BY THE SAME AUTHORS
Elementary Electro -Technical Series
COMPRISING
Alternating Electric Currents.
Electric Heating.
Electromagnetism.
Electricity in Electro-Therapeutics.
Electric Arc Lighting.
Electric Incandescent Lighting.
Electric Motors.
Electric Street Railways.
Electric Telephony.
Electric Telegraphy.
Cloth, Price per Volume, $1.00.
Electro-Dynamic Machinery.
Cloth, $2.50.
THE W. J. JOHNSTON COMPANY
253 BROADWAY, NEW YORK
ELEMENTARY ELEOTKO-TECHNICAL SERIES
.ELECTRIC
INCANDESCENT LIGHTING
BY
EDWIN J. HOUSTON, PH. D.
\\
AND
A. E. KENNELLY, Sc. D.
NEW YORK
THE W. J. JOHNSTON COMPANY
253 BROADWAY
1896
COPYRIGHT, 1896, BY
THE W. J. JOHNSTON COMPANY.
PREFACE.
THIS little volume has been prepared by
the authors with the view of presenting
both the art and science of practical electric
incandescent lighting to the general public,
in such a manner as shall render it capable
of being understood without any previous
technical training.
The necessity of some general knowl-
edge of the principles underlying the prac-
tical applications of incandescent lighting,
will be appreciated from the fact that at
the present time incandescent lamps are
manufactured in the United States at a
rate of about eight millions per annum.
This of course represents a large amount
iii
iv PREFACE.
of invested capital, only a small portion of
which is engaged in the actual manufacture
of the lamps, by far the greater part being
invested in the central stations where the
electric current is generated, and in the
streets and buildings where the conductors
and fixtures are provided. Since the whole
of this important industry has practically
come into existence since 1881, compara-
tively little opportunity has been afforded
to the public of acquiring a fair under-
standing of the subject.
It is in the hope of meeting the above
want that the authors have written this
book.
PHILADELPHIA,
June 1, 1896.
CONTENTS.
I. ARTIFICIAL ILLUMINATION, 1
II. EARLY HISTORY OP INCANDESCENT
LIGHTING, . . . . .18
III. ELEMENTARY ELECTRICAL PRINCIPLES, 43
IV. PHYSICS OF THE INCANDESCENT ELEC-
TRIC LAMP, ... .65
V. MANUFACTURE OF INCANDESCENT
LAMPS. PREPARATION AND CAR-
BONIZATION OF THE FILAMENT, . 83
VI. MOUNTING AND TREATMENT OF FILA-
MENTS, ...... 99
VII. SEALING-IN AND EXHAUSTION, . .118
VIII. LAMP FITTINGS, . . . .135
IX. THE INCANDESCENT LAMP, . .163
X. LIGHT AND ILLUMINATION, . .198
XI. SYSTEMS OF LAMP DISTRIBUTION, . 210
XII. HOUSE FIXTURES AND WIRING, . 237
Vi CONTENTS.
CHAPTER PAGE
XIII. STREET MAINS, .... 284
XIV. CENTRAL STATIONS, . . .304
XV. ISOLATED PLANTS, . . . .324
XVI. METERS, 334
XVII. STORAGE BATTERIES, . . .345
XVIII. SERIES INCANDESCENT LIGHTING, . 374
XIX. ALTERNATING - CURRENT CIRCUIT
INCANDESCENT LIGHTING, . .386
XX. MISCELLANEOUS APPLICATIONS OF
INCANDESCENT LAMPS, . . 402
INDEX, 419
ELECTRIC INCANDESCENT
LIGHTING.
CHAPTER I.
AETIFICIAL ILLUMINATION.
t
DOUBTLESS, the earliest artificial illumi-
nant employed by primitive man was the
blazing fagot, seized from the fire. The
step from this to the oil lamp marked an
era in civilization, since man's work was
then not necessarily limited in time by the
rising and setting of the sun. It is diffi-
cult, at this time, to estimate properly the
great boon to civilization this invention
afforded. Although at first sight it does
not seem to be a very great step from the
2 ELECTRIC INCANDESCENT LIGHTING.
flickering light of the torch to that of the
oil lamp, yet, when we consider the re-
quirements of an artificial illuminant, the
superiority of the latter is evident in what
may, perhaps, be regarded as the most essen-
tial requirement of such an illuminant ;
namely, its duration; i. e., the length of
time during which it can supply a
proper illumination without renewal. In-
stead of the fitful flickering of the torch
we have the comparatively steady glow
of the oil lamp ; instead of the evanescent
light of the torch, we have in the oil
lamp a means for furnishing light for
many hours.
Fortunately, for the sake of progress,
man's ingenuity was not arrested by this
great achievement. There followed in its
wake, various improvements in forms of
oil lamps, but this illumiuant sufficed to
ARTIFICIAL ILLUMINATION. 3
light the world for many centuries, as the
tombs, parchments and bas reliefs of the
remote past attest.
Passing by the many improvements in
various forms of oil lamps, perhaps the
next noted step in the production of arti-
ficial light followed in the discovery of
coal oil, and marked improvements were
made in lamps designed especially to burn
this natural illuminant. The next great
improvement in this direction was the
lighting of extended areas by means of
illuminating gas. Here, for the first time
in the history of the art, means were de-
vised for supplying an illuminant from a
central station, under circumstances which
permitted of its ready distribution over
extended areas.
The greatest step, however, in the pro-
4 ELECTRIC INCANDESCENT LIGHTING.
duction of an artificial illuminant was un-
doubtedly that which followed the inven-
tion of the voltaic pile in 1796. Then, for
the first time in the history of science,
means were provided whereby powerful
electric currents could be readily produced
and their effects observed. Naturally the
use of these currents soon led to the dis-
covery of the very powerful illuminating
properties possessed by the voltaic arc.
It must not be supposed, however, that in
these successive steps each new illuminant
completely supplanted its predecessors.
On the contrary, the very fact that night
could thus be transformed into day, stimu-
lated improvements in the pre-existing
forms of illuminants, and, in the emulation
thus produced, marked advances were
made in earlier methods. Thus, it was at
one time claimed, when gas was first intro-
duced into cities, that existing methods of
lighting would sot^ entirely repla^
by the new illummany'foj^
being the case, old methods were suffi-
ciently improved to fairly hold their own,
and even to require continued improve-
ments in the new illuminant, in order to
maintain its superiority. So, again, when
the electric light was introduced, it was
predicted, by enthusiasts that gas lighting
would now be entirely replaced by its new
rival, but, as is well known, this expec-
tation was not realized. So markedly have
improvements in different methods of
illumination gone hand in hand, that we
possess to-day all our original methods,
save, perhaps, the torch.
Before discussing the advantages pos-
sessed by the incandescent electric light, it
will be advantageous to consider in general
the requirements of any artificial illumi-
6 ELECTRIC INCANDESCENT LIGHTING.
nant. It is evident that the most satis-
factory artificial illuminant will be that
whose properties most nearly resemble
those of the sunlight it replaces. No
existing illuminant completely satisfies
our requirements from this standpoint.
The ideal artificial illuminant should
possess the following properties ; it should
be,
(1) Safe.
(2) Cheap.
(3) Hygienic.
(4) Steady.
(5) Eeliable.
(6) Akin to sunlight in color.
(7) Capable of ready subdivision.
(8) Cool.
(9) Readily turned on and off at a
distance.
(10) Amenable to the purposes of dec-
oration.
ARTIFICIAL ILLUMINATION. 7
As regards safety, it is evident that an
artificial illuminant should be safe both
to property and to life. All bare or naked
flames are open to the objection that com-
bustible material, coming in contact with
them, may start dangerous conflagrations.
A good artificial illuminant must neces-
sarily be cheap; for, unless it can be fur-
nished a"t a price which brings it fairly
within the reach of all, its use will be
very limited.
All artificial illuminants must necessarily
permit of continued use without any dele-
terious eifects on the health. Such defects
may arise either from products of combus-
tion vitiating the surrounding air, or, possi-
bly, in extreme cases, from injury to sight.
Steadiness is a prime essential of a good
8 ELECTKIC INCANDESCENT LIGHTING.
illuminant. If an artificial light produces
an unsteady, nickering illumination, an
injurious strain may be put on the eye, in
its endeavor to accommodate itself to the
varying intensity of illumination. More-
over, a good artificial illuminant must be
reliable. In other words, it must be able
to furnish light without constant attention,
and without danger of becoming acciden-
tally extinguished.
Ordinary sunlight, as is well known, con-
sists of a mixture of a great variety of
colors. The color of natural bodies is
entirely due to the light which falls on
them. For example, a green leaf, illumined
by sunshine, possesses the power of absorb-
ing nearly all the sunlight but the green
light, which it gives off. For such a green
leaf to appear of its natural color by arti-
ficial light, this light must possess not only
ARTIFICIAL ILLUMINATION. 9
the exact tints of the various greens which
it throws off in sunlight, but also the
various proportions of such tints. The
same is true for other colors. A good
artificial illuminant, therefore, to be able
to replace sunlight, should possess not only
all the colors of the sunlight, but should
also possess these colors in nearly the same
relative intensities as does sunlight.
A requirement of an artificial illum-
inant, which is very difficult to fulfil, is
that the light it produces shall not be
localized, but shall uniformly illumine the
spaces to be lighted. In order to meet
this requirement, the artificial illuminant
must readily yield itself to subdivision,
that is, instead of giving a great amount of
light at each of a few points, it should
give a smaller 'quantity of light at a great
number of separate points.
10 ELECTRIC INCANDESCENT LIGHTING.
The light of all artificial illmninants
is accompanied by heat, and, in nearly every
case, the amount of heat so accompany-
ing the light is very greatly in excess of
what is essentially necessary. Such is
true even in the case of sunlight. A nota-
ble exception, however, is found in the
phosphorescent light of the firefly and the
glow-worm, which yield light practically
devoid of wasteful heat ; i. e., non-luminous
Iwat.
•
Another requirement of an artificial
illuminant is that it shall readily yield
itself to control, that is to say, that it shall
easily be turned on or off at a distance,
otherwise, the location of the separate
sources of light would necessarily be
limited and thus prevent the most advan-
tageous distribution. Finally, a good arti-
ficial illuminant should be readily adapta-
ARTIFICIAL ILLUMINATION. 11
ble to the purposes of decoration, or
it will otherwise be shorn of many of its
advantages.
The principal artificial illuminants in
use at the present day, are, coal oils, gas,
candles, arc and incandescent electric
lamps.
t
The incandescent lamp is now so
generally employed as an indoor artificial
illuminant, and its advantages for this
purpose are so evident, that it is scarcely
necessary to show how much more fully it
meets the requirements of a good illuinin-
ant than any of its predecessors. It suffices
to say that it does not vitiate the air, is re-
liable, steady, can be maintained without
any trouble on the part of the consumer,
and readily yields itself both to sub-division
and to decorative effects. When used on
12 ELECTRIC INCANDESCENT LIGHTING.
a comparatively large scale it can compete
favorably as regards cost witli any other
artificial illuininant, and, even when em-
ployed on a small scale, its manifold advan-
tages will frequently more than compensate
for the disadvantage of a slightly increased
cost.
As regards the ability of an incandes-
cent lamp to produce a color of light akin
to sunshine, it must be confessed that it is
far from realizing this object, but this
objection also exists to even a greater
degree in nearly all other artificial illu-
minants. Coal oil, candles, gas, and oil
lamps generally, do not produce a light so
nearly akin to sunshine, as does the incan-
descent lamp. As we shall see later on
the light emitted by an incandescent lamp
can be made to approach more closely the
characteristics of sunlight by increasing
ARTIFICIAL ILLUMINATION. 13
the temperature of the glowing carbon
filament.
When the incandescent electric light
was first introduced on a commercial scale,
a belief existed that the extended intro-
duction of this illuminant would be
attended by many dangers, and it is true
that, when negligently installed, such
dangers do exist, but experience has amply
proved that when installed with ordinary
care, there is far less danger from the use
of incandescent lighting than from the use
of any other artificial illuminant.
So far as the safety of incandescent
lighting for fire risk is concerned, a re-
port of the Massachusetts Insurance Com-
missioner, giving data extending from 1884
to 1889, as to the origin of 12,935 confla-
grations which took place in Massachusetts
14 ELECTRIC INCANDESCENT LIGHTING.
during that time, will speak favorably for
this illuminant. Of this number of fires
only 42 were attributed to electric wires.
The breakage and explosion of kerosene
oil lamps produced in the same time about
22 times as many fires, while the careless
use of matches produced more than 10
times as many.
As to the danger to life, the evidence is
still more in favor of electricity as an arti-
ficial illuminant. While it is true that
fatal accidents do occur from contact with
high-pressure wires, yet the proper instal-
lation of a system practically precludes the
possibility of such accidents. Besides the
immunity from fire which the electric
lamp ensures, owing to the fact that the
filament is sealed in a glass chamber, thus
preventing contact with combustibles, it
also possesses the marked advantage that
ARTIFICIA
it dispenses entire! ;
dangerous friction matched
readily capable of being lighted and ex-
tinguished at a distance by the mere turn-
ing of a switch. To thoroughly appreciate
the danger to life from electricity, and to
ascertain the extent to which it can be
avoided, it will be necessary to understand
some of the leading elementary principles
of electrical science, and we will, there-
fore, postpone this consideration to a later
chapter.
An extended experience with the incan-
descent lamp, has clearly established its
advantages over gas or coal oil. Take, for
example, convenience of night work, in a
large textile manufactory. It is a well
known fact that the air of such buildings,
when illumined by gas, so rapidly heats and
becomes so rapidly vitiated during summer
16 ELECTRIC INCANDESCENT LIGHTING.
time, that the amount and character of the
work the workmen are capable of per-
forming, necessarily suffers as compared
with day work. Since the introduction of
the incandescent lamp, however, the ab-
sence of increase of temperature and
impure air, on prolonged runs, is so marked,
that in many instances it has been found
that the amount and character of the night
work compares very favorably with day
work during the summer season. This
circumstance has frequently occasioned
surprise to the public, since the tempera-
ture of the glowing filament in an incan-
descent lamp is quite high, but it must be
remembered that while an incandescent
lamps emits no gas, the filament not being
con sinned, a gas burner not only gives off
all the gas that would escape from it were
it unlighted, but, in addition, a much
greater volume of heated air. Every cubic
ARTIFICIAL ILLUMINATION. 17
foot of ordinary illuminating gas requires
for its combustion, the oxygen from about
20 cubic feet of air, so that a burner con-
suming 5 cubic feet per hour combines with
the oxygen of about 100 cubic feet of air
per hour. Besides vitiating such air by
the products of combustion, and raising its
temperature by the great amount of heat
given off, this vitiation and increase of
temperature requires much more thorough
ventilation than when the incandescent
lamp is employed. In cases where the
character of night work requires keen
sight, a highly heated vitiated air has an
injurious influence upon the eye.
CHAPTER II.
EARLY HISTORY OF INCANDESCENT LIGHTING.
ELECTRICITY was first employed as an
illuminant in the arc lamp. In this lamp,
as is well known, two rods of carbon are
first brought into contact, and then gradu-
ally separated while a powerful electric
current is passing between them. A
cloud of incandescent carbon vapor, called
the voltaic arc is thus established, and the
ends of the carbons, particularly that of
the positive carbon, or the one from which
the current flows, becomes intensely
heated, forming a brilliant source of light.
The greatest difficulty attending the
practical application of the arc system of
18
EARLY HISTORY. 19
lighting, arose from the fact that the arc
lamp was too intense a source of light for
most practical purposes within doors, and
could not readily be subdivided into a
number of smaller units ; for, even if the
space to be lighted would require all the
light emitted by a single arc lamp, yet
this light, coming from practically a single
point, would necessitate the production of
disagreeable shadows.
From very early times in the history of
the application of electricity to lighting,
the idea was conceived by various invent-
ors of employing continuous conductors,
instead of the discontinuous conductors in
arc lighting. These continuous conductors
were rendered incandescent by the pas-
sage through them of an electric current ;
or, in other words, the current heated
them to a white heat. Various substances
20 ELECTRIC INCANDESCENT LIGHTING.
were employed for this purpose, at first in
air, such as platinum or iridium wires, or
other metals. These lamps, however, were
not found to give practically useful
results, since if the current strength was
made sufficiently great to raise them to
a white heat, although the light they
would then emit would be quite satisfac-
tory, yet disintegration would occur, from
the free contact of the air, and would soon
result in the destruction of the lamp.
Moreover, the temperature at which the
glowing wire becomes white hot, is so
very nearly the temperature of its melting
point, that any accidental increase in the
-current, even to a slight degree, would
result in the fusion of the wire and the
rupture of the lamp. If to avoid these
difficulties, the lamp was burned at a
lower temperature, the light it emitted
was unsatisfactory, not only on account of
EARLY HISTORY. 21
its lessened intensity, but also because it
was distinctly red and dull.
During the year 1878, a great improve-
ment was effected in the platinum incan-
descent lamp, whereby platinum was ob-
tained in a condition in which it was
possible to employ safely much greater
current strength without rapid deteriora-
tion. This process consisted essentially in
sending a gradually increasing current of
electricity through the wire while in a
vacuous space. As is well known, plati-
num possesses in common with some
other metals, the power, of absorbing or
occluding within its mass, air or other
gases. When such a wire is heated by the
sudden passage of an electric current, this
occluded gas is liberated explosively, and
the wire thereby becomes cracked, fissured,
and rapidly disintegrates. It was dis-
22 ELECTRIC INCANDESCENT LIGHTING.
covered that if the wire, while in a vacuous
space, was subjected to the passage of a
gradually increasing current of electricity,
first beginning with a weak current, that
the occluded gas was slowly liberated and
that if, when a very high vacuum was ob-
tained, the wire was maintained for a few
moments at a temperature only a little
below that of its melting point, it became
physically changed, free from cracks and
had its point of fusion raised. Moreover,
the surface of the wire was altered, so that
a given current strength produced a greater
amount of light. The invention of a lamp
consisting of such a treated platinum wire
in a vacuous space was a marked step in
the production of an artificial electric illu-
niinant. Nevertheless, even the improved
conductor did not answer commercial re-
quirements, so that the improved platinum
lamp did not come into any extended use.
EARLY HISTORY. 23
To make the incandescent lamp practi-
cable it was necessary to enclose the wire
in a transparent chamber from which the
air had been removed. Such an improve-
ment would render the lamp more efficient
for the following reasons :
(1) The absence of air would prevent
the rapid disintegration of the wire.
(2) The absence of air would prevent
the re-absorption or occlusion of air by the
heated wire.
(3) The absence of air would prevent
loss of heat, and much electric power would
thereby be saved since a smaller current
would bring the wire to the incandescent
temperature.
(4) The absence of drafts of air would pre-
vent irregular coolings and heatings of the
wire, with consequent fluctuations of light.
(5). The wire would be protected from
mechanical injury.
24 ELECTRIC INCANDESCENT LIGHTING.
The early history of the art contains the
names of numerous inventors who applied
the foregoing principles for the purpose of
producing satisfactory incandescent lamps.
Without attempting to record in full all
these early attempts, we will briefly allude
to some of the more prominent of the early
inventors in this field of artificial illumina-
tion.
In 1841, Frederick De Moleyn patented
in England a process for the production of
an incandescent lamp, based on the incan-
descence of platinum wire placed inside an
enclosing glass chamber from which the
air had been exhausted.
In 1849 Petrie produced an incandes-
cent lamp in which short thin rods of
indium, or its alloys, were used as the in-
candescent material.
EARLY HISTORY. 25
Considerable excitement was created in
scientific circles in France, in 1858, by the
announcement to the French Academy of
Sciences by one of its members, of an
improvement in incandescent lamps made
by one De Changy. De Changy employed
an incandescent platinum lamp of the form
shown in Fig. 1, in which a wire of plati-
num G ', is heated to incandescence by the
passage of the current through it. The
wire was enclosed in an exhausted glass
vessel. The excitement caused by this
lamp was due to the claim made by De
Changy that he had solved the problem for
the successful sub-division of the electric
light. His lamp never came into com-
mercial use.
The great improvement in lamps of this
type was made by the substitution of
carbon wires for platinum wires. The
26 ELECTRIC INCANDESCENT LIGHTING.
FIG. 1. — DE CHANGY'S LAMP.
honor of this discovery appears to be due
to an American by the name of J. W.
Starr, who employed plates of carbon
EARLY HISTORY. 27
placed inside a glass vessel containing a
Toricellian vacuum. Starr associated him-
self with an Englishman of the name of
King, and an English patent was obtained
under the name of King. It is from this
fact that the Starr lamp is not infrequently
alluded to in literature as the King lamp,
or sometimes, as the Starr-King lamp.
Starr gave a successful demonstration in
England with a candelabra of 26 of his
lamps, the number of States then in the
American Union. His untimely death,
while on his return voyage to this coun-
try, retarded the progress of this inven-
tion. The Starr-King lamp is illustrated
in Fig. 2, where A, is a rod of carbon
clamped at its extremities to the rods G and
J9, and situated in a Torricellian vacuum
above the surface of mercury in a barometer
tube, the current passed through the rod A
raising it to the incandescent temperature.
28 ELECTRIC INCANDESCENT LIGHTING.
FIG. 2.— STARR-KING LAMP.
The next invention worthy of notice
was that of Lodygnine, who produced a
carbon incandescent lamp in 1873. This
invention was deemed by the St. Peters-
EARLY HISTORY. 29
burg Academy of Sciences, of sufficient
merit to warrant the award of a special
prize. Lodyguine's lamp used needles
of retort carbon terminating in blocks
and placed inside an exhausted glass
globe. In practice, in order to avoid the
expense of exhausting the globe, Lody-
guine sometimes left air within the globe
and then sealed it hermetically, depending
on the consumption of the residual air
by the heated carbons to produce a space
devoid of oxygen. This form of vacuum,
however, was not found suitable for com-
mercial purposes. Lodyguine's lamp was
improved by Kosloff, who introduced
modifications for the supports of the
carbon pencils.
In 1875, Konn produced a lamp very
similar to Lodyguine's. Konn's lamp is
shown in Fig. 3. The glass globe J3,
30 ELECTRIC INCANDESCENT LIGHTING.
FIG. 3.— KONN'S LAMP.
EARLY HISTORY. 31
closed at the top, rests with its base upon
a washer of soft rubber placed in the lamp
base, and pressure is brought by the screw
block Z, in such a manner as to maintain
an air-tight joint. The lamp was ex-
hausted through the orifice K. The
metallic base formed one terminal of the
lamp, while the rod D, passing through
the base A, in an insulating tube, formed
the other terminal. Between the terminals
F and D, were two rods of carbon, so
arranged that one only was in circuit at
any one time. The thinner central por-
tion of these rods was heated to incan-
descence by the electric current passed
through the lamp, while the thicker por-
tions O, O, served to cool the extremities,
at the points of contact with the support-
ing frame. After the first rod was con-
sumed, it dropped out of position, and
allowed the second rod to replace it. When
82 ELECTRIC INCANDESCENT LIGHTING.
FIG. 4.— BOULIGUINE'S LAMP.
EARLY HISTORY. 33
the second rod was consumed, the lamp be-
came automatically short-circuited. The
difficulty with this lamp was the too rapid
consumption of the carbons.
In 1876 an improvement was made by
Boulyguine, who produced a lamp intended
to obviate this difficulty. In Boulyguine's
lamp, the carbon is automatically fed
upwards as it consumes away. A section
of this lamp is shown in Fig. 4. Here the
lower holder of the carbon rod has a slide
in which guides press upwards under gravi-
tational force, and feed the carbon filament
up against the upper electrode as the short
rod or filament is consumed.
Fig. 5, shows a form of Sawyer lamp.
Here the light-giving medium consists of
an incandescent carbon pencil at the top
of the lamp. This consumes away slowly
34 ELECTRIC INCANDESCENT LIGHTING.
FIG. 5. — SAWYER'S LAMP.
EARLY HISTORY. 35
in operation at its contact with the upper
carbon block, at the rate of about ^th
FIG. 6.— FARMER'S LAMP.
inch per hour. The carbon pencil was
about eight inches in length and was
forced upwards as consumption took place.
The lamp was mounted in a glass globe
partially exhausted.
36 ELECTRIC INCANDESCENT LIGHTING.
Fig. 6, shows an early form of carbon
lamp introduced by Farmer in 1879.
Here a short horizontal carbon rod is
gripped between two large metallic blocks
in the exhausted globe.
All the lamps we have hitherto de-
scribed have proved commercially useless,
in the forms in which they were presented.
What might have been the effect of intro-
ducing slight modifications into such lamps
is beyond the province of this volume to
consider. There can, however, be but
little doubt that the want of a cheap
source of electric current stood as much
in the way of the progress of electric in-
candescent lighting, as any glaringly in-
herent difficulty in the structure of the
lamps themselves.
