architects, 1878, p. 178.

432 THE SCIENTIFIC PAPERS OF

structed of that material ; and amongst all the plates and angles
not one had been returned as defective in quality. Since then the
Admiralty had used steel for the construction of their boilers and
ship plates, almost to the exclusion of iron ; and although pro-
bably more than 10,000 tons had been used, he believed he was
correct in saying that no plate had been returned as being cracked
or unsatisfactory, except a very few from mere mechanical
blemishes. There had been no case of mild steel, properly made
and worked, giving way iu a mysterious and treacherous manner.

The demon of cheapness, however, seemed to be abroad, and it
had settled especially upon steel. The works with which he was
more particularly connected had introduced two qualities of mild
steel : one which was to compete for price in the open market ;
and another, called "special metal," which was vouched for as
being in every way reliable. This metal was not only made of
more expensive material, but it received greater attention, and
was worked to a greater extent : crop ends were cut off more reso-
lutely than one could afford to do under all circumstances ; and
consequently it was rather more expensive to produce than ordi-
nary steel. But he was sorry to say there were many engineers
who, for the sake of £2 or £3 per ton, preferred the unguaranteed
material for the construction of their boilers ; others, however,
took only the special or guaranteed material. It appeared strange
to him that there should be such a tendency to get cheap steel for
the construction of boilers, &c. ; because, when a good iron boiler
was wanted, Yorkshire iron was used, costing from £22 to £40 a
ton ; whereas reliable steel plates could be obtained for £14 to
£16 a ton. The special steel plates thus cost a great deal less
than what engineers were willing to pay for the best Yorkshire
plates ; but engineers would have the cheapest steel, at perhaps
£11 to £13 a ton, and took their chance. He thought it was a
very dangerous policy, and one that naturally led to the dissatis-
faction of which they had heard.

He might mention another form which steel now took, and
which had received his special attention : he referred to the pro-
duction of rivet-bars. For a long time engineers had continued
to use iron rivets for riveting steel plates together. He had
always considered this was like stitching a silk gown with a
cotton thread, making the stitching material a Aveaker thing than

.S7A- \\'1LUAM SIEMENS, l-.R.S.

433

tin- material to be stitched. Latterly a certain confidence had
been created in favour of steel for rivet-making ; and great care
u;i> taken in producing rivet-steel of such a quality as would make
rfectly reliable. But what was the fact? Since rivet-steel
had been brought into the market, it was not sold for so high a
price as was given for the best rivet-iron. The latter cost as
much as £19 a ton, but for rivet-steel engineers went down at
once to the very cheapest quality they could get, which he might
mention was steel rolled from crop-ends, and with all the defects
of crop-ends about it. It was therefore unfair to criticise steel
severely, unless all the circumstances regarding it were known,
and al>ove all things unless it was known what price had been paid
for it.

In the discussion of the Paj>cr

"ON HARDENING IRON AND STEEL; ITS CAUSES
AND EFFECTS," by PROFESSOR AKERMAN,

Du. SIEMENS* congratulated Professor Akerman on having
dealt with a subject which had puzzled all those who were practi-
cally engaged in the treatment of steel, and which had never yet
been sufficiently or scientifically explained. It would be impossible-
at a meeting like that to discuss all the matters brought forward
in the paper. There was, however, one leading idea running
through the whole, which was to the effect that carbon existed in
iron and steel, not in two conditions, as was now generally sup-
posed, but in three conditions, and that it was to this intermediate-
condition that the phenomena of hardening and tempering were
mostly attributable. Professor Akerman held that this inter-
mediate carbon travelled towards the graphitic condition when
steel was gradually cooled, and that this furnished the key to the
somewhat striking changes that arose if they took an ingot of
steel, hardened it, and then subjected it to the processes of

* Excerpt Journal of the Iron and Steel Institute. 1880, p. 437.

VOL. I. F F

434 THE SCIENTIFIC PAPERS OF

hammering and rolling. Dr. Siemens confessed that he had been
sceptical regarding this effect until he had had light thrown upon
it by Professor Akerman. If pressure was the cause why steel
became stronger in manipulation than in chilling the outside of a
heated ingot, greater pressure was thrown upon the interior metal,
and it was easily conceived that the effect produced would be
similar to that brought about by hammering and rolling. The
experiments referred to in the paper seemed to prove this very
clearly.

There was one point to which Professor Akerman also called
attention, and which Dr. Siemens could not allow to pass without
observation. They had heard much of the extension of mild steel,
and they knew that the results were of the most variable character.
Some eminent steel makers and users tested bars two inches long,
and obtained a marvellous elongation before rupture took place.
Others, again, including the British Admiralty, used bars eight
inches long, and the elongation, after fracture, was measured from
the distance of the two points that were originally eight inches
apart, after placing the fractured pieces end to end. Other ex-
periments, especially those of Mr. Barlow and the Committee
appointed by the Institution of Civil Engineers, dealt with very
long bars, and the rates of elongation were consequently much
less. Professor Akerman suggested they should exclude the drawn
portion of the bar from the total elongation, and thus obtain the
real elongation, and his suggestion deserved the greatest possible
attention. But why carry the test to rupture at all ? They knew
that up to a certain point the elongation continued very uniform
throughout the bar, and it would be quite sufficient to carry the
test to the point where partial tearing had set in. They should
thus obtain a measure of the force which a bar was capable of
resisting, because from the moment the slightest irregularity in
the line of contour occurred, the result was no longer a proper
test of the strength of the material, and there was a fault or error
created by the continuance of the general elongation after that
partial action had set in. But in whatever way the experiment
was conducted, it would be far more correct to reckon the elonga-
tion of a bar of steel by excluding that portion of it which had
been drawn, and therefore presented no portion of legitimate elon-
gation of the bar.

S7K WILLIAM SIEMENS, F.R.S. 435

Tlu-iv \vi-re many points referred to by the Professor which he
not observe upon. There was, however, the vexed question
of nomenclature, which they would yet have to deal with. In
(l.TiiKiiiy, Sweden, and Austria, the nomenclature recommended
iiy tin- committee at Philadelphia had been adopted and carried
out. In K upland they had seen considerable difficulties in the
way of adopting that nomenclature, and these had arisen from the
want of indications as to the limit at which a material ceased to
be ingot iron and commenced to be steel. It was clearly shewn in
the paper that all material containing only traces of carbon was
subject to hardening ; and Professor Akerman now suggested that
they should select an arbitrary limit for steel, namely, the limit at
which feldspar would no longer touch the hardened part of the
iron. This might be a limit which would keep on one side of the
boundary line the material known as mild steed, and on the other
what was now generally called rail or engineering steel. Still he
mst say he saw difficulties. In England some engineers used
rails of a harder character, and others insisted upon metal which
was softer. Then, again, there was a process by which phosphorus
took the place of carbon in steel of that description, and they
would, by employing phosphorus steel, obtain hardness without
temper. These limits would be very difficult to draw, but they
should very carefully reconsider the question, because it appeared
to him that as some countries had adopted the new nomenclature,
and England was adhering to the old one of calling by the name
of steel all malleable metal that had passed through the fused
condition, they would be in an isolated position. It would be better
certainly if a uniform system could be adopted, and a meeting
like the present would be a suitable place and occasion for doing
so. Before sitting down he wished to give expression to the
pleasure that all probably felt in Professor Akerman's most valu-
able paper, which, in his opinion, would be hereafter cited as a
standard exposition of the subject.

p p 2

436 THE SCIENTIFIC PAPERS OP

In the discussion of the Paper

ON "DEPHOSPHORISING IN THE CONVERTER,"

BY J. MASSANEZ (Hoerde), Westphalia,

DR. SIEMENS * said, that, after the speeches they had heard
regarding the economical conditions necessary for the basic
process, little remained to be said. Dr. "Wedding had pointed out
the importance of removing the slag before the final operation
took place, and he thoroughly agreed with him that that would
be a most important step to take. But when experimenting in
the open-hearth furnace with similar objects in view, he had
found it very difficult in practice to remove this slag thoroughly
from a large surface of fluid steel. The slag came away only
partially ; it hung about the furnace ; and although in the
Bessemer converter they had to deal with a smaller area than in
the open-hearth furnace, yet there was delay and practical in-
convenience in removing the slag, which was not of a nature,
however, to stand in the way of accomplishing the end. The two
causes of difficulty in the way of carrying out the process appeared
to be, first, the sulphur, and secondly, the silicon. Now as
regarded the elimination of the sulphur, as Professor Tiinner had
remarked they must depend chiefly upon the blast furnace, and in
order to accomplish such a result, in dealing with poor ironstone,
a difficulty arose on account of the great amount of basic material
that had to be added to the ore. If the ores employed contained
a considerable percentage of manganese, the difficulty regarding
the sulphur would be very much diminished ; but there remained
that of reducing the silicon to a very low percentage, which he
thought could hardly be effected with ores such as those of
Cleveland, which contained a large amount of silicon in their
composition. The use of hot blast would probably also have to
be very much restricted, involving an increase of fuel consumption
and a decrease in the output of the furnace. But he should like
to ask the gentlemen connected with the Rhenish Steel Works

* Excerpt Journal of the Iron and Steel Institute, 1880, pp. 556, 557.

WILLIAM SIEMENS, F.R.S.

437

\\h« ther he had been rightly informed the previous day that the
iron treated by the basic process contained no more than one-tenth
per cent, of silicon ? If this were so, it showed that they had
managed to obtain a brand of iron that did not exist on the other
side of the Channel, and it was very interesting that they should
know exactly the cause. Could iron be produced there containing
—phosphorus, '2% per cent. ; silicon, -^ per cent. ; and man-
ganese, '2"1 per cent. ? Such iron appeared to him to be most
exquisitely suited for carrying out this basic process, and it was of
the greatest interest to know whether they found it necessary at
the Rhenish Steel Works to obtain such a chemical composition
in order to be successful in the process.

HERR MASSENEZ, in replying to Dr. Siemens, stated that it was
not at all difficult to make the iron described, in Germany. They
had at Ilsede very large quantities containing no more silicon
than one-tenth per cent., and with white iron of that kind they
might well work the basic process They made iron con-
taining just the quantity of silicon, manganese, and sulphur
mentioned by Dr. Siemens ; all that was necessary was to add
a sufficient quantity of limestone to the ore

In the discussion of the Paper

"ON MACHINERY FOR STEEL-MAKING BY THE
BESSEMER AND THE SIEMENS PROCESSES,"

By BENJAMIN WALKER, M. Inst. C.E.,

DR. C. WILLIAM SIEMENS * said the author had wisely confined
himself to the manipulation of steel after it had been produced.
If he had attempted to deal with processes, his paper would have
assumed dimensions far beyond the scope of what could be dealt
with at one time, and there would also have been chemical con-

* Excerpt Minutes of Proceedings of the Institution of Civil Eng;neers, Vol.
LXIII. Session 1880-1881, pp. 22-24.

438 THE SCIENTIFIC PAPERS OF

sidcrations involved which would have led beyond the usual range
of the discussions. He had, however, mentioned two processes
now largely used : the Bessemer process, which, by the great
revolutions it had effected not only in the manufacture of iron
and steel, but in the engineering of the world, was too well known
to require any further comment ; and the process with which
Dr. Siemens's name was connected, which, although much less
used, had been in existence nearly as long as the Bessemer process ;
and which, although it had been for many years confined to com-
paratively few works, had lately, owing to some particular advan-
tages it possessed under certain circumstances, attained a con-
siderable extension. Setting aside the question of process, the
paper dealt very ably with the modes by which the ingot metal
should be treated in order to convert it into such forms as were
required. The author advocated economy in the construction of
the steam-engine, a subject which Dr. Siemens was glad to see
had been taken up at last seriously by the makers of that class of
machinery. At the time when the puddling furnace and the re-
heating furnace sent up flames of living fire into the air, the
question of economy in the engine was really a superfluous one ;
there was so much steam produced in using only a portion of
the heat that would otherwise have been entirely thrown away,
that the difficulty rather was how to keep it under, than to
economise it. The converter, however, and the regenerative gas
furnace had materially changed that condition of things. The
heat was now in a great measure utilized to do its work ; it was
employed for producing the effect required, and there was little
heat to be thrown away. If, therefore, steam was required to
work the metal after it was produced, that steam had to be
raised by means of special fires, and it became as important to
work mill-machinery upon economical principles as it was a
steam-engine on board ship. No wonder, then, that economical
engines had attracted the attention of those practically engaged
in working steel ; and he quite agreed with the author in
selecting a compound engine as the type which was perhaps
the most suitable for the purpose. It distributed the load very
uniformly upon the crank centre, and it carried out the system
of expansion without involving a maximum strain too much above
the mean strain on the working piston. The author proceeded to

A/A' WILLIAM SIEMENS, F.R.S.

439

recommend a form of engine working directly on the roll, and in
that respect he was obliged to differ from him. It was stated
that in working the steam cylinder with the connecting-rod
directly upon the axis of the roll, gearing was saved. That was
t nil-, but the gearing was saved at the expense of an enormous
strain on the craiik pin, an elaborate foundation, and a relatively
slow working of the piston. Admitting the advantage of sim-
plicity in connecting an engine directly to the working crank,
there were the drawbacks to which he had referred, which were in
fact drawbacks in regard to economical results, and also in regard
to the first cost of the engine. If the gearing was properly
constructed, and if the wheels and pinion were made of steel,
there was, in his opinion, nothing very objectionable in it. A
quick-working engine was more economical and more easily
repaired than a slow-working, long-stroke engine, and the pinion
and wheel were convenient pieces of mechanism to convert the
speed into the degree of reduction necessary to work a pair of
rolls : he therefore thought it would be better, on the whole, to
continue the system of having a quick-working reversing engine,
and to reduce the circular velocity by gearing. The author had
also dealt with the question of hydraulic arrangements for working
the pit both with the Bessemer converter and with the open-
hearth furnace, and in other parts of the machinery ; and the
question arose, what pressure would be most suitable for such a
purpose ? Sir William Armstrong was quoted as an authority for
fixing 700 Ibs. to the square inch as the proper pressure to be
used ; and for such purposes as had been named he believed it
was a very convenient working pressure. It would, however, be
wrong to assume that Sir William Armstrong advocated that
limited pressure under all circumstances. Dr. Siemens went to
him last year and asked whether he could make a cylinder
hydraulic ram of 3 feet diameter to work to 8,000 Ibs. pressure
per square inch. He was rather struck by the proposition, but he
did not say "No." Although his first idea was to use a coil for
the cylinder to give strength to it, a forged steel cylinder of 7-inch
wall thickness was eventually decided on. The accumulator and
other hydraulic arrangements were carried out by Sir William
Armstrong, but the hydraulic cylinder was furnished by Sir Joseph
Whitworth of compressed steel. He mentioned this case to show

440

THE SCIENTIFIC PAPERS OF

that there was almost no limit to the hydraulic pressure that could
be called into requisition. The special object he had in view was
to place telegraph cables under the equivalent of Atlantic pressure.
Into the plungers cores of insulated wire were placed, and a
pressure equal to the deep-sea Atlantic pressure was applied. It
was of great value for such a purpose, inasmuch as it ensured the
soundness of the insulating covering when it came to be subjected
to the enormous pressure which it had afterwards to bear. But
he thought that hydraulic arrangements of equal force would be
applicable with great advantage for metallurgical purposes in
pressing rather than beating steel into its definite form.

In the discussion of the Paper

" ON THE COMPAEATIVE ENDURANCE OF IRON AND
MILD STEEL WHEN EXPOSED TO CORROSIVE
INFLUENCES," by DAVID PHILLIPS, M. Inst. C.E.,

DR. C. W. SIEMENS * said, it perhaps would have been better
if the discussion had been commenced by persons more interested
in the use of iron and steel, than by those who, like himself, were
intimately connected with their production ; but in another respect
it might possibly save the time of the Institution, if he took that
early opportunity of referring to some conclusions in the paper
with which he could not agree. The author had given the results
of elaborate experiments on a subject which was of the utmost
importance to engineers ; and if his conclusions were to be relied
upon, engineers were daily committing a grave error in using a
material which gave so slight a guarantee of endurance. But
while he accepted every one of the experimental facts adduced
by the author, he thought he was in a position to prove, from
the author's figures alone, that his conclusions were entirely
erroneous. He had referred, in the first place, to a long series

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol
LXV. Session 1880-1881, pp. 98-101.

WILLIAM .sy /•;.)//;. v\, F.R.S.

441

of experiments made by the Admiralty, under the direction of
Admiral Aynsley, and as far as the collection of facts was
concerned, nothing could be more conscientious or thorough than
that series of experiments ; but as regards proof, they went no
further than to show what every experimenter ought to avoid, and
how he ought not to conduct his experiments in the future. The
author hud placed before the Institution the apparatus then used.
It consisted of thirty-eight tubes of iron and steel riveted in
metallic contact with the shell of a boiler, and exposed partly to
air and partly to hot water and solutions. Although the author
had a very poor opinion of electricity and its effects, Dr. Siemens
had a strong belief in electricity wherever it had a chance of
art ing for good or for evil, and he was convinced that the results
obtained in those experiments were rendered entirely unreliable
through galvanic agency. The results were most variable.
Whereas one iron (common iron seemed to be the best) gave
a corrosion of only 7 grains, on an average, per square foot,
Bessemer steel gave 21 grains, or three times the amount during
the time of exposure. The author was really most merciful when
he stated that the result was only 69'3 per cent, in favour of iron,
because it was really 300 per cent, in that instance. Notwith-
standing these experiments, the Admiralty had adopted a mode of
action which seemed strangely at variance with the conclusions to
which the experiments would point. They now used steel almost
to the exclusion of iron, and he hoped that some one connected
with the Admiralty would state the result of more recent experi-
ments undertaken with a better knowledge of the conditions under
which they should be made. He believed the conclusions since
arrived at were very different from those deduced by Admiral
Aynsley some years ago. At Table VII. the author had com-
pared Landore metal and iron, the one giving an average loss
per square foot of 506'24 grains, and the other of 483'1 7 grains,
showing a difference of only 4'8 per cent, against the steel,
and that was, at any rate, a great deal better than GO per cent.
On the same page, the author stated, "The only peculiarity
worth noticing in this experiment is that, while the two plates in
the feed-water heater lost 381*8 and 394'2 grains, the two in the
boiler fed from the heater lost only 8'0 and 3'4 grains respectively."
Therefore in the boiler the iron lost by corrosion about one forty-

442 THE SCIENTIFIC PAPERS OF

eighth, and the steel about one hundred and twentieth of what
they respectively lost in the feed-water heater, the loss of ron
in this case being about two and half times that of the steel.
In Table IX., set 98, the Y steel produced by the Siemens
process gave a corrosion of 1,220 grains, and the DD. Yorkshire
iron 1,221 grains, in each case per square foot of surface ; the
corrosion in those instances being practically the same. In the
next set, 99, the Y steel gave a corrosion of 259 grains ; Stafford-
shire iron 269 grains, and DD. Yorkshire iron 260 grains ; showing
that the steel came out best in that series. In set 102, the Y
steel gave 109'5 grains (in rain water), and the DD. Yorkshire
iron 144 grains. And yet the author followed up these facts with
the conclusion that steel corroded on an average 64'8 per cent,
more rapidly than iron. He entirely objected to the mode of
reasoning adopted ; he contended that averages were only applic-
able to errors of observation. If an observer was not certain of
his weighings, and he made a hundred weighings of the same
piece of iron, he would be perfectly justified in taking the average.
But nothing could be more unscientific or erroneous than
averaging several materials, one group of which he chose to call
iron, and another which he called steel. It was as if a moral
philosopher wanted to find out whether fair complexioned people
were more virtuous than dark complexioned people, and were to
take six fair people and six dark people promiscuously ; then
finding that they were all very well behaved, except one of the fair
people, who happened to have just escaped from gaol, and had
committed six murders, he were to draw his average and say,
" I find that fair people have committed, on an average, one
murder each, and should therefore not be trusted." That was the
kind of argument which the author appeared to have adopted.
There were substances, compounds of iron, manganese, and silicon,
sold for steel, which no doubt corroded very rapidly ; but it was
fur the consumer not only to select the proper material, but also
to see that it was properly used. He believed it was in regard to
the proper selection and use of the materials that the enormous
discrepancies with which they had to deal would be found. The
author stated that there was more cinder in iron than in steel, and
therefore that there was prima facie ground for supposing that
iron would corrode more than steel. There was, however, an

WILLIAM SIEMENS, J-'.K.S.