Before leaving this brief history of the
EARLY HISTORY. 37
early forms of incandescent lamps, it may
be well to discuss some of the many forms
that were devised for burning in the open
air. These lamps strictly speaking were of
a type which may be best described as semi-
incandescent lamps. They operate essen-
tially on the principle of sending a current
through a slender rod of carbon pressed
against a block or larger mass of the same
material. Under these circumstances the
slender rod is raised to intense incandes-
cence, especially at its point of contact
with the larger electrode, where, in reality,
a miniature arc is formed. Various de-
vices were employed in lamps of this type
for feeding the carbon rod or pencil as it
was gradually consumed. In some lamps
an enclosing glass chamber was employed
in order to reduce the consumption of the
carbon as far as possible. This type of
lamp passes insensibly into lamps of the
38 ELECTKIC INCANDESCENT LIGHTING.
purely incandescent form. In fact, as will
be observed, some of the preceding lamps
were of the seini-incandescent type.
FIG. 7. — REYNIEB'S LAMP.
One of the most successful of the early
lamps of the semi-incandescent type was
that of Keynier. In the Reynier lamp a
movable rod of carbon C\ Fig. 7, of small
diameter, supported as shown, rests against
EAEL'
a contact block
to limit the incandesce
lower extremity, a lateral
FIG. 8. — REYNIER'S LAMP.
is kept pressed against the rod <7, by
means of a spring R. The current, there-
fore, passes through the lateral contact
piece Z, through the slender rod to the
contact block of graphite J5, so that only
the portion of the rod between these
40 ELECTRIC INCANDESCENT LIGHTING.
points is rendered incandescent, by far the
greatest amount of light coming from the
tip or extremity J. Fig. 8, is a semi-dia-
grammatic view of the same lamp ar-
ranged, however, to be used in connection
with an external globe. The two binding
posts at the top of the lamp form its
terminals. No vacuum is necessary with
this form of lamp since the slender rod of
carbon is fed downwards as the point con-
sumes away.
In 1878 an Englishman, named Werder-
mann, took out a patent for a lamp
founded on a somewhat similar principle.
In the Werdermann lamp, shown in Fig. 9,
as in the Reynier lamp, means were devised
for pressing a slender carbon rod against a
fixed electrode, and feeding this rod up-
wards as it consumed away. Werdermann
employed for his fixed electrode a carbon
EARLY HISTORY. 41
disc and advanced a slender carbon rod by
the action of a weight or counterpoise. In
order to avoid the casting of shadows
downwards, a notable defect in the
FIG. 9. — WERDERMANN'S LAMP.
Reynier lamp, Werdermann inverted his
lamps, as shown in the figure. Lamps of
the combined Reynier- Werdermann type
at one time were in fairly successful practi-
cal use, and formed one of the features
42 ELECTRIC INCANDESCENT LIGHTING.
of the Electrical Exhibition of 1881 at
Paris.
The advent of the really successful in-
candescent lamp dates from about 1879,
and from this year, the growth of the in-
candescent electric lamp industry has been
extremely rapid. The year 1879, there-
fore, marks the entrance to the epoch of
contemporaneous history, into which we
are unable to enter at the. present time.
We will, therefore, close this extremely
brief history of the art, referring the
reader for further particulars to contem-
poraneous literature.
CHAPTER III.
ELEMENTARY ELECTRICAL PRINCIPLES.
THE light emitted by the glowing fila-
ment of an incandescent lamp is one of the
effects produced by the passage of the
electric current through it. Before pro-
ceeding to a discussion of the operation
of the incandescent electric lamp, it will
be necessary to consider briefly the lead-
ing elementary principles concerning the
production and flow of electricity.
An electric flow is always produced by
the action of. what is termed electromotive
foi*ce, generally contracted E. M. F. No
electric source produces electricity directly.
44 ELECTRIC INCANDESCENT LIGHTING.
What is produced is an electromotive
force, and an E. M. F., in its turn, pro-
duces an electric current when permitted
to do so. Thus, in such an electric
source as a voltaic battery, an E. M. F.
is produced, and this independently of
whether it is permitted to establish an
electric current or not. If an E. M. F. be
permitted to act, it will invariably establish
an electric current. In order to do this
a conducting path must be opened to it.
Such a conducting path is called a circuit.
For convenience, it is universally agreed
to regard electricity as leaving an electric
source at the point called the positive
terminal or positive pole, and re-entering
the source, after having passed through
the circuit, at a point called the negative
terminal or negative pole / that is to say,
electricity is regarded as flowing from the
ELEMENTARY ELECTRICAL PRINCIPLES. 45
positive to the negative pole, in the ex-
ternal portion of the circuit, and from the
negative to the positive pole, in the internal
portion of the circuit, that is within the
source.
Since all electric currents are due in
the first instance to the action of an E.
M. F., it is necessary to obtain definite
ideas concerning this force. E. M. F. is
to electric flow the analogue of ordinary
pressure to the flow of liquids or gases;
that is to say, a flow or current never
occurs in a liquid or gas, except as the
result of difference of pressure acting
on it, and the liquid or gaseous flow is
always directed from the point of higher
pressure to the point of lower pressure.
So in the electric circuit, a current of elec-
tricity never flows unless there be a differ-
ence of electric pressure, or an E. M. F.,
46 ELECTRIC INCANDESCENT LIGHTING.
and the electric flow is always directed
from the point of higher to the point of
lower electric pressure.
E. M. F. is measured in units called
volts. Thus, a voltaic cell, of the type
known as a Leclanche cell, has an E. M. F.
of, approximately, 1 1/2 volts. Such a
cell is shown in Fig. 10. A carbon plate
CGC, is connected to the positive terminal
JP, while a rod of zinc Z Z, is connected to
the negative terminal N. This cell pro-
duces, when in good order, an E. M. F.
of about 1 1/2 volts, even when on open
circuit, that is when the terminals jP, and
N, are disconnected as shown, and are,
therefore, producing no current. If, how-
ever, these terminals be connected through
a conducting path, or circuit, a current of
electricity will flow under the pressure of
1 1/2 volts, from the positive terminal jP,
ELEMENTARY ELECTRICAL PRINCIPLES. 47
through the external portion of the circuit
back to the negative pole N, and from the
FIG. 10. — LECLANCHE CELL.
zinc plate through the solution S xSJ to the
positive terminal, thus completing a cir-
cuital path.
48 ELECTRIC INCANDESCENT LIGHTING.
When an E. M. F. greater than 1 1/2
volts is required, from cells of this type, a
number of such cells are so connected to-
gether as to act as a single source. Such a
combination of cells is called a battery. If,
for example, two such cells be joined to-
gether, with the negative pole of one cell
connected to the positive pole of the other,
they would produce a battery of three volts
E. M. F., and 100 such cells connected in
this way, or connected in series, as it is
called, would produce a battery having
a total E. M. F. of approximately 150
volts.
Voltaic batteries are seldom employed
for supplying current to incandescent
lamps except on a small scale, for the
reason that the cost of the current so pro-
duced would be excessive. In almost all
cases of electric lighting, the E. M. F. is
ELEMENTARY ELECTRICAL PRINCIPLES. 49
obtained from a dynamo-electric generator,
a machine for producing electric power
by the expenditure of mechanical power.
Dynamo-electric machines, suitable for
incandescent lighting, will be considered
in a subsequent chapter.
When an E. M. F. is permitted to act
upon a closed conducting path or circuit,
the value of the current strength produced
therein, that is, the amount of electricity
which flows in a given time, will depend
upon two circumstances ; namely,
(1) Upon the value of the E. M. F.;
i. e.j the number of volts.
(2) Upon a property of the circuit
called its electric resistance, that is to say,
the opposition it offers to the flow or pas-
sage of electricity through it. Thus, if
the cell shown in Fig. 10 has an E. M. F.
of 1 1/2 volts, and produces a much
50 ELECTRIC INCANDESCENT LIGHTING.
greater current strength in fa one circuit
than in another, it is because the latter
circuit offers a greater resistance to
the flow of electricity than the former.
Some idea of the action of resistance, in
opposing the flow of electricity in an
electric circuit, can be had by consider-
ing the analogous case of the flow of gas
through a pipe or main. A long narrow
pipe will obviously offer a greater resist-
ance to the passage of gas through it,
from a reservoir, than a larger short pipe.
The resistance of an electric circuit is
measured in units of electric resistance
called ohms. The ohm is a resistance such
as is offered by about two miles of ordinary
overhead trolley wire, or about one foot of
very fine copper wire, No. 40 A. W. Gr., which
3
has a diameter of about — Tth inch.
PROPERTY Of';
ELEMENTARY ELECTI$fr$£ PRINCIPLES. 51 „>
The electric
very important quantity.
resistances will, therefore, be of interest.
An ordinary Bell telephone receiver has
a resistance of about 75 ohms.
An ordinary telegraph sounder has a re-
sistance of about 2 ohms.
An ordinary incandescent lamp, of 16
candle-power, intended for 11 5- volt circuits,
has a resistance, when lighted, of about
250 ohms.
The electric resistance of a conductor
depends upon its length, its cross-sectional
area, and upon the nature of the material
of which it is composed. The resistance
increases directly with the length of con-
ductor, and inversely as the cross-sectional
area. Thus, if a mile of trolley wire,
weighing 1,690 Ibs. has a resistance of
1/2 ohm, two miles will have a resistance
52 ELECTRIC INCANDESCENT LIGHTING.
of one ohm, and 10 miles, a resistance of 5
ohrns. If the above trolley wire be
doubled in cross-sectional area, and, conse-
quently, in weight, so as to weigh 3,380 Ibs.
per mile, its resistance will be only 1/4
ohm per mile, or 2 1/2 ohms in 10 miles.
The resistance of wires, each a mile long,
and of the same diameter as ordinary
trolley wire, (0.325") would be different
with different materials. Thus, while such
-a wire of copper, has a resistance of about
1/2 an ohm, an iron wire of the same size,
length and diameter, would have a resist-
ance about 6 1/2 times greater, or about
31/4 ohms, and when of lead, about 12
times greater, or about 6 ohms. Conse-
quently, in order to compare the relative
resistance of wires of different materials, it
is necessary to refer each material to a com-
mon standard of dimensions. This is done
ELEMENTARY ELECTRICAL PRINCIPLES. 53
by considering the resistance of a wire
having unit length and unit cross-sectional
area, that is to say, the resistance of a wire
having a length of one centimetre and a
cross-sectional area of one square centi-
metre. The resistance of a wire with such
unit dimensions is called its specific resist-
ance or its resistivity. The resistivity of
standard soft copper is 1.594 microhms;
i. <?., 1.594 millionths of one ohm. If then
a wire has a length of one kilometre (100,-
000 centimetres) and a cross-sectional area
of 1 square centimetre, it would have a re-
sistance of 1.594 x 100,000 microhms =
1.594 X 100,000
1,000,000 ai°94 ohm> at the tem-
perature of melting ice.
The resistivity of a wire is the scientific
standard for comparing its resistance with
that of other wires of the same length and
54 ELECTRIC INCANDESCENT LIGHTING.
cross-sectional area. In English-speaking
countries, where the centimetre is not in
general use as the unit of length, the stand-
ard called the circular-mil-foot is very
commonly employed. A circular-mil is
the area of a wire one mil, or
an inch, in diameter. This is not to be
confused with the area of such wire ex-
pressed in square inches. The number of
circular mils cross-sectional area in any
wire, is obtained by squaring its diameter
in mils. Thus a wire half an inch in
diameter, would be a wire of 500 mils
diameter, and the number of circular mils
cross-sectional area in such wire would
be 500 X 500 = 250,000. Such a wire
would have a resistance per foot of ^ '' ^
250,000
= 0.000,00414 ohm. A circular-mil-foot
of standard soft copper at 20° C. has a
resistance of 10.35 ohms.
ELEMENTARY ELECTRICAL PRINCIPLES. 55
The resistivity of a material varies with
its temperature. In the case of most me-
tallic substances, the resistivity increases
as the temperature increases. Thus, the
resistivity of copper wire is about forty-
two per cent, greater at the boiling point of
water, than at its freezing point ; or, a cop-
per wire would have about forty-two per
cent, more resistance, at the temperature of
the boiling point of water, than at its freez-
ing point. The resistivity of insulating
materials, however, diminishes as the tem-
perature increases. Carbon behaves in
this respect like an insulating material, its
resistivity diminishing as its temperature
increases. An ordinary incandescent lamp
has about twice as much resistance when
cold, as it has when heated by the electric
current to an incandescent temperature.
The current strength, which passes
56 ELECTRIC INCANDESCENT LIGHTING.
through any circuit, is measured in units
of electric flow called amperes. As in the
case of a current or flow of gas, we may
estimate the flow as so many cubic feet of
gas per minute, or per second, so an electric
flow, may be estimated as so many units
of electric quantity per second. The unit
of electric quantity is called the coulomb,
and is the quantity of electricity, which
would flow in one second through a cir-
cuit having a total resistance of one ohm,
when under a pressure of one volt. A
rate of flow equal to one coulomb-per-
second is the unit of electric flow or cur-
rent, and is called the ampere. A current
of 10 amperes, therefore, means a flow, or
transfer, of 10 coulombs of electricity in
each second. The ordinary incandescent
lamp of 16 candle-power, operated at a
pressure of 115 volts, requires to be sup-
plied with a current of about 1/2 ampere.
ELEMENTARY ELECTRICAL PRINCIPLES. 57
The relations existing in any circuit
under a given resistance and E. M. F. are
readily determined by reference to a law
discovered by Dr. Ohin, and named Ohm's
Law. This law may be expressed briefly
as follows :
The current strength in any circuit is
directly proportional to the total E. M. F.
acting in the circuit, and inversely pro-
portional to tlie total resistance in the
circuit.
If the pressure or E. M. F. be measured
in volts, and the resistance in ohms,
the current strength that passes in am-
peres may be briefly expressed as follows:
Volts
Amperes = ^-r-
Ohms
For example, if a circuit comprise a
dynamo and an external path consisting
of lamps and wires, so that the E. M. F.
in the dynamo is 100 volts, and the resist-
58 ELECTRIC INCANDESCENT LIGHTING.
ance of .. the circuit is 2 olnns, then the
current strength passing through the cir-
cuit will be - - = 50 amperes.
a
Again, if an incandescent lamp has a
resistance (hot) of 220 ohms, and is con-
nected to a pair of mains between which
the electric pressure is steadily maintained
under all circumstances at 110 volts, then
the current, which will pass through the
110 1
lamp irom the mains will be
ampere.
We have already alluded to the fact
that a current never flows in water unless a
difference of pressure exists therein. In
order to produce this difference of pressure,
the water has usually to be raised to a
higher level ; and, to do this, energy is re-
quired to be expended, or work performed,
f PROPERTY OF
ELEMENTARY ELECTK~ ~"
on the water. The
tricity. In order to produce an electric
flow, energy requires to be expended, or
work produced, by the electric source.
Work is never done unless force acts
through a distance. A force that is
merely producing a pressure on a body,
but no motion of the body, is naturally
performing no work. When, for example,
a block of granite is raised through a
vertical distance against the earth's gravi-
tational force, work is done. The amount
of such work is measured by the force
which acts, multiplied by the distance
through which it acts. For example, if
a block weighing 200 pounds be raised
through 10 feet, an amount of work will
be done expressed in a common unit of work
called the foot-pound, as being equal to
200 pounds X 10 feet = 2,000 foot-pounds.
60 ELECTKIC INCANDESCENT LIGHTING.
If this weight when raised be placed on
a shelf or other support, it would still
produce a pressure, of 200 pounds upon
the support, but would be doing no
work.
Another unit of work is called the joule,
and is approximately equal to 0.738 foot-
pound, so that a foot-pound is about thirty-
five per cent, greater than a joule, or 1
foot-pound = 1.356 joules. A man weigh-
ing 150 pounds, and raising his weight
through a distance of 100 feet by walking
upstairs/ necessarily performs an amount
of work against gravitational force equal
to 100 X 150 = 15,000 foot-pounds =
20,340 joules.
When an electric current passes through
a circuit, work is always done by the E.
M. F. which drives the current through the
ELEMENTARY ELECTRICAL PRINCIPLES. 61
circuit. The amount of work done by
the E. M. F. is expressed in joules, as the
product of the pressure and the quantity
of electricity it drives. Thus, if a quan-
tity of electricity equal to 100 coulombs,
passes through an electric circuit under
a pressure of 50 volts, then the amount
of work which will be expended in
the passage, will be 50 X 100 = 5,000
joules = 3,690 foot-pounds. A joule is,
therefore, equal to a volt-coulomb. This
work will always be taken from the
source of the driving E. M. F. Thus if
the dynamo in the last example, supplied
the E. M. F., then an amount of work
equal to 5,000 joules will have been taken
from the dynamo, and this amount of work
must have been delivered to the dynamo
through its driving belt. Or, if the E.
M. F. had been supplied by a voltaic
battery, this amount of work would have
62 ELECTRIC INCANDESCENT LIGHTING.
been supplied at the expense of the zinc and
the liquid in the battery ; that is to say a
certain amount of zinc would have been
consumed, thereby liberating at least 5,000
joules of work.
It is necessary to distinguish carefully
between the amount of work performed in
any case and the rate at which such work
is performed ; for example, so far as the
result attained is concerned, the same
.amount of work is done, when the man be-
fore alluded to, weighing 150 pounds, raises
himself through a height of 100 feet by
the stairs, whether he performs this work
in one minute or in ten minutes, but his
rate-of-doing-ioork, or his activity would be
10 times greater in the former than in the
latter case. In the former, his activity
would be 15,000 foot-pounds-per-minute,
or 250 foot-pounds-per-second, while in the
ELEMENTARY ELECTRICAL PRINCIPLES. 63
latter case it would be only 1,500 foot-
pounds-per-minute, or 25 foot-pounds-per-
second. Activity is commonly rated either
in terms of a unit of activity called a
foot-pound-per-second, or in a unit called a
horse-power, one horse-power being equal to
550 foot-pounds-per-second, or 33,000 foot-
pounds-per-minute. Thus, in the preced-
ing case, the man would be expending an
250
activity of - = 0.455 horse-power in the
first case, and 0.0455 horse-power in the
second case.
Electric activities are similarly measured
in units of electric power or activity called
watts, the watt being equal to an activity
of a joule-per-second, or a volt-coulomb-per-
second, or a volt-ampere.
Thus if a current of 1/2 ampere passes
64 ELECTRIC INCANDESCENT LIGHTING.
through an incandescent lamp under a
pressure of 115 volts, the electric activity
or rate of expending energy in the lamp,
will be 115 x 1/2 = 57.5 watts, or joules-
per-second = 42.4 foot-pounds-per-second.
CHAPTER IV.
PHYSICS OF THE INCANDESCENT ELECTRIC
LAMP.
BEFORE proceeding to a detailed descrip-
tion of the incandescent lamp and the
methods employed in its manufacture, it
will be advisable to obtain a general in-
sight into the physical laws controlling its
operation.
Briefly speaking, the operation of the
incandescent electric lamp is based upon
the principle of raising the temperature of
a thin thread or filament of some refrac-
tory substance, such as carbon, as far as
is consistent with practical working. In
G5
66 ELECTRIC INCANDESCENT LIGHTING.
order to avoid the oxidation of the carbon
filament, it is enclosed in a glass lamp bulb
in which a vacuum is maintained. Briefly,
the function of the electric lamp is to
convert electric energy economically into
.luminous energy or light.
Light may be defined in two distinct
senses.
First, subjectively, as the physiological
effect produced through the eye on the
mind by the radiations emitted by a
luminous body. In this sense, lights
differ in their color and intensity ; that is,
we are conscious of different perceptions
both of color and of intensity.
Second, objectively, as the physical
cause which produces the sensation of
light. In this sense light can have an
existence independent of the eye. Objec-
tively, light consists of rapid vibrations or
PHYSICS OF INCANDESCENT LAMP. 67
to-and-fro motions in a medium called the
universal or luminiferous ether, which
permeates all substances and spaces.
The luminiferous ether is believed to
be an extremely tenuous and highly elastic
medium, which not only fills interstellar
space, but which even permeates the
densest forms of matter. In this sense
any vibrations, or to-and-fro motions, of the
luminiferous ether, even though not capable
of affecting the eye, are properly spoken
of as light, but of course the light, which
it is the function of the incandescent lamp
to produce for the purposes of illumina-
tion, must necessarily be light in the
physiological sense.
The rapidity of the oscillations in the
ether which constitute light is usually
enormously great. This frequency has
68 ELECTRIC INCANDESCENT LIGHTING.
been indirectly measured up to over 800
trillions; i. e., over 800,000,000,000,000
double vibrations per second, and they
may exist down to comparatively low fre-
quencies, although they have only been
measured in heat radiation as low as about
100 trillions per second. Those frequen-
cies which lie between 390 trillions and
760 trillions, are capable of affecting the
normal eye as physiological light.
All bodies emit from their surfaces
radiations or waves into surrounding space.
A body at ordinary temperatures emits
only waves of a comparatively low fre-
qiiency, far below that which is capable
of affecting the eye physiologically as
light. As the temperature of the body is
increased, not only does it emit these
waves of low frequency more powerfully,
but, in addition, it also emits waves of a
PHYSICS OF INCANDESCENT LAMP. 69
higher frequency. When a tempera-
ture of about 500° C. is reached, the
highest frequency emitted by the body
will be just about 390 trillions of double
vibrations, or complete to-and-f ro motions,
per second, and will be just visible to the
eye. The body is then said to be at the
temperature of dull red. As the tempera-
ture is still further increased, the total
radiation of waves from the surface in-
creases, and at the same time higher fre-
quencies are introduced. These affect the
eye successively as orange, yellow, green,
blue, indigo and violet ; and, finally, all
these colors being present, the body is
said to be white hot, and has a tempera-
ture of roughly 1,500° C. If the tempera-
ture be still further increased, the lumi-
nous intensity ; i. e.9 the amount of visible
radiation per unit area of its surface, will
increase, and at the same time still higher
70 ELECTRIC INCANDESCENT LIGHTING.
frequencies will be introduced into these
vibrations, which are necessarily invisi-
ble to the eye. These are called ultra-
violet rays, or those rays above the
violet.
If we analyze sunlight, we will find that
it contains all frequencies from the lowest
that we can measure up to the ultra-violet.
The greatest intensity of these vibrations
falls within the limits of the invisible fre-
quencies, only about thirty per cent, of the
vibrations which we receive at the earth's
surface being capable of affecting the eye.
The visible frequencies combine to pro-
duce on the eye an impression, known as
white light. In other words we judge a
luminous body as white, when it emits
frequencies of the character and general
proportions which exist in sunlight, or the
relative proportions of red, green, yellow,
PHYSICS OF INCAN
and blue frequencies or
the same in this light as in
Generally speaking, nearly all sources of
artificial light emit frequencies which are
the same as those in sunlight, but the
amount of radiation in the different fre-
quencies or colors varies markedly from
sunlight. Thus a candle while emitting
all the visible frequencies of sunlight
has a marked preponderance of red rays
relative to the number of violet rays,
and consequently has a reddish yellow
tint.
The color of a body depends upon two
things ; viz. , first upon the quality of the
light it receives, that is upon the propor-
tionate distribution of the various frequen-
cies, and second upon the selective proper-
ties of its surface. When light falls on a
72 ELECTEIC INCANDESCENT LIGHTING.
colored body, the surface of the body pos-
sesses the power of absorbing some of the
frequencies and throwing off others unab-
sorbed. Thus, a blue body, illumined by
sunlight, absorbs practically all the frequen-
cies except those of the blues, which it emits.
For the blue body to be visible in its proper
tint, therefore, it is necessary that the light
which illumines it shall not only contain
blues, but shall also contain the same pro-
portionate quantities of the different tints
of blue as sunlight. If these colors be not
present in the artificial light, such a blue
body would fail to possess its character-
istic daylight color-value. Consequently,
an artificial illuminant, in order to replace
sunlight for the purpose of revealing the
proper daylight color-values of bodies,
must contain not merely the same fre-
quencies as" sunlight, but also the same
relative distribution of such frequencies.
PHYSICS OF INCANDESCENT LAMP. 73
All light serves either to distinguish the
form and shape of bodies, or their colors.
For the latter, as we have seen, an approx-
imate imitation of sunlight is necessary;
for the former, almost any frequency of
light will be sufficient. For example, if
differently colored bodies be observed in
a dark room, by means of an artificial
light containing practically but one fre-
quency, and called usually a monochro-
matic light, only the color of this light
will be visible, all other colored objects
will appear black, or devoid of color. A
yellow, monochromatic light, may be ob-
tained, for example, by the burning of
alcohol on a wick soaked in common salt.
In the pure yellow light this emits, all
yellow colored objects will appear in
nearly their true tints, while the reds,
blues, greens, and other tints will appear
devoid of color. All objects, however,
74 ELECTRIC INCANDESCENT LIGHTING.
whatever their color, will show their form
even when so illumined.