443

;itil difference U-lween cinder in iron and the scale of steel.
The cinder in iron was a glassy substance, which was a dielectric,
and therefore had no effect upon the corrosion of the metal,
us the scale on steel, which was produced in rolling, had
y deteriorating influence ; it was a magnetic oxide, which was
i \v to the steel, and wherever the metal was exposed in the
nee of such magnetic oxide, corrosion took place rapidly.
Au'iiin, if the scale should be rolled into steel plates, as was
sometimes the case, rapid corrosion ensued, for the same reason.
But he need hardly say that with proper care those causes of
undue corrosion could be and were prevented, and the extensive
use to which steel was now put proved sufficiently that there was,
at any rate in ordinary practice, no such destructive effect going
(in. The author stated that those interested in steel had been
singularly negligent in not following up the question of corrosion.
Being himself much interested in steel, Dr. Siemens had for some
years caused a running set of experiments to be carried out by
.Mr. Willis, the chemist at Landore, which told a very different
story from the author's. In one series, extending over six
months, made partly in a boiler supplied with salt water, and
partly by exposure in a tidal river — the plates being exposed to
the air for six hours, and then immersed for six hours in salt
water — the result was in some instances of open exposure slightly
in favour of iron, but, in the cases of boilers, always very
much in favour of steel. He had just received a report from
Mr. Willis, in which reference was made to a point of importance,
the perfect cleaning of the surfaces. It had been found that if
the surfaces were carefully cleaned of oxide by dipping the plates
in the first instance in an acid solution, the corrosion was always
much diminished. The evil effects of scale on steel were pointed
out at the Institution of Naval Architects,* by Mr. Barnaby, on
the 5th of April, 1871), when he clearly showed that the magnetic
oxide scale was very deleterious in its effects. He believed that
it was now the practice of the Admiralty to clean the scale on"
before using the plates for shipbuilding. During the past week
he had received a number of letters, quite unsolicited, from
gentlemen interested in the use of steel, all speaking in the most

* Vide Transactions of the Institution of Naval Architects, VoL XX. p. 225.

444 THE SCIENTIFIC PAPERS OF

definite manner in favour of steel as a metal not liable to corrode
under ordinary circumstances. One of them was from the Clyde
Bank Foundry, in which it was stated by Mr. Thompson that forty
steel boilers had been at work for more than two years, and that
their examinations had led to the conclusion that no active
corrosion was going on. He believed it would be found, from
general experience, that steel under proper conditions lasted at
least as well as iron. He hoped that the discussion would bring-
out such further facts as would put the question practically at
rest. That there was, under certain conditions, a very active
corrosion going on both upon steel and iron was clearly proved by
the paper, and by other experiments ; but the conclusions drawn
by the author in favour of iron were, he thought, unjustified by
the results of his own experiments as well as those 'of others.

In the discussion of the Paper

" ON THE PECULIARITIES OF BEHAVIOUR OF STEEL
PLATES SUPPLIED FOR THE BOILERS OF THE
IMPERIAL RUSSIAN YACHT ' LIVADIA,'" by W.
PARKER, Esq., Chief Engineer Surveyor of Lloyd's Register,

DR. SIEMENS*. said : The behaviour of the material composing
the boilers of the Livadia is so extraordinary, and the results so
contradictory, that it is certainly of the greatest interest to have
it sifted to the very bottom. When the defective condition of these
plates was first discovered, Mr. Parker kindly sent me a piece of
plate broken off short by a hammer. The analysis which I had
made showed that the metal was irregular in its composition, and
it gave a certain percentage of an element which I do not see
represented on that table, and that is silicon. The specimen

* Excerpt Transactions of the Institution of Naval Architects, Vol. XXII.
1881, pp. 27, 28.

•S7A' WILLIAM SIEMENS, l-'.K.S.

445

which I had given me contained certainly a very considerable per-
(•'•inage of silicon.

Tki' J're.tu/fnt. Could you state the percentage ?

Dr. Siemens. I can give it in the notes ; I think it was about
•".; per cent., but unfortunately I have not the analysis with me.
Other analyses by eminent chemists have given, I am aware,
different results ; they have shown no silicon. But if we look
upon the interesting table that has been placed before us, I think we
have quite sufficient ground given us to account for these differences
in the analyses. There we have in the same plate, carbon, differ-
ing in the proportion of at least one to two between the outside
and the inside of the plate ; we have phosphorus, a substance
that is not easily oxidised, when combined with metal, distributed
in the same irregular manner, and the same with regard to sulphur.
Now a question arises, how can that very great difference in those
constituents have arisen in the metal ? Is it that the metal
itself — the ingot — was not composed of the same mixture, that it
was imperfectly fused, and imperfectly distributed in the ingot
mould ; or has the metal, after it has been rolled into plates, been
subjected to treatment whereby its chemical nature has been so
completely disturbed as would appear from that table ? I can
hardly imagine that in the same ingot — although there may be
some differences of composition found between one portion and
another — differences of such extraordinary magnitude could possibly
have occurred. I am, therefore, forced to the conclusion that the
metal has been greatly injured in the after-process to which it has
been subjected. Mr. Kirk has brought before us very interesting
experiments indeed, showing how steel of undoubted good quality,
when heated injudiciously and heated for a length of time in a
furnace containing a large excess of oxygen, can be, what we call
in the trade, burnt. It would have been interesting if we could
have seen another column there, only I admit it would have been
very difficult to ascertain it, showing the amount of oxygen
absorbed in the metal. I think it would have been found that
the plate near the surface had absorbed a very considerable
amount of oxygen, which assisted in making it the brittle, un-
reliable substance which it became when it was expected to stand
its work. It is stated in the paper that every precaution was
taken, and that when the plates after punching showed defects,

446 THE SCIENTIFIC PAPERS OF

they were taken back to the works to be annealed. I am disposed
to take considerable objection to that procedure. Mild boiler
steel should bear any amount of rough work in a cold condition
without receiving the least injury. If by punching mild steel
you lose strength, I consider you have a clear proof that the
material is not what it should be. So far from diminishing its
strength, the very act of punching should increase the strength of
the material. In fact, any treatment to which mild steel is sub-
jected, straightening or punching it, or straining it in any one
place, so as to alter slightly its form, the effect invariably is
increase of strength, and not a slight increase of strength either.
It is, therefore, a practice with which I have little sympathy, that
of annealing plates after punching. I would prefer to see the
metal used, after it has been punched, without being annealed,
and if it is a proper metal it should not require annealing, except
in case it had been partially heated in a smith's fire. If the holes
are afterwards bored to make the edges smooth, that may be
desirable to give more elegant work, but for the mere sake of
getting strong and reliable work, I do not believe that it is necessary.
Punching alone should give you the full strength of the metal
when the work is put together. Therefore, what we should know,
in order to judge this question more fully, is the exact mode in
which the metal was treated after it had been made, what sized
ingots were cast, and were those ingots hammered before they
were rolled ; if they were reheated, as they were undoubtedly,
what kind of furnace was used, and were they allowed to remain,
as sometimes happens, a night in the furnace. Then, again, Mr.
Thornycroft has already alluded to the processes of annealing
which are sometimes adopted : I think there is great danger in
annealing steel plates ; only if a steel plate has been heated in the
forge fire, if it has been heated partially in order to bend it,
annealing is necessary, to put it as a whole in a natural condition ;
but if the metal has been worked cold, I believe it would be much
better to leave it unannealed. Then, again, annealing is not
always done in a proper and judicious manner ; if one plate only
could be put into a well-heated furnace, to take up rapidly the
temperature necessary in order to put the particles to rest towards
one another, and then could be taken out and allowed slowly to
cool, probably no danger would arise. But plates are heaped up

.Y/A' WILLIAM

F.R.S,

447

in the annealing furnace, and in order to bring the heat down to
the bottommost plate, a strong flame is applied to the uppermost,
and it is but natural that those plates are put in great danger.
The probability is that some of them will be overheated and that
will be heated all round at the edges, but will not be
uniformly heated. Such a process of annealing must be, at any
fraught with danger to the material produced. Now one
rord, my lord, with regard to the process used. It has been said
icre, that the plates last alluded to, those made at the Cumberland
Works, were undoubtedly made by the Siemens process. Now, no
doubt this statement has been received authentically, but was
there not perhaps some little mistake, inasmuch as those works
lave only lately put up one furnace according to my system, and I
was only told the other day by Mr. Snelus himself that it has not
yet been properly put to work. I do not, of course, wish to imply
that this could not have happened with steel manufactured by one
process or another.

DR. SIEMENS. Might I be allowed to say one word in explana-
tion ? There is one observation which I made which has been
perhaps rather severely criticised. It concerns a matter of great
importance. I do not wish to say whether plates should be
punched only, and left, or whether it would be better to clean out
the holes. But what I say is, that in taking a strip of mild steel
(and I refer not to the somewhat harder steel used in ship building,
but to the milder steel used in boilers) and punching a line of
holes across a plate of this steel and subjecting it to tensile strain,
the total strain that piece of metal will bear would be fully equal,
or, in the case of first-class metal, about 10 per cent, superior, to
the normal strength of the material multiplied into the remaining
cross-section. I base that view, not upon mere theoretical grounds,
but upon many experiments, which I have made and seen tried,
and I should be very happy if Mr. Martell would give me an
opportunity of testing it with him.

448 THE SCIENTIFIC PAPERS OP

In the discussion of the Paper

" ON THE USE OF MILD STEEL FOR SHIP-BUILDING
IN THE DOCKYARDS OF THE FRENCH NAVY,"

By M. MARC BERRIER-FONTAINE,

DR. SIEMENS * remarked : The paper that has just been read
before the Institution is one of very great interest to naval
architects, because it brings before us the practice developing in a
neighbouring country, in which steel was used at a very early date
indeed for naval construction. If I may be allowed to discuss the
end of the paper first, I would refer to the remarks of the author upon
the liability of steel to rust ; and I would here draw attention to
a veiy exhaustive discussion which took place only a week or more
ago at the Institution of Civil Engineers. There an alarming
account was presented, showing that steel rusted more rapidly
than iron ; but after two evenings' discussion the prevailing-
opinion arrived at was that with steel of a proper character this
was not the case ; and the facts brought before the Institution left
no doubt upon the subject. Now, with regard to the French
practice, I wish to mention that the open-hearth process used in
the production of steel plates in France differs essentially from the
process that has been more generally adopted in this country.
There the Siemens-Martin process is used, and steel is made by
the fusion of scrap steel or scrap iron with pig metal, and the
final addition of spiegel, whereas in this country the process with
which my name is more exclusively connected — whereby steel is
produced by the use of ore and pig metal only — is practised by
most of the large works which carry on that manufacture ; and I
am disposed to consider that there is an essential difference be-
tween the two kinds of steel ; for you get a more refined and
homogeneous metal by employing only very little scrap and using
the ore instead of the scrap process. This difference may account
for the less favourable results which the French Navy appears to have

* Excerpt Transactions of the Institution of Naval Architects, Vol. XXII.
1881, pp. 139, 140.

.s7A' U'lI.I.lA.M \//-;. !/A. Y.V. I'.R.S. 449

ol)t iiint'il jus regards thecorrosion of their ships. I would attribute tint
result also in some measure to their continuing to use iron rivets
in the construction of steel hulls. The discussion already alluded to
lit nut the most variable results regarding corrosion. In one
case steel seemed to corrode much more quickly than iron. In other
cases, in experiments continued over years, steel showed decidedly less
corrosion than iron. But one thing is certain, that when different
rials are brought into contact with one another, and with sea-
water, rapid corrosion of one or the other will ensue, and I could
not too strongly urge the desirability of using in naval construc-
tion one material only, be it iron or be it steel. As I have men-
tioned the process with which I am most intimately connected, I
wish it to be understood that I have no desire whatever to set that
process before any other. I do not think that users of steel should
inquire too much into the process involved. We had yesterday
brought before us facts regarding certain materials that had been
used, which speak for themselves ; and plates, different parts of
which give different analyses, are evidently materials that ought
never to be employed. It is in the users' power to ascertain
whether the material they propose to use, and which is supposed
to be homogeneous, is so or not. I was reproached for not having
gone into the causes of the failure of the material employed in the
construction of the LivadicCs boilers, but the facts speak for them-
selves. Metal that is brought into the ship-yard ought to be
above all things homogeneous. I mentioned yesterday a circum-
stance which gave rise to great difference of opinion. I said that
good mild steel should not lose strength in being punched, but
should rather gain strength. My lord, I have since then looked
up such experimental facts as I happened to have at my office in
support of that view. I have here a report which was made by
an assistant to Mr. Ilendel, who wished to see what the effect of
punching was, and those experiments were made in consequence
of observations I made at the Institution of Civil Engineers. The
result of the experiments, which are very accurate and perfect, is
as follows : In a strip which broke with 30f tons to the square
inch, when one hole was punched the strength diminished to
30 tons to the square inch ; when two holes were punched side by
side the strength increased per square inch to 31'G5 tons ; and
when three holes were punched the strength was increased to

VOL. I. GO

450 THE SCIENTIFIC PAPERS OF

32 -65 tons per square inch, showing an increase of more than
2 tons to the square inch after having been punched or disturbed
in the greatest possible manner ; and I hold that the most crucial,
the most searching test this homogeneous metal can be put to is
to punch it with several holes at close distances in a line. If it is
really high-class metal it will simply flow when it is put to a great
local strain, and the ultimate strength will increase, for the same
reason that if a bar is stretched to a certain extent its strength
per square inch is increased. But if it is metal that is not capable
of perfectly solid flow it will diminish in strength. I have here,
also, experiments by Professor Kennedy ; these experiments will
shortly be published by the Institution of Mechanical Engineers,
and also go in support of the view I stated yesterday, that of good
mild steel it might be said, what the old proverb attributes to
" a wife and a mulberry tree," " the more you beat it the better
it be."

In the discussion of the Papers

" ON THE USE OF A MECHANICAL AGITATOR IN THE
MANUFACTURE OF BESSEMER STEEL," by W. D.
ALLEN, Sheffield ; and

"ON THE DISTRIBUTION OF ELEMENTS IN STEEL
INGOTS," by G. J. SNELUS,

DR. C. TV. SIEMENS * said he wished to make a few remarks on
the second and third papers, which opened up questions of very
great interest. They dealt with two distinct questions, one being
that of mixing the metal before it was poured into the ingot
mould, and the other the action that took place within the ingot
mould when the steel was allowed to cool very gradually. Sir
Henry Bessemer had, at a very early period, directed his attention

* Excerpt Journal of the Iron and Steel Institute, 1881, 386-388.

WILLIAM SIEMENS, F.R.S. 451

to the first of these two questions, as ho had to most questions re-
iranling the production of steel in large masses by his process,
lie miirht say that he had himself occasionally given attention to
it also, but he had not been bold enough to resort to the mechani-
stirrer, which Sir Henry Bessemer had conceived, and which
now practically being carried out, as it appeared, with advan-
tage. He had tried sometimes to produce agitation by resorting
to what the copper-smelters called " polling," that was, putting
into the ladle poles of dried wood, which made a considerable
agitation without disturbing the chemical condition of the metal.
No doubt all these means were very beneficial, and he was not the
least surprised to hear of the good results which had been obtained.
The second question was one involving greater difficulty, because
the results placed before them by Mr. Snelus proved, what they had
already had occasion to observe, especially with regard to the plates
of the Livadia's boilers, that there was chemical change going on in
the metal when it first consolidated. The portion which first
solidified seemed to take with it such admixtures as manganese,
phosphorus, and carbon even, often leaving the metal in the centre
of a different composition. The only way to counteract such
action within the ingot mould would be that of very rapid cooling.
That would be attended with great difficulty, and particularly if
they dealt with very large moulds. But he believed he had
observed that it depended also upon the time which was allowed
to the fluid metal, when thoroughly mixed, to stand before it was
poured into the mould. There was such a thing as imperfect and
perfect chemical admixture. To use a popular illustration, they
might pump carbonic acid into water and make soda-water, but if
they used it immediately after it was made, they might just as well
save themselves the trouble of forcing the carbonic acid into the
water, because it came out instantly. If they had a second bottle
made in the same manner, and it were left for a week, they would
find a better chemical combination ; but if they left it for three
months, they would have the carbonic acid retained by the water
for a very considerable length of time after being poured out. He
had observed that considerable benefit accrued to the steel if, after
it had been poured into the ladle, it were allowed to remain there
under the protecting covering of slag to prevent decrease of
temperature. The objection to this mode of proceeding was, that

G O 2

452 THE SCIENTIFIC PAPERS OF

at this portion of the process time was valuable ; but he felt sure
that great practical benefit would accrue if the steel could be kept
in a quiescent fluid state for a quarter of an hour before pouring.
He considered that both papers were of great interest, as they
directed the attention of steel manufacturers to points where great
improvements, as regards uniformity of metal, might yet be looked
for.

In the discussion of the Paper

"ON THE METALLURGY AND MANUFACTURE OF
MODERN BRITISH ORDNANCE,"

By Colonel MAITLAND, Superintendent, R.G.F., Woolwich,

DR. C. W. SIEMENS,* F.R.S., in opening the discussion, said
that the observations which he proposed to address to the Insti-
tute had reference almost entirely to the first paper, that by
Colonel Maitland on the practice now followed at the Royal Gun
Factory. When a public department like the Ordnance brought
their practice before them, it was, no doubt, with the intention of
inviting criticism, and it was a duty for the Iron and Steel Insti-
tute to discuss it freely. It was satisfactory to find that the
Ordnance Department had at last seen the advantage of using steel
not only for the inner tube but also for the hooping of their
ordnance. To that extent he was pleased to express his entire
concurrence with the author. He did not, however, quite agree
with him that it could have been the absence of block powder or
compressed powder that was the cause of the long delay ; because
if steel was both stronger and tougher than iron, then whilst the
more dangerous small powder was used, there must have been all
the more reason to resort to it as the stronger and tougher
material. It was true, as Colonel Maitland had said, that with the
use of small powder the strain thrown upon the inner tube was so

* Excerpt Journal of the Iron and Steel Institute, 1881, pp. 483-488, 521, 522.

.sVA' WILLIAM SIEMENS, F.R.S. 453

great as to render the assistance of hooping comparatively small ;
hut. if that was the case, surely a lighter hoop of steel would have
ansuvivd the purpose of the larger hoop of iron. It might have
1 «.•!• n u !••_;•< •<!, " Oh, but we wanted the material for recoil purposes,"
hail not the author said the objects were to make the gun light,
j>o \\vrful, and safe. He thought that to realise these three
qualities, steel must, undoubtedly, have been better than iron.
Nor could he admit that large powder was not known twenty years
ago, because he held in his hand a very interesting paper that was
read before the Institution of Civil Engineers, in that very room,
in I860, entitled " Our National Defences," in which Rodman's
perforated cake powder was described and very fully discussed by
the meeting. They had, therefore, during the last twenty-one
years, been in possession, not only of large powder but of a state-
ment of the advantages obtained by its use in the United States.
Perhaps Colonel Maitland would favour them with further ex-
planation on this point.

Another and a purely scientific point connected with the state-
ment made by Colonel Maitland had reference to the temperature
to which the hoops were heated before they were shrunk on. The
author said that that temperature was of little importance provided
it was sufficiently high. He begged to differ from him ; he
believed that it was of the utmost importance that the tempera-
ture to which the hoops were raised should be adjusted with the
greatest nicety, and if they would allow him he would make his
meaning plain. (Dr. Siemens drew upon the blackboard a hoop
surrounding a tube.) He asked them to suppose that to be a
hoop which was bored so as to be a little smaller than the tube
upon which it had to be shrunk. Supposing they heated the hoop
considerably beyond the point that would allow it just to slip on,
what would be the result ? When the inner surface of the hot
ring came into contact, or nearly into contact, with the cold tube,
it would chill all round, and form in itself an abutment against
the further shrinkage of the heated ring ; and the metal behind,
being still hot, would readily yield by stretching tangentially
instead of contracting radially. It would be fallacious to suppose
that a ring cooled internally would shrink back to its original
diameter. A ring so cooled would, under no circumstances, shrink
back entirely to its original diameter, although not resisted by nn

454 THE SCIENTIFIC PAPERS OF

internal core, but the difference between the original and final
diameter would depend upon the temperature to which it had been
heated in the interval. He would undertake to take a ring of a
certain diameter, and by the mere treatment of heating and in-
ternal cooling increase the diameter to an extent very considerably
exceeding the external diameter of the tube upon which it was to
go. Therefore, he would submit that the temperature to which
the rings were heated should be adjusted to the greatest nicety, in
order to ensure the condition upon which the safety of the gun
depended. The subject of " Physical changes occurring in iron
and steel at high temperatures " was brought before the Iron and
Steel Institute by Mr. Wrightson, one of their members not long
ago,* and the conditions there referred to applied very forcibly to
the case now before them.

The paper under discussion went on to describe the process
followed at the Gun Factory to make the steel they used. The
apparatus there employed consisted of two things, a furnace, stated
to be Price's patent retort furnace, and a process called the open-
hearth process, but which was in reality as minute a description of
the one known by his name as though it had been taken from the
instructions which he was in the habit of sending out to his
licensees, and in accordance with which nearly half a million tons
of steel were annually produced in this country. Price's furnace,
it was stated, differed from the regenerative gas furnace in having
a retort gas-producer, and of that a considerable point was made.
He had with him a sheet which he had taken out of a patent he
had obtained as early as 1863, describing a very similar gas-
producer ; and if Colonel Maitland was present he should like him
to point out the essential difference, if any, between the two.
Although Dr. Siemens had used the retort gas-producer for some
time, he had come to the conclusion many years ago that it was
not productive of the best results ; that, in order to convert coal
into gas of the greatest heating power, it should be acted upon
energetically, so as to attain the highest possible temperature at
the point where the carbonic acid took up another equivalent of
carbon to form carbonic oxide. He felt inclined to challenge the
Department to make a comparative experiment, using their furnace

* "Journal," No. II. 1879, p. 418, and No. I. 1880, p. 11 et seq.

SfR WILLIAM SIEM1-:.\S, l-.R.S. 455

and their open-hearth arrangement, and see what the consumption
of fuel was, and what time was necessary for the charge as com-
pared with the furnace constructed entirely in accordance with his
pluus. If it should be proved that the results were in favour of
his own construction, he thought he should have a good claim upon
tin- Department to come to him for designs before erecting further
turn aces. He need hardly say that these furnaces at Woolwich
had been erected and the process set to work without any com-
munication with himself. The Department considered themselves
above any such liability as patent rights, and therefore appeared
bound to follow out their own detailed designs. The paper went
on, moreover, to describe the unfortunate results obtained with a
Siemens furnace erected nineteen years ago at the Arsenal. Any
unprejudiced person, reading the paper, would think that he had
gone there promising them wonderful results from his furnace, and
that they, finding it a failure, had abandoned it, and resorted to
devices of their own. But if such a conclusion were come to by
any member, it would be a very erroneous one. The Department
at that time, acting on the principle that the Government was in-
dependent of patentees, did net address themselves to him, but to
a former draughtsman of his who had established himself in
business as a contractor. That gentleman svas permitted not only
to contract for the erection of a furnace, but to furnish his own
designs. Possibly the former circumstance had influenced him to
some extent to put in the enormous culverts that might have
answered for eight or ten such furnaces, and must have rendered
the construction very costly indeed. At that period furnaces had
been erected by him (Dr. Siemens) for precisely the same purposes.
Sir William Armstrong had at that time not one but several of his
furnaces for heating gun coils at his works, and how highly he
approved them was shown by his own allusions to the matter when
President of the British Association, and again as President of
Section G of the same Association at Birmingham. Members of
the Institute would recollect, moreover, that when visiting the
Creusot Works, three years ago, they saw there four splendid fur-
naces erected by Messrs. Schneider for heating large gun and other
forgings. At Krupp's the same furnace was also used, particularly
in connection with the Gun Factory ; and the same could be said also
with regard to Sir Joseph Whitworth. Having all the gunmakers

456 THE SCIENTIFIC PAPERS OF

of great importance throughout the world using the regenerative
furnace as designed by him, why was it that at the Royal Gun
Factory, and only there, it was a failure, and was now for the first
time publicly so described ? The fact was, it was a failure there
because it had been put up without knowledge. It was true that
after the furnace had been erected and failed, he had received a
letter from the War Office to go and see what was the matter. He
expressed his surprise at being called upon at so late a date to see
to it, but he immediately went, and reported that, although he had
not seen the furnace in operation, it was evidently constructed iu
a manner to prevent its proper action ; amongst other things there
was no sufficient ventilation provided below the seat of the furnace,
although it was to work with a cinder bottom. The coil would
therefore stand in the fluid cinder, and this could not be altered
without pulling down the furnace, and re-erecting it in accordance
with approved designs. This was the course he recommended,
and he received instructions to prepare plans, but before these
were quite finished, he received another letter, stating that the
estimates for the year did not contain an item for rebuilding such
a furnace, and its consideration must be postponed ; and it had
been postponed to the present day.

He thought that the question of patents as regarded public
departments should be inquired into and cleared up. He was
quite sure that if the Government had not claimed immunity from
such liability, they Avould have come to him in the first instance,
and they would have had the benefit of all the experience which he
had gained at that time. A good furnace would have been erected
for much less money than they had paid, even if they had in-
cluded some few hundreds for royalty and personal charges which
might have been incurred. By this time they would have saved,
indeed, many thousands of pounds, while he would have been
richer not only by the amount of royalty he might have received,
but because the reputed failure in question had stood in his way
for years and years in the introduction of his furnace. He had
felt aggrieved regarding this for the last eighteen years, but now
was the first opportunity that had been given him for publicly
relating the circumstances of the case.

Before sitting down, he would say only a few words regarding
the very careful and practically valuable paper of Mr. Butter.

S/K WILLIAM X/AM/A-.Y.s, J'.R.S. 457

Mr. liuttcr incidentally described hydraulic compression for check-
ing the recoil of gun carriages, but without reference to its
origin. II civ, again, he considered himself aggrieved to a
certain extent, because hydraulic compression had been adopted
by tiic l)cjiartnu't)t on his recouiniendation. He had been called
down to Woolwich to advise whether electricity could not be
employed for checking the recoil of gun carriages. At that time
he knew too much of the weakness of electricity, arid then and
there suggested the plan of hydraulic compression, which, modified
and improved no doubt by the Department, had been generally
adopted throughout the service. He muse add, in justification of
Colonel (now General) Clarke, who was then chief superindent,
that he gracefully acknowledged his claim regarding this improve-
ment in a paper read by that officer before Section G of the
I'.ritish Association at Exeter ; but as regards the Department, he
never received so much as a line of acknowledgment for either the
or the time he had devoted to the matter.

DR. C. WILLIAM SIEMENS, F.R.S., said that before the discussion
closed he would like to say a few words in explanation of his obser-
vations of the previous day. From what had fallen from Mr.
Markham, he was almost afraid that he had conveyed the idea to
the members that he wished to make a personal attack, either upon
the gallant Colonel who had brought the paper before them, or
upon other gentlemen connected with Woolwich ; but he meant
no reflection either upon their honourable disposition or their skill.
In fact, nothing was further from his thoughts. He found fault
with Colonel Maitland when he presented himself before the
Institute as a historian, and he thought it his duty to correct an
erroneous impression. His attack was not directed towards
individuals, but towards the system. No one had a right to com-
plain if it were understood that the Government were resolved not
to recognise patent rights. He had never preferred a claim against
the Department, and therefore he had no right to complain in that
respect ; but he considered that a patentee was the guardian of
his invention while it was in its infancy. It was interfering very
much with his interests if a Government Department took up his
invention without consulting him, and then, perhaps, through an
oversight of the essential conditions, brought about a failure which

THE SCIENTIFIC PAPERS OF

acted very detrimentally to the introduction of that invention. It
was with regard to such unauthorised applications of inventions
that he felt rather strongly, and he therefore wished to say some-
thing on that subject.

In the discussion of the Paper

"ON STEEL FOR STRUCTURES,"
By MR. E. MATHESON,

DR. SIEMENS* said, he thought the author had brought forward
one of the vexed questions of the day, viz., to what extent steel
should be trusted for engineering structures. In consequence of
rules which had hitherto been followed by the Board of Trade
with regard to steel, that material had not come unto use for
engineering purposes except under special circumstances ; and
although he believed that at the present time the officers of the
Board of Trade were anxious to study the material, and gain from
their own observation the experience necessary to give confidence,
the public were still without the result of their labours in that
direction. Steel differed very materially from iron in regard to
the variety of strength and other qualities which it was able to
assume. There was steel which resisted a tensile strain of
nearly 80 tons, and whose elastic limit exceeded 40 tons; with a
very slight modification of its composition, it would assume an
ultimate strength of only 27 tons ; but on the other hand it
would have acquired a power of elongation amounting to nearly
25 per cent, before breaking took place. It was, therefore, an
important question, which of the endless varieties of steel between
those limits was the best for engineering purposes. He thought
it would be impossible to give one satisfactory answer to a
question of such scope. It required a careful study of the

* Excerpt Minutes of Proceedings of the Institution of Civil Engineers, Vol.
LXIX. 1882, pp. 36-39 and 58.

A/A' WILLIAM SIEMENS, F.R.S. 459

precise purpose for which the material was required. If it was

wanti'd to make a chain-Krid-v, sici-1 of so tons Invakinj: strain
per square inch was the proper material ; if, on the other hand,
a suspension link bridge were constructed, the links could hardly
be made of steel of such hardness, and the engineer would
have to be content with a tensile strength of from 35 to 40 tons.
The reason why it would not be safe to take steel of a higher
ultimate strength was that its elongation would be insufficient
to give a sufficient factor of safety ; in case of unequal strains
between the parts, accidents might arise, whereas in dealing with
steel of from 35 to 40 tons per square inch, there would be an
elongation of something like 10 per cent, before breakage could
occur, and within that range the unequal strains could adjust
themselves. Again, if it were intended to put steel under a
compressive strain, a high limit of elasticity was the chief
desideratum, and it was possible in that material to vary the
limit of elasticity without interfering with the ultimate strength.
Two bars of steel, both with an ultimate strength of 35 tons to
the square inch, might have different elastic limits, but within those
elastic limits the ductility would be influenced. It might appear
that, owing to the great variety of conditions that were attainable,
steel was an uncertain material, but that was not the case ;
whereas, in the manufacture of iron a good deal depended upon
the personal skill of the worker, in steel it was rather a chemical
test which determined the qualities ; and it was quite possible to
ensure a uniformity in steel such as could not be approached in
iron, only it was necessary that the engineer should understand
clearly what was the quality of steel required for certain purposes,
and then it was for the manufacturer to produce that quality
uniformly and regularly. If, then, structures like links of a chain
should have a tensile strain of, say 38 tons per square inch, to
what extent might such a material be trusted ? In the case of
wrought iron, engineers thought it generally safe to trust a struc-
ture to one-fourth of the breaking strain, and the same rule might
still hold good for steel of that description, because the elongation
before rupture would be very much the same as in the case of iron.
But in considering the case of very mild steel the same rule should
no longer hold. If they constructed a bridge of cast iron they
would consider it unsafe to load it to more than one-sixth or even

460 THE SCIENTIFIC PAPERS OF

one-eighth of its breaking strain, whereas, in the case of wrought-
iron, it was thought safe to go to one-fourth. Why was this the
case ? Because wrought iron would elongate before it broke ; but
in dealing with mild steel, an elongation to more than double that
in the case of wrought-iron took place before fracture, and there-
fore the same reason that would make an engineer trust wrought-
iron to a higher proportion of its ultimate strength than cast-iron,
would lead him to trust steel to a still higher limit as compared
with its own breaking strain. There was another reason why mild
steel could be trusted to a limit exceeding half the elastic limit,
namely, that in riveting or punching, its strength was not dimin-
ished in anything like the same extent as in the case of wrought
iron or of harder steel. He had caused a great many experi-
ments to be made with mild steel, and had found that the strength
per square inch of the remaining section of material, after a line
of holes had been punched in it, was nearly equal, and in some
cases, even superior, to the original strength of the bar per square
inch of section. That seemed strange but it was perfectly con-
sistent. Taking a bar of this material, and putting it under
tensile strain nearly equal to the elastic limit, for a length of time,
it would be found that not only was the elastic limit raised, but
that the ultimate strength was raised, and that very materially.
It was possible to teach a bar to carry at least 10 per cent, more
than it did on the first trial, by simply keeping it under strain for
a length of time. In the same way in subjecting the material to
a kind of solid flow — any kind of strain — its ultimate strength
and its elastic limit would be raised ; and the reason why punching
or ill-treatment of any kind did not injure such material was that
it obeyed the law of solid flow ; whereas, an inferior material,
would not in its entirety, obey that law, but would be liable to
rupture in its ultimate parts. He, therefore, submitted, that in
using steel for riveted structures only the mild kind should be
employed. It might be said that in that case it would be neces-
sary to give up the strength of the material and approach the
ultimate strength of iron ; and what was the use of the material
if no more could be done than that with it ? His answer would
be that it was a safer material when put together than the stronger
steel, and much safer than iron (although its ultimate strength
was not much greater,) for the simple reason that there was a

.S7A1 WILLIAM .V/AM/A'.V.V, I'.K.S. 461

reliable strength in every part of it, even in the lines of riveting,
that a material of that sort would stretch more than 20 per cent.
li.-i'i ire it broke, and that for that reason the unequal strain originally
set up in the structure would be set right by straining it. In
fact, he could see no reason why one-third of the breaking strain
should be considered as the limit to which mild steel could be
safely trusted. If the shell of a boiler could be rolled in one
piece without riveting he would advocate the use of a much
stronger steel, of from 85 to 40 tons, but so long as riveting was
i'd to he would give the preference to the milder steel, and
notwithstanding its mildness, would trust it to the limit of one-
third of its breaking strain, or, say, 8 tons per square inch, which
he thought, would be a safe limit. He hoped the discussion would
materially contribute to the obtaining of a safe but progressive
rule which would enable engineers to use steel not only to the same
extent to which it had already been applied in other countries, but
to a still greater extent.

DR. SIEMENS, in explanation, said the strength of the steel
before punching, alluded to by Mr. Parker, was rather high for
the quality. The weakening by punching was very great indeed. It
had brought the teusile strength of the remaining metal down to 17
tons per square inch, whereas the bar, after being annealed, had a
tensile strength of 24 tons per square inch ; but even that was
apparently (> tons below the original strength of the bar. As
Mr. Campbell had remarked, punching was an art, and ought to
be done properly ; but he felt satisfied that, in punching mild steel
properly, engineers need not look for a reduction in tensile strength
of 24 per cent., such as the results referred to indicated. He had
the particulars of experiments showing that the strength of mild
steel was not materially influenced either way by punching without
annealing. In one experiment communicated to him by Mr.
Rendel, three bars were prepared from the same plate and in the
same manner. The first one was not punched at all, and it broke
at a tensile strain of 31* tons per square inch. When one hole
was punched in the middle of it, the tensile strength was reduced
to 30 tons per square inch. On punching two holes side by side,
whereby the remaining metal was more equally strained, the strain
\vas again brought up to 81 '65 tons per square inch, and similar

462 THE SCIENTIFIC PAPERS OF

results were obtained by Professor Kennedy and by himself over and
over again in testing good mild steel. Of course the stretching was
ever so much less after the bar was punched through the middle,
because it did not matter how long the bar was, the strain fell
only in the line of least sectional area. In the first bar there was
a stretch of 2 inches, in the second f inch, and in the third
inch.

In the discussion of the Paper

"ON THE INFLUENCE OF THE BOARD OF TEADE
EULES FOE BOILEES UPON THE COMMEECIAL
MAEINE," by J. T. MILTON, Esq.,

SIR WILLIAM SIEMENS * said : — My Lord, when gods fall out,
mortals suffer, and I present myself to-day as one of those
sufferers, in the character of a shipowner. Nine years ago I had
a ship constructed for cable purposes — the " Faraday." I took
great care myself of all the arrangements necessary to make that
ship efficient for laying and picking up cables, but as regards
boilers and engines I relied entirely upon the Board of Trade
Eules and Lloyd's Eules as they then existed, my instruction
being simply, " Make the boilers as safe and the engines as effi-
cient as they can be made." The result was undoubtedly a suc-
cess. That ship never failed to do its arduous duties, being out
on the Atlantic sometimes in the winter months for weeks
together, engaged in very rough work. But in the course of a
few years the Board of Trade Eules were altered, and it was
intimated that these boilers were really no longer sufficient under
the new Eules laid down. We had the boilers carefully inspected
after each voyage, and they were pronounced both by the Board
of Trade surveyors and Lloyd's surveyors to be in perfect condi-

* Excerpt Transactions of the Institution of Naval Architects, Vol. XXIV.
1883, p. 88.

\/A- WILLIAM SIEMENS, r.R.S. 463

tion. X' verthcless, in order to induce us to fall in with the new
Rule, the pressure was reduced after each voyage :> Ihs., and we
should verv so. m liiive been in tho happy condition of depending
upon our sails. The operations that have to be carried on on
l>o;u.l this ship arc so important, that I decided to have steel
boilers put in, at an expense of £10,000, although the inspection
of the boilers proved that they were in perfect condition. Now,
this is a hard case to me as a shipowner, because if the two
authorities had been agreed regarding the Rules, I should not
have been put to this expense, and have had to sell good boilers
second-hand, which have gone into other ships not requiring to be
under the rules of the Board of Trade. As an engineer, and one
interested in the manufacture of materials, I should say that the
Board of Trade Rules give us excessive thicknesses, especially in
cases where first-class material is used. I see on page 79 that
certain strengths of iron are assumed, not according to the tests
this iron can stand, but it is taken to be equal to 47,000 Ibs. per
square inch in the one direction and 40,000 Ibs. in the other.
Now this seems to involve a condition of things which does not
exist — that of all materials being equally good. I would rather
say, have each plate, whether it be steel or iron, tested, and if it
does not stand an elongation of 20 per cent, in a bar 8 inches
long, it ought to be considered unfit for being put into a boiler.
Mr. MacFarlane Gray has given us his reasons why the Board of
Trade insist upon a thickness of plate exceeding, as he admits, the
necessities of the case. He has given us an instance where, at a
pressure of only 17 Ibs., a plate had cracked all along the seam,
although the boiler was expected to stand 70 Ibs. But I fail to
follow Mr. MacFarlane Gray in his argument. The forces that
come to bear (viz., the strains) upon the material through expan-
sion by heat increase enormously with the resisting power of that
material, and therefore the very fact which Mr. MacFarlane Gray
has brought before us, appears to me to tend rather in the direc-
tion of avoiding excess of thickness, because with excess of thick-
ness you get always, whatever be the nature of the material you
employ, an inferior plate. But, I say, have sufficient thickness,
and be more careful regarding the quality of the material which is
employed. If the material of which that boiler was constructed
had been such as would elongate 20 or 30 per cent, before break-

464 THE SCIENTIFIC PAPERS OP

ing, I am quite certain that the strain put upon it by expansion
would have had for its results simply a local elongation, but there
would have been accommodation ; whereas rigid material has no
accommodation, but must yield when, through local action at any
one point, the strain put upon it exceeds its absolute strength.
This, therefore, appears to me to furnish a very powerful argu-
ment in favour of a material of high ductility, and the fault I
would find with all these rules is, that the factor of ductility is
not introduced as part of the factor of safety. Now, Sir Joseph
Whitworth, who has given much attention to this subject, has laid
it down as a rule that the value of a material is composed of the
sum of its ultimate strength and of its yielding power under strain.
I believe that this is a very excellent rule. Therefore if the factor
of safety of 5 is none too much for a material that would break
when extended 10 per cent., if it could be made to extend 20 per
cent., the factor 5 ought to come down to the factor 4 ; and you
would get with that factor at least as much safety as you would
get with the factor 5, and the lower extensibility of the material.
This is a question which I hope will be very fully and carefully
discussed, and I trust that it will lead to a Eule in which both the
Board of Trade and Lloyd's surveyors concur, in order to prevent
the recurrence of such incidents as I have given you.

In the discussion of Uw Paper

"ON THE USE OF STEEL CASTINGS IN LIEU OF
IRON AND STEEL FORGINGS FOR SHIP AND
MARINE ENGINE CONSTRUCTION," by MR. WILLIAM
PARKER, Chief Engineer Surveyor of Lloyd's Register of
Shipping,

SIR WILLIAM SIEMENS * said he thought the Institute should
accord their most hearty thanks to Mr. Parker for his very prac-

* Excerpt Journal of tlie Iron and Steel Institute, 1883, pp. 89-90.

WILLIAM SIEMENS, F.R.S. 465

paper. As an official referee of Lloyd's it would have
been much easier for Mr. Parker to continue on the old lines,
which would have saved him a great deal of trouble and a great
deal of responsibility ; and he (Sir William Siemens) thought it
was a most welcome thing to find a gentleman in Mr. Parker's
position corning forward and asserting the advantages in favour of
a new mode of practice. There could be no doubt that a steel
casting, if made without blow-holes, and if properly annealed, was
more reliable than a forging, if the form was at all of a compli-
cated kind. It had already been shown to them in forgings of
large shafts, and especially where a large shaft was suddenly con-
tracted down to a narrow throat or changed its direction in
order to form a crank, that the act of forging was productive
of a very great evil. A blow could not be struck sufficient
to penetrate the whole of the mass, and a light blow must
necessarily have the effect of extending the surface and thereby
producing cavities in the interior where strength was wanted.
On the other hand, if a casting be cooled in the mould, it must
contract unequally, and must have within itself the germs of
destruction ; and it must therefore be most advantageous to cool
it right down, whereby these differences of tension would be
brought to a maximum and very likely cause an evil that was
preventable. In order to anneal the casting to the best advan-
tage, it should be removed from the mould to the annealing stove
without allowing it to cool ; they would thus give time for the
different parts, while at a uniform temperature, to adjust them-
selves to one another.

Oil hardening had been alluded to, and very remarkable results
had been obtained through it which had never been satisfactorily
explained. Why steel should be increased in its ultimate strength
without losing its ductility was a matter requiring further investi-
gation. The oil did not seem to enter chemically into the process
at all ; the result could only be due to that degree of surface
cooling, which, while it did not amount to cracking or injuring
the surface of the metal, produced the effect of compression on
the inner portions of it, a compression which they knew was
favourable to a dense structure of the metal such as they found
suited best for resisting great strains. Nor could he agree quite

VOL. I. H H

466 SCIENTIFIC PAPERS OF SIR WM. SIEMENS.

with his friend Mr. Walker that the same result would be obtained
by fluid compression.