The light emitted by an incandescent
electric lamp is not capable of giving true
sunlight color-values. An analysis of the
light from the glowing filament, shows a
relative preponderance of red and yellow
rays and a deficiency of the blue and
violet, or high-frequency waves. This is
owing to the fact that the temperature
of the glowing filament cannot be raised
sufficiently high to emit the sunlight pro-
portions of the higher frequencies. Con-
sequently, reds and yellows, viewed by
incandescent lamp light, give nearer ap-
proach to their sunlight values than other
colors. It is true that by raising the
temperature of the carbon filament we
obtain a closer approach to sunlight radia-
tion, but, at the same time, for reasons
PHYSICS OF INCANDESCENT LAMP. 75
which will be subsequently explained,
the life of the lamp is greatly shortened.
A hot body loses its heat in one of four
ways ; viz., by radiation, by conduction,
by convection and by molecular transfer.
Were the glowing filament in an abso-
lutely vacuous space, it is evident that being
supported on a comparatively slender
base, apart from the trifling loss of heat
through this base or support there would
be no other means for losing the heat save
by radiation. Once admitting within the
lamp chamber even a minute trace of gas
such as would make the pressure one-
millionth part of that in the atmosphere,
then pure radiation would cease to form
the sole means of losing heat, and
molecular transfer would begin to act.
That is to say the air molecules coming
76 ELECTRIC INCANDESCENT LIGHTING.
into contact with the heated filament
would be shot off from its surface in
straight paths, and, in a highly attenuated
atmosphere, the molecules would nearly
all fly to the chamber walls without
mutual collision. Heat energy is required
to produce this motion, and the loss of
energy so occasioned forms an additional
means whereby a heated body in a rarified
atmosphere parts with its heat.
As more air is admitted to the interior
of an exhausted lamp, the collisions of the
air molecules, flying from the heated surface,
become more frequent, and the frequency
with which they are returned to the hot
surface also increases, increasing thereby
the loss of heat by molecular transfer. As
soon, however, as the mean free paths, or
uncollided path, of the molecules becomes a
small fraction of the distance between the
PHYSICS OF INCANDESCENT LAMP. 77
filament and the wall of the chamber,
the frequency with which the molecules
return to be shot off from the hot surface
increases very slowly, so that beyond this
point there is scarcely any increase in loss
of heat by molecular transfer as air is
admitted into the chamber. As more air
is gradually admitted into the chamber,
however, the loss by simple convection
increases, i. e., by a thermal stirring of the
air owing to differences of density, or local
winds, within the lamp chamber.
The radiation emitted by the ordinary
incandescent electric lamp is principally
non-luminous; that is to say the greater
portion possesses a frequency below the
inferior limit of visibility. An ordinary
16-candle-power lamp has an activity,
when in operation, of about 50 watts.
Of this activity about 48 watts are ex-
78 ELECTRIC INCANDESCENT LIGHTING.
pended in non-luminous radiation, and
only about 2 watts, or four per cent., in
luminous radiation. The problem that still
remains to be solved in the incandescent
lamp, as it does indeed in the case of all
artificial illuminants, is to produce a radia-
tion which shall lie wholly, or almost
wholly, between the limits of visible fre-
quencies. As it exists to-day, in the case
of the incandescent lamp, the energy is
expended in producing about ninety-six
per cent, of objectionable heat radiation,
and four per cent, of light radiation. Even
this, however, has a luminous efficiency
superior to that of gas and oil.
It has been found that the light emitted
by the firefly and the glow-worm is prac-
tically confined to the visible limits of
frequency. Could an incandescent lamp
be made to restrict its radiation to such
PHYSICS OF INCANDESCENT LAMP. 79
frequencies, the problem of cheap light
might be solved. Unless, however, such
light could be made to possess these fre-
quencies in sunlight proportions, it is
question able whether the problem of an
efficient illuminant would be solved from
the color standpoint.
Passing by the question of the produc-
tion of lamps which shall yield fre-
quencies characteristic of the light of the
firefly and glow-worm, there would appear
to remain but one direction in which the
same result might be reached; namely,
by obtaining a substance which possessed
marked powers of selective radiation ; that
is, a high ability to radiate waves of high
frequency, and but little for radiating
those of low frequency.
When an electric current is sent through
80 ELECTRIC INCANDESCENT LIGHTING.
the filament of an incandescent lamp, the
activity expended in the lamp will be the
product of the pressure at the lamp termi-
nals in volts, and the current, in amperes,
passing between them. This activity will
be entirely expended as heat in the sub-
stance of the filament, and this heat will
be liberated from the surface in virtue of
the increase of temperature of that surface.
As the amount of heat liberated in the
filament increases ; i. e., as the current
strength passing through it increases, the
temperature of the filament is compelled
to rise in order to emit the activity which
is developed in its mass. If the surface of
the filament be large, it will take a large
total amount of activity in the lamp to
maintain a large radiation per square
inch, or per square centimetre ; whereas,
if the surface of the filament be small, the
opposite result will be produced. Con-
PHYSICS OF INCANDESCENT LAMP. 81
sequently, a certain relation must exist
between the surface, the length, and the
cross section of the filament, in order that,
at the pressure intended for the circuit,
the radiation per square inch, or per square
centimetre, shall be sufficient to bring the
lamp to the proper temperature. The
nature of the surface of the filament deter-
mines, to a great extent, the character and
amount of the radiation which it will emit.
It might be supposed that the same activity
per square centimetre of surface would be
attended by the same temperature and rela-
tive distribution of vibration frequencies.
Such, however, is not the case, the charac-
ter of the surface having a marked influ-
ence upon its emissivity, one type of carbon
producing, with a given activity of sur-
face, a greater amount of light than another.
The temperature at which ordinary in-
82 ELECTRIC INCANDESCENT LIGHTING.
candescent lamp filaments are operated,
is estimated at about 1,345° C. If this
temperature be 'exceeded, by only a few
degrees, although the candle-power of the
filament will be materially increased, yet
the disintegration of the carbon, by a
process akin to evaporation, will be rap-
idly brought about. Thus, an increase of
2° C. is believed to be accompanied by
an increased candle-power of about three
per cent., but this gain is at a marked
decrease in the life of the lamp.
^nw
PROP,
'ffry OF
\ki<V
CHAPTER V.
MANUFACTURE OF INCANDESCENT LAMPS.
PREPARATION AND CARBONIZATION OF
THE FILAMENT.
THE most important step to be taken
in the manufacture of an incandescent elec-
tric lamp is the preparation of the fila-
ment. Of all substances which have been
employed for filaments, carbon alone has
been found to meet the requirements of
use. In the first place, carbon is highly
refractory ; that is, capable of withstand-
ing a high temperature before reaching its
point of volatilization. In the next place,
its resistivity is high, so that a high resist-
ance can be readily given to a short length
84 ELECTRIC INCANDESCENT LIGHTING.
of filament, with a correspondingly high
electric pressure at the terminals, for a
given activity in the lamp.
Carbon is universally employed for
lamp filaments, not only for the reasons
above pointed out, but because this
material readily lends itself to being
fashioned and shaped into the filament,,
prior to being subjected to various proc-
esses of carbonization, intended to ensure
a nearly pure form of hard high-resisting
carbon.
A great variety of materials have been
employed for producing the carbon fila-
ments of incandescent lamps. All these
substances agree in that they are of such a
nature as will yield a nearly pure carbon
when subjected to carbonization, i. e., to
the action of heat while out of contact
PREPARATION OF THE FILAMENT. 85
with air. Substances suitable for this pur-
pose may be divided into two sharply
marked classes ; namely,
(1) Carbons of fibrous origin, such as
bamboo.
(2) Structureless carbons, such as pastes
or mixtures of finely ground carbon incorpo-
rated with some suitable carbonizable liquid.
Among substances of fibrous origin, that
have been employed for incandescent fila-
ments, may be mentioned paper, bamboo,
bass fibre, cotton thread, and silk thread,
both of the latter substances being first
subjected to a process which is called the
parchmentizing process, by treatment with
sulphuric acid. Cellulose, treated in such
a manner as to be converted into a variety
of gun-cotton and subsequently carbonized,
has also been employed. This material,
although of fibrous origin, would by its
86 ELECTRIC INCANDESCENT LIGHTING.
treatment be rendered capable of classifica-
tion as structureless material. Without
here entering into a full description of the
various processes required for the manu-
facture of filaments from the above
materials, it will suffice to describe in
detail the process adopted in a few cases.
In the manufacture of a bamboo fila-
ment, carefully selected bamboo is em-
ployed, from which both the softer por-
tions in the interior, and the hard silicious
parts near the surface, have been removed.
The material is then cut or fashioned, by
the aid of a cutting tool, into filamentary
strips, care being taken to obtain as nearly
as possible the same area of cross section
by passing the filaments through gauges.
The filaments so prepared are then con-
verted into hard carbon by means of a car-
bonizing process.
PREPARATION OF THE FILAMENT. 87
In the production of a filament from
loosely spun pure cotton, the thread is first
cleansed from grease by boiling in soda or
ammonia, this cleansing being necessary for
ensuring uniformity of action on the part
of the sulphuric acid used in a subsequent
process. The thread is then thoroughly
washed in water and afterward soaked in
sulphuric acid of specific gravity 1.64.
The time of immersion is exceedingly
short, varying with the thickness and char-
acter of the thread, from three to fifteen
seconds. On removal from the acid, the
thread is 'again washed in water. After
removal from the water, care must be
taken to avoid warping. The dried thread
in this condition has the appearance of
catgut and is called amyloid.
The parchmentized thread has a rough
surface, and is too irregular in diameter to
88 ELECTRIC INCANDESCENT LIGHTING.
permit it to be subjected in this condition
to the carbonizing process. In order to
ensure uniformity in its diameter, and
smoothness of its surface, it is passed
through a series of draw plates, until a
sufficient amount has been removed from
it to permit the cutting tool to act on all
portions of its surface. These draw plates
are made either of steel, or of jewels, the
latter being the most frequently employed.
The thread so produced is then subjected
to the carbonizing process.
Another process, which also produces an
amyloid thread from pure cellulose, is
sometimes employed. Here pure cellulose
is dissolved at a temperature of about the
boiling point of water, in zinc chloride of
specific gravity 1.8. The viscous mass so
obtained is then forced or squirted under
pressure through a die of suitable diameter
PREPARATION OF THE FILAMENT. 89
into a vessel containing alcohol, which
causes a hardening of the filament. If
this process is properly carried on it pro-
duces a filament of uniform cross-sectional
area, so as to dispense with the necessity
for passing the thread through shaving
dies. Care has to be taken to avoid the
formation of air bubbles in the viscous
mass, which would either result in a break-
ing of the thread, or in a lack of homo-
geneity of the filament. This process is
now in very extensive use.
Another process has been invented for
the production of an artificial material
suitable for cutting or shaping into fila'ments
for incandescent lamps, and subsequent car-
bonization. This process consists essen-
.tially in converting cellulose, obtained from
cotton, into a substance known as pyroxy-
liue,.,or gun-cotton, whence it is converted
90 ELECTRIC INCANDESCENT LIGHTING.
into celluloid. The celluloid is rolled into
thin sheets. The sheets in this condition
are not yet fit to be subjected to the car-
bonizing process, and must first be again
converted into cellulose. This change is
effected by treating them with am-
monium sulphide, which produces struc-
tureless, cellulose sheets. These are then
treated with bisulphide of carbon, or tur-
pentine, to remove all traces of sulphur,
and are then ready to be cut or shaped
into filaments and carbonized.
The filaments so cut or shaped by
the preceding processes, must be sub-
jected to a carbonizing process. During
carbonization a number of changes occur
in the filament. In the first place the
filament becomes hard and brittle, and,
to a certain extent, acquires a definite
shape or form, so that it is necessary be-
PROPERTY CF
91
PREPARATION
fore carbonizing it
which it is to permanently retain. In
the next place the filament shrinks per-
ceptibly during carbonization, so that
means must be devised for permitting
this shrinking to take place without
rupture.
The means employed for the carboniza-
tion of the filament, will vary with the
shape and character of the material of
which it is made. The degree and dura-
tion of the heat employed in carbonization
will also vary with the character and
method of preparation of the material.
It will, therefore, be necessary briefly to
describe the methods employed for the
carbonization of particular characters of
filaments. In this connection, however,
it must be remarked that the manufac-
ture of the incandescent lamp, as it is
92 ELECTRIC INCANDESCENT LIGHTING.
carried out to-day, is, to a great extent,
a secret process. Therefore, such descrip-
tions must necessarily be limited to
known processes which have been tried
and which have been found successful in
practice.
In the case of the bamboo filaments, the
material, shaped as already described, is
placed in a suitable box or receptacle
formed of carbon, or other refractory sub-
stance, under such conditions that the fila-
ment shall be permitted to contract freely
while being subjected to a constant and
even tension. In Fig. 11, such a box is
shown with the filaments to be carbonized
placed in position. This box is in two
parts as shown, an outer air-tight chamber,
and an inner forming plate. The fila-
ments to be carbonized are placed in the
inner groove around the block E II, with
PREPARATION OF THE FILAMENT.
93
their extremities secured beneath the
bridge piece a h, which is fixed in an
outer frame. As the filaments contract
FIG. 11.— CARBONIZING Box.
during the process, since they cannot
draw up into the groove from the bridge,
they pull the whole block EH, forward
under the bridge, and so maintain a steady
94 ELECTRIC INCANDESCENT LIGHTING.
and uniform tension upon the filaments.
A suitable cover, provided with a flange
fitting into the outer groove, is placed on
the box. A number of such boxes are
packed together into a suitably closed flask
which is placed in the carbonizing furnace.
When the filaments to be carbonized
are prepared from cotton thread, by the
process already described, owing to the
pliability of the material, it is necessary
to give it the required shape before it is
set or hardened during carbonization.
This is accomplished by winding the
thread on a suitable block or form. As
in the case of the bamboo filament, means
must be provided for allowing shrinkage
to take place. This is accomplished by
means of a carbonizing frame, which con-
sists essentially of a suitably shaped block
and an end piece of carbon, connected to-
PREPARATION OF THE FILAMENT. 95
gether by sticks of wood. These sticks
are firmly fixed to the end piece and
pass loosely into openings in the main
block which rests on them. The threads
to be carbonized, are wrapped tightly
around the block and end piece, which
is so placed in the carbonizing box that
on the shrinkage of the thread, the
wooden sticks, whose dimensions have
been carefully selected, shrink to the same
extent, thus permitting the upper block to
slowly descend on to the end piece. A
number of such frames are packed to-
gether in a carbon box, which is itself
placed in a crucible. The space between
box and crucible is filled with powdered
carbon.
In the carbonization of amyloid threads,
it has been found that in order to obtain
the best results, the heat must be very
96 ELECTRIC INCANDESCENT LIGHTING.
gradually applied. Should the crucibles
be exposed to a sudden increase of tem-
perature, the filament contracting too
suddenly, may be broken. Moreover, the
gases produced by distillation might possi-
bly deposit sufficient carbon on the sides
of the filaments, to interfere with their
homogeneity and also to cause them to
cohere. Consequently, a pyrometer is fre-
quently used in connection with the
furnace in order to regulate its tem-
perature. The time required to effect the
carbonization will vary with the size and
character of the filaments. Ordinary
thread filameuts may require eighteen
hours for their proper carbonization, al-
though a longer time may be employed.
The period required for carbonization,
necessarily varies with the size of the fila-
ments, their character and the nature of
the furnace.
PREPARATION OF THE FILAMENT. 97
Incandescent lamp filaments are now
sometimes manufactured by a squirting
process. This is in the main a secret
process, but it consists essentially in ob-
taining an exceedingly intimate mixture
of carbonaceous materials, forming them
into a plastic mass and subjecting the
same, while in the plastic condition, to
powerful pressure, whereby they are forced
or squirted through molds in die plates.
Means are devised for preserving the shape
which it is desired that this material
should take, and then, after carefully dry-
ing the same, it is submitted to a suitable
carbonizing process. Experience has shown
that squirted filaments are capable of being
rendered perfectly uniform. In all cases
after the filaments have received the proper
carbonization, it is necessary to wait until
the furnace and its contents are cooled,
before removing them from the carboniz-
98 ELECTRIC INCANDESCENT LIGHTING.
ing boxes, both on account of their fragile
character, and for the sake of the furnace,
which is apt to be injured by the sudden
chilling.
CHAPTER VI.
MOUNTING AND TREATMENT OF FILAMENTS.
HAVING traced the filaments up to the
completion of their carbonization, we will
assume that they are ready to be removed
from the carbonizing box. As already
stated, the box and its contents are sup-
posed to have cooled down to the tem-
perature of the room. Care must be exer-
cised in removing the filaments from the
box, since the carbonization has changed
the physical nature of the material, and the
carbons are now brittle, though hard and
elastic. •
The next step in the manufacture of the
lamp now begins viz., the mounting of
00
100 ELECTRIC INCANDESCENT LIGHTING.
the filament, or placing it upon a glass sup-
port through which the leading-in wires,
or the conductors which carry the current
to the filament are sealed. The ends of
the filament are attached to the extremi-
ties of the leading-iu wires by any suitable
means. Much ingenuity has been dis-
played in obtaining a suitable mounting
for the filament, and for a long time this
mounting constituted the weak- point in
the lamp. Any collection of incandescent
lamps, embracing specimens from an early
period in the history of the art, would show
how marked the evolution has been in this
particular.
In order to obtain an idea of the rela-
tion of the various parts of «the mounted
filaments reference may be had to Fig. 12,
which shows one of the common forms of
mounting. A glass tube T, has a shoul-
TREATMENT OF FILAMENTS.
101
der blown on it at 8. Two copper wires
MI, w2) ordinarily 0.015" in diameter, are
welded to short pieces of platinum wire,
\ wt
FIG. 12. — THE MOUNTED FILAMENT.
the method generally adopted being to
hold the end of the copper wire against
the end of the platinum wire in a flame,
when they fuse together. These two
102 ELECTRIC INCANDESCENT LIGHTING.
wires constitute the leading-in wires.
These wires are then laid in the glass
tube T, and the glass is fused around the
platinum wires in a flat seal at the point
/SJ so that short projections of the plati-
num p, p, extend through the glass seal.
The glass seal is then carefully annealed.
The filament f, is now connected with its
ends to the wires p, p.
When it is remembered that the opera-
tion of an electric lamp necessarily brings
the filament to a white heat, it will be evi-
dent that means must be provided either
for preventing the joints at p, p, from
attaining a high temperature ; or, if such
temperature be attained, that the character
of the joint must be such that it will not
suffer. In any event it is clear that the
seal S9 must not be exposed to an incan-
descent temperature, and, therefore, the
TREATMENT OF FILAMENTS. 103
platinum extremities p, p, must remain
comparatively cool. The method which
is always adopted is to provide such a
cross section at the joints p, p, that, taken
in connection with the thermal conductiv-
ity of the platinum wires, the tempera-
ture of the joints will be much below the
incandescing temperature of the rest of the
filament. In the earlier forms of lamps,
in which very large currents were usually
employed, this result was accomplished by
providing massive terminals connected
with the carbon filament, possessing so
great a radiating surface that their tem-
perature was necessarily, comparatively
low. Moreover, since most of these early
forms of lamps did not employ a solid
glass seal, it was still further necessary
to reduce the temperature of the leading-
in wires at the points of entry by means
of cement.
104 ELECTRIC INCANDESCENT LIGHTING.
Platinum is employed where the glass
is fused around the leading-IB wires, for
FIG. 13.— HORSESHOE LAMP.
the reason that the expansion and con-
traction of platinum is almost the same
TREATMENT OF FILAMENTS. 105
as that of glass. The expansion of glass
rods varies from 0.0007 to 0.001 per cent,
of their length, for each degree Centigrade,
or just about the mean value of the
expansion of platinum wires. When,
therefore, glass is sealed around a plat-
inum wire, and the seal is heated, the
glass and the platinum expand together,
and, on cooling contract together. If
this were not the case, there would be
a continual tendency to shear the glass
over the platinum surface, and break
the seal both on expansion and con-
traction. Among ordinary metals iron
comes next to platinum as regards its
expansion, and iron has been employed to
some extent for the seal in incandescent
lamps.
Fig. 13, represents an early form of
lamp in which this method of connection
106 ELECTRIC INCANDESCENT LIGHTING.
was employed. A filament is connected
to large terminals of copper supported on
an insulating bridge. These terminals are
further connected with long zig-zag strips
of copper, extending to the base of the
lamp, for the purpose of ensuring a low
temperature where they pass through the
base.
In some of the earlier lamps of the
modern type the ends of the filaments
were made of enlarged cross section, so that
they were not only more readily secured
to the platinum leading-in wires, but also
by their increased mass prevented the cur-
rent from raising it to the temperature of
incandescence. In Fig. 14, a filament of
bamboo is provided with enlarged ends.
In some cases the carbon filaments, after
they had been subjected to the carboniz-
ing process, and while still in place in
TREATMENT OF FILAMENTS. 107
the carbonizing box, had their ends thick-
ened by deposits of carbon, produced from
FIG. 14.— BAMBOO FILAMENT.
108 ELECTRIC INCANDESCENT LIGHTING.
the decomposition of a- carbonizable gas,
forced into the mold at a certain stage
in the process.
A very early form of joint consisted of
a platinum bolt and nut, as shown in
Fig. 13. A small screw lamp was some-
times employed for the same purpose as
shown in Fig. 15. At one time, small
metal blocks, O, C, were employed
fastened upon the enlarged extremities of
the filament, as shown in Fig. 16. Some-
times a small socket was formed in the
end of the wire, and the carbon was
placed in this receptacle and the socket
pressed around it. In another form, the
socket joint was secured more firmly to
the carbon by covering it with an elec-
trolytic coating of carbon, or the wire was
wrapped around the carbon and subse-
quently secured to it in the same manner.
TREATMENT OF FILAMENTS. 109
FIG. 15. — EARLY FORM OP JOINT LAMP.
110 ELECTRIC INCANDESCENT LIGHTING.
This proved an excellent form of joint,
and was employed extensively for a num-
FIQ. 16.— EARLY FORM OF CLAMP.
TREATMENT OF I fc^ AMENTA 'PL fil^Y £, f
ber of years. It has, n^^er, been re-
placed by a still simpler r^gcln^wbich '
no socket is employed, but one en3ToT"t1ie
filament is abutted against the end of the
platinum wire, and carbon is deposited
around the joint. This joint is obtained
by dipping the abutting end into a suit-
able carbonizable liquid and sending a
powerful current through the abutment
wThile in the liquid. Under these con-
ditions, a decomposition of the liquid
occurs and hard carbon is deposited
on the joint, thus effecting a thorough
seal. At the present time the still simpler
method is usually adopted of cementing
the two together by a lump of carbon
paste or dough.
In the early history of the modern in-
candescent lamp, the filaments, produced
by substantially the processes already de-
112 ELECTRIC INCANDESCENT LIGHTING.
scribed, when placed in the exhausted
lamp chamber and rendered incandescent
by the passage of a current through them,
were frequently found to glow irregularly,
that is to say, there were parts wThich were
rendered vividly incandescent while other
parts were only dull red. This was due
to the fact that the process did not pro-
duce homogeneous carbon filaments. In
other words, either the diameter varied at
different parts, or the resistivity of the
material varied, or both. Consequently,
when the incandescing current was sent
through the filament, and the temperature
gradually increased, the high resistance
parts of the filament first began to glow,
while the others remained comparatively
cool, the heat being developed by the cur-
rent in proportion to the resistance en-
countered. If such a spotted filament
were employed in the lamp and the tern-
TREATMENT OF FILAMENTS. 113
perature raised by increasing the cur-
rent, so as to make all parts of the fila-
ment glow, then the temperature of the
high resistance portions would probably
be raised beyond the limit of safety
and the life of the lamp would conse-
quently be much shortened ; while, on the
contrary, if the higher resistance portions
of the filament wrere limited to the safe
temperature, the lower resistance portions
would be at so low a temperature, that the
candle-power of the entire lamp would be
unduly low.
This difficulty was happily overcome by
an exceedingly ingenious process, gener-
ally called the flashing process, which
consisted essentially in a means whereby
carbon was caused to be deposited only
upon those portions of the filaments whose
resistance was higher than the rest. In
114 ELECTRIC INCANDESCENT LIGHTING.
other words, if the filament were unduly
narrow at some particular spot, or had an
undue resistivity at such spot, then carbon
would be deposited upon this spot only.
The flashing process is carried on sub-
stantially as follows. The mounted fila-
ment is placed in a suitable chamber from
which the air has been removed, and
which is subsequently filled with a hydro-
carbon vapor. An electric current, whose
strength is gradually increased, is then
sent through the filament. A hydro-
carbon gas or vapor, suffers decomposition
in the presence of a heated surface as soon
as a certain temperature is reached. As
the current gradually increases, the carbon
filament begins to glow at its point of
greatest resistance, and this point, conse-
quently, receives a deposit of carbon, thus
decreasing the resistance locally. If this
TREATMENT OF FILAMENTS. 115
current strength were maintained, the car-
bon would cease to glow at these points.