There was one more point which he might allude to. Mr. Parker
had shown on a table that the best results, as regards wear and
tear, were produced from a comparatively mild steel. He thought
it should be borne in mind that for all working parts of machinery
the ultimate strain which the metal had to endure did not exceed
two or three, or at most four tons to the square inch, for it was
inconceivable for a shaft, while it did its regular work, to be
strained in any part above four tons to the square inch. If the
steel was of the mildest description, the limit of elasticity would
be about fifteen tons. There was no object to be gained in raising
that limit up to twenty tons, and the ultimate strength to forty
tons, if they lost thereby that most valuable of all qualities in all
metals, that of ductility, or yielding before fracture took place.
All moving parts of machinery, such as shafts and cranks, he
thought should be made of steel of the mildest description.

INDEX TO VOLUME I.

HEAT.

AEEO-STEAM ENGINE.

AEBO-steam engine, Warsop's, 136.

Air-engine, 36 ; comparison of, an
essential difference between steam-
engine and, 39, 46, 138 ; difficulty
with joints of, 162 ; drawbacks in,
164 ; (Ericsson's, 26, 36 ; action
of, 26 ; compared with steam-
engine, 49 ; description of, 42 ;
economy of, estimate of, 43, and
means of improving, 27 ; experi-
ments on, 59 ; failure of, cause of,
58 ; friction great in, 44 ; heated
cylinder of, drawback of, 44 ;
heating surface too small in, 28, 44 ;
. imperfections of, 44, 163 ; indi-
cated horse-power of , 45 ; practical
arrangement of, 28 ; pressure effec-
tive in, 43 ; radiating surfaces
large of, 44 ; regenerator applied
to, 42 ; theoretical diagram of, 42 ;
theoretical result impossible in,
28 ; weight and bulk of, 44) ;
expansive, theoretical superiority
and practical inferiority of, 38, 39 ;
important for small power, 164 ;
joints of, cement for, 162 ; re-
generative principle applied to,
161 ; regenerator in, misconception
of use of, 163 ; (Stirling's, 36, 39 ;
advantages of, 39 ; and Cornish
engine compared, 41 ; description
of, 40 ; failure of, cause of, 41 ; and
perfect engine compared, 41 ;
practical defect of, 42 ; pressure,
working of, 40 ; real effect of,
41 ; theoretical diagram of, 40 ;

P.LAST.

theoretically inferior but actually
equal to Ericsson's, 46).

Air expansion. See Expansion of air.

Air injected into steam, 136 ; action
taking place when, 137 ; prevent-
ing sediment and priming in
boiler, 136, 137 ; work increased
when, 136, 138.

Allen governor, 159 ; essential dif-
ference between Siemens's liquid
governor and, 160 ; mode of action
of, 159 ; simplicity of , 159 ; throttle
valve essential to, 161.

Anthracite coke, 100.

Appold oven, 100.

Atmosphere, closeness of indwellings
heated by steam, 192.

Auxiliary propulsion, 124.

BIDDER. S. P., coal working and
breaking, machines for, discus-
sion of paper by, 128-130.

Black, experiments of, on latent heat
of steam, 18.

Blast furnace, chemical action in,
133 ; fuel, minimum for smelting
in, 134 ; limit of economy in, 134 ;
opinions of iron-masters regard-
ing, 134, 135.

Blast furnace gas, application of, 182.

Blast of high temperature, 132 ;
duties to be performed by, 134 ;
question considered, 133 : reduc-
tion in quantity of required, 135 ;
temperature of, 135.

H H 2

468

L\DEX TO VOLUME I.

BLOOMING FUBNACES.

Blooming furnaces, application of
regenerative gas furnace to, 93.

Boiler of Watt's engine, 51.

Boilers, application of Landore mild
steel to, 170 ; bursting steel, im-
possibility of, 173 ; encrustation,
how formed in, 136 ; iron riveting
injurious in steel, 171 ; riveting
plates of steel in, mode of, 170.

Boiler feeding, 78, 80.

Boulton and Watt's condensing
engine compared with expansive
air-engine, 38.

Boyd, W., steam-boilers, experiments
relative to, discussion of paper by,
170-173 ; steel, failure of in
boilers, experiments on, 170.

Brix, experiments on quantity of
heat in steam, 19.

Bunsen on dissociation, reference to,
204.

Bye-products from illuminating gas,
186.

CALOEIC engine, 28 ; theoretical
consumption of, 28.

Carbonic oxide and carbonic acid,
relative heating powers of, 133.

Carnot on heat, reference to, 32.

Carnot's theory, 59.

Cavendish Society, publication of
Eegnault's experiments by, 20.

Chance's glass-works, application of
regenerative gas-furnace at, 89.

Chemical energy, production and
application of. 177.

Cheverton, B., heated air, use of as
motive power, discussion of paper
by, 26-28.

Chronometric governor. See Gover-
nor, Chronometric.

Chubb, C.J., coal-getting machinery,
discussion of paper by, 128-130.

Circuit system of pneumatic trans-
mission, 147 ; advantages of, 155.

Clapeyron on heat, reference to, 32.

COMBINED VAPOTJE ENGINE.
Clansius on conversion of heat into
mechanical effect, 59 ; on steam
condensation, 33, 34.
Clerk, D., gas-engine, theory of. dis-
cussion of paper by, 203-206.
Cleveland ironstone, reduction of,

in blast-furnace, 134.
Clock regulated by liquid in rota-
tion, 115 ; action of, 116 ; descrip-
tion of. 116 ; formula to determine
velocity of cup with continuous
overflow, 118 ; table of comparison
of with chronometer, 119.
Coal, casualties in getting, 128 ; cost
of per horse-power per annum, 52 ;
economy in use of. 52 ; expendi-
ture and possible saving per annum
of steam, 52 ; getting, force ex-
pended in, 128 ; machines for
working and breaking down, 128 ;
wedging and blasting, 128.
Coke, advantage of, as fuel, 181 ;
making treated too lightly,
101.

Coke ovens, 99 ; Appold, 100 ; con-
version of ordinary into close, 99 ;
for production of heating and il-
luminating gases, 101 ; primitive,
99 ; quality of coke in, 99 ; regene-
rative, 100 ; semi-cylindrical hori-
zontal, 100.

Coleman, J. J., air refrigerating
machinery and its applications
discussion of paper by, 192-203.
Combined steam, 76 ; advantage

of, 77

Combined steam and ether engine,
45 ; additional effect of, how
otherwise attainable, 45 ; descrip-
tion of, 45 ; practical difficulties
of, 45.

Combined vapour engine, 69 ; advan-
tages of. 71 ; causes influencing
result in, 70 ; comparison of. with
expansive steam-engine, 70, 71, 72 ;
disadvantages of, 71 ; extra power
of, obtained by expansive work-

1XDEX TO VOLUME 1.

469

COMBUSTION.

ing, 72 ; theoretical considerations
of, 70.

Combustion, complete, 204 ; partial,
•J"l : [M.-rfect in regenerative gas-
furnace, 96.

Combustion of vegetable substances,
168 ; continuous, systematic and
equal feeding in Head's, J., sys-
tem, 169 ; Qarrett's, F., system,
169.

Combustion, waste products of, used
in, regenerative gas furnace, 81.

Condensation, depends on, 3 ; due to
development of power, 74 ; how
effected, 3 ; perfect, 3 ; prevented
by mixing air with steam, 138 ;
surface, 69 ; Dr. Ure's experiment
on, 14.

Condenser, advantages of effective
surface, 12 ; Hall's, 12 ; historic
sketch of, 11 ; Hornblower's sur-
face, 12 ; for land and other steam-
engines, 1 ; loss of mechanical
effect in, 52; (Me Carter, 167,
compared with ordinary, 167 ;
compensating, advantage of, 167 ;
worked on Newcomen principle,
167) ; Newcomen's, 11 ; object of,
3 ; regenerative, see Regenerative
condenser ; surface, 12 ; water re-
quired for, 7 ; Watt's, 12, 51.

Condensing water, required for
ordinary condenser, 7 ; required for
regenerative condenser, 2, 6 ; tem-
perature of, 4.

Conical pendulum, principle of,
107.

Contre-vapeur system, Le Chatelier's,
165.

Convertibility of physical forces.
See Physical forces.

Cooke, C. W., Wenham's heated air-
engine, discussion of paper by,
161-164.

Copper, great heat-conducting power
of, 14 ; thick pan of evaporates
more quickly than thin, 14,

EJECTOR-CONDENSER.

Cornish engine, action of, 68 ; reason
of economy of, 68 ; superheated
steam used in, 24.

Cowper, E. A., expansive steam-
engines, experiments on, 24 ; hori-
zontal pumping-engines, descrip-
tion of, discussion of paper by, 68,
69 ; hot-blast stoves, application
of regenerative system to, 94 ;
regenerative hot-blast furnaces,
recent improvements in, discus-
sion of paper by, 132-135 ; steam,
suggestions as to experiments on,
18.

Cylinder, steam, of Watt's engine,
51 ; uncovered, loss by condensa-
tion in, 138.

DA.LTON on expansion, 21.

Davy, heat, material theory of, dis-
proved by, 30.

Dean, experiments to burst steam-
boilers, 173.

Desprez, quantity of heat in steam,
19.

Diagram of gas production in gas-
retort, 185 ; theoretical of Erics-
son's, 42 ; and of Stirling's engine,
40.

Disassociation, 179, 204 ; Bunsen on,
179, 204 ; Ste.-Claire Deville on,
179, 204 ; temperature of, 204.

Dulong. See Expansion of air.

Dynamical curve, 36.

Dynamical theory. See Heat.

EATON,R.,Warsopaero:steam engine,
discussion of paper on, 136-139.

Ebelmen's, blast-furnace gases, re-
searches on, 182.

Efficiency of mechanical . con-
trivances, how adjudged, 133.

Ejector - Condenser, 157 ; annular
water- jet in, 158 ; increasing con-
densing surface of, 157, 158.

470

INDEX TO VOLUME I.

ELASTIC MEDIUM.

Elastic medium in perfect heat
engine, theoretically and practi-
cally considered, 59, 60.

Electric arc for high temperatures,
181.

Ellissen. See Gas retorts.

Engines. See Air, Caloric, Combined
vapour, Cornish, Gas, Heat, Per-
fect, Pumping, Kegenerative
steam, Steam, Steam-steering.

England, J., Giffard injector, dis-
cussion of paper by, 101-106.

Ericsson. See Air-engine, Regenera-
tor.

Ether-engine, advantages and dis-
advantages of, 70 ; and steam-
engine compared, 70.

Evaporation of sugar, 150 ; high
temperature in, ill effects of, 151 ;
steam- jet exhauster applied to,
151 ; vacuum pan for, 151.

Expansion of air and gases by heat,
21 ; curve of, 22, 162 (Determina-
tions by Dalton, 21 ; Dulong and
Petit, 21 ; Gay Lussac, 21) ; Sie-
mens's, C. W., views regarding, 22.

Expansion of steam. See Steam.

Expansive steam-engine. See Steam-
engine.

Expansive valve-gear, Joy's, theoreti-
cally and practically superior to
old link motion, 174.

FARADAY'S gas-burners, reference to,
190.

Fire-syringe for converting force
into heat, 54.

Flame, regulation of in regenerative
gas furnace", 96.

Flint-glass melting in closed pots in
regenerative gas furnace, 92.

Formula for liquid in rotation, 111,
112 ; for power of heat-engine, 37.

Foucault, governor of, 110.

Fox-Henderson, regenerative steam-
engine manufactured by, 16.

GAS-PEODUCEE.

Franklin Institute, experiments on
iron of, 58.

Freezing, operation of, differs essen-
tially from power producing, 195,
197.

Frost on expansion of isolated
steam, 21.

Fuel. Sec Coke. Gaseous fuel, Liquid
fuel, Peat, Regenerative furnace,
Regenerative gas furnace.

Fuel, evaporative power of, Rankine
on, 126 ; heat energy in, 178 ;
saving of, by regenerative gas fur-
nace, 65 ; wasted, 62.

Furnace. See Regenerative furnace,
Regenerative gas furnace.

GAIGNAED DE LA TOUE'S state of
vapour, 54.

Galton, Capt. D., steam-heating
for towns and villages, combina-
tion system of, discussion of paper
by, 190-192.

Garrett, F., on combustion of vege-
table substances, 169.

Gas-engines, 203 ; mechanical ar-
rangements of, 203 ; modern, 205 ;
Siemens's, of 1860, 205 ; smaller,
206 ; theory of, 203 ; three types
of, 204.

Gas fire. See Regenerative gas fire.

Gas from blast-furnace, application
of, 182.

Gas furnace. See Regenerative gas
furnace.

Gas, illuminating. See Illuminating
gas.

Gas lamp. See Regenerative gas
lamp.

Gas-producer, action of, 87 ; carbonic
acid changed into carbonic oxide
in, 87 ; description of, 86 ; evolu-
tion of hydro-carbons in, 8i
excess of pressure in, how effected,
88, 89 ; fuel used in, 86 ; gases
proceeding from, 87, 91, 95 ; hori-

JM)EX TO VOLLM1: I.

OA8-PBODUCEB8.

•ital cooling tube in, 89 ; pro-
i'i<.ii of sulphuring due to
chemical transformation in, 92 ;
(Sievunu'g, C. W., new form of,
• inscription of, 183 ; high temper-
ature in centre of mass of, 183 ;
modus opernndi, 1 83 ; water de-
composed to enrich gases from,
183) ; successful use of, 86 ; under-
ground use of, suggested, 95 ; uni-
form quality of gas from, 88 ;
water decomposed in, 87.

Gas-producers, 175; concentrating
heat in, 175 ; historical record of,
175 ; Liirmann's, 175 ; Price's
-imilarto Siemens's old form, of,
175 ; retort action too slow for,
177 ; Siemens's, C. W., patents of
1863 and 1864 for, 176 ; Siemens's,
86, 182, 183.

Gas-retorts, method of producing
.separate heating and illuminating
gas in, 185, 186 ; M. Ellissen's and
M. Regnault's experiments on,
185 ; variation of quality and
quantity of gas from at different
periods of operation, 185, 186.

Gasworks, regenerative gas furnace
applied at, 94.

Gaseous fuel, 177, 184 ; advantages
of, 178, 179 ; arrangement to pro-
ducefrom retorts with illuminating
gas, 185 ; for distant supply, 184 ;
heating power of, 178 ; prevention
of smoke by use of, 179 ; producers
for, 86, 182 ; reason of heat-energy
of, 178 ; retorts for, 181 ; Siemens's
new producer for, 183 ; Siemens's
producer for, 86, 182.

Gay Lussac, expansion of air, 21.

Gay Lussac and Clement, steam,
quantity of heat in, 19.

Giffard injector, 78, 101 ; action of,
102, 103 ; action of, how paralyzed,
103 ; advantageously used when
water has to be heated and raised,
106 ; (application as

GOVERNOR.

79 ; economy in, reason for want
of, 80 ; inferior to a pump in, 80) ;
boiler feeding, economical for, 80 ;
comparison of theory and practice
in, 79, 104, 105 ; entire condensa-
tion of steam jet necessary in,
103 ; failure, cause of partial, 105 ;
jets of water and steam, relative
areas of, 104 ; marine engines,
applicability to, 104 ; motor, in-
applicability of, as, 104, 106 ;
quantity of water which may be
forced into boiler by, 103 ; refer-
ence to, 143 ; and steam jet,
difference of action between, 144 ;
theoretical considerations regard-
ing, 102; water, increased tem-
perature of, passing through, 78,
103, with higher pressure of steam
in, 79.

Glass-melting in regenerative gas
furnace, 89, 90, 91.

Gorrie, Dr. See Refrigerating ma-
chinery.

Governor (Chronometric, 108; ap-
plication to astronomical and
chronographical instruments, 109 ;
cross-armed, good action of, 1 40 ;
instantaneous and regular action of,
109, 110; leading idea in and how
realized, 109 ; uniform rotation
obtained by, 109) ; Foucault's, 110;
(Gyrometric, description of, 120,

121 ; formula for speed of cup of,

122 ; rapidity of readjustment of,

123 ; uniform velocity of engine
governed by, 123) ; ( Watt's cen-
trifugal, 107 ; action of, 107 ;
defect of, 108 ; depends on per-
manent change of speed, 108 ;
improved suspension of, 108 ;
throttle valve, how worked by,
108).

Governor for steam engine, 139 ;
advantage of addition of spiral
spring to, 140 ; chronometric ac-
tion of, 139. &c Alleo.

472

INDEX TO VOLUME 7.

GRAY, J. MACFARLANE.

Gray, J. McFarlane, steam-steering
engine for Great Eastern, dis-
cussion of paper by, 124, 125.

Gyrometer, liquid, 112 ; action of,
113; applications of, 115; auto-
matic dip of cup in, 114 ; correc-
tion for lower orifice of cup of,
117 ; description of, 113 ; power
absorbed in, 114; range of power
in, 115 ; rotation of liquid in outer
vessel, how prevented, 113 ; uni-
form rotation produced by, 114 ;
velocity of, excess of calculated
over actual, 118 ; velocity of,
formula to compute, 118 ; vertical
position indifferent to, 120.

Gyrometric governor. See Governor,
gyrometric.

HALL, condenser of, 12.

Head, Jeremiah, steam-engine go-
vernor, with nearly perfect action,
discussion of paper by, 139-141.

Head, John, combustion of refuse
vegetable substances under steam-
boilers, discussion of paper by,
168-170.

Heat, Carnot on, 32 ; Clapeyron on,
32 ; consumption of, in expansion,
28; conversion of , into mechanical
effect, 29 ; curve of equal, 33 ;
(dynamical tlieary of, 30, 62 ;
Helmholtz, Mayer, Eankine,Thom-
son on, reference to, 32 ; practical
application of, 52 ; Kegnault's ex-
perimental proof of, 54) ; econo-
mical application of, 62 ; (engine,
formula to express power of, 37 ;
Holtzmanon, 32; relative economy
of, 38) ; identity of, and dynamical
effect, 53 ; lost, 62 ; (material
theory of, 29, 30 ; disproved by
Davy, Dulong, Joule, 30 ; how
explained, 30) ; scientific applica-
tion of, 62 ; specific, proportionate
to expansion, 34 ; theoretical views

JOY, D.

regarding, 29 ; undulatory theory
of, 29 ; (unit of, 31, 53 ; Joule's, 31,
74 ; results by different observers,
32 ; unaffected by material, 31).

Heating by steam. See Steam, Heat-
ing by.

Heating furnaces, application of
regenerators to, 93.

Helmholtz on heat, reference to, 32.

High pressure engines, reference to, 4.
See Eegenerative Condenser for.

Holtzman on heat, reference to, 32.

Hornblower. surface condenser of, 12.

Hot-blast stoves. See Kegenerative
hot-blast stoves.

ILLUMINATING gas, 181 ; first em-
ployed in 1792, 181 ; illuminating
power of increased by drawing
from bottom of retort, 100, 101 ;
value of bye-products from, 186.

Injection condenser. See Eegenera-
tive condenser.

Injector. See Giffard injector.

Interchanger of temperature, Sie-
mens's, 0. W., for refrigerating
machinery, 193, 194, 201.

Iron, temperature of maximum
strength of, 58.

Irvine, latent heat, experiments on,
18.

Isolated steam. See Steam.

JAMESON, J. W., combined vapour-
engine and expansive engine, dis-
cussion of paper by, 69-72.

Jet. See Steam jet.

Jones's coal-getting machine, 129.

Joule, material theory of heat dis-
proved by, 30.

Joule's unit of heat, 31, 74 ; views
regarding elastic pressure, 31.

Joy, D., reversing and expansive
valve-gear, discussion of paper by,
173, 174.

INDEX TO VOU 'Ml-. /.

473

KILNS.

va, brick and pottery, applica-
f regenerative gas furnace to,

VI.

!i, F. W., Allen governor and
thiuttlc-viilve for steam-engines,
discussion of paper by, 159-161.

LAMPADIUS, gas investigations of,
181.

atelier, his method of freeing
steam from water, 80 ; invention
of Contre- vapour system, 165;
reference to labours of, 165.

Link motion doomed, 174.

Liquid fuel, 125 ; arguments in
favour of, 126 ; combustion per-
fect with, 131 ; distilled from coal,
127 ; evaporative power of, 126 ;
grates for, 131 ; practical and
theoretical results with, 125 ; price
of, 127 ; progress of, 130 ; sources
of supply of , 131; superiority of ,
to solid, 131.

Liquid gyrometer. See Gyrometer,
liquid.

Liquid in rotation, 111 ; curve
formed independent of specific
gravity, 111 ; equations of, 112 ;
formula for curve formed by, 111 ;
paraboloid produced by, 111.

Lloyd and Summerfield's application
of regenerative gas-furnace, 175-
177.

Low-pressure engines. See Regene-
rative condenser for.

Lurmann, gas-generating furnaces,
discussion of paper by, 175-177 ;
gas-producers criticised, 175, 177.

MAKSHALL, W. P., suggestions as to

experiments on steam, 18.
Material theory. Sec Heat.
Mayer on heat, reference to, 32.
McCarter condenser, criticism of, 167.
Meadows, J. McC., peat fuel ma-

chincry, discussion of paper by,
ir,r, in;.

Mil<l steel. N^v Steel, mild.

Morin, Gen., experiments with re-
generative steam-engine, 162.

Morton, A., ejector condenser for
steam-engines dispensing with
air-pump, discussion of paper by,
157, 158.

Motion, heat manifestation of, 30.

Murdoch, J., illuminating gas pro-
duced by, in 1792, 181.

NEWALL & Co.'s trial of regenera-
tive steam-engine, 61.

Newcomen's condenser, 11, 167.

Newton. See Refrigerating ma-
chinery.

Non-condensing steam-engines, 4.

OPEN fireplace, method of heating
by, 191 ; regenerator applied to,
188 ; ventilation by, 191.

PAMBOUR'S expansion curve of
steam, based on Watt's law, 24.

Parker, W., experiments to burst
steel boiler, 173.

Paul, B. H., liquid fuel, discussion
of paper by, 125-127.

Peat, application of, 166 ; charcoal
from, 166 ; conversion of into gas,
166 ; drying of, 166 ; fuel, import-
ance of, 167 ; use of, in locomo-
tives, 166.

Perfect engine, 47, 54, 162 ; and
actual engine compared, 38 ; cha-
racteristics of, 46, 54 ; cylinder, ca-
pacity of working for, 55; difficulty,
practical, of, 71 ; expansion, infi-
nite in, 47, 55 ; impractical nature
of, reason of, 55 ; losses of heat in,
reduction of, 47 ; regenerator, ap-
plication of, to, 47 ; theoretical

474

JNDEX TO VOLUME I.

PERKINS.

minimum consumption of fuel in,
163.

Perkins, steam-guns, experiments
on, 35.

Pernplet, coal distillation, coke
manufacture, utilization of bye
products, discussion of paper by,
99-101.

Petit, expansion of air, 21.

Physical forces, conservation of, 53

. (convertibility of, 53 ; illustrations
of, viz., falling weight, 53 ; ham-
mer striking anvil, 53; fire syringe,
54).

Plate-glass melting, application of
regenerative gas-furnace to, 89, 90.

Pneumatic dispatch, circuit system,
Siemens's, 146 ; coal consumption
in, 147 ; speed of carriers in, 147 ;
vacuum in, 147

Pneumatic despatch tubes, details
of, 153. See Steam-jet, Siemens's.

Pole, W., Cornish engine, treatise
on, rei'erence to, 34 ; expansive
steam-engine, experiments on, 34.

Power of heat-engine, formula for,
37.

Preston, F., McCarter condenser for
steam-engines without air-pumps,
discussion of paper by, 167-168.

Price's gas-producers, 175.

Priming in boilers prevented, by in-
jecting hot air into water, 136.

Producers. See Gas-producers.

Puddling. See Kegenerative furnace,
Regenerative gas furnace.

Pumping, 68 ; Cornish engine, eco-
nomy of, for high lifts in, 68 ;
crank and fly-wheel, employment
of, for low lifts in, 68.

Pumping-engine, theoretical effect
of, 80.

Punch, helical, 172.

Punching, 171 ; experiments on, by
Kiley, J., 172 ; by Sharp, H., 172 ;
tension in metal produced by, 172.

Pyrometer, copper or platinum ball

REGENERATIVE CONDENSER.

application of, 97 ; description of.
97 ; measuring high temperatures
by, 97.

QUANTITY of heat, 18 ; in steam, 18.

RAISING water. See Steam-jet,
Siemens's.

Rankine, dynamical theory of heat,
32 ; fuel, evaporative power of,
126 ; steam, expansion of, 33 ;
steam-jet, investigations on, 144.

Refrigerating machinery, 192 ;
power-producing machinery and,
essential difference between, 195 ;
Siemens's interchange!' of temper-
ature for, 193, 194, 201, cost of
ditto, 202 ; (Dr. Gorriv's <»• ~
ton's, 193 ; action of, 194 ; concep-
tion of, error in, 194 ; description
of, 195 ; engine power, expenditure
of, examination of and report on,
195 ; expanding apparatus in, in-
quiry into cause operating against
performance of, 199 ; experiments
made with, 196 ; theory of, 197 ;
trial of, 193).

Regenerative coke-ovens, 100 ; no
secondary products with, 101.

Regenerative condenser, action of,
1, 3, 6 ; addition of, to injector
condenser, 5 ; applications of, 2, 8,
16 ; (for High-pressure engines, 4;
action of, 5 ; advantages of, 7 ;
applications of, 8, 16 ; description
of, 4 ; dimensions of, 8 ; indicator
diagram of, 6 ; saving of fuel with,
8 ; ) (Injection, 15 ; present form
of, 15 ;) for locomotives, 9, 16, 17 ;
(for Low-pressure engines, 11 ;
advantages of, 11 ; description of,
11 ); Society of Arts' Gold Medal
awarded to, 2 ; (Surface'. 13 ;
description of, 13, 15 ; essential
features of, 13, 15 ; rationale of,
13) ; water required for, 2, 6.

INDEX TO VOLUME /.

475

KKUENRRATIVE FURNACE.

nerative furnace, 62 ; action of,
til ; alternating principle of, 66,
applicable for intense heat,
\<-l. i.l, 66 ; brickwork of, durability
of, 65 ; description of, 63 ; diffi-
culty in connection with, 83 ;
economy greatest in, 67 ; gaseous
fuel, first use of, in, 84 ; heat in,
accumulation of, 64 ; invention of
I. Siemens, 63; smokeless, 65
(puddling iron with, 65; advan-
tage of, (56 ; experience limited
with, 67 ; heat retained long in,
67 ; quality of iron improved in,
65 ; temperature high in, 67 ;
time employed in, 66 ; yield large
in, 66).

Regenerative gas fire for ordinary
grates, 188 ; cost of, 189 ; descrip-
tion of, 188.

Regenerative gas furnace, 81 ; advan-
tages of, 93 ; applications of, 81 ;
(application to puddling and

. welding iron, 92, 96 ; plate-glass
melting, 89 ; round flint glass
furnace, 91 ; steel melting, 94) ;
cost of construction of, 98 ;
essential features of, 95 ; flame,
regulation of, in, 96 ; fuel em-
ployed in, 81 ; gas and air heated
by waste products of combustion

. in, 81 ; gas-producer, separate, for,
85, 86 ; gasification of solid fuel
for use in, 82, 85 ; heat, effect
produced by, 81 ; heat, trans-
fer of, in, 82 ; pots, open, used
in, 92 ; principle of, 82 ; siege
kept cool in, 89 ; temperature
of regenerators of, 91, 97.

Regenerative gas lamp, description
of, 189 ; intense light of, 189.

Regenerative hot-blast stoves, 132 ;
differ from ordinary hot-blast
stoves, 132 ; reason for economy
of, 132 ; save heat in blast fur-
nace, 134 ; temperature, high, at-
tainable with, 132.

ROBINSON, J.

Regenerative principle, 02 ; universal
applicability of, 63.

Regenerative steam engine,
applications of, 61 ; cylinder, re-
generative, description and object
of, 55 ; essential parts of, viz., fur-
nace and working and regenera-
tive cylinders, 55 ; essentials of
heating vessel in, 58; respirator,
use of, in, 56 ; similarity to Stir-
ling's and Ericsson's engines only
apparent, 59 ; working of, 55, 56.

Regenerator, 27, 47 ; action of, 49 ;
conflicting opinions regarding, 42,
48 ; discovery of principle by Stir-
ling, 39, 57, 82 ; employment of
for high temperatures suggested by
F. Siemens, 82 ; Ericsson's views
regarding, 27 ; experimental in-
vestigations by Siemens, C. W., on
action of, 47 ; heating surface in-
sufficient in Ericsson's, 49 ; sus-
picions regarding, 49, 57 ; useful
application of, 49.

Regenerators, composed of, 83, 84 ;
reversing valves in, 84, 93 ; simpli-
city and permanency of, 63.

Regnault, disapproval of Watt's law
by, 17, 19 ; (experiments of, 32 ;
Cavendish Society published, 20) ;
quantity of heat in steam, 19. See
Dynamical theory, Gas retorts,
Watt's law.

Respirator, 48, 57. And see Re-
generator.

Retorts. See Gas retorts, Gaseous
fuel.

Riley, J., on punching, experiments
by, 172.

Riveting. See Boilers.

Robinson, J., Giffard injector for
feeding steam-boilers, discussion
of paper by, 78-79 ; modern loco-
motives, designed with a view to
economy, durability, and facility
of repair, discussion of paper by,
165.

476

INDEX TO VOLUME I.

KtTMFORD.

Rumford, steam, quantity of heat in,
19.

Euthven, re-action propeller, refer-
ence to, 124.

Ryder, J. N., superheated steam,
application of, discussion of paper
by, 72-75.

ST. HELEN'S Glass Works, applica-
tion of regenerative gas furnace
at, 89.

Saltley Works, experiments on re-
generative condenser at, 16.

Saturated Steam. See Steam, satu-
rated.

Selwyn, Captain J., liquid fuel, pro-
gress of, discussion of paper by,
130-131.

Sharp, H., punching experiments
by, 172.

Siemens, C. W. See Expansion of
air, Gas-engines, Gas-producers,
Gaseous fuel, ( Governors, Chrono-
metric, Gyrometric), Gyrometer ;
papers by, 1-2, 3-17, 17-26,29-50,
50-61, 62-65, 81-95, 107-123, 141-
153, 177-190; Refrigerating ma-
chinery (Regenerative Condenser,
furnace, gas fire, gas furnace, gas
lamp, steam engine), Regenerator,
Steam, Steam-jet ; suggested coal-
getting machine, 129.

Siemens, F., regenerative furnace,
invention of, 63.

Southern, latent heat of steam, 19.

Society of Arts' Gold Medal awarded
to regenerative condenser, 2.

Specific heat of elastic fluid pro-
portionate to rate of expansion by
heat, 34.

Speeds, alternation of, affecting
steam governors, 140.

Steam, ancient knowledge regarding,
50 ; combined, 76 ; conducting
power of, 61 ; expanded is super-
heated, 24 ; (expansion of behind

STEAM ENGINE.

piston, causes condensation, 33,
34 ; Clausius, Rankine, referred to,
33) ; (expansion of isolated, 21 ;
Frost's experiments on, 21 ; Sie-
mens's apparatus for determining,
21 ; and law of, 21 ; and curve of,
22 ; and experiments on, 22 ; and
table of, 25) ; (expansion of satu-
rated. 23 ; agrees with that of air.
23 ; curve of, 33 ; variation of
pressure arid density of, 23) ;
heat of, increased with density,
60 ; (heating by, 190 ; closeness of
atmosphere with, 192 ; criticism
of system of, 191 ; gas preferable
to, 191 ; large pipes necessary for,
191 ; objections to, 191) ; higl
pressure, cooling effect of expan-
sion of, how explained, 24 ; high
pressure contains excess of heat
which superheats expanded, 20,
24 ; latent heat of, observed by
Black, 18 ; mixed with air, 136 ;
quantity of heat in, 18, 19 ; pre-
ferable to other elastic mediums
for caloric engines, reasons why,
60 ; rise of temperature with pres-
sure of, 78 ; saturated, great capa-
city for heat of, 76 ; specific gravity
of, 60 ; (superheated, advantages
attained with, viz., complete evai-
poration, 73, condensation pre-
vented, 73, 74, 77, and increased
bulk, 73 ; trial of, 72, 73 ; con-
siderations in experimenting with,
73 ; economy by use of, not great,
76 ; heat advantageously applied
in, 74 ; joints for, difficulty with,
and cement for, 75 ; regenerative
steam, versus, 75 ; saving by use
of, 75) ; (total heat of, 17 ; Sie
mens's, C. W., apparatus and ex-
periments for determining, 20, 26 j
Watt's law for, 19).
Steam engine, air engine, comparison
and essential difference between,
39, 46, 138 ; application of, recent,

INDEX TO VOLUME /.

477

STEAM-ENGINE GOVERNOR.

50 ; efficiency of Watt condensing,
: (cj-pantive, 24 ; condensation,
.•ts not shewn in diagrams of,
;> t ; condensing, example of, and
unlimited horse-power of, 87 ;
whi'ii theoretically perfect, but
impracticable, 3f>) ; first suggests!
use of, 50 ; high pressure, loss in,
4 ; low-pressure condensing, dia-
gram of, 35 ; power of, depends
on, 29 ; principle of, laid down by
Watt, 50 ; ( Watt's, 50 ; improve-
ments, recent, of, 51 ; organic
parts of, viz., furnace, boiler,
cylinder, and condenser, 51).

Steam-engine governor. See Go-
vernor.

Steam jacket, G9 ; advantage of, 69 ;
condensation in cylinder prevented
by, 72; cylinder, sides and end
require, 69, 138 ; Watt's engine
supplied with, 72.

Steam jet, 141 ; air delivery by,
dependent on surface contact
between air and steam, 143, 154 ;
air delivery in inverse ratio to
weight of air acted on, 143 ; air
pressure, limit of, attainable with
given pressure of steam, 143 ;
elastic force changed into onward
motion in, 153 ; investigations re-
garding, by Rankine, 144, and
Zeuner, 144 ; parts, judicious ar-
rangement of, 141 ; principle of
action of, 143 ; simplicity of, 141.

Steam jet, Siemens's, 142 ; air de-
livered proportionate to surface
contact in, 154 ; annular jet of
steam in, 142, 143, 156 ; applica-
tions of (to evaporation of sugar,
150 ; costly apparatus saved by,
use of, in, 151 ; description of ar-
rangement of, 151 ; successful
experiments with, 152) ; (to Gas-
Producers, 152 ; advantages of,
153 ; description of, 152) ; (to
Pneumatic Dispatch Tubes, 146 ;

SUPERHEATED STEAM.

description of, 147 ; intercepting
apparatus for, 147, 148); (to
raining water, 148; apparatus,
description and action of, 149 ; lift
attainable with, 150 ; preliminary
results with, 150) ; compression
or exhaustion proportionate to
steam pressure in, 143, 154 ; con-
clusions derived from experiments
with, 143 ; delivery tube in, para-
bolic curve, best form of, 143, 155 ;
description of, 142 ; details of,
considered, 156 ; exhauster exhi-
bited in action, 155 ; (experiments
with as blower, 146 ; as exhauster,
145) ; rationale of, 142 ; steam
orifice, increase of area of, with
increase of compression or vacuum,
157 ; and Zeuner's apparatus com-
pared, 144, 145.

Steam steering engine, 124 ; auxili-
ary propulsion, useful for, 124 ;
submarine telegraph ships, impor-
tant for, 125.

Steel boilers. See Boilers.

Steel melting. See Regenerative gas
furnace.

Steel, mild, application to boilers,
170 ; homogeneity of, 171 ; punch-
ing plates of, 172 ; tearing of, 171 ;
yielding property of, 170.

Stirling, regenerator or respirator,
invention of, 57.

Stirling's engine. Sec Air engine,
Stirling's.

Straw, analysis of, 169 ; ash in, 169 ;
heating power of, 169 ; water in,
169.

Ste.-Claire Deville, dissociation, re-
ference to, 204.

Submarine telegraph ships, difficulty
of manoeuvring, 125 ; steam steer-
ing apparatus for, 125.

Sugar, evaporation of. See Steam-
jet, Siemens's.

Superheated steam. See Steam, super-
heated.

478

INDEX TO VOLUME I.

SURFACE CONDENSER.

Surface condenser. See Condenser,
Condensation, Regenerative con-
denser.

TABLE of comparative merits of
steam and air engines, 46 ; (of
experiments on expansion of
isolated atmospheric steam, 25 ;
with steam jet as a blower, 146,
and as an exhauster, 145) ; of
quantity of heat in steam, 18, 19.

Temperature of combustion limited,
179, 203 ; of condensing water, 4 ;
of regenerative furnace, 67 ; of
regenerative gas furnace, 91, 96,
97, 180.

Thomson on heat, reference to, 32.

Total heat of steam. See Steam.

Tresca, experiments with regenera-
tive steam engine, 162.

UNIFORM rotation, 107 ; liquid in
rotation applicable to production
of, 112 ; obtained by Chronometric
Governor, 109.

Unit of heat. See Heat, Joule.

Ure's experiments on condensation,
14 ; on quantity of heat in steam,
19.

ZEUNER.

VALVE gear, Joy's reversing and
expansive, 173 ; best expansive.
174 ; perfect cut-off with, 174.

Vegetable substances, combustion
of, 168.

Vena contracta. for steam jet, 156.

WARDLE, C. W., Giffard injector,
elevator for colliery drainage
discussion of paper by, 79-81.

Water, absorption of heat, slow, by,
14 ; condensing, temperature of, 4
conductive power low of, 14 ;
specific, heat of, high, 14.

Water gas, reference to, 133.

Water raising. See Steam-jet, Sie
mens's.

Watt's condenser, 3, 12 ; governor
107, 108, 140 ; law of total heat of
steam, 19 ; steam engine, 50, 51 ;
steam jacket, 72.

Welding. See Regenerative gas
furnace.

Wenham's heated air-engine, 161.

Wethered, J., combined steam, dis-
cussion of paper by, 76, 77.

ZEUNER'S experiments on steam-
jets, form of apparatus, 144.

INDEX TO VOLUM1 I.

479

METALLURGY.

ABSORPTION.

ABSORPTION of carbon by iron rules
••numl'or1' of pig-metal, 265; of
ln-at in blast furnace by trans-
fer of oxygen of ore to carbon of
coke, 262.

il consumption of fuel to smelt
;i ton of iron, 264.

A lamson, D., iron and mild steel,

mechanical and other properties

of, discussion of papers by, 373-

378, 409-412.

Adamson's test piece, objection to,

873.

Additional height of blast furnace.
256 ; consumption of fuel, reasons
why theoretical, not attainable by,
256, 257, 267 ; fails to produce
economy beyond neutral point,
257.
Advantages of open top to blast

furnace, 332.

Akerman, Prof., hardening iron and
steel ; its causes and effects, dis-
cussion of paper by, 433-435.
Akerman, Prof., rotary furnace,

report on, 342 ; on steel, 375.
Allen, W. D., Bessemer steel, use of
mechanical agitator in manu-
ture of, discussion of paper by,
450-452.

Ammonia used in making soda, 330.
Analysis of cinder from puddling
furnace, 241 ; of puddled bar,
illustrating separation of im-
purities in puddling, 244 ; of
puddling furnace contents, by
Messrs. Calvert and Johnson, 238.
Ancient metallurgical methods
wasteful of fuel and material, 330.
Ancient steel, hardness of, 379 ; high
quality of, 379 ; process, descrip-
tion of, 378, 379.

BASIC LINING.

Annealed steel reduced in strength,
427.

Annealing, irregular heating in,
447 ; overheating in, 447 ; process
criticised, 446 ; after punching,
371 ; of steel, 407 ; of steel cast-
ings, 465.

Annealing hard steel, necessity of,
421, 427.

Anthracite coke, 325 ; Appold ovens
for, 327 ; binding material for,
326 ; in blast furnace improves
quality of pig-iron, 332 ; height of
blast furnaces for, 326 ; in hori-
zontal coke ovens, 327 ; Landore
blast furnaces designed for. 326 ;
manufacture of, at Crensot, 326 ;
proportions of material in, 327 ;
sulphur disappears with use of,
327.

Architectural purposes, steel for.
404.

Armstrong system of ordnance, 398.

Attwood, C., attempt to make steel
on hearth of regenerative furnace,
388.

Averages, applicable to errors of
observation, 442 ; illustrations of,
442.

BALLING, iron prevented from, by
copper, 306, 331.

Barnaby, N., naval construction, use
of steel in, 409-412 ; shipbuild-
ing, iron and steel for, 313-315,
discussion of papers by.

Basic lining, transition from silica
lining to, in Bessemer, puddling
and open-hearth steel furnaces,
414.

480

INDEX TO VOLUME I.

BASIC PROCESS.
Basic process, 436 ; pig-metal suit
able for, 437 ; sulphur should b
removed in blast furnace for, 436
tapping slag in, 436.
Bauxite, analysis of, 217, 297, 298
converted into emery by intense
heat, 296 ; infu'sibility of, 296
Le Chatelier's proposal to use, 217
lining of calcined, and plumbago
powder, 296 ; resists fluid cinder
296 ; supply of, 304.
Beam of steel v. beam of wood, 405.
Bell, I. L., address of, discussion of

309-310.

Bell, I. L., blast furnaces of different
dimensions, development and ap-
propriation of heat in, 264-268 ;
blast furnaces, economy of fuel in,
conditions of, 276-279 ; Price's
retort furnace, 324-325 ; separa-
tion of impurities from iron in
different processes, 353-360 ; sepa-
ration of phosphorus from pig-
iron, 362-365, discussion of papers

ty.