If, however, the current strength be fur-
ther increased, the carbon would begin to
glow at the point of next highest resist-
ance, and this in turn, receiving a deposit,
would cease to glow at this current
strength. It will be readily seen, there-
fore, that as the current strength is gradu-
ally increased, the filament receives a
deposit at those portions of its mass only
where it needs increase in conducting
power, and soon the entire filament will
glow with a uniform intensity of light.
It must not be supposed, that the car-
bon is now absolutely of the same area
of cross section, or of the same thickness
throughout. The flashing process has
rendered it electrically, but not mechani-
cally, homogeneous. We have correctly
116 ELECTRIC INCANDESCENT LIGHTING.
described the process as consisting of suc-
cessive steps reached by gradually increas-
ing the current strength. In point of fact,
these steps follow one another so rapidly
that the process at first sight may seem
to be almost instantaneous, only a few
seconds being required for an exceedingly
spotted carbon to emit a uniform glow.
Although at the present day improve-
ments in manufacture have resulted in the
production of filaments, which are so
nearly uniform in their resistance that
they will glow uniformly when placed in
the lamp, and, therefore, do not require
to be subjected to the flashing process,
nevertheless, since this process results in
giving to the filament other valuable prop-
erties, it is still generally practiced. Not
only are the surfaces of flashed carbon
filaments harder than those which have not
TREATMENT OF FILAMENTS. 117
undergone tins process, but the amount of
liglit which they emit for a given current
strength is markedly increased.
The flashing process is sometimes
carried on in liquids, such as beuzine, the
filaments being dipped in the liquid and
the current, as before, supplied in gradu-
ally increasing strength. In such cases,
however, the decomposition of the liquid
produces an atmosphere of gas around
the filament so that the difference in the
process is rather in appearance than in
reality.
CHAPTER VII.
SEALING-IN AND EXHAUSTION.
THE mounted and flashed filament has
now to be inserted in an enclosing glass
chamber, in which it is hermetically sealed.
This sealing- in, is preferably accomplished
by the actual fusion of the glass stem to
the lower part of the globe. It will be
interesting, therefore, to examine in detail
the method generally employed in the
manufacture of the incandescent lamp
chamber and its hermetical closure on
sealing-in.
Fig. 17, represents the successive steps
that are generally taken in the sealing-in
118
SEALING-IN AND EXHAUSTION.
119
of the mounted filament in the lamp
chamber. The glass lamp chamber A,
FIG. 17.— STEPS OP SEALING-IN PROCESS.
has the form shown, the open tubular pro-
jection being left at £, for the exhaustion
of the chamber. The open end of the
120 ELECTRIC INCANDESCENT LIGHTING.
chamber A, is of such dimensions that the
mounted filament can be introduced into
it up to the shoulder d, which then rests
in contact with the lower end a, of the
chamber. The stem is then grasped by
the glass-blower in one hand, and the
tubular end of the chamber in the other,
and the two revolved together as one
piece, in a suitable blow-pipe flame di-
rected upon the shoulder or joint, until
the fusing temperature is reached, and the
edge a, becomes hermetically sealed with
the shoulder d. By this means it will be
seen that an enclosing chamber, made en-
tirely of glass, is provided with leading-in
wires passing through the support at />, p.
In the early history of the art, it was
necessary that this delicate operation
should be performed by a skilled glass-
blower, but during recent years, machines
have been introduced which grasp the
SEALING-IN AND EXHAUSTION.
globe aud stein and revolve them in the
blow-pipe flame with the requisite amount
of pressure. The machine seal, so ef-
fected, is made as swiftly and neatly as
that of the most expert workman. The
sealed-in lamp is then carefully annealed,
by subjecting it to the action of a gradu-
ally diminished heat, while under the
action of a roller. Great care is neces-
sary that the annealing of the joint should
be thoroughly effected.
Attempts have been made, at different
times, to produce a lamp in which a 'me-
chanical seal was effected between the
stem and the globe, instead of a seal by
fusion. Such a seal possesses the advant-
age of permitting the lamp to be readily
repaired on the breaking of the filament.
Figs. 18, 19, and 20, show a form of Kich
stopper-lamp, as it is generally called.
122 ELECTRIC INCANDESCENT LIGHTING.
Fig. 18, represents the mounted filament,
which is connected to the extremities of
two iron wires sealed into the neck or
FIG. 18. — STOPPER-MOUNTED FILAMENT OF INCAN-
DESCENT LAMP.
stem L. We have already pointed out
that iron is sometimes employed for this
purpose. The stopper portion of the stem
SEALING-IN AND EXHAUSTION. 123
at /SJ is ground by machinery to fit a simi-
larly ground seat in the opening of the
lamp globe, which is shown at A, Fig. 19.
The mounted filament is inserted into the
lamp chamber, and the stopper secured in
its seat by a flexible cement. A brass
shell is then secured around the base B of
the lamp, enabling connections to be main-
tained with its terminals, as shown in
Fig. 20.
The mounted filament, having thus been
introduced into the lamp chamber, and the
base of the lamp hermetically sealed, the
next step is the exhaustion of the lamp
chamber. An exceedingly small quantity
of air left in the chamber will contain
sufficient oxygen to cause rapid destruction
of the lamp filament. Consequently, it is
necessary to remove, as far as possible, all
traces of air from the interior. This is
124 ELECTRIC INCANDESCENT LIGHTING,
FIG. 19.— GLOBE OF STOPPER LAMP.
SEALING-IN AND EXHAUSTION. 125
accomplished by means of pumps. The
ordinary mechanical air-pump, of the auto-
matic valve type ; i. e., in which the
valves are opened and closed automatically
by the motion of the piston, is capable
of producing fairly high vacua, but not
sufficiently high for the purposes of being
used alone in the exhaustion of the lamps,
since the residual air would still be detri-
mental. Mechanical pumps, however, are
often used for producing a rapid exhaus-
tion of the lamp chamber, the final exhaus-
tion being accomplished by means of a
mercury pump. The operation of the mer-
cury pump, however, is so simple in practice
and efficient in action, that in some cases, the
use of the mechanical pump is dispensed
with, and the entire process of exhaustion
is carried on by the mercury pump alone.
A great variety of mercury pumps have
126 ELECTRIC INCANDESCENT LIGHTING.
FIG. 20.— COMPLETED STOPPER LAMP.
PROPERTY
SEALING-IN ANJ^xJ^QIAUSTION. 127
been devised, all of
niently divided into two classes"?1
those of the Geissler type, in which the
vacuum is obtained by utilizing the prin-
ciple of the so-called Torricellian vacuum,
of the barometer tube, and those of the
Sprengel type, in which the vacuum is
obtained by the fall of a stream of mer-
cury. When a column of mercury is per-
mitted to fall through a vertical tube,
connected near its upper end by a branch
tube with the chamber to be exhausted,
the air will be carried away from the
chamber by becoming entangled as bub-
bles in the falling column. Mercury
pumps of this character are well adapted
to the exhaustion of lamps, owing to the
simplicity and efficiency of their action.
When the proper degree of vacuum is
obtained, the lamp is sealed-off by fusing
128 ELECTRIC INCANDESCENT LIGHTING.
the tube at tlie top of the lamp chamber,
with a blow-pipe flame, a constriction being
provided in the tubulure I, at &1 as shown
in Fig. 17, for facilitating this process.
In the early state of the art this sealing-off
was effected while both the lamp and the
filament were cold. It was found, when
thus sealed-off, that, although the proper
vacuum had been obtained, and the lamp
operated satisfactorily for a while, yet the
vacuum soon invariably deteriorated, so
that the life of the lamp was unduly short-
ened. The explanation was at last found
in the fact that gases were occluded or
absorbed in the carbon filament, as well as
condensed on the inner surface of the globe.
These gases adhered to the filament, or to
the globe, with too great a force to per-
mit them to escape into the lamp chamber
during the process of exhaustion. When,
however, the lamp was lighted after con-
SEALING-IN AND EXHAUSTION. 129
eluding the process of exhaustion, the in-
tense heat of the filament disengaged
these gases which reduced the vacuum in-
juriously. The remedy is simple. As
soon as a fairly good vacuum is obtained
in the lamp, an electric current is sent
through the filament and the last stages of
the pumping process are carried on while
the filament is aglow. Immediately be-
fore sealing-off, the current is increased
beyond the strength intended to be em-
ployed in practice, and the lamps sealed
while the pumping is carried on. By this
means the occluded gases in the filament
and on the globe are carried off, and this
process has done much to improve the lamp.
When incandescent lamps were first
manufactured on a large scale, an effort
was made to obtain as high a. vacuum as
possible, and pumps were employed which,
130 ELECTRIC INCANDESCENT LIGHTING.
it was claimed, left a residual atmosphere
in the lamp chamber of but 1,000,000th of
its original amount; that is that 999,999
parts of air out of every million had been
removed. Even at the present time, while
it is generally considered that residual
atmospheres of air, containing as they
necessarily must traces of oxygen, result in
a decreased life of the lamp, yet it may be
asserted, as a result of actual experience,
that residual atmospheres of certain gases,
such as chlorine or bromine, or mixtures
of the same, may not only be innocuous
but actually advantageous.
It is quite possible, during the opera-
tion of a lamp, the sealing-off of which
has been thoroughly made, that the vacuum
may actually improve rather than dete-
riorate ; for, in the disintegration of the
carbon, to which we shall allude in a sub-
SEALING-IN AND EXHAUSTION. 131
sequent chapter, a deposit of finely divided
carbon takes place inside the lamp cham-
ber. This carbon possessing, as it does,
the power of occluding or absorbing resid-
ual atmospheres, would naturally tend to
improve the vacuum during use.
The exhausted and sealed lamp must
now be provided with a base, whereby it
can be readily placed in a socket or sup-
port. The lamp base is so arranged that
two metallic portions, suitably insulated
from one another, are connected to the
ends of the leading-in wires. These por-
tions on the base are adapted to connect
the lamp with the circuit wires by the
mere act of placing it in the lamp socket.
Various forms of lamp bases are employed
as we shall see hereafter. The lamp base
is attached to the lamp by a cement, gen-
erally of plaster of Paris.
132 ELECTRIC INCANDESCENT LIGHTING-.
Some of the more usual forms of lamp
bases are shown in Fig. 21 at A, B, C
and D. The two separate metallic pieces,
which are electrically connected to the
ends of the leading-in wires, are, in all
FIG. 21.— LAMP BASES.
cases, indicated by the letters a and Z>.
An inspection of the figure will show that
A^ has a central contact pin as one termi-
nal, and a concentric brass ring as the
other terminal. B, has a central pin as
one terminal, and an external concentric
cylinder as the other. At (7, two concen-
tric cylinders are employed, the inner one
has, however, a screw thread for securing
it in its socket. D, has a screw shell as
SEALING-IN AND EXHAUSTION. 133
one terminal, and a central cap as the
other terminal.
It is sometimes convenient to fit a lamp
of one manufacture into a socket of
FIG. 22. — LAMP ADAPTERS.
another manufacture. For this purpose a
device called a lamp adapter is employed.
An adapter consists essentially of an exte-
rior base and an interior connection piece.
134 ELECTRIC INCANDESCENT LIGHTING.
In Fig. 22, four adapters are shown suit-
able for attachment to a stopper lamp, and
provided with bases corresponding to the
lamp bases shown in A, B, C\ D, of Fig.
21, similar letters corresponding in the
two figures. The use of stopper lamps has,
however, been abandoned.
CHAPTER VIII.
LAMP FITTINGS.
the removal of the lamp from the
pumps, it is tested for candle-power, and
marked in volts for the pressure which
should be supplied to it in operation.
Some examples of finished lamps are
shown in Fig. 23. These lamps are all of
the same type and differ only in the form
of base. The bases of these lamps differ
so as to permit each lamp to be used on
some particular socket.
One of the simplest forms of sockets
intended for inexpensive work, especially
135
136 ELECTRIC INCANDESCENT LIGHTING.
FIG. 23. — COMPLETED INCANDESCENT LAMPS.
LAMP FITTINGS.
137
where the lamps are not open to direct ob-
servation, such as at the foot or side lights
of theatres, is shown in Fig. 24. This
FIG. 24. — SIMPLE FORM OP SOCKET.
socket is intended to receive a lamp with a
screw base. A wooden shell Z, is screwed
to the wall or other support, by ordinary
screws passing through the screw holes,
one of which is seen at S. A and ^?, are
brass screws intended for the reception of
138 ELECTRIC INCANDESCENT LIGHTING.
the circuit wires or mains, and are con-
nected respectively to the brass screw shell
FIG. 25. — KEYLESS WALL-SOCKET.
(7, and a central cap beneath. These will
make contact with the two insulated por-
LAMP FITTINGS. 139
tions of the base of the lamp when the
lamp is screwed in.
FIG. 26. — KEY WALL-SOCKET.
Fig. 25, shows a more sightly form of
keyless socket, that is, a socket which is not
140 ELECTKIC INCANDESCENT LIGHTING.
provided with a key or switch, and which,
therefore, constantly maintains connection
between the mains and the lamp. The
base 13 J?, is of porcelain, and is fixed in
position by screws passing through screw
holes C\ C. Supply wires, connected with
the mains, pass beneath through the
grooves W, W. Connections are main-
tained through the interior of the shell.
A socket of a similar character, but pro-
vided with a key K, is shown in Fig. 26.
In this case the lamp is lighted or extin-
guished by the turning of the key. In
the case of keyless sockets the lamps are
turned on or off by the action of a distant
switch.
Socket keys, open and close the circuit of
a lamp at one point in a variety of ways.
Two forms of socket keys are shown in
LAMP FITTINGS.
141
Figs. 27 and 28. Fig. 27, shows a socket
suitable for use with a lamp base of type
B, Fig. 21; and Fig. 28, shows a socket
suitable for use with the lamp base of type
PIG. 27.— DETAILS OP SOCKET.
Oj in that figure. In Fig. 27, the turning
of the key K, makes contact between the
brass segments b, Z>, through the interme-
diary of the cam, on the extremity of the
key axis. In Fig. 28, a similar method is
142 ELECTRIC INCANDESCENT LIGHTING.
employed. Here the movement of the
key closes connection between contacts
by ^ through the metal piece C.
FIG. 28. — DETAILS OP LAMP SOCKET.
Fig. 29, represents in cross-section a
lamp with a screw base in its socket.
The current is turned on or off at the key
K. 6r, is the globe, and T, the tip at which
the lamp was sealed on removal from the
pump ; Pj is the filament ; 89 the seal of
the leading-in wires ; W, the welds be-
tween the platinum and the copper lead-
"
JP
FIG. 29.— LAMP AND SOCKET SHOWING CONNECTIONS.
144 ELECTRIC INCANDESCENT LIGHTING.
ing-in wires; $', the seal of the lamp stem
with the glass chamber. Z, Z, are projec-
tions of the glass on the surface of the
shoulder, intended to aid in securing the
/ o
lamp in its plaster of Paris cement. 6Yand
12, are the cap and screw metallic pieces of
the base, each in soldered connection with
one leading-in wire as shown. A and J3,
are the cap and screw connections of the
socket, each in connection with one of the
external wires entering the socket through
the pipe or metallic support P. One of
these wires is connected directly with the
brass shell A, while the other is connected
with the cap />, only through the interme-
diary of the key K. M M, is the external
brass shell of the socket, insulated from
the interior portions by the hard rubber
ring n n.
When an incandescent lamp is supported
LAMP FITTINGS.
145
by a flexible cord, it usually requires both
hands to turn the key at the socket on or
off, one to hold the lamp, and the other to
FIG. 30. — Pusn BUTTON KEY SOCKET.
turn the key. Fig. 30, shows a device by
which the turning on or off can be accom-
plished by the hand which holds the
146 ELECTRIC INCANDESCENT LIGHTING.
socket. This is accomplished in replacing
the ordinary key by two pressure buttons.
Pressure upon the stud AJ forces it home
until it is locked by the trigger B, thus
PIG. 31.— SPRING SOCKET FOR SCREW BASE.
turning on the lamp and keeping it turned
on. Pressure on the trigger B, releases
the push A., which returns to its original
position under the action of a spring.
Fig. 31, shows a form of spring socket for
a screw lamp base. For temporary work,
LAMP FITTINGS.
147
such as exhibitions, etc., where the expense
of permanent fixtures would not be justi-
fied, a temporary socket is sometimes em-
ployed, as shown in Fig. 32. It consists of
FIG. 32.— TEMPORARY SOCKETS.
a spiral spring holding the screw base of
the lamp, and connected with one supply
wire W, while the brass screw in the centre
of the spiral is connected with a second
supply wire beneath the wooden base bar
148 ELECTRIC INCANDESCENT LIGHTING.
Such construction is, of course, not re-
garded as safe for permanent use in
buildings.
Where lamps are placed in positions
exposed to the weather, it is necessary to
FIG. 33. — WEATHER-PROOF SOCKETS.
employ some form of weather-proof socket.
Two forms of such sockets in common use
are shown in Fig. 33, that on the right hand
side being of glass, and that on the left,
LAMP FITTINGS. 149
of hard rubber or composition. These
sockets are intended for the reception of
screw bases.
Various forms of reflecting surfaces are
employed in connection with the lamps so
FIG. 34.— METALLIC SHADE FOB REFLECTING LIGHT
DOWNWARDS.
as to throw the light in any desired direc-
tion. Two such forms of lamp shades,
devised to throw the light downwards,
150 ELECTRIC INCANDESCENT LIGHTING.
are shown in Figs. 34 and 35. The depth
of the shade will vary with the character
of the illumination required. That shown
in Fig. 34, is suitable for the illumination
of desks from above. Fig. 35, is suitable
FIG. 35. — METALLIC SHADE FOR REFLECTING LIGHT
DOWNWARDS.
for the illumination of a billiard table.
Fig. 36, shows a form of half shade some-
times employed for desk use. Here one
half of the lamp only is covered by the
LAMP FITTINGS. 151
shade which in outline conforms with the
shape of the lamp, the inside of the shade
being provided with a good reflecting sur-
face to throw the light downwards.
FIG. 36.— METAL HALF SHADE FOR DESK USE.
It is important in order to ensure
good illumination for reading, that all
portions of the printed page shall be
equally lighted. Although this is gener-
152 ELECTRIC INCANDESCENT LIGHTING.
ally secured by the aid of aiiy good shade,
yet it is often preferable to employ for
this purpose the form of shade and en-
closing globe shown in Fig. 37.
FIG. 37.— REFLECTOR SHADE.
i Instead of forming the lamp shade out
of a single surface, a number of sur-
faces are sometimes employed, either plane
or curved. Figs. 38 and 39 represent
LAMP FITTINGS. 153
panel reflectors j i. e., reflectors composed of
strips or panels of silvered glass. These
FIG. 38. — CONCAVE PANEL SHADE AND REFLECTOR.
reflectors are suitable for store windows or
railway stations. The shade in Fig. 38,
is designed to throw light downwards from
FIG. 39.— CONCAVE PANEL SHADE AND REFLECTOR.
154 ELECTRIC INCANDESCENT LIGHTING.
its concave surface, and Fig. 39 to throw
the light outwards from its convex surface
as well as downwards. Finally, reflectors
of similar forms are also arranged to oper-
FIG. 40. — PANEL REFLECTOR AND SHADE FOR CLUSTER.
ate with clusters of lights to illumine
larger halls. A concave panel reflector of
this type is shown in Fig. 40.
Sometimes corrugated silvered glass is
employed in various forms for reflecting
LAMP FITTINGS. 155
purposes. Fig 41, shows a reflector of
this type made in the form of a shade.
FIG. 41. — CORRUGATED REFLECTOR AND SHADE.
Even transparent or translucent sub-
stances may at times be employed for
reflectors. In such cases they aid in
scattering light downwards as well as in
156 JflLKCTKIO INCANDESCENT LIGHTING.
reflecting it. Materials employed for this
purpose are glass and porcelain, either
plain or corrugated, transparent or trans-
FIG. 42. — GLASS SHADES.
lucent. Fig, 42, shows a variety of shades
employed for such purposes.
Owing to the fragile nature of the in-
candescent lamp chamber, it is necessary,
LAMP FITTINGS.
157
when lamps are placed in exposed posi-
tions, to protect them from accidental
destruction by blows. For this purpose
wire gratings or shields are arranged, of
FIG. 43.— HALF WIRE-GUARD.
such form and outline as shall not seriously
interfere with the light. Such grat-
ings would, of course, be inadmissible in
any situation where marked shadows
158 ELECTRIC INCANDESCENT LIGHTING.
are objectionable, and should, therefore,
only be employed in cellars, underground
passages, or in similar situations. Fig. 43,
shows a form of half wire-guard, which,
FlG. 44.— FCLL, WlRE-GUAKD.
as its name indicates, only surrounds
a portion of the lamp proper. Fig. 44,
shows a full wire-guard surrounding the
entire lamp. In Figs. 45 and 46, some
LAMP FITTINGS. 159
forms of wire-guards are shown, suitable
O '
for portable or hanging lamps.
Sometimes, instead of employing a wire
guard to protect the lamp, the lamp is
FIG. 45.— WiRE-GuAKDS FOR SUSPENDED LAMPS.
entirely surrounded by an air-tight or
a steam-tight glass lamp chamber as in
Fig. 47. This globe consists of two parts,
a cylindrical glass cover with a rounded
160 ELECTRIC INCANDESCENT LIGHTING.
end, provided with a screw thread metallic
cap, capable of tightly fitting into the base
FIG. 46. — PORTABLE LAMP GUARD.
as shown. A steam-tight lamp chamber is
employed under circumstances where it is
either desired to protect the lamp from cor-
LAMP FITTINGS.
161
FIG. 47. — STEAM-TIGHT LAMP CHAMBER.
rosive vapors, from spray at sea, or for the
purpose of avoiding any possible accident
which might result, from the explosion of
162 ELECTRIC INCANDESCENT LIGHTING.
inflammable gases or clouds of dust, on the
accidental breaking of the lamp chamber
and globe. It will, of course, be under-
stood that the use of any form of wire
grating, or steam-tight globe, must neces-
sarily be attended by a marked decrease in
the useful illumination of the lamp.
CHAPTER IX.
THE INCANDESCING LAMP.
WE have now traced the manufacture of
the lamp up to the time when it is ready
to be connected with the supply wires and
actuated by the electric current. With this
the manufacture of the incandescent lamp,
that is a lamp capable of being rendered
incandescent, now ceases, and the story of
the incandescing lamp, and the theory of
its operation begins. We will, therefore,
trace in this chapter such theoretical con-
siderations as enter into the action of the
incandescing lamp.
An incandescing lamp may be regarded
as an electro-receptive device, wherein the
164 ELECTRIC INCANDESCENT LIGHTING.
energy of the electric current is being con-
verted into heat energy, part of which
is luminous. It is, therefore, necessary,
at the outset, to determine the amount
of energy which the lamp is absorb-
ing. This, as we have already seen, is
equal to the product of the number of am-
peres, or the current strength supplied to
the lamp, and the pressure in volts at the
lamp terminals. If the resistance of a
lamp, when hot, is 200 ohms, and the pres-
sure between the mains 100 volts, then the
current which will pass through the lamp
will be - - — 1/2 ampere, and the activity
200
supplied will be 1/2 X 100 = 50 watts or
h horse-power.
The temperature which an activity of
50 watts, will .produce in a lamp fila-
THE INCANDESCING LAMP. 165
merit, depends both upon the extent
of the surface area of the filament and
upon the character of its surface. If
the surface of the filament be large, the
activity per square inch, or per square cen-
timetre, will be comparatively small, and
the filament will not be highly heated. If,
on the other hand, the surface area of the
filament be small, the intensity of its sur-
face activity will be great, and the tem-
perature of the filament will be high.
The surface activity of an incandescing
filament is, approximately, from 70 to 100
watts per square centimetre, or 450 to 650
/»
watts per square inch ; i. e., about T7)^ns
o
to T^ths of a horse-power per square inch
of surface.
The electric arc lamp has in its crater, a
166 ELECTRIC INCANDESCENT LIGHTING.
surface activity of, approximately, 3 KW,
or 3,000 watts per square centimetre of sur-
face; i. e.j 19.35 KW per square inch, or
27.15 HP per square inch. This shows
that the surface area of the crater is very
small, since an arc lamp takes only about
3/5ths horse-power. The apparent surface
activity of the sun is, approximately,
10 KW, or 13.4 HP per square centi-
metre; i. e,, 64.5 KW, or 86.5 HP per
square inch. Since, as we have shown, the
highest frequency of the waves which are
emitted by a heated surface increases with
the temperature, it is evident that the
highest frequency reached in the solar light
waves will be greater than in the arc light
waves, and this in its turn will be greater
than in the incandescent light waves, since
the intensity of surface activity in watts
per square centimetre differs so markedly
in these three surfaces.