Bell, I. L., on blast furnace, 264 ;
diagram of, on distribution of
temperature in blast furnace,
290; discussion with Siemens,C.W.,
on blast furnace, 358, 359 ; pro-
cess of dephosphorization, 362 ;
standard paper on blast furnace,
265 ; views on hot blast for and
capacity of blast furnace, 276.

Bending test for steel, maintenance
of, 420.

Berkley, G., iron and steel, strength
of, &c., discussion of paper by,
269-271.

Berrier-Fontaine, M., shipbuilding,
steel for, discussion of paper by,
448-450.

Bessemer, H., reference to, 375 ; re-
ference to remarks by, 305.

Bessemer converter and open-hearth
steel furnace compared, 375.

Bessemer medal, presentation of

BLAST FURNACE,
to Prof, von Tunner, 360 ; re-
marks on occasion of, 360.
Bessemer steel, early, 411 ; for gene-
ral purposes, 376 ; pouring of,
ebullition in, 322.

Bessemer steel process, 215, 438 ;
application extended, of steel due
to, 215 ; cheap steel produced by,
215, 380 ; difference between ore
process and, 287 ; difficulty of re-
moving sulphur and silicon in, 436 ;
f erro-mangancse or spiegeleisen in,
212 ; high quality of steel not always
producible with, 411 ; high tem-
perature of fusion produced by
oxidation of iron in, 286 ; im-
purities concentrated in dimin-
ished weight in, 376, 411 ; in-
dustries revolutionized by, 381 ;
first introduction of. 380 ; Mushet's
addition of spiegeleisen to, 212,
380 ; occlusion -of oxygen in metal
made by, 287 ; oxidation of carbon
and silicon in, slight oxidation of
manganese in, non-oxidation of
phosphorus and sulphur in, 286 ;
results of, 381.
Birmingham Sample Steel Works,

advantage of, 218.
Blair's process, criticism of, 319 ;

spongy iron by, 365.
Blast furnace, additional height of,
256, 257, 266; Bell's, I.L.,a standard
paper on, 265 ; boshes of, increased
temperature at, 267 ; capacity of,
considerations determining, 274,
275 ; capital expenditure v. other
means of saving in, 275 ; charging
hopper of, advantage and economy
by closing, 256 ; equality of heat
brought in by hot blast, and car-
ried away by escaping gases from,
263, 264 ; equality of ore charged
and smelted in, 257 ; amount of
gases, CO and C02, passing from,
260 ; gases issue at reduced tem-
perature from, with increased tem-

INDEX TO VOLUME I.

481

BLISTER STEEL.

p«T:iturc of hot blast, 267 ; teraper-
of, unaffected by tempera-

tun- of hot blast, 268.
lUi>t.-r steel, 328.
Board of Trade limit of strength for

mild steel, 375, 408.
Boilers, careful inspection of, 462 ;

mild steel for, 367 ; steel not

elongating 20% in 8" unfit for, 4(!3 ;

IrM ing of, 462 ; thickness of plates

for, excessive, 463 ; Trade, Board

of, rules for, 462.
Bridges, American suspension, made

of steel, 404.
Brittleness of hard steel rails caused

by cold, 272.

CALCULATION of heat necessary in
blast furnace to melt cinder, 261 ;
to melt iron, 261.

Calculation of iron oxide required in
puddling, 243.

Calculation of metal produced from
cinder in puddling furnace, 241,
242.

Capacity of blast furnace, 291 ;
considerations determining, 274,
275 ; depends on nature of coke
and ore, 274 ; economical heating
due to large, 275 ; gaseous flow
slow with large, 274 ; time for
chemical reaction due to large,
275 ; work done in relation to, 275.

Capital expenditure ». other means of
saving in blast furnace, 275.

Carbon, chemical tests of, in steel,
340 ; chemically combined in hard
steel, 339 ; combination of, in
steel, mode of, 339 ; mechanically
mixed in soft steel, 339; phosphorus
and sulphur in puddling furnace,
order of removal of, 238 ; in
puddling removed with ebullition,
and evolution of carbonic oxide,
£39 ; in steel, percentage of, 210 ;

VOL. I.

CHBXOT'H PROCESS.
table of, in cast and wrought iron
and steel, 253.

Carbonic oxide, absorbed by fluid
steel, 322 ; heat rendered latent in
formation of, 262.

Carburization of steel, degree of,
419.

Caron on steel, 211.

Casson Dormoy puddling furnace,
IM3; circular puddling chamber in,
334 ; cooling puddling chamber
of, mode of, 334 ; dandy in, use
of, 334 ; experiments with, 333 ;
fettling in, 334 ; time required to
bring iron to nature in, 334.

Cast iron, danger of, for buildings,
406 ; tables of carbon and silicon
in, 253.

Cast steel, manufacture of, 209 ; by
Sheffield process, 211.

Casting, moulds divided longitudi-
nally for, 336 ; moulds filled from
below for, 336 ; oxidation of metal
in, 335 ; running metal through
refractory material in, 336 ; Sie-
mens-Martin process, arrange-
ments for, in, 335.

Catalan forge, 214.

Cementation steel process, 214, 328.

Chalk, a thorough cinder, 310 ; re-
sult of double combustion of cal-
cium and carbon, 310.

Charging hopper of blast furnace,
advantages and economy by
closing, 256.

Cheap constructive material, steel
as, 405.

Cheapness injuring steel, 432.

Chemical admixture, perfect and
imperfect, 451 ; analysis of steel,
importance of, 424 ; problems in
puddling, 238 ; work required to
smelt ironstone, calculation of
amount of, 260.

Chemistry, value of, 209.

Chenot's process of manufacturing
spongy iron, 365.

I I

482

INDEX TO VOLUME I.

CHILLING.

Chilling effects similar to those of
rolling and hammering, 434.

Cinder, in blast furnace, heat neces-
sary to melt, 261 ; from puddling
furnace, analysis of, 241 ; in
puddling furnace protects metal,
243.

Clay's experiments with spongy
metal, 292, 365.

Cleveland pig, difficulty of elimina-
ting phosphorus in puddling,
283 ; unsuitable for steel-making,
282.

Coal pits, gas-producer at bottom of,
224.

Cochrane, C., blast furnaces, waste
gas from, and increased capacity
of, discussion of paper by, 255-
264.

Coke in blast furnace produces only
Jrd heat-producing power, 290.

Cold affects steel, 364, 428.

Cold-shortness in steel, 212, 288.

Columns of steel, 407.

Combining oxygen of ore with carbon
of fuel, calculation of heat re-
quired for, 262.

Combustion, in blast furnace of fuel
into | CO and \ C02, 260, 262 ; in
blast furnace and hot-blast stove
compared, 279, 290 ; hindered by
dissociation, 384 ; theoretical and
practical limit to temperature of,
384.

Compressed metal, zone of in punch-
ing steel, 370.

Compressed steel, Whitworth's fluid,
320, 321, 429.

Compression of steel, 429 ; by gas
pressure, 429 ; honeycombed struc-
ture reduced or removed by, 429,
430 ; by Whitworth's process,
429.

Conductivity of metals affected by
foreign ingredients, 330.

Conflagration, ordinary temperature
of, below that of fused steel, 40G.

DAVIS, A.

Consumption of fuel in blast fur-
nace, diminution of by hot blast,
291 ; in blast furnace, regarding,
256, 257, 267 ; to smelt a ton of
iron, 263, 264 ; in steel production,
302.

Cooling steel gradually in ingot, 450.

Cooling tube for gas-producer, 222,
245, 386.

Copper, blast furnace yield increased
by, 331 ; contaminated by silver,
330 ; eliminated by direct process,
305 ; iron prevented from balling
by intermixture of, 306, 331.

Cornish boilers, economy due to
large heating surface of, 274.

Corrosion, 449 ; Admiralty early ex-
periments 011, 441 ; cinder in iron
affects, 443 ; diminished by re-
moving oxide, 443 ; experiments
on, at Landore, 443 ; galvanic
action affects, 441 ; recent ex-
periments on, 441 ; relative, of
iron and steel, 440 ; scale in steel
affects, 443 ; variable opinions re-
garding, 441.

Corrosion of steel, 374 ; compared
with that of iron, 440 ; due to
manganese, 365, 368, 392 ; general
experience on, 444 ; by riveting
steel plates with iron rivets, 449.

Creusot, manufacture of anthracite
coke at, 326.

Crucibles, fusion of steel in, 229 ; in
regenerative gas furnace, 229,
381.

DAMASCUS steel, 379.

Danks's rotary puddling furnace,
difficulty of puddling overcome
in, 280 ; lined with metallic oxides,
281 ; inapplicable to Siemens-
Martin process, 281 ; yield in,
283.

Davis, A., steel-compression, dis-
cussion of paper by, 429-433. .

INDEX TO VOLUME I.

483

DEAD MELTING.

Dead melting, cuused by oxygen in
iron going to carbon, 337 ; caused
mining steel in fluid condition
at comparatively low temperature,
illustration of, from com-
pleting Bessemer blow, 33s ;
Sheffield practice for, 337 ; Snelus's
view of, criticism of, 337.
Decarburization steel process, 211.
Di-1'mition of steel, criticisms of, 31(1,
SI 7 ; Hackney's, as malleable
fused iron, 316 ; Percy's, as iron
capable of hardening, 31 U ; Sic-
mens's, C. W.,as iron compounded
to possess superior strength, 316,
:5-J3 ; want of good, 402 ; Whit-
worth's, by its strength and tough-
ness, 317.
Denny, W., shipbuilding, steel for,

discussion of paper by, 42(i- 1:>S.
Dephosphorization, 413 ; bauxite
bricks for, 413 ; Bell's, I. L., pro-
cess for, 362 ; caused by basic
material added to the bath, 415 ;
of Cleveland pig, by rich iron
oxide, 363 ; cost of, by Bell's, I. L.,
process, 3(53 ; depended on basic
cinder, high temperature, and
oxygen, 416 ; experiments by
Siemens, C. W., in 1863 on, 413 ;
ferrous oxide used in, 415 ; iron
destroyed in, 415 ; Le Chatelier's
bauxite lining for, 413 ; lime
lining for, 363 ; magnesia lining
for, 413 ; in open-hearth process
by addition of silicon-iron and
spiegeleisen, 417 ; in rotator, 362 ;
by tapping off first slag before
balling in rotator, 418 ; Thomas
and Gilchrist's paper on their pro-
cess of, 413 ; Williamson, Prof.,
suggests use of other bases than
iron oxides for, 363.
Die should be larger than punch in

punching mild steel, 371.
Direct process, advantages of, com-
bined with pure and intense heat,

ran.

341 ; copper eliminated by, 305;
for iron and steel manufacture,
303, 340 ; practical ; objections to,
viz. rich ores, pure fuel, and much
labour, 341 ; pure iron by, 341 ;
for reducing iron by forcing car-
bonic oxide into it, 303 ; simplicity
of, 341.

Direct steel process, 289.

Direct wrought-iron process, 289.

Dirt, Palmerston's definition of, 330.

Dissociation hinders combustion. 384.

Distillation of coal in gas-producer,
221, 245, 385.

Drilling and punching compared,
410.

Ductility of steel, 459 ; of steel
castings, 466.

ECONOMICAL and wasteful puddling
compared, 283.

Economy of blast furnace affected
by use of hot blast, 276 ; of fuel
in regenerative gas furnace, 224,
225, 247.

Elasticity, calculating, experiments
for, 270 ; definite limit of, in
metals, 269 ; limit of, no elonga-
tion by strain within, 269, 270 ;
limit of, more important than
ultimate strength, 271, 423 ; of
metals not affected by time, 269 ;
of steel, 420, 459 ; of steel, experi-
ments wanting on influence of
temperature on, 424,

Elimination of carbon, manganese
and silicon in ore-reducing pro-
cess, 286 ; of sulphur and phos-
phorus in ore-reducing process
compared with Bessemer steel
process, 304.

Elongation tests must exclude drawn

portion, 434.

Engine working directly en rolls,
criticism of, 439.

I I 2

484

INDEX TO VOLUME I.

ENGINEERING "STRUCTURES.

Engineering structures, steel for,
375.

Equality of heat carried away from
blast furnace by escaping gases,
and brought in by hot blast, 263,
264.

Equality of ore charged and smelted
in blast furnace, 257.

Errors of observation, averages ap-
plicable to, 442.

Experiments on puddling by Price
and Nicholson, 238.

FACTOR of safety, calculated erro-
neously in terms of ultimate
strength, 271 ; ductility as affect-
ing, 464 ; for steel, 428, 459, 460.
Failure of Chenot, Clay and Kenton

to puddle spongy iron, 242.
Failure of steel, causes of, 409 ; due
to faults in treatment, 410 ; due
to silicon, 377.

Falling weight test of steel, 271.
Faraday described regenerative gas

furnace in 1862, 210.
Final action in blast furnace, 290.
Fluid steel, carbonic oxide absorbed

by, 322.

Forth bridge, steel for, 420.
Fremy on steel, 211.
Fuel, in blast furnace, should live
down to crucible, 327 ; available
for gas-producer, 219, 245.
Fuel consumption in blast furnace,
minimum 1\ cwt., error in cal-
culating, 256, 257 ; minimum, re-
capitulation of arguments regard-
ing, 259.

Fuel , theoretical minimum, in pro-
duction of steel, 302.
Fused steel and other steel, distinc-
tion between, 323.
Fusion of steel, importance of, 403.
Fusion of steel in crucibles. 229 ;
(in regenerative gas furnace, 229,
381 ; bed of coke-dust for, 229 ;

GASES.

saving and arrangement of cru-
cibles in, 229 ; reduction of cost
of, 229 ; melting chamber for,
229 ; saving of fuel in, 229.)
Fusion of steel processes, 215 ; in
crucibles, 229 ; Hindoo, recently
revived by Heath, Price & Nichol-
son, Gentle Brown, and Atwood,
216 ; Huntsman's, 215 ; open-
hearth, q. v. ; Reaumur in 1722
used, 215 ; regenerative open-
hearth, q. v. ; Uchatius's, 216.

GAS-producer, 219, 385 ; air, how
admitted to, 220, 245 ; aqueous
vapour, utilization of, in, 221 :
charging arrangement for, 219,
385 ; C02, how reduced to CO,
in, 220, 245, 385 : coal pits, bot-
tom of, for, 224 ; composition of
gases from, 219 ; cooling gas from,
advantage of, 223, 386 ; cooling
tube elevated, for, 222, 245, 386 ;
damper, sliding, for, 245 ; descrip-
tion of, 219, 245 ; distillation of
tar and hydrocarbons in, 221, 245,
385 ; any fuel available for vola-
tilization in, 219, 245 ; gas, low
heat of combustion of, 220 ; grate
of, 220 ; grouping of, and leading
gas from into main flue, 224, 246,
387 ; indraughts into, how pre-
vented, 221 ; losses in, considera-
tion of, 324 ; objections to, refuted,
223 ; plenum of pressure within,
how maintained, 222, 223, 245 ;
retorts for, 325, 454 ; steam de-
composed in, 221 ; surplus heat of,
how utilized, 221 ; syphon ar-
rangement of, 222, 246, 386 ; tem-
perature of gas rising from, 222,
385 ; total loss in, 12 per cent.,
324 ; waste heat of. how utilized,
221.

Gases set free in moulds in congela-
tion, 335.

INDEX TO VOLUME I.

485

GENTLE BROWN.

le I'-rown, revival of Hindoo
' process by, 216.

Gilchrist, P. O., phosphorus, elimina-
tiini of, discussion of paper by,

li:,.
rs of steel, 407.

Government tests of inventions with-
out patentee's co-operation an in-
terference with his rights, 457.

Graphite unconsumed in blast fur-
nace, 327.

Guns, effect of gunpowder on, 453 ;
internal chilling and subsequent
gradual cooling proposed to equa-
lise strain in, 400 ; steel better
than iron for, 453 ; use of steel
ensures lightness and strength in,
453 ; temperature, necessary ad-
justment of, for shrinking hoops
on, 453, 454.

HACKNEY'S definition of steel, 316.

Hackney.W., anthracite coke, manu-
facture of, 325-328 ; steel, manu-
facture of, 315-320, discussion of
papers by.

Hard coke comes out with slag from
blast furnace, 327.

Hard and mild steel equally strong
within limits, 395, 420, 423, 424.

Hard steel rails, brittleness of, from
cold, 272.

Hard steel rails used in France, 318.

Hard untempered steel, 435.

Hardness of steel, 313.

Heat, absorption of in blast furnace,
262 ; application of, 209 (of com-
bustion of coke in the blast fur-
nace, 290 ; of gas-producer, low,
220) ; evolution of, in relation to
volume, 261 ; generation of in-
tense, 209 ; rendered latent in for-
mation of carbonic oxide, 262 ; (to
melt cinder in the blast furnace,
calculation of, 261 ; iron in the
blast furnace, calculation of, 261 ;

IMPURITIES.

iron in cupolas, 261 ;) in smelting
operations, viz., melting iron, melt-
ing ore, and transferring oxygen of
ore to carbon of coke, 260, 261.

Heated steel, deflects before failing,
407.

Heaton and Hargreaves' steel pro-
cess, 215.

Heath, application of manganese due
to, 212, 380 ; Hindoo steel process,
revival of, by, 216 ; patents of, con-
tested, 212.

High class steel, purity of, 377,
378.

High pressure, multitubular boilers
unsuitable for, 372 (vessel to resist,
description of, 372 ; joints of, 372 ;
risk of explosion avoided in, 372 ;
system applicable to steam boilers,
373).

Honeycombed structure reduced by
compression, 429, 430.

Hot blast, 265 ; blast furnace econo-
my affected by, 276 ; blast fur-
nace gases affected by, 267 ;
stoves, combustion in, contrasted
with that in blast furnace, 279,
290 ; increases bulk of heat in
blast furnace, 267 ; increased tem-
perature of, improves economy of
blast furnace, 266, 267 ; increased
temperature of, causes reduction
in temperature of waste blast fur-
nace gases, 267.

Hot-blast stoves, regenerators for,
258.

Huntsman's steel process, 215.

Hydraulic arrangements in steel-
making, 439.

Hydraulic cylinder of compressed
steel, 439.

IMPURITIES, added to pig iron in
blast furnace, 290 (in steel, 211 ;
effect of, 338 ; experiments on,
difficulties of, 214).

486

INDEX TO VOLUME I.

INCREASED TEMPERATURE.

Increased temperature at boshes of
blast furnace, negative effect of,
267.

Indian Wootz made by steel fusion
process, 215.

Ingots of steel, non-homogeneity of,
339.

Intelligence required in using and
working steel, 314.

Iodine used in silver extraction, 330.

Ireland, J., iron sponge, recent im-
provements in manufacture of by
Blair process, discussion of paper
by, 365-366.

Iron, in blast furnace, heat necessary
to melt, 261 ; slow oxidation of,
in dry atmosphere, 277.

Iron ores, temperature of reduction
in blast furnace, high for peroxide,
magnetic and dense, 277 ; low for
hydrated and spathose, 277 ; not
high enough for Ilmenite and Mar-
bella, 278.

Iron oxide, required in puddling,
calculation of, 243.

KENNEDY, A. W. B., mild steel, dis-
cussion of paper on, 402-408.

LANDORE open hearth steel process.

388 ; bath, final adjustment of, in.

389 ; charging pig metal and scrap
in, 390 ; ferro-manganese used in,
391 ; furnace bed, Le Chatelier's
mixture for, 231 ; limestone added
to, 389 ; ore. advantages of, in,
391 ; oxidation with ebullition of
gas in, 389 ; preliminary opera-
tions in, 389 ; sampling bath in,
389 ; scrap, use in, advantages of,
391 ; slag from, 390 ; modification
of, for special quality of steel, 391;
stirring bath and tapping, 390 ;
temperature requisite for, 389 ;
time occupied in, 390 ; yield by,
390.

MANGANESE.

Larkin, H., steel direct from ore,
manufacture of, discussion of
paper by, 328-329.

Le Chatelier's proposal to use bauxite.
217.

Lefroy, Sir H., remarks by, regard-
ing Siemens, C.W.. 378.

Lester, J., Danks's rotary puddling
machine, discussion of paper by,
280-283.

Lime for furnace linings, 304 ; and
magnesia linings, Siemens, C. W.'s,
experiments on, 305 ; used in ore-
reducing process, 304.

Lime lining, dephosphorization by,
363 ; used in ore-reducing pro-
cess, 304.

Limestone added to open hearth
steel process, 411

Limit of elasticity in metals, 269,
270, 271, 423,424,459.

Lining, basic,usecl in open-hearth steel
process, 412 ; of calcined bauxite,
and plumbago powder, 296.

Livadia's steel plates, 444 ; burnt,
445 ; carbon varying in, 445 ;
chemical nature of, 445 ; irregular
composition of, 444. 445 ; silicon
in, 445.

Longridge, J., views on ordnance cri-
ticised, 399.

MACHINERY for steel-making, 437 ;
compound engine for, 438 ; econo-
my in steam engine for, necessity
of, 438 ; hydraulic arrangements
in, 439 ; quick-working reversing
engine best for, 439.

Magnesia lining, difficulty of repair-
ing, 413, 414.

Maitland, Col., British Ordnance,
metallurgy and manufacture of,
discussion of paper by, 452-458.

Manganese, action of, non-chemical,
288 ; in blast furnace removes
sulphur from ore and fuel, 288 ;

INDEX TO VOLUME I.

487

MANGANKSK IN STKICL. |

oxiilisos outof str, l. :u; I ; red-short-
ness prevent. •<! i.y. L':;J ; in Sie-
ii) "n- 1 n-. H •••-. :ifii< Hi of,.') 12 ; traces
only of, in LTHIM! siccl. .Si; I.
Manganese in steel, 212 ; action of,
1 ; acts as an alloy, 287 ; advan-
tages of, 212 ; application of, due
Heath, 212, 380 ; when bene-
ficial effect of ceases. 213 ; chemi-
cal action of, doubtful, 212 ; cor-
rosion due to, 365, 392 ; dangerous
ingredient, 374 ; distribution of,
not uniform, 364 ; effect of, due to,
213 ; heterogeneity caused by, 368;
neutralises sulphur, 312 ; Parry
on, 213 ; phosphorus counteracted
by, 339, 364 ; report on by Hack-
ney and Willis, 287 ; steel re-
melted without, 213 ; rolling as-
sisted by, 368, 374, 377 ; sulphur
counteracted by, 212 ; treacherous
ingredient, 364, 392.

Manufacture of cast steel, 209.

Manufacture of steel, advance, re-
cent, in, 379 ; by Bessemer process,
q. r. ; compared with that of
wrought iron, 403 ; by fusion,
209 ; by Huntsman's process, 379 ;
liquefaction necessary in, 214 ;
Reaumur's proposal for, 379 ; by
Martin's steel process, 218, 388 ; by
ore or Siemens process, 284, 448 ;
by scrap or Siemens-Martin pro-
cess, 284, 448 ; Swedish ore used
for, 212.

Marche, E., steel, use of, matters
affecting, discussion of paper by,
373-378.

Martell, B., shipbuilding, steel for,
discussion of paper by, 367-368.

Martin, open-hearth steel process,
218, 388 ; gold medal awarded at
French exhibition, 218, 388 ; at
Sireuil Works, 218.

Massanez, J., dephosphorizing in the
converter, discussion of paper by,
43(5-137.

MILD STEEL.

Material in blast furnace not com-
parable with regenerators of \\n\-
blast stoves as heat absorber, 2.>.

Matheson, E., steel for structures,
discussion of paper by, 458-462.

.Mayiiard, A. N., bridges, steel in
construction of, discussion of
paper by, 409-412.

Melting, calculation of heat neces-
sary for, cinder in blast furnace,
261 ; iron in blast furnace, 261 ;
iron in cupolas, 261.

Metallic iron, in steel, 398 ; in
wrought iron, 398.

Metals, conductivity of, 330.

Mild steel, 407 ; annealing not re-
quired for, 427 ; arguments in
favour of, 421, 427; Board of
Trade limit of strength for, 375,
408 ; for boilers, 397 ; boilers of,
cannot burst, 431 ; careful manu-
facture and working of, 432 ;
castings, 464 ; chemically pure
iron, 369 ; compared with hard as
regards strength and ductility, 428 ;
compared with wrought iron, 398 ;
contains 99'75 per cent, of metallic
iron, 398 ; differs physically from
iron, 369 ; ductility of, 421 ; for
gun-making, 323 ; and hard steel
equally strong within limits, 395,
420, 423, 424; and hard steel,
tables of tests of, 396 ; homo-
geneity of, 449 ; not injured by
sudden cooling, 421 ; Iris and
Mercury buiit of, 431 ; mainly
made by ore process, 416 ; by ore
process, 448 ; plates, relative cost
of, and Yorkshire plates, 432 ;
punching action in, 370 ; punching
or straining, strength increased by,
371, 397, 446, 447, 449, 460, 461 ;
rails, preference for, 272 ; re-
liability of, 430 ; rivet bars of,
432 ; for riveted structures, 460 ;
riveting plates of, 369 ; for ship-
building, Hi) 7. 448 ; by Siemens-

INDEX TO VOLUME I.

MILTON, J. £.

Martin process, 448 ; solid flow a
quality of, 422, 450, 460 ; special,
432 ; steel rivets for plates of,
409 ; for structural purposes, 315 ;
containing sulphur must con-
tain manganese, 305 ; tearing
of, due to its homogeneity and
uniformity, 412; tears easily,
370 ; tests of, for tin plates,
422 ; uniform in quality in large
masses, 315 ; working of, 408.

Milton, J. T., commercial marine,
influence of Board of Trade rules
for boilers on, discussion of paper
by, 462-464.

Mushet's addition of spiegeleisen to
Bessemer steel process, 212, S80.

NATURE of steel, 21 0.
Nitrogen in steel, 211.
Nomenclature of steel, 435 ; various
views regarding, 435.

OBJECTIONS to blast furnace, theo-
retical and practical, 290.

Objections to gas-producer refuted.
223.

Occluded gases in steel, action of
pressure on, 430.

Occlusion of gas, diminished with
increasing temperature, 337 ; in
steel, 337, 338.

Open-hearth furnace and Bessemer
converter compared, 375.

Open-hearth steel, chemical com-
position of, 411 ; for special ap-
plications, 376.

Open-hearth steel process, 216 ; basic
lining used in, in 1863 and 1865,
412 ; capability of regenerative gas
furnace to produce steel by, 382 ;
a chemical or cooking process,
318 ; description of, 318 ; dephos-
phorization by, 417 ; difficulties

ORE.

of, 217 ; Heath's, 216 ; limestone
added in, 411 ; Martin's, q.v. ; re-
fractory material required in fur-
nace for, 318; regenerative, <?.r. ;
report on, by Ste.- Claire Deville,
Beauclerc, and Caron, 216 ; Sie-
mens's, q.v. ; Sudre's experiments
on, under patronage of Emperor
Napoleon III., 216 ; temperature,
high, required in, 318 ; time for
sample-taking and adjusting com-
position in, 411.

Open top to blast furnace, advan-
tages of, 332.

Ordnance. Armstrong system of, 398 ;
hard material outside, weak inside,
wrong system of constructing,
399. 401 ; Longridge, J., views re-
garding criticised, 399 ; mode of
construction of, to suit mild steel,
398, 401 ; steel for inner tube,
of, 400 ; strains, distribution faof,
in manufacture of , 400 ; Woolwich
system of, 398.

Ore charged into blast furnace
limited in quantity and capacity
for heat, 256, 257.

Ore-reducing steel process, 284 ;
bath of fluid pig metal in, 284 ;
bauxite objectionable for, 285 ;
compared with Bessemer steel pro-
cess,286,287; metal" deadmelted"
in, 287 ; decarburizing agent in, 287 ;
disadvantage of raw ores for, 285 ;
elimination early of manganese
and silicon, and late of carbon in,
286 ; elimination of sulphur and
phosphorus in, 304 ; manganese,
oxidation of in, 286, 287; mild
steel made by, 416 ; output quick-
ened by use of scrap in, 416 ;
refractory material suitable for,
285 ; and scrap process, difference
between, 448 ; silica chemically
objectionable in, 285 ; special or
prepared ores for, 284 ; spiegelei-
sen, necessity for addition of in,

INDEX TO VOLUME I.

489

OXIDATION.

287 ; steel of high quality pro-
duced by, 285 ; sulphur and phos-
phorus oxidation of in, 287.

ation of silicon and carbon,
slight of manganese, and non-
oxidation of phosphorus and
sulphur in Bessemer steel process,
286.

PARKER, W., marine boilers, steel for,
369-373; steel plates of Livadni.
444-447 ; use of steel castings in
lieu of iron and steel forgings,
464-466, discussion of papers by.

Parry on manganese in steel, 213.

Patentee, guardian of his invention,
457.

Patentee's rights, and Government
departments; 455 ; interference
with, 457.

Percy, Prof., on puddling, suggestion
that sulphur and phosphorus are
removed by liquation in, 243 ;
puddling, conclusions of on experi-
ments in, 238 ; steel, definition of,
by, 316 ; on steel processes, 214.

Perfect steel should be aimed at, 419.

Pernot's circular furnace, criticism
of, 318, 319.

Phillips, D., iron and steel, compara-
tive corrosion of, 440-444 ; J. A.,
recent metallurgical processes,
330-331, discussion of papers by.

Phosphorus, difficulty of eliminating
from Cleveland pig, 283.

Phosphorus in steel, 211, 212, 288,
320 ; amount of permissible, 363 ;
causes brittleness or cold short-
ness, 212, 288, 363, 425 ; counter-
acted by manganese, 339, 364.

Phosphorus and sulphur innocuous
in steel under certain conditions,
319.

Picton, J. A., architectural construc-
tion, iron as a material for, 402-
4(J8 ; constructive materials, pro-

PUDDLINQ.

gross of iron and steel as, 419-
422, discussion of papers by.

Pig-iron, improvement in steel de-
pendent on improvement in, 289 ;
improvement in quality of by use
of anthracite coke in blast furnace,
332 ; " number " of ruled by
absorption of carbon by iron, 2fi5.

i'inks's process of casting, 336.

Piping of steel in moulds, 312.

Platinum, variation of electric con-
ductivity in cast and welded, 236.

Pourcel, A., dephosphorization of
iron and steel, discussion of paper
by, 416-419.

Pouring steel, agitation before, 450,

451 ; ebullition in, 312 ; kept
quiescent at high temperature,

452 ; stopping moulds in, 322.
Precipitation process, bed for, 295 ;

experiments ^proving successful
reduction of iron by, 293, 295 ;
furnace for, description of, 295 ;
at Landore, table of results of,
293, 294 ; Mokta ore used in,
analysis of, 295 ; manual labour
and skill involved in, 295 ;
rationale of, 293 ; results at
Blochairn by, 294 ; slags from,
analyses of, reference to, 295.

Prejudice against steel, 404.

Price, J., testing rails, discussion of
paper by, 271-273.

Price's retort furnace, 324 ; and
regenerative gas furnace, criticism
of Bell's, I. L., comparison of, 324 ;
used at Woolwich, 325.

Price and Nicholson, revival of
Hindoo steel process by, 216.

Properties of steel due to chemical
admixture, 379.

Puddled bar, analysis of, 244 ;
theoretical consumption of fuel
in production of, compared with
actual, 301, 302.

Puddling, chemical problems in,
238 ; economical and wasteful

INDEX TO VOLUME I.

PUDDLING.

compared, 283 ; important branch

of metallurgy, 237 ; Siemens's,

C. W., views regarding, 239, 241,

and confirmation of, 280 ; silicon

removed in, 239 ; waste due to

oxidation in, 242.
Puddling furnace, action in, popular

views of, 239 ; analysis of con-
' tents of, 238 ; cooling puddling

chamber of, by open vessel, 334 ;

order of removal of carbon.

sulphur, and phosphorus in, 238 ;

yield from, actual and theoretical,

242.
Puddling process, usually wasteful,

laborious and incomplete, 244.
Punching, annealing after, 371 ; die

should be larger than punch in,

371 ; and drilling compared, 410 ;
, or straining increases stength of

mild steel, 371, 397, 446, 447, 449,

460, 461.
Punching of steel, 460 ; action in,

370.
Pure steel, little influenced by

temperature variations, 425 ;

production of, 340.

QUALITIES of steel, affected by
treatment, 316, 317 ; determined
by chemical test, 459 ; have to be
studied, 314 ; various, 340.

Quantity of blast required in blast
furnace, 279. «

KAILS, actual test of on railway line
important, 272 ; testing machine
for, 272 ; wearing quality of, test
of, 272 ; wheels on, propelling and
sliding action of, 272.

Reaumur made steel by fusion, 215.

Recoil of gun carriages, checking
by hydraulic compression, 457 ;
(Siemens, C. W.'s, suggestion for,
referred to by Col. Clarke at

REGENERATIVE GAS FURNACE.

Exeter meeting of British Associa-
tion, 457 ; used without acknow-
ledgment by Government, 457).

Red shortness in steel, 211 ; pre-
vented by manganese, 232.

Reduction of temperature of gases
from blast furnace, how effected,
278.

Regenerative gas furnace, 209, 455 ;
Sir Win. Armstrong's allusion to,
455 ; arrangement of for lower tem-
peratures, 387 ; cool chimney in, 224.
247 ; description of, applicable to
ore and scrap open-hearth pro-
cesses, 219, 382 ; most economical
for high temperature, 387 ;
economy of fuel in, 224, 225, 247 ;
economy of steel melting in, 387 ;
failure of at Royal Gun Factory,
cause of, 455, 456 ; failure of
early Sheffield experiments, cause
of, 217 ; Faraday described in
1862, 210 ; (flame in, concentra-
tion of, 227 ; regulation of length
of, 227) ; fusion of steel in crucibles,
229, 381 ; gas and air separately
heated in, 224 ; gas in, advantages
of using, 227 ; heat accumulation
in, 226, 247 ; heat, action of, in,
illustrations of, 226 ; heat ample
in, 218 ; heat of combustion,
almost entirely available in, 387 ;
heat below temperature of work
utilized in, 224, 225, 247; heat
nearly all retained in, 224, 225,
247 ; high temperature, effect of,
in, 236 ; invented jointly by Sie-
mens, C. W. and F., 210 ; out-
ward pressure in, advantages of,

228 ; gas-producer for, 245, and
see Gas-producer; rate of com-
bustion of gas in, regulation of,
227 ; refractory material for, 226,
236 ; regenerators for, 224, 246 ;
regulation of, by chimney-damper,

229 ; retorts formerly used with,
325 ; success of at gun factories of

INDEX TO VOLUME I.

491

Ifl !•[ Dlll.IXO.

mstrong, Krupp, Schneider and
Whitwdi-tli, l.V) ; temperature in,
practical limit cf, iisl ; visible
.i[i|>. malice of flame in, depending
On U-inpfratuiv. :M ; Woolwich
Arsenal type of, I .">."•. l.'ii'i.

•lenitive gas puddling funiace,
~1 1 1 ; advantages of, detailed,
L'lS; applications of, 252, 309;
auxiliary heating chamber for,
i1 1 ^ : bed of, with water bridges,
248 ; best iron from, 251 ; cost
of, 254 ; criticisms on, reply to,
•-'."••I. -'>'>; description of, 244,

246 ; flame in, command of tem-
perature and quality of, 247, 383 ;
gas and air regulating valves in,

247 ; heat not wasted in, 254 ;
heats increased in number with,
2.">i>; improvement of iron in, to
what due, 255 ; and ordinary
puddling furnace, tables of com-
parison of, 249, 250 ; power of,
tested with inferior pig metal,

• '< ; puddler's labour reduc'.'d in,
252 ; rabble applied to, 2~>'2 ;
saving of fuel in, 251 ; water
bridges in, consumption of fuel
saved by, 252 ; workmen well
disposed to, 254 ; yield of puddled
bar, exceeds charge of pig metal
into, 251.

Regenerative open hearth ore pro-
cess, 233 ; apparatus for, present
form of, 233 ; charging in, manner
of, 234 ; experiments over several
years with, 233 ; Grand Prix at
French Exhibition for, 233 ;
heated gas forced among ore in,
234 ; qualities of ore for, 235 ;
reduction of ore into sponge and
dissolution in bath in, 234 ; ores
used in, steel dependent on, 235 ;
testing of metal in, 235.

Regenerative open hearth steel pro-
cess, 217, 389 ; Attwood's experi-
ments, 217, 388 ; bauxite used in,

217, 218, 388; Boiguc, Rambour
and Co. 'a experiments at Mont-
lucon with, 217 ; bottom repair ..f
in, 389 ; charging materials, 231,
389 ; experiments on, at Birming-
ham Sample Steel Works, 218, 388 ;
experiments with, incomplete, 233;
high temperature in, method of
maintaining, 231; ferro- manganese
used in, 391 ; Le Chatelier's expe-
riments on, 217, 388; Martin's, 218,

388 ; Martin's, French Exhibition
gold medal awarded to produce by,

218, 388 ; ores charged in, 389 ;
rail metal by, 390 ; reagents em-
ployed in, 233 ; running metal from
blast furnace in, 332, 333 ; sand
used in, 218 ; spiegeleisen added in,
_:52, 284 ; steel of special quality
by, 391 ; tapping furnace in, 231,

389 ; testing samples from metallic
bath in, 232, 284 ; time available
for chemical reaction in, 232 ; time
required to melt charge in, 232, 390.

Regenerators, action of, 224, 383 ; air,
224, 382 ; bricks for, arrangement
and size of, 228; brickwork actually
required in, calculation of quantity
of, 227 ; chambers of loose fire-
bricks, 224, 246, 382 ; cooling of,
224, 383 ; current in, advantages of
regular reversal of, 224, 225, 247 ;
detailed description of action in,
247,383; gas, 224, 382; experiments
by Siemens, C. \V., on, reference to,
228 ; how heated, 224, 383 ; heated
from above downwards, 228 ; hot
gas supplied to, produces hot
chimney, 386 ; opposite currents
passing through, should have equal
capacity for heat, 259 ; placed
below heating chamber, 228 ; re-
versing valves in, action of, 246,
247, 383 ; temperature of, 224,
383 ; temperature constant in,
225 ; temperature range in, 386 ;
vertical, advantages of, 228 ; waste

492

INDEX TO VOLUME 1.

BEMOVAL OF IMPUEITIES.

products of combustion passing
through, 224, 383.

Eemoval of impurities in puddling
depends on high temperature,
244.

Kesearch, encouragement of, 359.

Reversing valves for regenerators,
246, 247, 383.

Eise of temperature of blast furnace
gases, cause of, 258, 260.

Riveting of steel, 460.

Rolling of steel, assisted by man-
ganese, 368, 374, 377 ; tires, 374.

Rotary furnace, Siemens's,C.W., 292,
304, 393 ; abandonment of, in
connection with steel-melting fur-
nace, 292 ; advantage of over blast
furnace, 308 ; analysis of iron, 348 ;
analysis of slag, 351 ; anthra-
cite in, 299 ; apparatus and pro-
cess, description of, at Towcester,
343 ; balling in, 299. 395 ; bauxite
used in, gee Bauxite ; and blast
furnace compared on theoretical
grounds, 300 ; cast steel obtain-
able in, 300 ; charges from, work-
ing results of, at Towcester, 347 ;
charging of, 299 ; Cleveland pig
carefully puddled in and melted
on open hearth made good steel,
359 ; carbonic oxide consumed at
moment of generation in, 307 ;
carbonic oxide developed in, 394 ;
comparison of results with those
from blast furnace and puddling
furnace combined, 310, 356 ; com-
plete combustion in, explanation
of, 301 ; consumption of coal in,
at Towcester, 345 ; consumption
of fuel in, theoretical and actual,
302 ; cost of hammered blooms
from, 345 ; cost of process, 344 ; de-
oxidation of ore and fusion of
earthy matter in, 301 ; description
of, 298, 393 ; discussion between
Bell, I. L., and Siemens, C. W., on,
356, 357 ; economy of fuel in, 310 ;

SAHUELSON, B.

energetic action in, necessity
for, 307 ; experimental furnace no
criterion of economy of, 306 ;
experiments with at Sample Steel
Works, Birmingham, 309 ; fettling
for, 309 ; fuel in, 306 ; heat ab-
sorbed in, 301 ; heat of high tem-
perature attainable in,305 ; heated
gas and air introduced into, 298 ;
iron dephosphorized in, 354 ; lining
for, 306 ; loss in, 344 ; material
from, suitable for steel-making,
308 ; mechanical tests of iron from,
349, 350 ; modus operandi in, 394 ;
process in, 301 ; production of
CO and C02, within, 301, 307;
products of combustion withdrawn
from, 298 ; purity of iron from,
343, 395 ; reaction in, 299, 301, 354,
355 ; reaction and fusion combined
in, 307, 319 ; reports on, by Profes-
sors Akerman and von Tunner,
342 ; rotation of, 292 ; saving of
fuel in, 300 ; sulphur and phos-
phorus not added to iron in, 308 ;
tapping of, 299 ; temperature high
in, 394 ; theoretical consumption
of fuel for production of metallic
iron in, 307 ; throat of not at-
tached, 304 ; time occupied in
working charge in, 299, 344 ; varia-
tion of process with different ores,
358 ; -wood in, 299 ; working of,
298,306; working for a time with-
out gas, 299, 308 ; yield from, 344.

Rotary puddling furnace, 280 ;
Danks's, 280 ; steel-melting tem-
perature not attainable in, 281.

Running metal direct from blast
furnace for steel manufacture,
332.

STE.-CLAIRE DEVILLE, report on
open-hearth steel process, 216.