THE
Collecting these res^ik$^abularly, we
Solar surface activity, . . . 10,000 watts pei
Arc-crater surface activity, . . 3,000 watts per square cm.
Incandescing-filament surface activity, 70 to 100 watts per
square cm.
If the surface activity of an incandes-
cing filament be increased, by passing a
stronger current through the filament, that
is, by subjecting it to a higher electric
pressure, the temperature of the filament
will increase, and with it the amount of
light given off per square centimetre.
This increase may be carried up to the
point of destruction of the carbon fila-
ment. The duration or life of an incan-
descing lamp, depends very markedly
upon the temperature and surface activity
of the filament. At a low temperature,
or at a dull red heat, an incandescing lamp
will last almost indefinitely ; at a vivid
168 ELECTRIC INCANDESCENT LIGHTING.
incandescence, or very high temperature
and intense surface activity, its life may be
only a few minutes. Between these two
extremes lies a mean temperature and sur-
face activity, at which it has been found
in practice most profitable and desirable
to operate the lamp. The object, therefore,
of the lamp maker is so to proportion the
dimensions of the filament, that, when con-
nected across the mains, the surface activity
will reach the amount necessary to give
the proper temperature to the filament, as
well as the desired total quantity of light.
It should be carefully remembered that
the surface activity, which determines the
brightness of the filament, is a quantity
altogether distinct from the total- candle-
power, or the total quantity of light given
off from the lamp. The brilliancy depends
only on the surface activity, while the
THE INCANDESCING LAMP. 169
total candle-power depends upon the total
surface area as well as upon the brilliancy.
It is common to find that a person looking
at two lamps, one of which may have a
high surface activity; i. e., a great bril-
liancy, but which only gives say 5-candle-
power, and the other of which has a low
surface activity, or small brilliancy, but
which gives 16-candle-power, will judge
that the brighter lamp is giving the greater
amount of light. In other words, the eye
is very sensitive to relative brightness or
brilliancy, but is by no means sensitive to
differences in total-candle-power ; i. e., total
intensity of the light. It is very essential,
in all artistic groupings or arrangements
of lamps, that their surface activity and
brilliancy should be as nearly equal as
possible, since, otherwise, appearing to the
eye unequally bright, they will fail to pro-
duce pleasing effects.
170 ELECTRIC INCANDESCENT LIGHTING.
Having given a certain temperature and
surface activity to the filament, the total
amount of light will depend upon the
total surface area. For example, a 16-
candle-power lamp, operated at a given
temperature, or at an efficiency of say 1/3
candle-per-watt, would take 48 watts of
activity. A 32-caudle-power lamp, at the
same brightness, surface activity, and
quality of carbon filament, and therefore,
with the same efficiency of 1/3 candle-
power per watt, would require to be sup-
plied with 96 watts and would, therefore,
require double the surface from which to
radiate the doubled activity. Such a lamp,
working at the same pressure, would re-
quire to be both larger in diameter and
longer. Broadly speaking, a lamp of high
candle-power will have a thick filament,
and a lamp of low candle-power a thin
filament, when working at a common pres-
THE INCANDESCING LAMP. 171
sure. Fig. 48, represents the relative sizes
of incandescent lamps of the same manu-
facture, voltage and efficiency, intended
for 16, 32, and 100 candles. Here the
gradually increasing lengths and diameters
of the filaments may be observed.
It may be well to explain in greater
detail the meaning of the term efficiency,
as used in the preceding paragraph.
Since energy must be expended in an in-
candescing lamp, in order to produce a
certain candle-power, it is evident, from
the standpoint of economy in energy, that
the greater the number of candles which
can be obtained per horse-power, or per
watt of activity, the greater will be the
efficiency of the lamp. We speak, there-
fore, of the efficiency of an incandescent
lamp as being l/3rd or l/4th of a candle
per-watt, meaning that an 8 candle-power
THE INCANDESCING LAMP. 173
lamp would take, in either case, 24 or
32 watts respectively, representing 248.7
or 186.5 candles per electrical horse-power.
In common usage, however, the term
efficiency is often unfortunately misap-
plied, so that the same lamps would be
spoken of as having an efficiency of 3 or 4
watts per candle, respectively, from which
it would seem that as we increase the num-
ber of watts to the candle we increase the
efficiency, whereas, it is evident that the
reverse is true. It is preferable, therefore,
to use the word efficiency in the less popu-
lar but more correct signification.
When an incandescent lamp is operated
at a constant pressure, a series of changes
takes place which it is important to follow
as closely as the knowledge we possess will
permit.
The temperature reached by the fila-
174 ELECTRIC INCANDESCENT LIGHTING.
ment of an incandescing lamp has been
estimated from a series of measurements, to
be in the neighborhood of 1,350° C., slightly
varying, however, with the surface activity
and brightness.
Thus, at an efficiency of 1/3 candle per
watt, the temperature is estimated to be
1,345° C.
At an efficiency of 1/4 candle per watt,
the temperature is estimated to be 1,310°
C.
And at an efficiency of 1/4.5 candle per
watt, the temperature is estimated to be
1,290° C.
If we increase the activity of an incan-
descing lamp one per cent.; i. e., if we in-
crease the pressure at its terminals to such
a point that the number of watts it re-
ceives increases by one per cent., the tem-
perature is believed to increase about 2° C.
THE INCANDESCING LAMP. 175
and the candle-power is believed to in-
crease about three per cent.
When the electric current passing through
a lamp produces the surface activity and
temperature for which the lamp is de-
signed, although, in general, the lamp
exhibits a steady diminution in tempera-
ture, surface activity and candle-power,
which continues while the lamp is used, yet
it frequently happens, that for the first
few hours these quantities actually in-
crease, so that a 16-candle-power lamp,
after the first fifty hours of its life, may
give 17 candles, and a greater brightness
than at the start. Even when this rise
occurs, however, at the end of the first
hundred hours the lamp will usually have
fallen in caudle-power, brilliancy, surface
activity and temperature, all these quanti-
ties being associated, to an amount varying
176 ELECTRIC INCANDESCENT LIGHTING.
with the type of lamp and the carbon from
which it has been manufactured.
We shall now examine the causes which
bring about the progressive decay of the
lamp above referred to. When an in-
candescent lamp is operated, it is found
that the negative half of the filament
throws off or projects carbon particles
from the surface, in all directions in
straight lines. It is generally believed
that this effect is due to a species of
evaporation. This evaporation takes place
with greatest activity near the negative
extremity of the filament, or the point
where the filament is united with the lead-
ing-in wire on the negative side. From
this point of maximum evaporation, the
effect diminishes to the centre of the fila-
ment and the positive side of the filament
shows but little evaporation. The presence
THE INCANDESCING LAMP. 177
of electric evaporation from negatively
charged surfaces is recognized at all tem-
peratures, and even under atmospheric
pressures, but, like all evaporation, is aided
by a high temperature and vacuum. Conse-
quently, its effect is pronounced in an incan-
descent lamp. '
The existence of an evaporation of the
filament can be detected in a variety of
ways. For example, if a metallic plate
be supported in the lamp chamber, mid-
way between the two legs of an ordinary
horse-shoe filament, it is found that the
evaporation of the carbon particles from
the negative side of the filament, is actually
capable of carrying an electric current to
the plate. That is to say, the stream of
negatively charged carbon particles im-
pinging against the surface of the metallic
plate, delivers up to it the electricity with
178 ELECTRIC INCANDESCENT LIGHTING.
which they are charged and so results in
the passage of an electric current.
The continued evaporation of carbon
from the surface of the filament, produces
a gradually increasing blackening of the
surface of the globe, since the projected
carbon particles adhere to the walls of the
chamber where they strike. The con-
tinued bombardment of the glass walls
slowly coats them with a layer of carbon,
which being opaque, reduces the amount
of light emitted by the lamp. An old
lamp, if examined against a white surface,
such as a sheet of paper, will be seen
to be distinctly blackened over its in-
terior surface. This blackening, or, as
it is sometimes called, age coating of the
lamp chamber, takes place, other things
being equal, most rapidly in lamps that
have been burned at a high tempera-
THE INCANDESCING LAMP. 179
ture, since in these the evaporation is
more rapid.
Another proof that the particles of car-
bon leave the surface in straight lines is
to be found in the "shadows" produced
on the surface of the glass, when the
filaments are straight horse-shoes, or lie
wholly in one plane. In such cases the
bombardment from any portion of the
negative leg is necessarily intercepted by
the positive leg in the plane of the fila-
ment, so that the globe is protected in this
plane by the interposition of the positive
leg. The result is, that after the lamp
chamber is visibly blackened, a distinctly
marked line can be traced on the sur-
face of the glass opposite the negative
leg. The term shadow is sometimes ap-
plied to this. There will, however, be no
shadow on the side of the glass nearest to
180 ELECTRIC INCANDESCENT LIGHTING.
the negative leg, nor will the shadow be
produced if the polarity of the supply
mains is occasionally reversed, or, in the
case of lamps supplied by alternating cur-
rents where each leg is positive and neg-
ative alternately.
Not only does a lamp filament give less
light after being operated at high pres-
sures for a considerable length of time,
owing to an increase in its resistance and the
blackening of the globe, butialso to a change
in the surface nature of the filament,
whereby it gives less light for a given
surface activity. In technical language
its emissivity increases, so that the tem-
perature, which is attained by a given sur-
face activity, is reduced. In other words?
the lower the emissivity of a filament, the
higher the candle-power and brilliancy for
a given surface activity. Of these three
THE INCANDESCING LAMP. 181
causes for decreased candle-power with age ;
namely, diminished current strength and
activity, diminished translucency of the
globe, and increased emissivity, the loss in
candle-power is about equally affected by
each.
Fig. 49 shows curves representing the
change in candle-power of an ordinary
incandescent lamp when initially operated
at various efficiencies. Thus, at an efficiency
of 0.3 candle-per-watt, the lamp gives 14.5
candle-power. At an efficiency of 0.4
caudle-per-watt, 22.5 candle-power, and
at 0.5 candle-per-watt, 32 candle-power.
Roughly speaking, if we double the effi-
ciency we treble the candle-power and
brilliancy of the lamp. On this account
it would obviously be advantageous to
increase the efficiency of a lamp as far as
possible.
182 ELECTRIC INCANDESCENT LIGHTING.
In the preceding diagram we have
plotted the relation between candle-power
and efficiency. If, however, we plot the
relation between candle-power and activity,
0.1 0.2 0.3 0.4
EFFICIENCY. CANDLES PER WATT /"
0.5
FIG. 49. — CURVE OF HELATTVE CANDLE-POWER OR
BRIGHTNESS FOR A PARTICULAR CHARACTER OF CAR-
BON FILAMENT OPERATED AT DIFFERENT EFFICIENCIES.
183
we find, rotigid^|^^^^lfiS^ the candle-
power increases as the cube of the activ-
ity, so that if we double the activity of
the lamp ; i. e., increase the pressure at its
terminals until the product of this in-
crease of pressure and the increased cur-
rent is double what it was originally, the
candle-power of the lamp will be in-
creased about eight times, giving about
four times as much light per horse-power
or per watt expended in the lamp.
Fig. 50 shows the relation, which has
been experimentally found to exist, be-
tween the average lifetime of lamps of the
same make as that represented in Fig. 49,
and the efficiency at which such lamps are
burned. A study of this curve will show
Why it is not, at present, possible to obtain
in practice an efficiency beyond a certain
value ; for, at an efficiency of 0.2 candle-
184 ELECTRIC INCANDESCENT LIGHTING.
per-watt, corresponding to 8 candles in the
particular lamp of Fig. 49, the mean dura-
tion of life is over 16,000 hours; while at
0.5 candle-per-watt, corresponding to 32
1(5,000
15,000
13,000
12,000
11,000
10,000
9,000
8,000
7,000
6,000
5,000
o
X 4,000
hi 3,000
2,000
1,000
0 0.1 0.2 0.3 0.4
EFFICIENCY. CANDLES PER WATT
FIG. 50. — CURVE SHOWING THE MEAN DURATION OP LIFE
IN A PARTICULAR CLASS OF INCANDESCING LAMPS AT
VARIOUS EFFICIENCES.
THE INCANDESCING LAMP. 185
candles in Fig. 49, the average duration of
life is only 100 hours. In the first case, we
should have had an 8-candle-power lamp
lasting, say 18,000 hours, and representing a
total output of 144,000 candle-hours, while
in the second case we have a 32-candle-
power lamp, lasting only 100 hours, and
representing a total output of 3^200 candle-
hours. In the first case, however, the
candle-power would be obtained from the
lamp at a comparatively heavy expense in
energy, 2 1/2 times more energy, in fact,
than that necessary for a candle in the
second case. Moreover, the lamp in the
first case would be very dull in color and
unpleasant to the eye.
Between the above two extremes of
long life and low efficiency, and short life
and higli efficiency, there exists a certain
mean value most suitable for commercial
186 ELECTRIC INCANDESCENT LIGHTING.
purposes. This mean value, as lamps are
now constructed, is in the neighborhood of
0.3 candle-per-watt, representing an aver-
age life time of 2,000 hours. This dura-
tion must, however, only be considered as
the average lifetime, since it frequently
happens that among a number of lamps
manufactured by the same process, and
with equal skill, some may only last about
100 hours at this efficiency, while others
may greatly exceed 2,000 hours.
With any installation of electric lamps,
it lies, therefore, in the power of the
engineer in charge of the plant, to so oper-
ate the lamps that their life may be bril-
liant but short, or dull but long; as
will depend entirely upon the pressure
maintained at the lamp terminals. Thus,
if a 16-candle-power incandescent lamp,
intended for an activity of 50 watts, at 115
THE INCANDESCING LAMP. 187
volts, and therefore, to be operated at a
1 (\
normal activity of — or 0.32 candle-per-
ou
watt, be steadily operated at a pressure
one volt in excess, or 116 volts, its initial
candle-power will be raised to 17 candles,
but its lifetime will be abbreviated about
seventeen per cent. Again, if the pressure
be steadily maintained at 2 volts excess, or
117 volts, the initial candle-power will be
normally 18 candles, but the probable life-
time will be reduced about thirty-three per
cent.
When lamps are placed in somewhat
inaccessible positions, such as near the ceil-
ings of high halls, where there is some
inconvenience in reaching them, it becomes
particularly objectionable to have to renew
these lamps too frequently. In order to
avoid this, the plan is sometimes adopted
188 ELECTRIC INCANDESCENT LIGHTING.
of operating the lamps at a comparatively
low efficiency and brilliancy, thus necessi-
tating some extra expense in power, but
greatly prolonging the average lifetime.
t
Fig. 51 represents the rate of variation
of candle-power in similar samples of the
same type of lamp when operated at dif-
ferent steady pressures. Curve No. 5 rep-
resents a 108-volt, 1 6-candle-power lamp
of 0.286 candle-per-watt normal efficiency,
operated at 108 volts. The candle-power
slightly rises to about 16.6 candles, in 90
hours, and finally falls to 11 candles after
500 hours. The efficiency of the lamp at
this time will, of course, be materially
reduced.
Curve No. 4 of Fig. 51, represents the
behavior of a similar lamp operated at 110
volts pressure. Here the initial candle-
THE INCANDESCING LAMP.
189
power is raised to 17 1/2 candles and after
500 hours burning, the candle-power is
200 300
HOURS
Fm. 51. — CURVES OP CANDLE-POWER IN LAMPS OF SAME
TYPE OPERATED AT DIFFERENT FIXED EFFICIENCIES.
about 10 3/4 candles. Curve No. 3 shows
similar results for a pressure of 112 volts.
190 ELECTRIC INCANDESCENT LIGHTING.
Curve No. 2 for 113 volts and Curve No.
1 for 114 volts. Here the candle-power
\
100 200 300 400 600 600 700 HOO 900
HOURS
FIG. 52.— CURVES OP CANDLE-POWER OP THE SAME TYPES
OP LAMP OPERATED AT DIFFERENT EFFICIENCES.
commences at 24, but is less than 14 after
200 hours.
Fig. 52 represents the behavior of four
similar types of lamps, operated steadily
THE INCANDESCING LAMP. 191
at various efficiencies, instead of at various
pressures. Curve No. 1, represents the
behavior with 0.25 candle-per-watt ; Curve
No. 2, 0.286 candle-per-watt; Curve No. 3,
0.333, and curve No. 4, 0.4 candle-per-
watt. It will be seen that the high effi-
ciency lamp falls to fifty per cent, of its
initial candle-power in 600 hours, while
the lowest efficiency lamp only loses ten
per cent, of its candle-power in the same
time.
It is evident, therefore, that no matter
what the initial efficiency of a lamp may
be, a time will come in its life when its
efficiency must be low. When this point
is reached, the amount of activity absorbed
by the lamp, at constant voltage, is less
than it was at the outset, seeing that the
resistance of the attenuated filament is
increased. On the other hand, the effi-
192 ELECTRIC INCANDESCENT LIGHTING.
ciency, or candles-per-watt, has diminished,
and as regards its light, the electric power
is more wastefully applied. It becomes,
therefore, a question whether it would not
be advisable to discard or break the lam}),
and replace it by a new one, having a
greater efficiency. By so doing we incur
the expense of a new lamp earlier than
if we waited for the old lamp to break
naturally, but we utilize the power of the
central station more economically.
The question of the smashing point of a
lamp, or the point in the life of a lamp at
which it may be deemed more economical
to replace it by a new one, or its economical
age, may be considered from three distinct
standpoints ; namely :
(1) From the central-station point of
view.
(2) From the consumer's point of view.
THE INCANDESCING LAMP. 193
(3) From the isolated-plant point of
view.
The central station has usually to re-
place broken or useless lamps, and since
the charge for service is based upon the
ampere-hour, or the watt-hour, so long as
the lamps burn and the consumer is fairly
satisfied, the smashing point may be indef-
initely extended. It is to the station man-
ager's advantage, however, to maintain a
steady pressure over the system of mains,
so that the lamps are not forced above
candle-power and their lives unduly
abridged. In the best central stations
great care is, therefore, always taken that
the pressure is maintained as closely as
possible to the normal. Should the pres-
sure become markedly increased, the cost
of renewing lamps will be rapidly aug-
mented. Should it become markedly
194 ELECTRIC INCANDESCENT LIGHTING.
diminished, the consumers would be dis-
satisfied.
From the consumer's point of view the
lamps would require to be operated at a
high efficiency regardless of their lifetime,
since he would thus obtain, from the power
for which he pays, the highest brilliancy
and the maximum quantity of light. More-
over, lamps which will not break before
becoming seriously dulled, thus necessitat-
ing their renewal, are a disadvantage.
From the standpoint of the consumer, there-
fore, the smashing point of a lamp is reached
as early as possible, or at the opposite ex-
treme to that of the station manager.
From the standpoint of the owner of the
isolated plant, who is both producer and
consumer, the smashing point will neces-
sarily occupy some intermediate position;
195
Desires to
light for the
activity produced or coal consumed, on the
other hand he wishes to reduce the cost of
lamp renewals as far as possible. No pre-
cise rule, however, can be laid down.
There are in general, two purposes for
which light is ordinarily employed ; namely,
for actual use, as in reading, or other work,
and for the aesthetic purposes of ornamen-
tation. Regarding the latter purpose as
a luxury, lamps may be changed as often
as taste may dictate, but high-efficiency,
high-brilliancy, short-lived lamps will be
preferable. On the other hand, so long as
a lamp fulfills the utilitarian purpose of
enabling work to be conveniently and
healthfully performed, it is waste of money
to throw the lamp away. The smashing
point, therefore of lamps intended to
196 ELECTRIC INCANDESCENT LIGHTING.
illumine working rooms depends largely
upon the total candle-power installed. If
this is ample in the first instance, a very
considerable falling off in candle-power
may be permitted without interfering with
the usefulness of the light, and economy is
rarely pushed to such a degree as to limit
closely the candle-power installed for pur-
poses of reading or working.
It is found, however, that in central
station practice the best commercial results
are secured with ample satisfaction to the
consumers, if the efficiency at which the
lamps are operated on the system, is such
that the cost of lamp renewals is approx-
imately fifteen per cent, of the total opera-
ting expenses of the station. Where the
pressures between the mains can be closely
regulated, it is preferable to employ high-
efficiency lamps, but where the reverse is
THE INCANDESCING LAMP. 197
the case, low-efficiency lamps are desirable,
since a low-efficiency lamp can stand an
accidental increase in pressure with less
detriment to its length of life than a high-
efficiency lamp ; for, being normally oper-
ated at a lower temperature, an increase of
temperature may not be dangerous.
CHAPTER X.
LIGHT AND ILLUMINATION.
THERE are two technical words which
are very apt to be confused in their
meaning, and misused in their applica-
tion; namely, " light " and "illumination."
When correctly used, the word light signi-
fies the flow or flux of light emitted by a
luminous source, irrespective of the sur-
face on which it falls, while the word
illumination means the quantity of light
received on a surface, per unit of area,
whether received directly from the lumin-
ous source, or indirectly by diffusion and
reflection from surrounding bodies. The
words " light " and " illumination " are unf or-
198
LIGHT AND ILLUMINATION. 199
tunately often used synonymously, whereas
it is evident that they denote distinct
ideas.
By the candle-power of a source of light,
we mean the luminous intensity of the
source as measured in units of luminous
intensity. The unit of luminous intensity,
commonly employed, is the British candle,
and is equal to the intensity of light pro-
duced by a candle of definite dimensions
and composition, burning at the rate of 2
grains, or 0.1296 gramme, per minute. If
we speak of a source of light as having,
say a luminous intensity of 20 British
standard candles, we mean that if that
source were reduced to a mere point, it
would yield as much light as 20 standard
candles all concentrated at a single point.
A standard of luminous intensity very
200 ELECTRIC INCANDESCENT LIGHTING.
generally adopted, except, perhaps, in
English-speaking countries, is the Frencli
standard candle, called the bougie-decimale,
or the 1/2 Oth of the intensity of an
international standard unit called the
Viotte. The Violle is a unit of lumin-
ous intensity produced in a perpendicular
direction, by a square centimetre of plati-
num, at the temperature of its solidifi-
cation. The British standard candle is
slightly in excess of the bougie-decimale,
one British candle according to Everett
being 1.012 bougie-decimale.
If we imagine a point source of light, of
unit intensity ; i. e., one standard French
candle, or bougie-decimale, to be placed at
the centre of a hollow sphere, of the radius
of one metre, or 39.37", then the total
internal surface of the sphere will receive
a definite total quantity of light. Each
gJiBBttfljggs.
c^X ffy'
4 PROPERTY OF 1JI9
yMGHT AND ILLUMINATION. ^'// 201
»V4b
unit of
spherical nrimTTpwffl-TT^^ unit of
light ; and, since the total interior surface
of such a sphere contains 4 X 3.1416 =
12.566 square metres, the total quantity of
light received on the interior surface will
be 12.566 units, called lumens. Conse-
quently, if a point source of unit intensity
emits 12.566 lumens, a source, of say 20
bougie-decimales, would yield an intensity
of 251.32 lumens.
It is obviously not necessary that the
surface, which surrounds the unit point
source, should be spherical. The same
amount of light; namely, 12.566 lumens,
will be given off by the source independ-
ently of the shape of the receiving surface,
so that when such a source is placed, for
example, alone in a room, the total
quantity of light which falls directly upon
202 ELECTRIC INCANDESCENT LIGHTING.
the walls, ceiling and floor of the room-
from the source, will be 12.566 lumens.
This will be true in fact if there are other
sources of light in the room at the same
time. The quantity of light which each
point source will emit will be 12.566
times its luminous intensity.
If one lumen falls perpendicularly and
uniformly over the surface of one metre,
the illumination of that surface will be
one lux, the lux being the unit of illumi-
nation. If, for example, a bougie-decimale
is located at the centre of a sphere of one
metre radius, then each square metre of
the interior surface of the sphere will
receive one lumen of light, and this light
falls everywhere perpendicularly upon its
surface. The illumination on the interior
surface is everywhere one lux. A bougie-
decimale produces, therefore, when acting
LIGHT AND ILLUMINATION. 203
alone, an illumination of one lux at a dis-
tance of one metre.
The mistake is not infrequently made,
that because a surface receives light
directly from a given source of known
intensity, its illumination can be deter-
mined by mere calculation of its distance
from the source. It must be remembered
that it also receives reflected or diffused
light from all neighboring surfaces, which,
consequently, tend to increase its illumina-
tion.
The law of illumination from a single
point source, acting alone ; i. e., in a space
where all reflected or other light is ex-
cluded, is that the illumination varies
inversely as the square of the distance
from the source. Thus, we know that a
bougie-decimale produces an illumination
204 ELECTRIC INCANDESCENT LIGHTING.
of one lux upon a surface held perpen-
dicularly to the rays of light at a distance
of one metre. If the surface be held at
a distance of 2 metres from the point
source, in a room otherwise dark, the
illumination will be l/4th of a lux, or
(l/2)a ; at a distance of 3 metres the
illumination would be similarly l/9th lux,
or (1/3)2. Generally, a point source, of say
50 bougie-decimales, would produce, at a
distance of say 5 metres, an illumination of
50
-^ = 2 luxes.
In practice, if a surface were placed in
a room at a distance of 5 metres from
a source of 50 bougie-decimales, the illu-
mination received, if the rays be allowed
to fall perpendicularly upon the surface,
will be more than 2 luxes, because this
amount of illumination will be produced
LIGHT AND ILLUMINATION. 205
by the direct action of the source in a
space from which all other light was ex-
cluded, whereas, reflection from the walls
of the rooms, mirrors, ceilings, etc., will
increase this amount of illumination to,
perhaps, 10 luxes, and, if it were possible
to have the surfaces of the walls perfectly
reflecting, the illumination which would
be produced in all parts of the room
would be indefinitely great. Conse-
quently, the amount of illumination
received upon the surface of a desk or
table, depends not only upon the number
of lamps in a room, on their candle-power
and arrangement, but also upon the char-
acter of the surfaces of the walls and
furniture. Therefore, the question of
lighting a room is not altogether a ques-
tion of its dimensions and of the total
candle-power placed in it, but also depends
upon the arrangement of the lamps, and
206 ELECTRIC INCANDESCENT LIGHTING.
the character of the decoration and furni-
ture, and the nature of their reflecting
surfaces.
The illumination required for comfort-
able reading is from 15 to 25 luxes on
the surface of a printed page. Any
illumination less than 10 luxes is fati-
guing, if long continued. The illumina-
tion produced by full moonlight is about
l/8th lux, and that by full sunlight
80,000 luxes. The illumination in a
street as ordinarily lighted by arc lamps,
is, perhaps, 50 luxes near the ground
below an arc lamp, and about 1 lux
near the ground midway between the
lamps.
The luminous intensit}^ of an incan-
descent lamp is not the same in all direc-
tions, owing to the fact that the filament
is not a
positions,
is exposed th n TTtff-nttiTrry — JffiffTn simph
horse-shoe filament, with both legs in
one plane, gives less light in this plane
than in any other plane passing through
the vertical, because each leg intercepts
the light from the other. The mean
spherical candle-power of a lamp, in bou-
gie-decimales, is the quantity of light it
emits in lumens, divided by 12.566. In
other words, the mean spherical candle-
power of a lamp is the equivalent point
source which emits as much light in all
directions as the actual lamp does.
The spherical candle-power of an incan-
descent lamp is usually about twenty per
cent, less than its maximum horizondl in-
tensity, so that when we speak of a 16-
candle-power lamp, a point source of about
13 candles intensity would supply the same
208 ELECTRIC INCANDESCENT LIGHTING.
total number of lumens as is emitted from
the actual lamp. A point source would
Lave no base or socket and would disperse
light equally in all directions.
It may be of interest to note that a
lux-second, that is to say, the total time-
illumination produced by one lux for one
second of time, has been accepted by the
International Photographic Congress of
Brussels as the unit of time-illumination,
under the name of the pilot. A phot is
a lux-second, and is a unit of time illumina-
tion employed in photography. It is well
known in photography that the actinic
effect of light, that is its power of effect-
ing chemical decomposition as utilized in
photography, depends, for a given quality
of light, both on its intensity and on the
duration of its action. In photography,
therefore, this practical unit was required
LIGHT AND ILLUMINATION. 209
to represent tlie product of illumination
and time.
The caudle-power of incandescent lamps
varies from 1/2 candle up to 100 British
Standard candles, although both larger
and smaller candle-powers have been
specially prepared. The 16-candle-power
lamp is generally employed in the United
States, and the 10-candle-power lamp is
generally employed in Europe. The sizes
usually manufactured are 1/2, 1, 2, 3, 4, 5,
6, 8, 10, 16, 20, 32, 50, 60 and 100.
CHAPTER XI.
SYSTEMS OF LAMP DISTRIBUTION.
BROADLY speaking, there are two gen-
eral methods by means of which lamps
may be connected with a generating source
of electricity ; namely, in series, so that
the current passes successively through
each lamp before it returns to the source,
and in multiple or parallel; so that
the current divides and a portion passes
through each lamp.
Fig. 53, represents three lamps con-
nected in series, and Fig. 54, represents
three lamps connected in parallel. In the
case of the series connection shown in Fig.
210
SYSTEMS OF LAMP DISTRIBUTION. 211
53, the three lamps are so connected that
the current from the source, entering the
line at +, flows in the direction indicated
by the arrows, passes successively through
the lamps A, B, and C\ returning to the
FIG. 53.— SERIES-CONNECTED LAMPS.
source at the negative end of the line. In
Fig. 54, the three lamps A, B, Cf, are con-
nected as shown, to the positive and nega-
tive leads respectively, and the current
passes through them in the direction indi-
cated by the arrows. The arterial system
of the human body furnishes an example
212 ELECTRIC INCANDESCENT LIGHTING.
of a parallel or multiple system, since the
blood flow divides into a very great num-
ber of different channels or capillaries,
and, after passing through these independ-
FIG. 54. — MULTIPLE-CONNECTED LAMPS.
ent channels, finally unites in the veins
and returns to the heart.
In order to compare the relative advan-
tages of the series and multiple methods
of electric distribution, let us suppose that
a house has to be lighted electrically at a
distance of a mile from the dynamo, and
that for this purpose an amount of light
represented by 1,000 candles is required,
SYSTEMS OF LAMP DISTRIBUTION. 213
distributed in 50, 20-candle-power lamps.
Further, let us assume that the same
efficiency is secured in the operation of
these lamps, whatever system we may
adopt, or whatever dimensions the lamp
filament may take, a supposition in accord-
ance with general practice. Let us sup-
pose that this efficiency will be 1/4 candle
per watt. We shall then require to ex-
pend 4,000 watts of electric energy in the
lamp filaments in the house. This amount
of activity might be electrically expended
in a great variety of ways, as regards the
pressure and current of delivery, but it
will suffice to compare two ways only;
namely, the delivery of 4 amperes at a
pressure of 1,000 volts, and the delivery
of 1,000 amperes, at a pressure of 4 volts.
In each of these two cases the activity
delivered will be the same; namely 4
KW.
214 ELECTRIC INCANDESCENT LIGHTING.
If it be required to expend only 1,000
watts in the main conductors carrying the
current to the house, when all the lamps
are turned on, 5,000 watts or 5 KW will
have to be supplied at the generator termi-
nals, of which twenty per cent., or 1,000
watts, is permitted to be lost in transmis-
sion in the leads, as heat. We know that
this loss will be the product of the current
strength and the drop in volts in the two
conductors. In the first plan of 1,000
volts and 4 amperes at the house, the drop
of pressure in the two wires must be 250
volts, in order that 250 volts X 4 am-
peres = 1,000 watts, so that the resistance
of the two wires together, which shall
produce a drop of 250 volts, with a cur-
rent of 4 amperes, will be 62 1/2 ohms, or
31 1/4 ohms to each wire one mile in
length. The nearest size of wire to that
which has a resistance of 31 1/4 ohms to
SYSTEMS OF LAMP DISTRIBUTION. 215
the mile, is, No. 18 B. & S. or A. W. G.,
having a diameter of 0.0403", or, approxi-
mately, the ^th of ail inch. Such a wire
2f)
would weigh about 26 Ibs. and the two
wires forming the complete circuit would
weigh about 52 Ibs.
Considering the second plan of 4 volts
and 1,000 amperes, the drop in the wires
would have to be only one volt. In order
that the activity expended in them should
be 1 KW since 1 volt x 1,000 amperes =
1 KW the resistance in the two wires
together, to permit of a drop of but 1
volt with 1,000 amperes, must be /
of an ohm, since 1,000 amperes x
ohm = 1 volt. The two wires together
must, therefore, have a resistance of
216 ELECTRIC INCANDESCENT LIGHTING.
ohm, or each must have a resistance of
— th ohm. A wire which would have
this resistance would have 62,500 times
the cross-section and weight of the wire in
the preceding case ; so that, instead of re-
quiring 52 Ibs. of copper, in all, for the
two miles of conductor, we should require
approximately 3,250,000 Ibs., or 1,625
short tons of copper.
It is, therefore, evident that although a
given electric activity can be expended in
incandescent filaments either at a high
pressure or at a low pressure, the advan-
tage of a high pressure is very great, when
the power is to be transmitted electrically
to a distance. If we double the pressure
of transmission ; i. e., if we double the
number of volts between the two main
conductors, we require four times less cop-
SYSTEMS OF LAMP DISTKIBUTION. 217
per for a given percentage of loss of
activity in them. Thus, in the preceding
case, when we increased the pressure of
delivery from 4 volts to 1,000 volts, we
increased it 250 times, and, therefore, wre
diminished the amount of copper which
was required for twenty per cent, loss, 250
X 250 or 62,500 times. Consequently, the
first essential for economical distribution
of electric power to a distance, either for
lamps, or for any other purposes, is high
electric pressure of delivery.
If then we adopt provisionally, the plan
of supplying the house above considered
at a pressure of 1,000 volts and 4 am-
peres, this would appear to be most readily
carried out at first sight, by connecting
50 lamps in a single series through the
house, each lamp being intended for 20
volts and 4 amperes. When all the lamps
218 ELECTRIC INCANDESCENT LIGHTING.
are lighted, the total current would be 4
amperes, when the total pressure of de-
livery amounted to 50 x 20, or 1,000 volts.
Though such a system could, doubtless, be
installed, yet it would possess several dis-
advantages. In the first place, unless some
device were provided whereby a faulty
lamp became automatically short-circuited,
the failure of any single lamp would
interrupt the entire circuit and extinguish
all the other lamps. Moreover, the pres-
sure which would have to be supplied to
the house between the main conductors
would depend upon the number of lamps
employed. When a single lamp only was
lighted, the pressure required would be 20
volts and the current 4 amperes. This
would mean that the generating dynamo
in the station supplying the house could
only be used for that particular house,
since, if two houses were supplied from
SYSTEMS OF LAMP DISTRIBUTION. 219
the same dynamo, one might have all the
lamps turned on and thus require 1,000
volts and 4 amperes, while the other
might have only half its lamps turned on,
thus requiring 500 volts and 4 amperes.
For this and other reasons it is now uni-
versally considered that incandescent light-
ing on any extended scale must necessarily
be conducted by a multiple-arc system.
It would be practically impossible to
construct incandescent lamps capable of
being operated in parallel at the pressure
of 1,000 volts assumed in this case; for,
each lamp would have to be capable of
4
taking a current of — ths ampere, at a
pressure of 1,000 volts. The resistance
would have to be 12,500 ohms hot, so that
the filament would have to be very fine
and long. Such a lamp would be quite
220 ELECTRIC INCANDESCENT LIGHTING.
impracticable, and, moreover, the pressure
of 1,000 volts is • not considered safe to
introduce into a building. The maximum
pressure for which it has been possible,
until recently, to construct incandescent
lamps has been 120 volts, so that incandes-
cent lighting between a single pair of
conductors has been practically limited to
a pressure of 115 volts at the lamp
terminals.
The lack of economy of 11 5- volt pres-
sures for incandescent lighting at a dis-
tance, as regards the amount of copper
required, was early apparent, and a method
was invented and introduced which is
practically a compromise between the
series and parallel systems ; that is to say
a method was invented whereby the ad-
vantages of the parallel connection of
lamps were secured, together with the
SYSTEMS OF LAMP DISTRIBUTION. 221
advantage of higher pressure obtained by
coupling lamps in series. This method is
fundamentally what is called a series-mul-
tiple system, and in practice is what is
generally called the three-wire system.
A +
c— c
FIG. 55. — THREE- WIRE SYSTEM, SERIES-MULTIPLE CON-
NECTION.
The three-wire system is illustrated in
Fig. 55. Here there are two multiple-arc
circuits, one between the mains A and B,
and the other between the mains B and C.
Between each of these pairs of mains the
pressure is 115 volts, supplied by a sepa-
rate dynamo as shown. The positive
terminal of the dynamo D2, being con-
222 ELECTRIC INCANDESCENT LIGHTING.
nected to the negative terminal of $ the
dynamo Dly it is evident that between the
mains A A and C C, there will be a pres-
sure of 230 volts. In the case shown
there are 12 lamps in all, or 6 on each
side of the system, so that about 3
amperes will be flowing through the mains
A A and C O, and no current will pass
through the neutral conductor B B. If
the number of lamps on the two sides of
the system be unequal, the difference be-
tween the current strengths will return
by the neutral conductor ; for example,
if all the lamps between B and C, are
turned off, 3 amperes will flow along the
positive main A A, and return by the
neutral main B B, no current passing
through the negative main. Since, on the
average, when the Aviring is judiciously
carried out, the loads on the two sides of
the system will usually nearly balance each
223
other, ^^^ei^tjral conelucton^ifjxjpfily have
to carry ast^afli^ua^oi^f^Te current in
the outside mains, and may therefore be
much lighter. If the neutral conductor
could be entirely dispensed with, we
know that the copper required to sup-
ply the system with a given percentage of
loss in transmission would be four times
less on the three-wire system, than on
the two-wire system, since the pressure
of distribution would be doubled. The
three-wire system would, therefore, ensure
a saving of seventy -five per cent, in the
amount of copper required for the mains.
In practice, however, the amount of copper
in the neutral conductor averages, over
an entire system, about sixty per cent,
of that in the outside conductor, so the
neutral is made a little more than half
as heavy as either of the outside mains.
Under these conditions, the actual saving
224 ELECTRIC INCANDESCENT LIGHTING.
in weight of copper throughout the supply
conductors of a three- wire system is about
67 1/2 per cent, over that necessary for a
two-wire system having the same loss in
transmission.
A-f
) <§>
c
FIG. 56.— MULTIPLE-SERIES CONNECTION.
Fig. 56 represents a multiple-series sys-
tem equivalent to a three-wire system with
no neutral, and supplied at 230 volts pres-
sure. The disadvantage of such a system,
however, is that two lamps have to be
turned on and off simultaneously. The
three- wire system of Fig. 55 gives inde-
pendent control over every lamp.
SYSTEMS OF LAMP DISTRIBUTION. 225
The principle of the three-wire system
has been extended to four- and five-wire
systems. Four-wire systems are very rare.
Five-wire systems are employed in Europe
but have never come into favor in the
United States. A five-wire system saves
about ninety per cent, of the copper re-
quired for a two-wire system, but requires
four dynamos in series at the central sta-
tion, five sets of conductors and complica-
tion in house wiring and meters.
The three-wire system is in very
extended use in the United States. It
commonly happens that one three-wire
central station will distribute light and
power over an area whose radius is some-
what greater than one mile, whereas, with-
out the use of the three- wire system, the
radius of commercial incandescent lighting
from a central station would be probably
226 ELECTRIC INCANDESCENT LIGHTING.
only about one-half a mile or eight times
less area.
The drop of pressure, which is permitted
in incandescent lighting, does not depend
entirely upon the activity uselessly ex-
pended in the main conductors. For ex-
ample, in cases where capital would be
difficult to secure, and the interest upon
the capital invested would be large, it
would be desirable to employ compara-
tively small conductors, and waste a com-
paratively large percentage of the total
power in them. This would necessitate a
comparatively great difference of electric
pressure between the lamp terminals and
the generator terminals. In the case of a
single house supplied with 115-volt lamps,
it would not be a matter of much conse-
quence whether the pressure at the central
station were 116 or 166 volts, provided
SYSTEMS OF LAMP DISTRIBUTION. 227
the lamp pressure remained constant, but
where incandescent lamps are distributed
along street mains, in a city, and have
to be supplied at all distances from a few
yards to a mile or more from the central
station, it is absolutely necessary that the
pressure shall be nearly uniform through-
out the system, since, otherwise, the lamps
in or near the station will be at an
unduly high pressure, and will conse-
quently *be brilliant and short-lived, while
the more distant lamps will be burned at
an unduly low pressure, and be dull and
long-lived. The drop of pressure permis-
sible in the supply conductors is, conse-
quently, as much a matter of regulation
of pressure, and of uniformity of candle-
power, as it is a consideration of economy
in the transmission of electric power. If
lamps were less sensitive to changes in
pressure than they are, the amount of cop-
228 ELECTRIC INCANDESCENT LIGHTING.
per which would be employed in incan-
descent lighting would be less than it
actually is, but all the improvements made
of recent years in incandescent lamps have
been improvements in their efficiency,
whereby a smaller amount of activity is
required for a given production of light,
and this, as we have seen, is attended by
the development of a higher tempera-
ture and a greater sensibility to variations
in pressure, so that the most economical
lamps are also lamps which, other things
being equal, require a closer regulation of
pressure at their terminals.
The difficulty of maintaining a nearly uni-
form pressure over all parts of the mains of
a large incandescent system has been largely
overcome by the use of what are called
feeders. A feeder differs from an ordinaiy
supply conductor, or main, in that no lamp
SYSTEMS OF LAMP DISTRIBUTION. 229
or receptive device is directly connected
with it ; its sole purpose being to supply
the mains from the central station at some
distant point as indicated in Fig. 57.
Here D, is the dynamo at the central
station. F F, are feeders carrying the
current from the dynamo to some cen-
J )
— i — i
F
FIG. 57.— FEEDER DISTRIBUTION.
tral point in the mains A Aly B B^
In this way the difference in pressure be-
tween the various lamps depends only
on the drop of pressure in the mains, and
not on the drop of pressure in the feeder.
Thus, if the pressure at the lamps at A and
Al9 be 115 volts, the pressure at the feed-
ing point F, may be 116 volts, while the
pressure at the dynamo may be 150 volts.
230 ELECTRIC INCANDESCENT LIGHTING.
If the same lamps were supplied without
feeders as shown in Fig. 58, and the same
limiting difference of pressure maintained
between the lamps as in Fig. 57, namely 1
volt, it would be necessary to have prac-
tically 116 volts at the dynamo terminals
and 115 volts at the most distant lamp.
B •
FIG. 58.— TREE DISTRIBUTION.
This probably would require much more
copper in supply conductors than when
feeders are employed. The conductors in
Fig. 58 are shown as tapering or diminish-
ing in size towards the distant end.
Feeders may equally well be applied to
three-wire systems. Fig. 59 represents
diagrammatically the supply-mains of a
city district containing four blocks, 1, 2, 3
SYSTEMS OF LAMP DISTRIBUTION.
231
and 4. Here the three- wire mains extend
round the block-facings in one continuous
network. These mains are supplied from
the central station at S9 by the three- wire
FIG. 59. — DIAGRAM OP THREE- WIRE FEEDER
DISTRIBUTION.
feeders represented by the dotted lines, at
the feeding points, A, B, C and D. In
this way the pressure in the network of
mains may be within two per cent, of the
mean value, of say, 115 volts, while the
pressure at the central station may be 130
volts, representing a drop in the feeders of
232 ELECTRIC INCANDESCENT LIGHTING.
15 volts. If the lamps were connected
across the feeders they would be subjected
to a total difference of pressure, over the
entire system, amounting to 17 volts, but,
by connecting the lamps to the mains only,
they are rendered entirely independent of
the drop which occurs in the feeders.
If the system of mains be unequally
loaded, as for example, when the area over
which they extend, comprises both a resi-
dence district and a business district, so
that the load shifts in the morning and
afternoon to the business district, and in
the evening, almost entirely to the resi-
dence district, it may happen that the
load on some particular feeders may be
much greater than the load on others.
Consequently, the drop in the loaded
feeders will be in excess of that on the
comparatively idle feeders. Under these
SYSTEMS OF LAMP DISTRIBUTION. 233
circumstances, the pressure at the mains,
near the terminals of the idle feeders, will
be higher than that at the terminals of the
loaded feeders, thus bringing about an
inequality of pressure, prejudicial to the
life and proper performance of the lamps.
The difficulty arising from the inequal-
ity in the feeder load may be overcome in
one or more of four ways :
(1) By disconnecting certain feeders
from the bus-bars, or main terminals in a
central station, so as to increase the load
and drop on the remaining feeders.
(2) By introducing artificial resistances,
called feeder regulators, into the circuit of
the idle feeders; so as to increase artificially
the drop of pressure which exists in them.
(3) By employing more than one pres-
sure in the central station, that is to say,
by having one set of dynamos operating at
234 ELECTRIC INCANDESCENT LIGHTING.
FIG. 60. — FEEDER EQUALIZER RESISTANCE.
LAMP DISTRIBUT^W 235
for the
a pressure
supply of the shorter feeders to tlie area
AN7ithin the vicinity, and another set of
dynamos delivering, perhaps, 135 volts, for
FIG. 61.— EQUALIZER SWITCH.
the supply of longer feeders connected to
the outlying districts.
(4) By introducing more copper into
the system, either in the form of additional
236 ELECTRIC INCANDESCENT LIGHTING.
feeders, so as to share and equalize the
load, or in the form of more numerous
mains to distribute and equalize the pres-
sure.
Fig. 60, represents a form of resistance,
suitable for feeder regulation. Here a
number of spirals of heavy iron wire are
mounted in a fire-proof frame and so
arranged that under the influence of the
handle and switch, shown in Fig. 61, they
may be inserted in the circuit of a feeder
either in parallel or miseries.
In modern large central stations feeder
equalizers are rarely employed. The best
practice employs more than one pressure.
CHAPTER XII.
HOUSE FIXTURES AND WIRING.
THE incandescent lamp, when located in
a house, is either installed as a fixture, or a
certain freedom of motion is given to it,
so that, within certain limits, the lamp is
portable. This portability is effected by
maintaining the lamp in connection with
the mains by means of a flexible conductor
or lamp cord, usually called a flexible cord.
The lamp is then portable to the extent
of the length of the cord. Various forms
are given to portable lamps, two of which
are shown in Figs. 62 and 63. In Fig. 62,
the flexible cord c c, is attached to a read-
ing lamp, mounted on a stand as shown.
238 ELECTRIC INCANDESCENT LIGHTING.
FIG. 62. — PORTABLE LAMP FOII DESK USE.
This lamp can be raised and lowered
within a limited range, as well as turned
HOUSE FIXTURES AND WIRING. 239
FIG. 63. — FLEXIBLE LAMP PENDANT WITH ADJUSTER.
240 ELECTRIC INCANDESCENT LIGHTING.
about its axis without shifting the base.
Fig. 63, shows a form of movable lamp, in
which a limited portability is obtained by
what is generally known as a flexible
pendant. Here the lamp is hung from
FIG. 64. — FLEXIBLE SUPPORT FOB LAMP.
the ceiling by a flexible lamp cord. By
means of an adjuster J^ the lamp can be
raised or lowered.
The limited portability given to a lamp,
by attaching it to a sufficiently long
HOUSE FIXTURES AND WIRING. 241
flexible pendant, enables the light to be
applied to a variety of purposes, such, for
example, as the lighting of a music stand,
FIG. 65. — FLEXIBLE SUPPORT FOR DESK LAMP.
as shown in Fig. 64, or the lighting of a
desk, as shown in Fig. 65. Fig. 66, shows
a device for tilting a flexible pendent lamp
in any desired direction.
Fixed lamps, as their name indicates,
are lamps attached to electric fixtures, and,
242 ELECTRIC INCANDESCENT LIGHTING.
therefore, cannot be moved. They take a
great variety of forms, such as the bracket
lamp shown in Fig. 67, designed for
Jflfe
FIG. 66.— TILTED LAMP.
For ceiling
attachment to the wall,
attachment, lamps are either made of the
simple pendant type, as shown in Fig. 68,
or several lamps are placed
together
HOUSE FIXTURES AND WIRING. 243
in a cluster in an electrolier, as shown in
Fig. 69. As in gas lighting, the incandes-
cent bracket lamp is sometimes given a
movable arm so as to permit the lamp to
FIG. 67. — BRACKET LAMP.
be moved in one plane, within a certain
radius. Such a lamp is shown in Fig 70.
The size of the wire employed inside
a house will depend upon the amount of
current which the conductor is designed to
244 ELECTRIC INCANDESCENT LIGHTING.
FIG. 68.— PENDANT LAMP.
HOUSE FIXTURES AND WIRING.
245
carry. When of small size, the conductor
is given the form of a single wire, but,
FIG. 69.— ELECTROLIER.
in order to secure greater flexibility, larger
sizes are almost invariably stranded, that
246 ELECTRIC INCANDESCENT LIGHTING.
is, composed of several independent wires.
The former are called solid wires and the
latter stranded wires. In Fig. 71, the
solid conductor is marked <?, and has one
coating of insulator d, which is afterward
FIG. 70. — BRACKET LAMP WITH MOVABLE ARM.
covered by a braiding b. The stranded
conductor shown at C\ consists of seven
wires, twisted together as shown, and is
covered by two coatings of insulating
material, I) and E, respectively, and
finally by a coating of braid B.