Samuelson, B., blast furnaces, dis-
cussion of paper by, 273-276.

INDEX TO VOLUME /.

493

SCIENCE.

Science, sign of advancement of.:.' i;».

Scientific attention paid to puddling,
slight, 238.

Scott, M., casting arrangements for
Sir-iii'Ms- Martin process, discussion
of paper by, 335-338.

Scrap and ore process, difference
botween, 448.

Scrap process in regenerative gas
furnace, 230.

Sheffield steel process, 214, 328.

Shipbuilding, iron and steel for, 313 ;
steel for, 313, 367, 420.

Shocks, influence of, on steel, 271.

Short test pieces, objection to, 373.

Shrinking hoops on guns, action
taking place in, 453 ; temperature
of, 453, 454.

Siemens, C. W., definition of steel,
316, 323 ; gas-producers, q. v.,
lime and magnesia linings, experi-
ments on, 363 ; mild steel, q. v. ;
open-hearth steel process, q. v. ;
ore-reducing process, q. v. ; papers
by, 209-237; 237-255; 283-302;
303-308 ; 340-352 ; 378-401 ; pre-
cipitation process, q. v. ; views on
puddling, that carbon and silicon
are replaced by iron from fluid
oxide of iron, 239, 241 ; views on
puddling confirmed by experiment
at Sample Steel Works, Bir-
mingham, 240, 280, and results in
Danks's furnace, 280 ; regenerative
gas furnace, q, v. ; regenerative
gas puddling furnace, q. v. ; re-
generative open hearth ore pro-
cess, q. v. ; regenerative open
hearth steel process, q. v. ; re-
generators, q. v. ; rotary furnace,
q. v.

Siemens-Martin steel process, 284 ;
casting arrangements for, 335.

Siemens steel, weekly production of,
289.

Siemens, Werner, researches on
tungsten in steel, 213.

SPONGY IRON.

Silica iron, H'.U.

Silicon, non-absorption of by fluid
cast metal in contact with silica
or silicates, 240 ; removal in
puddling by reducing flame of low
temperature, 239 ; in steel, effect
of, 213 ; table of, in cast and
wrought iron and steel, 2.">:{.

Slag, mode of mixing with iron, 398.

Smelting ironstone, chemical work
in, 260.

Smelting operations, heat necessary
in, 260, 261.

Smith, E. F., Casson-Dormoy
puddling furnace, discussion of
paper by, 333-335.

Smith, J. T., Bessemer steel rails,
315-320 ; use of molten metal
direct from blast furnaces for
Bessemer purposes, 331-333, dis-
cussion of papers by.

Snelus, G. J., spiegeleisen, manu-
facture and use of, 311-312 ; steel
ingots, distribution of elements in,
450-452, discussion of papers by.

Snelus, view of dead melting,
criticism of, 337.

Soda, ammonia used in making, 330.

Solid flow, a quality of mild steel,
422, 450, 460.

Solidifaction of steel, chemical
change during, 451.

Spiegeleisen, addition of, to open-
hearth process, 232, 284 ; man-
ganese, percentage of, in, 311 ;
manufacture of, 311 ; on steel,
effect of, 311; sulphur, why not
found in, 311.

Spongy iron produced in blast
furnace, 290, 366 ; by Blair's pro-
cess, 365 ; Chenot's process of
manufacturing, 365 ; difficulties
in connection with, 366 ; failure
to puddle, 242 ; by Ireland's pro-
cess, 365 ; sulphur absorbed by,
366 ; trials of Chenot, Clay, and
Yates with, 292.

494

INDEX TO VOLUME I.

STEAM.

Steam blown into molten metal, 418.

Steel, ancient, q. v. ; annealing, q. v. ;
Bessemer, q. v. ; boilers, q. v. ;
boilers riveted with steel rivets,
367 ; carbon in, q.v.; (castings, free
from blow-holes, 465 ; in lieu of
forgings, 464 ;) chemical analysis
of, importance of, 425 ; a cheap
constructive material, 405 ; cold,
effect of, on, 364, 428 ; compared
with iron and wood for structures,
405 ; compressed, q. v. ; com-
pressed in centre causes porosity
towards outside through which
gases escape, 322 ; compression of,
by Whitworth's process, q.v., 391 ;
corrosion of, q. v. ; crucible, q. v. ;
definition of, q. v. ; ductility of,
459, 466, and q. v, ; elasticity of,
q. v. ; for engineering structures,
375 ; factor of safety of, q. v. ;
failure of, 377, 410, and q. v. ;
forgings, unequal strength of
complicated, 465 ; fusion of, q. i\ ;
gases cannot be driven out of,
by pressure, 321 ; guns, q. v. ;
(hardening ; Akerman's views
regarding, 433 ; due to carbon in
intermediate condition, 433 ; with
oil, 397, 465) ; impurities, q. v. ;
and iron, comparison of methods
of producing, 314 ; machinery,
q. v. ; making, Cleveland pig un-
suitable for, 282 ; manganese,
q. v. ; manufacture, q. v. ; mild,
q, v. ; open-hearth process, q. v. ;
ore-reducing process, q. v. ; phos-
phorus in, q. v. ; (plates, scale
rolled into, causes corrosion, 443 ;
steel rivets for, 367) ; pouring,
q. v. ; (processes, Bessemer, q. v. ;
Catalan forge, 214 ; cementation,
214, 328 ; decarburization, 214 ;
fusion, q. v. ; Heaton and Har-
greaves, 215 ; open hearth, q. v, ;
ore-reducing, q.v. ; Percy on, 214 ;
precipitation process, q. v. ; regene-

TABLE.

rative open hearth, q. v.; Sheffield,
214,328; Siemens, q.v.; Siemens-
Martin, 284, 448 ; Styrian, 214) :
production of, operations involved
in, 302 ; properties of, 317 ;
punching, q. v. ; rails, 378 ; red
shortness of, 211 ; rivets of, advan-
tages of, 367 ; for shipbuilding,
313, 367, 420 ; silicon in, effect of,
213 ; specification of chemical
constitution and temper of, 317 ;
spiegeleisen affects, 311 ; (strrtiijtli
of, varies, 313 ; versus ductility,
428 ; limit of, for different pur-
poses, 459) ; strengthened by
straining, 424 ; supply of, by
weight instead of dimensions, 367 ;
tires, rolling of, 374 ; toughness
and tensile strength of, 317 ;
uncertainty of, 316 ; varies in
strength and quality, 459 ;
weakened by occlusion of gases,
322 ; what is, 313 ; worked hot
and cold, experiments on, 376.

Stoppered moulds, uncertainty of
filling, 335.

Strains produced in manufacture,
how equalized, 270.

Stretching and tearing, difference
between, 410.

Styrian, ferro-manganese. 361 ;
furnace, 361 ; iron, 361 ; ore, 361 ;
steel process, 214.

Sulphur, disappearance of, with
anthracite coke, 327 ; (in steel,
211, 212; counteracted by man-
ganese, 212 ; when advantageous,
212).

TABLE, of carbon in cast and wrought
iron and steel, 253 ; of carbon in
steel, 210, 211 ; of comparison of
regenerative gas puddling furnace
and ordinary puddlingfurnace, 249;
250 ; of results of precipitation

INDEX TO VOLUME I.

495

TAPPIN".

|.r s-i, L".>:s, 294; of silicon ami

carbon in cast iron and steel, •-'.">:{ ;
of Si' ;nni>'- direct process result s
;H7— :«.~>2; of tests of mild and
hard steel bars, 396.
Tapping siliceous slag with addition

of lime, 417.

Tearing of steel, 370, 407, 412.
Tearing strain in homogeneous ma-
terial, 271.
Temper of steel, 210.
Temperature, of combustion, theo-
retical and practical limit to, 384;
diagram of distribution of, in
blast furnace, 290 ; of gases from
blast furnace as affected by tempe-
rature of blast, 278 ; high, of hot
blast, improves blast furnace eco-
nomy, 266, 267 ; high, required in
open-hearth steel process, 318 ;
low, of reduction of Cleveland ore
in blastfurnace, 277 ; of reduction
of ores in blast furnace, 277 ; steel
affected by variations of, 423 ;
steel unaffected by, as regards
absolute structure, 423.
Terre Noire steel, 377.
Test, falling weight, of steel, 271 ; of
iron from rotary furnace, 349, 350.
Test pieces, D. Adamson's ten-inch,
objections to, 373 ; standard eight-
inch, 373 ; objections to short, 373.
Tests for steel should not be dimin-
ished, 419.

Testing machine, 401.
Thomas, S. G., phosphorus, elimina-
tion of, discussion of paper by,
413-415.
Thomas and Gilchrist on dephospho-

rization, 413.

Titanium in steel, effect of, 213.
Tool steel without manganese, 305.
Toughness of steel, 313.
Toweester, application of rotary fur-
nace at, 342. See Rotary Furnace.
Treatment of steel affects its appli-
cations, 313.

WELniNO STEEL.

Trial of closed versus open tops in
blast furnace, 332.

Tungsten in steel, effects of, 213,
I'll; Siemens, Dr. Werner, re-
searches on, 213.

Tungsten steel, experiments on in-
creased-magnetic retaining power
of, 213.

Tunner, Professor von, Bessemer
medal, presentation of to, 360 ;
literary works of, 361 ; metallur-
gical researches of, 361 ; personal
qualities of, 361 ; report of on re-
generative furnace, 342.

UcHATirjs's steel fusion process, 216.

Ultimate strength, test of, not neces-
sary, 434.

Utilizing heating and chemical power
of gases in blast furnace, 275.

VARIETIES of steel, 403.

Variety of steel for different struc-
tures, 420.

Vessel to resist high pressure, 372,
373.

Volatilisation and separation of oxy-
gen from ore in blast furnace, 257,
262.

WALKER, B., steel-making ma-
chinery, discussion of paper by,
437-440.

Waste due to oxidation in puddling
after metal brought " to nature "
by removal of carbon, 242.

Water bridges, expenditure of heat
due to, 252.

Webster, J. J., iron and steel at low
temperatures, discussion of paper
by, 422-425.

Weight of structure, reduction of, by
use of steel, 4or>.

Welding steel, 374.

496

INDEX TO VOLUME I.

WELSH DINAS SILICA BEICK.

Welsh Dinas silica brick, analysis of,
226.

"West, J., shipbuilding, steel for, dis-
cussion of paper by, 426-428.

Wheels, action of, on rails, 272.

Whitworth, Sir J., fluid compressed
steel and guns, discussion of paper
by, 320-323.

Whitworth, Sir J., 321 ; definition of
steel, 517 ; rifling, mechanical
gunnery, gun material considered
by, 321 ; process for compression
of steel, 321.

Williams, K. P., permanent way of
railways, discussion of paper by,
338-340.

Williamson, Prof., regarding de-
phosphorization, 363.

YIELD FROM PUDDLING FURNACE.
Woolwich system of ordnance. 398.
Work done and heat developed in

blast furnace, 273.
Wrought iron contains only 96 or 97

per cent, of metallic iron, 398 ;

what it is, 403.
Wrought iron and steel, table of

carbon and silicon in, 253.
Wrought steel from Creusot, 380.

YATES'S trials with spongy metal,

292.
Yield from blast furnace increased

by copper, 331.
Yield from puddling furnace, actual,

241,282 ; difference between actual

and theoretical, 242.

END OF VOLUME I.

BHADBURY, AUM5W, & CO , PRINTERS, WHITEFR1ARS

— 5^~say k-i^

' ! 1 ' 1

| 1 | 1 : ! ! ! .

CB

I'll

1 1 1 1 | 1 1

|l|

V

m

i 1 1 1 1

s

//. / Ml-] ./„/,: 1851.)

' J.Couk V Htmuomi.I.uh.Er, «dwiy,W«atx.r

REGENERATIVE CONDENSER

Tig. 3. JwticAJtor Diagram, from Sigh Pressure ^Engine

VACUUM LINE

Ktcerpt froc. Inst. M.E. July. JS51.)

EXPANSION OF STEAM

Apparatus for Measuring the
Expansion of Isolated; Steam

„.-,

' '/•/// /Vw. Inst. N.E. J85Z.)

EXPANSION OF STEAM

/•.'/•//<////>•//>// <>t' Lmlutt-il Sti-uin ,li\ lion
at Cmuitunt Almusphenr rr&uurc

I'l.ltl-

STIRLING S ENGINE .

100 26S 330 06 VO XSS 2SO 2Uf 3<>(> ft\v

•

V-l !

REGENERATIVE FURNACE.

Hut, t:

!;i%>' 1 . ii-niiitn.hn.il

Section

1 -1M1 Jl^ll- Jf-^ll j'PJI JMi J^I^'JU-I" J !'! L. f\

\ p!!j!jujt3C| u r" :L/F'4'

$ j.. LaizpjCaCpjpjpat J"Jq //'

-^u_ ^ ;u£^gi3~MR£;^\;v^bCp

_i_jL_ji_ir_iar_1j.__j._j-' / fc.Lj,,.,,

_l . J ^ u „ U ' P Z, i-J .... — — ^fcf Y't*. -p C 3

P^PMRPPt-iPi-H^1--^ T^R^

irtr

Kg: 2 . Sectional Plcav.

Kg. 3 . Transverse Section

through
Headed, Chamber

.g. 4«. Transverse Section,

through
fireplace and Regenerator

Jurlun n 6 0

*urpt five, tut: jf.E. ]857 ftqc KM I

Plate

Effn-f't /hr. bust.. ME. 1862, Page 21)

Vol I.

Vol I

//./

/'/,!/. It

M.E.18BZ Page 2).)

s

S

x<rr,«

, Payr "1 )

v. : i

REGENERATIVE GAS FURNACE

I'l.M'i: f.A.i.s.s MI 1 i i.\i, it i

Ctiinim-\
Fin,-

Fig. 8.

Chi/itin-v
flue '

^4.

^

L_J 1 , « 1

E :

•— r— ^ •••••

KK 1 ' Chimney
1 Ftue

'P

Fig. 9.
| .Sn-tianal flan
~T «<• Jfe/vAj

'^II jl
-*

— •* GisFhie

I

llllnd.ln? 1 1 1 L__ 4

Scalt

J.i 1

. ! .

»<•«-<•

• ^«)*

, , /C , , . V

r. In.st. M.E JMZ /'<,.,, ;V

I

v< i :

M.E. 1062. Page 2Jj

flaU

UJ

U

CO

<

o

UJ

>

»-
<

IT

LJ

o

UJ

QC

.£. 186Z.fagt,

//.

REGENERATIVE GAS FURNACE
LCB

,

: i :

C H

I;i^ 14. Settumal flan, of Puddhng Chamber

Fig 15. SerAiojval, Flan- below fuddling Chamber

Chimney
flue

• i f i

'JL L.

•MM

<vn/ /Vvv /ruit Af.K.J^di' I'naf 21.)

REGENERATIVE GAS FURNACE

I'lntt- 11

l;i>l 1(1 Tr,ni.\\-ri:tt Srrtwn through inuring i-li ///»/.// <>

[ LJULJ

plLT

hjJLL

LLTJ1

Kg.17.
Vertical Section at G (Fig.23)

Fig.18.

yrrtt

/ /•/-,., //,.v/ .»//•; /^;;-. /;,/// ;v

Vol.1.

(Excerpt Proc.Inst.M3:. 1862. Page Zt.)

Vol.1

UNIFORM ROTATION.

J

_._£ ,

~~ra

I "- rpt, Phil TrtuiA J866.)

GYROMETRIC GOVERNOR

FOR STEAM ENGINES

!•:>,, T/>/ I'lnl Trmi.s JMX.)

Vol.1

8

GYROMETRIC GOVERNOR.

FOR STEAM ENGINES

ig. 6.

Flan with ewer aft

dusmng Oip

Scalf-

*•:'<>•/•/,(. Ptnl Trans. 1866.)

1

o

o

_l
o

u

£X

f-

o

UJ

_i

UJ

00 1

g *

~

/'//// TrnitA. JM%,

STEAM JET

Kg 2.

Loiufitudinnl .

STEAM JET

.1,1 l\.th.ni..tfi ,; .

fig 3 . Jransrrrst Section at IX .

Plata

i£. 4. I'nmsvrrsr \-<-t,,-/i at YY.

.5. Irarufrer.ve Section at ZZ.

I'',; fn*t,..V.I':.Jfl7?.)

UJ

-)

$

/'/•„, ///.v/ .»//•; Af/;

< *
*

erpt !',„, Inat ME

V.,1 I

/•'

/•/„,, ;vt

. JW, ME. 1372.)

R*t J.Cook tt H.maond.U* Bn>«4**j.W«Air

:
:

I
\

• Proc Jhst.Jf.EJS7Z.)

1 J. Cook >• Hin.io

STEAM JET.

II., i, .u>

i; .v, „//, ./,/ />/,.„//

<//•/.//»•// If I'.n.^ I'li-.lini-i

SCQ&A

Fig. 18. XongiUulwaJ Secti-on, of Blower.
ScaU J/M>*

Proc .!„,<<( M.KJ872.

» J.Cook It Hi«noo<I.Lilh Biv<d<ny.Wkit«ir

HEATED AIR ENGINE

/•/.,/, D

l.lfHIII.fll'll t'lll'\< III All

flll-M I.-. I III' ll\l:ll-l:-ll< ,.,/,•:„ .,.•!! ,T,,.,
llllllllllllll.il 111 II H'll.tlailt !•< .

Tlir full nir\, I., lln- ,,•!?••, i. •! • , • /><

.iiit-it i n</ I /< i i. -nli in in-ii ••/ >.•/«,//./ ,. •!,..,. jiirnl

HfMVI irJlK-lH'll !•!' I* Illl'l I tlllll I !'• l.tfKllllllllt/.

T/if two ntrvf.t «/•«• tli-avm t" i.-nnuli .<i
irljni'.-ydif'ti fn-r.\.inrf.

lift

:< h,; I,,*! M K.

Ml J.CeokfclUi»i»»«<!.liih b, ,..-.,.*•>'-••

Vol I

HEATED AIR ENGINE

^300-

....,

Expansion, Curyt, of Steam .

being thr dynamical, expansion curve
i<f .•nitiinitf'1 ttfum. alUm-mg for the
ZWB <'/ /«•«'' 'A.i/ M omrertfd mte
mffjutfucal rffret

Tkf ahitiled arm, or an) «Our rrrtangle
dmtrn to 0u> otrvr, rtpre^enta the ponrr
alitained from a ctmdaiamg atyinr trith
perfect vacuum but wutiaul ejcptnuum

>v/

^^—~— 1___L

j

> MMIO 800(1 «•«•

•rr f f 1*' J I

Voaatiea of .*>te<im
(Wafer-l)

i jit h-tt-

Robl J.Ccok ik tU»

Vol. I

I",:',

CIRCULAR QA8 PRODUCER.

(Kxrrr/if Jour. Sot: f'lt,-i». /nth/., 1881.")

V.,1. I.

L T

1

r5>

r

(Excerut Jour. Soc. Chem. Indu., 1881.")

Yul. !.

REGENERATIVE GAS FIRE AND GAS LAMP.

(Excrrpt Jmir. Sac. Chmi. /«////., 1881.)

CO

5

{tixcerpt t>r«c. ln*t.. t'.K.. 1881-82. K^. LXVlll.)

Vol. I.

t. t\K.. 1881-82. Vol. LMX.)

V..!. I.

(Kfn-rpt JOH>\ Cinm. S.T., 1868.)

Vol. 1.

I'!,,; .('...

STEEL MELTING FURNACE.

(Excerpt Jour* Chem* S«c., 1868.)

Vol. I.

STEEL MELTING FURNACE

{Evcerpt Jour. Chem. Soc., 1868.)

Vol. I.

/"'!>' 11.

GAS PRODUCER.

(Exi-rrpt Jour. Chem. Sac., 1868.)

V.il. I.

PUDDLING FURNACE

(Excerpt Jmir. Brit. Atsoc. 1868.)

Vol. I.

CENTBELINE BETWEEN TWO FURNACES

(Excerpt Jtntr. Chem. Sue., 1873.)

Vul. I.

l"ur. H.

(Excerpt Jaw Clit-m. Soc., 1873.)

BLAST FURNACE DIAGRAM.

ll.-.-itm/.

Koiluctinii.

Limestone.

Absorption.

Ditto.

VERY BR! CHT RED

Fio. 2. — BLAST FUBNACK.
Excerpt Jour. Chem. Soc., 1873.)

Fusion.

Flui.l Slag.
Fluid Pig Metal.

YMJ I

SECTION OF STEEL MELTING FURNACE

A /.'«•// ••/'
B . Tap JMf

do

:!,

G .1" r,,jiil<itin,;

SOI 1

M

•H •

rH -d

H 0)
•H

H O

«H

'-^ -H
^t -P

•H fl

•> 03
CO

a a)

0) X!
S 6-1

E

•H

CO

CO

H

IN

iO

CO CO

University oi Toronto
Library

DO NOT

REMOVE

THE

CARD

FROM

THIS

POCKET

Acme Library Card Pocket
LOWE-MARTIN CO. LIMITED