Fig. 72, shows two other forms of
stranded conductors ; the wire marked A,
HOUSE FIXTURES AND WIRING.
247
is provided with a highly insulating
material called okonite ; that marked B,
has in addition, a coating of braid. The
wire at A, is equivalent to No. 6 A.W. Gr.
d b
FIG. 71. — SOLID AND STRANDED CONDUCTORS.
in cross-sectional area. The insulation re-
sistance of a mile of this wire, when sub-
merged in water, is 1,000 million ohms,
that is, one billion ohms, or a begohm.
A flexible cord, such as has already
been referred to in connection with port-
able lamps, is necessarily a double con-
248 ELECTRIC INCANDESCENT LIGHTING.
due tor, since the current must be passed
both into and out of the lamp. These
two conductors are separately insulated,
and are then either twisted together, form-
ing what is called a twisted double con-
FIG. 72.— OKONITE-COVERED STRANDED WIRES.
ductor, or are laid side by side, and laced
together by a covering of braid, forming
what are then called parallel or twin con-
ductors. Fig. 73, shows some forms of
double flexible conductors. These are
first separately insulated and are then silk-
covered. They are sometimes technically
called silk lamp cord. In order to attain
HOUSE FIXTURES AND WIRING. 249
the flexibility required in such cords they
are composed of a comparatively large
number of fine copper wires stranded to-
gether.
FIG. 73. — FORMS OF DOUBLE FLEXIBLE CONDUCTORS.
We will now trace the network of con-
ductors in a house which we will suppose
receives its current from the street
mains, to the lamps in the different por-
tions of the house. First; proceeding
from the street to the house, we find a set
of conductors leading into the house,
250 ELECTRIC INCANDESCENT LIGHTING.
called the service wires. These in a two-
wire system consist of two conductors, and
in a three-wire system of three conductors.
Within the house the system of conduct-
ors may be arranged under the following
heads; namely,
(1) The risers, or the supply wires
which carry the current up from the ser-
vice wires to the different floors of the
house. They may be a single set or a
multiple set, but each set will be double
or triple according as the house is wired
on the two- or three-wire system.
(2) The mairts, or the principal supply
conductors running from the risers or ser-
vice wires along the different corridors or
passages. There are usually as many sep-
arate systems of mains as there are floors.
(3) The sub-mains or the supply-con-
ductors which branch off from the mains
along the side passages.
(4) jBT§$&$®t fft.jfffl^ qpj|$fed^m>m the
mains into i^T^uuIiIi?"uT **to fixtures in
halls. Roughly speaking, therefore, the
risers correspond to the trunk of a tree
through which the sap is fed ; the mains
correspond to the boughs ; the sub-mains
to the smaller boughs ; and the branches
to the twigs. The lamps or fixtures cor-
respond to the leaves and flowers.
Risers are usually of larger cross-section
than the mains; the mains are of larger
cross-section than the sub-mains, and the
sub-mains, in their turn, are larger than
the branches. This must naturally be the
case, since the risers must carry all the
current, the mains divide the current
among themselves, and the branches carry
only the current of the few lamps wired
upon them. We may imagine that each
lamp has a certain size of wire connected
252 ELECTRIC INCANDESCENT LIGHTING.
with it from the street mains, direct to the
socket, and, that since these wires run side
by side, they may be regarded as collected
into a single larger wire, such as a main or
riser.
The smallest size of wire which is per-
mitted to be used in wiring a building,
in the United States, is No. 18 A. W. G.
wire, having a diameter of 0.040".
The wiring of a building should be de-
signed in such a manner, that when all the
lamps are burning at any one time, the drop
in pressure between the street mains and the
most distant lamp shall not exceed a cer-
tain small percentage, usually three per
cent. This drop is calculated by deter-
mining the total amount of current in
amperes which will pass through the vari-
ous mains, sub-mains and branches, deter-
HOUSE FIXTURES AND WIRING. 253
mining for each such a resistance as will,
when carrying this current, produce drops
of pressure, the maximum sum of which
along any line shall not exceed the re-
quired percentage.
In very large buildings, a feeder system is
sometimes employed ; that is to say, the
service wares are connected by feeders
to centres of distribution, from which
mains extend both upward and down-
ward.
Two supply wires, carrying the full lamp
pressure between them, are never per-
mitted to remain in contact with each
other, even though both are insulated,
except in cases of flexible conductors,
which are in plain view and which never
carry, under normal circumstances, a pow-
erful current.
254 ELECTRIC INCANDESCENT LIGHTING.
Insulated wires are never allowed to
come into contact with concealed wood-
work, but when passing through wooden
beams or floors should be protected by in-
sulating tubes of porcelain or hard rubber.
There are three methods of carrying
out interior wiring; namely,
(1) Cleat work.
(2) Moulded work.
(3) Concealed work.
Cleat work is the simplest and cheapest,
but least ornamental type of wiring. The
wires are carried in insulating receptacles,
of wood or porcelain, in plain view on the
ceilings or upper part of the walls. The
wires should never rest directly upon the
walls or ceilings, but should be supported
by the cleat, a full half inch away from
the same.
HOUSE FIXTURES AND WIRING. 255
Fig. 74, shows two forms of wooden
cleats. Fj f] are the front pieces which
FIG. 74. — WOODEN CLEATS.
clamp the wires, and B B, are the brackets
which support the wires free from the
256 ELECTRIC INCANDESCENT LIGHTING.
walls. S Sj are the screw holes by which
the cleats are secured in place and clamped
together. The wires of opposite polarity
are always separated by such cleats to a
FIG. 75. — FORM OF SCREW CLEAT.
distance at least 2 1/2" apart. Cleat work
is only suitable for indoor work in dry
localities. Figs. 75 and 76 show forms of
screw cleats employing respectively wood
and glass insulation around the wire.
HOUSE FIXTURES AND WIRING.
257
Moulded work is more expensive than
cleat work, but presents a more sightly ap-
pearance. Fig. 77, shows different forms of
three-wire moulding, suitable for different
FIG. 76. — FORM OF SCREW CLEAT.
sizes of conductor. The moulding is made
in two parts ; namely, the base or mould,
with the grooves formed in it, and the
upper part, or capping, which covers the
mould. These moulds and cappings are
258 ELECTRIC INCANDESCENT LIGHTING.
usually made of soft pine, and are cut into
lengths of about ten feet. The moulds are
FIG. 77. — SECTIONS OF MOULDINGS.
first screwed in position on the walls or
ceilings ; the wires are then laid in them,
and, finally, the cappings are secured by
HOUSE FIXTURES AND WIRING.
259
screws over them, care being taken not to
injure the conductors in screwing on the
cappings. The mouldings are usually
FIG. 78. — PICTURE AND ORNAMENTAL MOULDING.
painted of a color to conform with the
ornamentation of the walls or ceilings on
which they rest. Fig. 78, represents at Py
260 ELECTRIC INCANDESCENT LIGHTING.
a form of moulding suitable for hanging
pictures around the walls of a room, and
at 0, a type of ornamental moulding.
The most difficult problem connected
with the distribution of wires by moulding,
lies in the connection of the electroliers
with the wires in the passages without
presenting an unsightly appearance. This
is sometimes accomplished by the use of
dummy moulding, or ornamental mouldings
symmetrically arranged on the ceiling from
the centres of the electrolier, in one only of
which mouldings the wires are placed.
The best solution of the problem of
avoiding the unsightly appearance of wires
is obtained by concealed work, where the
conductors are buried under floors or in the
walls and ceilings. This method should
not be adopted unless the wires besides
HOUSE FIXTURES AND WIRING. 261
their insulating cover, are provided with
a moisture-proof sheet or tube of papier-
FIG. 79. — INTERIOR CONDUITS.
mache or metal. Such protecting tubes
form in reality a conduit employed inside
262 ELECTRIC INCANDESCENT LIGHTING.
the building and generally called an
interior conduit.
Interior conduits may be made in a
variety of ways, one of which is shown
in Fig. 79. Conduits made of tubes of
papier-mache are soaked in a bituminous
m
FIG. 80.— BRASS-COVERED CONDUIT.
solution, which serves the double purpose
of rendering them insulating and practi-
cally water-proof. Moreover, when house
wires are carried through a complete
system of interior conduits, the wires can
be withdrawn and replaced at any time
without disturbing the walls or ceilings.
Fig. 80, shows a conduit tube sheathed
with a thin layer of brass so as to be
HOUSE FIXTURES AND WIRING.
263
water-tight. Fig. 81, shows a brass tube
joint connecting different lengths of con-
duit. Aj is a joint for connecting ordinary
conduit, and B, a joint connecting brass-
covered sheathed conduit. Where several
FIG. 81.— INTERIOR CONDUIT JOINTS.
conduits are united together, a junction box
is provided containing a number of outlets,
corresponding with the number of conduits
as shown in Fig. 82. These boxes are
closed by a metallic cover. Fig. 83, shows
two forms of connecting boxes for holding
264 ELECTRIC INCANDESCENT LIGHTING.
joints between the mains and branches of a
two- wire or three-wire system running in
FIG. 82. — INTERIOR CONDUIT JUNCTION BOXES.
interior conduits. The branch wires in
this case proceed from the box through
the smaller apertures. In cellars the con-
HOUSE FIXTURES AND WIRING.
265
duits are frequently heavily sheathed with
brass or iron.
FIG. 83. — CONNECTION Box FOH INTERIOR CONDUITS.
The cost of wiring a building depends
in a great measure upon the number of
266 ELECTRIC INCANDESCENT LIGHTING.
outlets required, that is on the number
of points at which the wires must be
brought out for connection to switches and
lamps. Where the lamps are installed in
groups, as in electroliers, fewer outlets are
required and the cost is less than if the
lamps were installed singly.
Concealed work is generally much more
readily and cheaply installed during the
construction of a building than at a sub-
sequent period. It has become customary
in large cities to wire all new buildings,
even though no arrangement has been
made to supply them immediately with
electric current. In some cases interior
conduits are placed in a new building with
the intention of subsequently wiring the
building.
It is often convenient to be able to turn
G. 267
some dis-
(Ts purpose, be-
sides the key frequently provided in the
socket for turning on or oft* the individual
lamp, lamp-switches are placed in the
branches, or in the mains, whereby all the
lamps supplied by said branches or mains
may be lighted or extinguished at once.
The size and character of the lamp switch
will depend upon the current strength it
is intended to control. Switches may be
divided into two general classes ; namely,
single-pole switches and double-pole switches.
In a single-pole switch, as the name indi-
cates, the connection is broken on one side
only of the two supply conductors; in a
double-pole switch, it is broken on both
sides. Switches are made in a great
variety of forms, a few of which are illus-
trated in the accompanying figures. Fig.
84, shows a form of simple single-pole
268 ELECTRIC INCANDESCENT LIGHTING.
switch. By turning the key K, the spring
/SY, is brought into contact with the supply
spring P, thereby ensuring the closing of
the branch circuit of the lamp through the
FIG. 84.— SIMPLE FORM OP SINGLE-POLE SWITCH.
switch terminals A and B. Single-pole
switches should never be employed, except
for a small number of lamps. Fig. 85,
shows a different type of switch. $, being
a single-pole switch, with two terminals,
and Dj a double-pole switch with four.
HOUSE FIXTURES AND WIKING. 269
s
FIG. 85.— SINGLE- AND DOUBLE-POLE SWITCHES.
In the latter case, two of these terminals
are connected with the branch wires and
the other two with the supply mains.
270 ELECTRIC INCANDESCENT LIGHTING.
Fig. 86, shows a larger form of double-pole
switch, intended for a comparatively large
number of lamps, say 100. Generally, care
FIG. 86. — LARGER FORM OF DOUBLE -POLE SWITCH.
is given to make the contacts of considera-
ble surface area, in order that no undue
heating may arise from imperfect contact.
Fig. 87, represents a different type of
HOUSE FIXTURES AlSTD WIRING.
PIG. 87. — DOUBLE-POLE SWITCH.
272 ELECTRIC INCANDESCENT LIGHTING.
double-pole switch. Here the turniDg of
the handle forces a lever into or out of a
groove, whereby the contact pieces, A B
and CD, are either insulated or are con-
nected together.
Fig. 88, represents a form of switch
which is sunk into the wall, so that its
plane outer surface is flush with the sur-
face of the wall. The switch is, therefore,
usually called a flush switch.
In normal operation, the current which
passes through the conductors in a build-
ing is only that which is necessary to sup-
ply the high resistance lamps connected
with them. If, however, an accidental
short-circuit, or direct connection, were to
take place between the positive and nega-
tive mains in any part of the building, the
supply wires connected with those mains
HOUSE FIXTURES AND WIRING. 273
would be apt to receive a rush of current
that might render them white hot. In
order to prevent the danger arising from
this overheating, a form of automatic
FIG. 88. — FLUSH SWITCH.
switch has been designed, which im-
mediately opens the overloaded circuit,
and thus interrupts the current. The
automatic switch invariably employed in
incandescent mains is quite simple in its
construction and operation. It is called
274 ELECTRIC INCANDESCENT LIGHTING.
a fuse cut-out, or safety fuse, and consists
essentially of a strip or wire of lead, or
fusible alloy, interposed in the circuit.
The area of cross-section of the fuse strip
is such that, while it will readily carry the
normal current intended to be supplied by
the conductors with which it is connected,
it will immediately fuse on the passage of
an abnormal current.
Fig. 89, shows a form of branch cut-out;
i. e., a fuse block inserted between a pair
of branch wires and the mains supplying
them. These branch Hocks consist of a
glazed porcelain base, provided with two
grooves in which are four terminals A, B,
O and Dj one pair of which, say A and B,
are connected to the branch wires, while
the other pair, O and D, are connected to
the mains. A, is connected to 6y, and B,
to Z>, through two strips /SJ S9 of fusible
FIXTURES AND WIRING. 275
alloy, clamped at the ends under smaller
screws, as shown. A suitable porcelain
FIG. 89.— BRANCH BLOCK.
cover is screwed over the whole, so that
the fuse is completely protected. If the
current passing into the branch wires be-
276 ELECTRIC INCANDESCENT LIGHTING.
comes dangerously strong, the fuses S, S9
are melted or blown, and thus automatically
interrupt the branch circuit.
Various forms of fuses are employed.
A common type is represented in Fig. 90.
Here the fuse consists of lead-alloy wire
inserted in a small glass plug P, somewhat
resembling the socket of a lamp. This
plug screws into receptacles in the cut-out,
in such a manner that when a fuse is
melted or blown, it is only necessary to
remove the plug and screw in a new one.
The forms of cut-out shown are designed
for use in connection with two-wire or
three-wire mains.
A complete system of house wiring
includes a number of fuses. Large fuses
are inserted in the service wires, where
they enter the cellar, these being usually
HOUSE FIXTURES AND WIRING. 277
FIG. 90.— PLUG CUT-OUTS.
called the main fuses. The main cut-out
fuses are usually placed close to a main-
switch, where the current can be turned
278 ELECTKIC INCANDESCENT LIGHTING.
on or off from the house at will. All con-
nections between risers and mains, or
between mains and -sub-mains, or sub-mains
and branches, are usually provided with
fuse cut-outs of the size corresponding to
the current which they have to carry. In
the circuits of the various fixtures small
fuses are frequently inserted. Fig. 91,
represents a few types of such fixture cut-
outs, as they are called, shaped to conform
with the fixtures for which they are
intended. If a short circuit should take
place, close to an individual lamp, the
small fuses in the lamp circuit would melt,
either at the fixture cut-out, or at the
branch cut-out supplying the same. If a
short circuit take place in a sub-main, the
fuse at the junction between the main and
sub-main would likewise melt, while
finally, if a short circuit should occur in
one of the larger mains, the fuses in the
PRCPERTY CF i
CHOUSE FIXTURES AND
279
FIG. 91. — FIXTURE CUT-OUTS.
main cut-out, where the service wires enter
the building, would, probably, be instantly
blown.
In all large installations, it is important
280 ELECTRIC INCANDESCENT LIGHTING.
to be able to detect quickly the location of
a melted fuse in order that it may be
FIG. 92.— DISTRIBUTION Box.
replaced and tlie extinguished lamps
restored as soon as possible. This is fre-
HOUSE FIXTURES AND WIRING.
281
quently done by collecting all the fuses
belonging to some particular portion of
the system at a point called a distributing
FIG. 92A. — DISTRIBUTION Box.
point, usually at a junction of risers and
mains, or mains and sub-mains. The in-
take wires ; i. e., those which feed the box,
are usually brought to a pair of metal bars
282 ELECTRIC INCANDESCENT LIGHTING.
in a box called a distribution box lined
with some fire-proof material let into the
wall. The out-put wires; i. e., those which
take their supply from the box, are at-
tached to separate terminals which main-
tain connection with the metal strips,
through safety fuses of the proper size.
Figs. 92 and 9 2 A, illustrate a particular
form of distribution box, which is let flush
into the wall and lined with fire-proof mate-
rial. A B, are the two intake wires. This
box is intended for use in connection with a
two- wire system. A, is placed in electric
connection with the metal strip g li, and B,
with the metal strip & Z, through two safety
fuses. The out-put wires are a b, c d and
ef; a, c and e, being positive and b, d and/,
negative. These wires are all placed in elec-
tric connection with their respective strips,
each being provided with a safety fuse, as
HOUSE FIXTURES AND WIRING. 283
shown, s, s, s, are three switches, for con-
trolling their independent circuits. Fig.
92 A shows the box with its cover in posi-
tion, but with the door opened. At the
top of Fig. 92 is a cross-section of the box.
In many cases these boxes have glass doors
through which the condition of the fuses
can be readily inspected.
CHAPTER XIII.
STREET MAINS.
IN small towns, systems of incandescent
distribution are effected by means of over-
head wires or overhead main conductors;
that is to say, both feeders and mains are
insulated wires supported on poles. This
method of construction is adopted, both on
the score of economy and the ease of inspect-
ing, repairing and connecting the Avires.
In large cities, however, where the number
of such conductors is necessarily greatly
increased, and where, moreover, multi-
tudinous conductors are required for other
than electric-lighting systems, the wires
are, to a greater or less degree, necessarily
buried underground.
STREET MAINS. 285
Three methods are applicable to under-
ground conductors ; namely, subways, con-
duits and tubes. Of these methods, the first
two provide means whereby the wires can
be replaced or removed with greater or less
readiness. By the third method, the tubes
are actually buried in the ground and need
excavation for examination or repair.
A subway differs from a conduit, in that
it consists of an underground tunnel of
sufficient dimensions to permit the passage
of a man. Underground subways, unques-
tionably provide the readiest means for
operating an extended system of con-
ductors. There are, however, two serious
difficulties that lie in the way of their
extensive adoption ; namely, expense, and
want of room. In our larger cities, an
unfortunate lack of uniformity has existed
in the mode of use of the space beneath
286 ELECTRIC INCANDESCENT LIGHTING.
the streets, and pavements, for the location
of the sewer, gas and water-pipes, steam
heating pipes, and the various systems of
electric conductors which are to-day so
imperatively needed in a modern city.
Unfortunately, in too many cases, the
location of these lias been placed under
the control of different and frequently
antagonistic officials. A lack of space has
consequently resulted, so that in most of
our large cities, the construction of a sub-
way system would require an entire recon-
struction of the systems of sewers, water,
gas and electric mains.
Where a subway is employed, it is
necessary to ensure its complete drainage
and also to provide for an efficient system
of ventilation, whereby the accidental
leakage of gas into the subway shall not
produce explosive mixtures with air.
STREET MAINS. 287
The conduit affords a much readier and
more easily applied system for under-
ground mains or conductors. A conduit
differs from the subway in that it merely
provides a space for the wire or cable.
Various forms of conduits have been
devised, but all consist essentially of
means whereby tubes, intended for the
reception of cables, and generally of some
insulating; material, are buried in the
o /
ground. They are provided with man-
holes, or a free space in the street, extend-
ing below the level of the conduits,
large enough to admit a workman.
When it is desired to introduce a wire
into a conduit, or to replace an injured
wire, the wires are drawn into or from
the conduits at the manholes.
Figs. 93 and 94, illustrate in section a
conduit formed of creosoted wood, laid
288 ELECTRIC INCANDESCENT LIGHTING.
together in sections, so as to leave cylin-
drical spaces between them, through
which the wires or cables may be drawn.
This system is frequently used for the
FIG. 93. — CONDUIT OP CREOSOTED WOOD.
reception of lead-covered high-tension
wires, and telephone cables.
While overhead conductors may be un-
objectionable in small towns or villages,
yet, in large cities, where the need for
wires is great and, moreover, is con-
STREET MAINS. 289
stantly increasing, a condition of affairs
might readily be brought about such as
is shown in Fig. 95, which represents
the condition of a street with the
FIG. 94. — CONDUIT OF CREOSOTED WOOD CUT AWAY TO
SHOW STRUCTURE.
many aerial wires that are to be ex-
pected. Contrast this with the altered
appearance of the same street, when the
wires were placed in underground con-
duits as shown in Fig. 96, and the advan-
tage of underground wires, from an aesthetic
standpoint, is manifest.
290 ELECTRIC INCANDESCENT LIGHTING.
FIG. 95. — VIEW OP CITY STREET WITH OVERHEAD WIRES.
STREET MAINS. 291
FIG. 96. — VIEW OF CITY STREET AFTER REMOVAL OF
OVERHEAD WIRES.
292 ELECTRIC INCANDESCENT LIGHTING.
Fig. 97, represents the third method for
underground conductors, viz.; the under-
ground tube. Here an iron pipe is em-
ployed, containing three insulated conduc-
tors, and intended for use in a three- wire
J
FIG. 97.— TUBE CONTAINING THREE SEPARATELY
INSULATED CONDUCTORS.
system of distribution. The iron pipe is
provided for the purpose of protecting the
conductors from mechanical injury. Fig.
98, shows cross-sections of different sizes
of these tubes. A, B and C, are the cross-
sections of the copper conductors, sur-
rounded and supported by a bituminous
insulating material. The outer ring is the
section of the iron pipe.
As the above form of underground tube
is to-day in extended use, a description of
STREET MAINS. 293
its manufacture will not be out of place.
The iron pipes are made up in lengths of
twenty feet. The copper conductors, cut
off to the right length, are prepared for
FIG. 98.— MAIN TUBES.
placing in the tube by wrapping each
with a loose or open spiral of rope.
Three rods are then assembled and held
together, in the position shown in the
cross-section, by wrapping them with a
tight wrapping of rope. The rods so
assembled, are now placed in a length of
tube, and one end of the tube is closed
294 ELECTRIC INCANDESCENT LIGHTING.
with a plug. The tube is then filled
with hot bituminous insulating material.
The tube is finally closed by a second
block. There will thus be provided,
FIG. 99. — SECTION OP STREET TRENCH CONTAINING A
FEEDER AND MAIN TUBE.
lengths of pipe twenty feet long, con-
taining three insulated conductors with
their extremities projecting at the ends.
The underground tubes so formed are
£// 295
V5> h
Y- //
sent dBJbCftoni the factory inc me r ' twenty -
T^>! 4*' ff?v-r~M^ 'r
foot sectionfeteseril)ed. They are subse-
«/
quently connected to one another, while
in position in the underground trench
prepared to receive them. Fig. 99, shows
a section through a trench provided for
a feeder and a main tube. The trench
is usually 30" deep, as shown, and is
situated in the street a short distance
from the curb, the pipes being laid end to
end at the bottom of the trench.
It now remains to connect the separate
lengths of the underground tubes. For
this purpose coupling boxes are provided
as shown in Fig. 100. A, shows a coup-
ling box suitable for a street connection,
and B, a coupling box for a right-angled
connection. The coupling box is formed
of a cast iron shell made in halves,
connected together by bolts passing
296 ELECTRIC INCANDESCENT LIGHTING.
through flanges, and clamped over collars
secured at the ends of the tubes. The
FIG. 100. — COUPLING BOXES.
collars are first secured in place. The
lower half of the box is then fitted and
the three flexible copper stranded con-
STREET MAINS. 297
nectors are forced on the ends of the con-
ductors as shown. By the aid of a torch,
the joints are all heated to a sufficiently
high temperature, and solder is melted
into them, thus forming a good metallic
junction. The upper half of the box is
then fitted into place, the two parts
clamped together by the bolts, and melted
bituminous compound is poured into the
box through an aperture in the top. The
coupling box is then closed as shown at J
in Fig. 97.
Fig. 101, shows a form of branch coup-
ling box suitable for a house-service con-
nection with the mains. In this case, A.
and B, are main tubes passing along the
street, while C\ is the house service tube.
Here the coupling box is shaped to con-
form to the requirements of the triple
connection as shown.
298 ELECTRIC INCANDESCENT LIGHTING.
Fig. 102, shows cross-sections of feeder
tubes. In these tubes, the neutral con-
ductors are of smaller size than the two
outside conductors, since the system is
FIG. 101.— BRANCH COUPLING Box.
arranged to be nearly balanced as to load
with reference to the neutral. The three
small black wires shown, are called pressure
wires; i. e., small insulated copper con-
ductors which are returned from the feed-
ing point, or point of junction between
STREET MAINS. 299
feeder and mains, to the central station, so
as to indicate in the central station, the
pressure which is supplied to the mains.
Thus, if the drop in the feeders, be say 15
FIG. 102.— FEEDER TUBES.
volts, and the pressure at the bus-bars, in
the central station, be 130 volts on each
side of the system, or 260 volts across out-
side bus-bars, the pressure in the mains at
the feeding point will be 115 volts on each
side, and this will be the pressure carried
by the small pressure wires back to the
indicators at the central station.
300 ELECTRIC INCANDESCENT LIGHTING.
FIG. 103. — JUNCTION Box.
Fig. 103, shows the interior and Fig. 104,
the exterior of a junction box ; i. e., a box
situated at the junction of two streets, or
STREET MAINS. 301
at a feeding point where a feeder joins the
mains. This box is of cast iron, and is
buried with its surface flush with the
street level. The tubes enter the sides of
FIG. 104. — JUNCTION Box.
the box at the lower level. The conduct-
ors are connected by flexible connections
with brass pieces supported on insulating
rings. These pieces are marked + and -
according to the polarity of the conductors
connected with them.
302 ELECTRIC INCANDESCENT LIGHTING.
Three metallic rings, insulated from eacli
other, are provided in the box corresponding
to the three conductors in the tubes. All
the positive conductors are connected across
to the positive ring, through fuse strips ; all
the negative conductors are connected to
the negative ring ; and all the neutral con-
ductors, to the neutral ring. In this way all
the positive conductors are placed in elec-
trical connection with each other, and also
the negative and neutral conductors. The
box is made water-tight by lowering a
cover over the bolts seen in the ring in the
upper part, screwing down nuts upon
these bolts, and pouring in bituminous
compound in a melted state over the bolts.
J?, is a feeder tube containing the pressure
wires and five conductors, a pair of nega-
tives, a pair of positives, and a neutral.
MM, are main tubes ; p, is a slab or strip
of insulating material, supporting the con-
STREET MAINS.
nections for the pressure wires, which are
connected through fuses with the three
rings in the box. Fig. 104, shows the
appearance of the junction box when closed
with an ornamental cover.
CHAPTER XIV.
CENTRAL STATIONS.
IF we could follow the buried conduct-
ors through the streets, up-stream, that is,
toward the supply, we would finally reach
a building toward which all these wires
converge. This building constitutes what
is called a central station. In it we will
find the means for generating the current
which is sent through the street mains and
feeder wires, through the service conduct-
ors into the house, and finally through the
risers and house mains and branches, to
the lamps. Limiting our description of
the station to one in which continuous
electric currents only are generated; i. e.,
304
CENTRAL STATIONS. 305
currents which always flow in the same
direction, and which are suitable for use in
the three-wire system of buried conductors
just described, and supposing, as is gener-
ally the case, that steam power is employed,
we will find that the apparatus can be
readily grouped into three general classes ;
namely,
(1) The dynamos.
(2) The engines.
(a) The boilers.
Directing our attention in the station
first" to that part of the building at which
the feeders enter from the street, we will
find them connected to a device called a
switchboard. This consists essentially of a
fire-proof frame supporting a number of
metallic terminals provided for connection
with the feeders and also with connections
designed to receive conductors from the
306 ELECTRIC INCANDESCENT LIGHTING.
generators. A number of instruments are
mounted on this switchboard, consisting of
ammeters to show the strength of current,
and voltmeters to show the pressure on the
various feeders, while switches are pro-
vided for opening and closing the various
circuits.
Fig. 105, shows a partial view of a
switchboard in a central station. /SJ /S7, S,
are rows of massive switches mounted
upon fire-proof slabs of insulating material.
Above the switches are the voltmeters and
ammeters, while at F, are the field regulat-
ing boxes for controlling the pressure of
the generators. During the daytime, the
load on the mains is principally due to
motors arid is considerably less than the
night load. As evening approaches, the
load increases, as is shown on the ammeters
at the station. As soon as it becomes
PRCFCRTY OF
308 ELECTRIC INCANDESCENT LIGHTING.
necessary to introduce another pair of
dynamos, the engine driving them is
started, the pressure of the dynamos is
brought up to that required, and the
switches are then closed at the switch-
board, thus connecting the new generators
with the feeders. The reverse process is
adopted as the load diminishes, and it is
unnecessary to any longer maintain the
extra generators in the circuit.
Fig. 106, shows a type of generator unit
frequently met with in large central
stations, for low-pressure incandescent-light
distribution. This figure represents a por-
tion of the engine room in a large central
lighting station, and shows two generator
units at A and B, respectively. A, is a
vertical condensing, triple-expansion engine,
Ej E, E, whose main horizontal shaft
drives at each end the armature of a dy-
310 ELECTRIC INCANDESCENT LIGHTING.
namo or generator 6r, to be presently
described. This engine has two platforms
P, P. This engine runs at 120 revolu-
tions per minute, developing a maximum
of 800 HP; or, approximately, 600 KW,
the two generators being of 200 KW
capacity each. At B, is a smaller generat-
ing unit of similar construction, provided
with a single platform />, and. also driving
two smaller dynamo armatures, each of 100
KW capacity, the maximum out-put of the
engine being 300 KW, or about 400 HP,
at 172 revolutions per minute.
It is evident, under these circumstances,
that when the engine is developing its full
load, the dynamos will be overloaded about
forty per cent. Owing to the fact that
the full load on a central station is of short
duration, this has been found to be an
economical practice.
CENTKAL STATIONS. 311
In order to provide power to drive the
engines and dynamos a battery of boilers
is installed. Boilers are of various types.
Since an incandescent lamp requires an
activity of 50 watts, or about tpth of one
horse-power at its terminals, and since
losses occur in the feeders, mains and
house wires of, perhaps, ten per cent, on
an average, the activity per lamp at the
dynamo terminals, in a central station, will
be about 55.5 watts. Moreover, since an
average loss of say fourteen per cent,
occurs in the engine and dynamo, the
activity per lamp, generated by the engine,
must be about 64.5 watts, or about the
T—^th of a horse-power, so that on an aver-
1 l.D
age, one indicated horse-power at the
engine, represents 11.5 sixteen-candle power
incandescent lamps of 50 watts each. In
312 ELECTRIC INCANDESCENT LIGHTING.
other words, the average commercial effi-
ciency of such a system of distribution,
starting from the indicated horse-power
of the engine, is, approximately, 77.5 per
cent. About 17 pounds of steam, at 160
pounds pressure, are required per indicated
horse-power-hour with engines of the type
shown. This represents about 2.9 pounds
of coal per horse-power-hour delivered
electrically in consumers' lamps.
A large central station may readily sup-
ply 60,000 incandescent lamps or more.
Assuming that 60,000 is supplied at maxi-
mum load, the boiler power needed will be
correspondingly great. Take, for example,
the central station that was required to
supply the buildings and grounds of the
World's Columbian Exhibition at Chicago,
in 1893, with electric light. There were
employed for lighting this exhibition,
CENTRAL STATIONS. 313
about 100,000 incandescent lamps and
about 5,000 arc lamps. There was
naturally required for this purpose, as
well as for driving the machinery in the
buildings, a very great amount of power.
Fisf. 107, shows a view of the main
O '
boiler plant in the above exhibition. Here
batteries of boilers, representing an aggre-
gate of 24,000 horse-power are shown.
_/), Z>, are the main doors, d, d, the fire
doors, and beneath these latter the ash-pit
doors. 6r, is the steam gauge to show the
steam pressure in the boiler above that of
the atmosphere ; g, the water gauge ; M, the
steam drum, and P, the main steam pipe.
To the student of the incandescent
lamp, the most important features of the
central station are the dynamos or genera-
tors designed to supply the current to the
314 ELECTRIC INCANDESCENT LIGHTING.
lamps. Limiting our attention now to
continuous-current dynamos, we will find
these to be of a variety of types, although all
operate on essentially the same principle.
Broadly speaking a dynamo-electric ma-
chine or generator, is a device whereby
electromotive forces, and from these, elec-
tric currents are produced, generally by
the revolution of conductors through
magnetic flux. The magnetic flux is pro-
duced by field-magnet coils. E. M. Fs. are
set up in the coils on the revolving por-
tion, called the armature. These E. M.
Fs. are generated in the armature coils
in successively opposite directions, as they
pass each pole ; consequently, they need
to be commuted, or caused to assume the
same direction in the external circuit.
This is effected by a device called a com-
mutator.
316 ELECTRIC INCANDESCENT LIGHTING.
Thus Fig. 106, represents muJMpolar
generators, there being fourteen magnets or
poles Jt/J Mj My on the large dynamo, and
eight magnets or poles, on the small dynamo.
These magnets form part of the massive
stationary iron frames F F F,f f f, sup-
porting the outboard bearings O, o. In
the cylindrical bore of these field-magnet
poles, run the armatures, which are rigidly
secured to the main engine shaft. A light
metallic frame supports a number of pairs
of brushes B, B, upon the surface of the
commutator, there being as many pairs of
brushes as poles, so that A, has fourteen
pairs of brushes upon its commutator and B,
has eight pairs. These pairs of brushes, or
double brushes, as they might be called, are
insulated from their supporting frame, but
are connected in alternate pairs with the
main terminals T, T. Thus the first, third,
fifth, etc., brushes are connected to one
CENTRAL STATIONS. 317
terminal, and the second, fourth, sixth, etc.,
with the other. H, h, are handles for rock-
ing the entire brush-holder frame to-and-
fro within a small angular range around
the axis, so as to cany all the brushes for-
ward or backward upon the surface of the
commutator. This adjustment is made
to prevent sparking at the brushes from
the current which they carry at different
loads. The main terminals T, T, are con-
nected to switches on the main switch-
board of the station.
Fig. 108, shows in greater detail another
form of central station generator of the
same type. F, F, F, is the field frame.
M, M, My are the field fcoils, six in num-
ber, consisting of coils of insulated wire
wound upon soft steel cores bolted radially
to the field frame, as shown. A., is the
armature whose surface runs within the
318 ELECTRIC INCANDESCENT LIGHTING.
cylindrical polar space formed by the poles
of the field magnets. (7, C, is a commu-
FIG. 108. — SIX-POLE GENERATOR.
tator, consisting of copper strips, rigidly se-
cured in a cylindrical frame, and insulated
from one anotl;fc«:^^
conducting loops wound upon the arma-
ture. 13, It, are the brushes, of which there
are six pairs, connected to two metallic
rings, one ring being in connection with
three alternate pairs. These rings are
finally connected to the main terminals T,
Tj by cables, as shown. The current sup-
plied by the armature is collected by each
brush, leaving the armature at the three
positive brushes, and, after traversing the
external circuit of the feeders, mains,
house wires and lamps, returning to the
armature through the three negative
brushes, thereby completing the electric
circuit. H, is a handle for advancing or
retreating the brushes over the surface of
the commutator. The machine shown has
50 KW capacity, that is to say, it will
deliver at its terminals activity to the
amount of 50 KW, (400 amperes at a
320 ELECTRIC INCANDESCENT LIGHTING.
pressure of 125 volts, or 400 X 125 =
50,000 watts). It requires about 75
horse-power or about 56 KW to drive it
at full load. Its gross weight is 6,500
pounds, representing an output of, approxi-
mately, 7.7 watts per pound of total
weight. It has a shaft five inches in di-
ameter, and occupies a space of 30" X 61"
X 51" in height.
The armature of the preceding machine
is illustrated in greater detail in Fig. 109.
It consists of a metal cylinder, which car-
ries on its outer surface a laminated iron
core, built up of annular sheet-iron discs,
clamped side by side, so as to form, when
assembled, an almost complete cylindrical
external surface of iron, A A. Gaps are
left, at suitable intervals between adjacent
sets of discs, to provide for the ventila-
tion of the armature, by means of the pas-
CENTRAL STATIONS.
321
sage of air by centrifugal force from within
outward, to aid in the cooling of the ar-
FIG. 109. — ARMATURE OF GENERATOR.
mature, when operated. The outer sur-
faces of the iron discs are provided with
322 ELECTRIC INCANDESCENT LIGHTING.
longitudinal slots, parallel to the axis of
rotation. These slots are designed to re-
ceive the armature conductors. One hun-
dred and sixty-eight of these slots are thus
provided to receive one hundred and sixty-
eight sets of conductors.
The winding adopted in the armature is
conducted as follows : At A, a pair of
rods or wires comes through a slot across
the armature surface. A pair of flexible
copper strips, insulated from their neigh-
bors, are soldered to the rods at #, and run
down to b. These connect two adjacent
commutator bars at b. From this they run
back behind the connections W, IF, to the
slot at c, and then across the armature sur-
face underneath the two rods or wires
which are seen to approach at c. Having
reached the opposite side of the armature
at c'y they descend obliquely on the other
CENTRAL STATIONS. 323
face to a point opposite d, and then ascend
to the armature surface at a point c. Here
the rods connected therewith return across
the armature surface to the point 0, de-
scending again to the commutator at the
point f. In this manner a pair of con-
ductors zig-zag across the armature, via
g i ~k, emerging again one slot in advance
of a, and so on. In this manner it must
always happen that the wires in the slot
a, pass under one pole, the wires c e g i Jc
will also be passing under the other poles.
The effect of the commutator is to enable
the brushes to collect 'all the currents
which are being generated in the various
wires passing under the different poles and
to unite them in the external circuit.
CHAPTER XV.
ISOLATED PLANTS.
FOR the general purposes of house light-
ing, where comparatively few lamps are re-
quired, it is more economical for the house-
holder to rent his electric power from a
central station, than it would be for him to
erect and maintain his own plant. Since
a plant requires boilers and engines, it
would follow, unless a fairly considerable
number of lamps is required, that the
extra expense of the installation as well as
the services of an engineer, or engineer
and fireman, would make it much cheaper
in such cases to take the service from the
street mains as supplied by the nearest
ISOLATED PLANTS. 325
central station. There are many cases,
however, in which either the number of
lights, or the circumstances are such, as to
warrant, in point of economy, the main-
tenance of what may be called an isolated
lighting plant, in contradistinction to a
central-station lighting plant.
It is clear that when the number of
lights required reaches a certain limit, it
may be preferable to maintain an isolated
plant, rather than to rent the light, since
a large building or plant would thus
possess the advantage of being indepen-
dent of the running of the central station,
or of accidents which might occur to the
street mains. Moreover, a large isolated
plant, being necessarily circumscribed in
the area of its distribution, would require
a much smaller outlay or expenditure in
copper, and, consequently, may be built to
326 ELECTRIC INCANDESCENT LIGHTING.
produce light cheaper thaii a central
station.
But even in cases where the number of
lights is not very great, circumstances may
arise where it would still be economical to
establish an isolated plant. Such cases
would be found, for example, in manufac-
turing establishments, where boilers and
steam engines have necessarily to be main-
tained in action during the time that light
is required. Again, isolated plants arc
necessary in sections of country remote
from central stations. A brief description
of isolated plants for the supply of in-
candescent lighting will, therefore, be of
interest.
Any suitable form of dynamo can be
used for an isolated plant. The dynamo
may be driven either directly from the
engine shaft or by means of leltimj. The
ISOLATED PLANTS.
327
direct connection requires less floor space,
is somewhat more efficient, and saves wear
FIG. 110. — BELT-DRIVEN GENERATOR FOR ISOLATED
PLANTS.
330 ELECTRIC INCANDESCENT LIGHTING.
of belting, but possesses the disadvantage
of requiring the engine to run at a com-
paratively high speed, and the dynamo at
a comparatively low speed. This means
that for all sizes of generator below 50 KW
at least, the cost of a direct-driven plant
is greater than that of a belt-driven plant,
because a slow-speed dynamo means a
heavier and, consequently, a more expen-
sive dynamo, and a high-speed engine is
more difficult to maintain in running
order.
Fig. 110, represents a bipolar, or two-
pole, generator, driven by a belt from a
small vertical engine with its governor
inside the fly wheel.
Fig. Ill, represents a quadripolar or
four-pole direct-driven generator, suitable
for isolated plants. In this case the en-
332 ELECTRIC INCANDESCENT LIGHTING.
gine is completely enclosed in a cast-iron
shell and runs in oil.
Fig. 112, represents a type of isolated
lighting plant employed on board ships and
supplied to several vessels in the United
States Navy. This quadripolar generator
delivers 200 amperes at 80 volts pressure,
or 16 KW, at a speed of 400 revolutions per
minute. It is directly connected as shown
to the vertical marine engine.
Fig. 113, represents an isolated three-
wire plant, consisting of a 100-horse-power
horizontal steam engine, driving two 32
KW quadripolar generators at a speed of
270 revolutions per minute. The magnet
poles in this case are inside the armature
and the outer surface of the armature has
its insulated conductors bare, so as to form
a commutator, upon which the sets of
ISOLATED PLANTS.
333
brushes shown in the figure can rest.
The switchboard for controlling the vari-
ous circuits is represented behind the
machine on the right hand side of the
figure. The apparatus required for such a
switchboard is similar in kind to that of
a central station, but usually is smaller and
less complex.
CHAPTER XVI.
METEES.
THE sale of any product requires some
suitable unit of measure. Since electric
power is undoubtedly a product requiring,
as it does, the establishment of an expensive
plant for its production, a unit of measure
is necessary, as well as an apparatus
w^hereby the number of units delivered to
the customer by the producer may be
measured.
Various plans have been devised for the
measurement of electric supply. Of these,
however, only two forms are in general
use for the measurement of continuous-cur-
334
METERS. 335
rent supply. One of these forms meas-
ures the quantity of electricity delivered
to the consumer in units of supply called
ampere-hours, while the other measures the
energy in watt-hours. Each of these forms
of apparatus is called a meter.
Fig. 114, shows a form of electrolytic
meter indicating the supply in ampere-
hours. This apparatus depends for its
operation on the fact that an electric cur-
rent, when sent through a solution of zinc
sulphate, will effect a decomposition of the
solution, depositing metallic zinc on a zinc
plate connected with the negative termi-
nal, and dissolving or removing an equal
quantity of zinc, from a plate of zinc con-
nected with the positive terminal. The
indications of this meter are obtained by
carefully weighing the plates, before and
after the supply they are to measure has
336 ELECTRIC INCANDESCENT LIGHTING.
passed through them. After the meter
has been in use for some time, it will be
found that the zinc plate connected to the
positive terminal has decreased in weight,
FIG. 114. — ELECTROLYTIC METER.
and that connected with the negative
terminal has increased in weight. This
difference of weight is a measure of the
number of ampere-hours that have passed
through the cell.
METEES. 337
In the interior of the meter box, in the
upper part, are two large strips of ger-
man silver 7t, J?, carrying the current to
be supplied to the lamps, and offering a
definite small resistance to its passage.
The positive and negative supply wires
enter the box at P and N, and are se-
cured to the terminals P and N, inside.
As the meter shown is a three-wire meter,
it consists of two meters, in one box, one
being for the supply on the positive main,
and the other for the supply on the nega-
tive main. If the number of lamps
lighted in a house, on each side of the
system, were always the same, the records
of these two meters would be equal.
When the current passes through the re-
sistance J5 .7?, it establishes a certain drop
of pressure, as already explained in Chap-
ter III. This drop amounts to about 0.4
volt at full load. In a pair of derived or
338 ELECTRIC INCANDESCENT LIGHTING.
shunt circuits, connected across the ex-
tremities, of each strip of resistance R, is
placed a pair of small glass bottles with
resistances wound on a spool s, behind
them. The total resistance of the bottle,
or plating bath, and the spool in its circuit,
bears such a relation to the resistance of
the shunt R, that each milligramme of zinc
electroplated on the surface of the nega-
tive zinc plate 2, represents a definite
number of ampere-hours of current sup-
plied through the meter according to its
size. The plates 2, 2, are made of zinc and
mercury alloy, and are separated from
each other by hard rubber washers.
Copper rods extend upwards from these
plates through the corks <?, <?, to the
copper clips p, p. There are two bottles
for each side of the meter, so that a dupli-
cate record is kept on each side of the
supply. It is usual to renew the bottles
METERS. 339
once a month. The bottles after being re-
moved are emptied, the zinc plates washed
and dried, and weighed in a chemical bal-
ance, and the amount of the supply deter-
mined from the difference in weight dur-
ing the month. The solution employed is
of pure zinc sulphate in water, having a
density of 1.11 at 60° F. The advantage
of this meter is its simplicity, and the fact
that all its essential working parts are re-
moved and replaced once a month. The
disadvantage of the meter, is that it does
not show directly to the consumer, the
amount of supply which has been de-
livered.
When such meters are placed in ex-
posed situations, where the solutions might
be liable to freeze, a thermostat is em-
ployed to maintain the temperature of the
bottles above the freezing point. Such a
340 ELECTRIC INCANDESCENT LIGHTING.
thermostat is shown in Fig. 115. Here a
metallic strip, composed of two unequally
expansible metals riveted together, is so
arranged, that, when exposed to a suffi-
p
FIG. 115. — THERMOSTAT FOR ELECTROLYTIC METERS.
ciently low temperature, the unequal con-
traction of the metals will cause a bending
or warping, which will close the circuit of
a lamp through the set screw s. The
lamp will then burn until the temperature
rises sufficiently to allow the strip p p, to
straighten and break the circuit.
METEKS.
341
Another form of meter in general use is
shown in Fig. 116. In this form of in-
strument, the number of watt-hours deliv-
ered to the consumer is registered. It
FIG. 116.— INTERIOR OF RECORDING WATTMETER.
342 ELECTEIC INCANDESCENT LIGHTING.
consists of a small motor, the field magnet
coils M, M, of which are in the direct cir-
cuit of supply. The armature A, placed
within the field coils, revolves upon the
vertical shaft S9 8, with a speed propor-
tional to the activity delivered ; i. e., to the
product of the volts and amperes. For
example, if the pressure at the house
mains be 110 volts, and the current be 2
amperes, then the activity delivered would
be 110 X 2 = 220 watts; and, in one hour,
this rate of delivery would result in a
supply of 220 watt-hours, while the rotary
speed of the armature shaft would be pro-
portional to this value 220. A small com-
mutator is seen just above the field coils
with its long slender brushes running
back to supports behind the apparatus.
The armature is placed in a circuit of high
resistance across the mains, so that it ab-
sorbs a constant small amount of activity,
METERS.
343
FIG. 117. — RECORDING WATTMETER.
whether the meter be running or not.
The shaft engages by means of an endless
screw with a pinion wheel, forming part
344 ELECTRIC INCANDESCENT LIGHTING.
of a train work of dial-recording median-
o
ism, similar, to that of a gas meter.
The above form of the apparatus, when
completely enclosed, is seen in Fig. 117.
The cover is secured to the base by a wire
sealed with the leaden seal s. The supply
and output wires pass through the meter
beneath. The advantage of this meter is
that it enables the customer to observe
the amount of power he consumes. Its
disadvantage is that it constantly absorbs
a small amount of power. A meter should
always be installed in a dry place, and
inserted in the service wires between the
street main cut-out and the risers.
CHAPTER XVII.
STORAGE BATTERIES.
IF the supply of electric current re-
quired for incandescent lamps was uni-
form throughout the twenty-four hours, it
would be easy to determine the most eco-
nomical generator units of boiler, engine
and dynamo, in order to meet this require-
ment economically. Unfortunately, how-
ever, in nearly all cases the variations in the
load are very great. A few hours of the
twenty -four require a load greatly in excess
of the average. If, in order to meet this
load economically, the generating plant be
subdivided into a number of small units,
so as to permit them to be readily with-
drawn and added as required, both the
345
346 ELECTRIC INCANDESCENT LIGHTING.
expense arid the complexity of the gener-
ating system would be necessarily in-
creased. If, on the other hand, a large
generating unit be installed, it would
have to be operated for the greater part of
the twenty-four hours at a very small
load, and, therefore, uneconomically.
The above difficulty is sometimes met
by the employment of storage batteries,
which are charged during the hours
of light load, and discharged during the
hours of heavy load, thereby equalizing
the load on the station. Since, at the
time of full load, the output is obtained
both from dynamos and storage batteries,
it is evident that a much smaller generat-
ing plant of boilers, engines and dynamos
is rendered necessary.
In order to determine whether it would
STORAGE BATTERIES.
347
be economical to install a storage battery,
it is necessary to ascertain the load diagram
of the station ; that is, the curve which rep-
resents the output required to supply the
FIG. 118.— LOAD DIAGRAM.
lamps during the twenty-four hours of the
day. Such a load diagram is shown in Fig.
118. In this figure the hours of the day are
marked off horizontally, and the current
348 ELECTRIC INCANDESCENT LIGHTING.
strength delivered to the feeders is
marked off vertically in amperes. An in-
spection of the figure will show that the
peak of the load, that is, the maximum load
of the curve, occurred at 5.30 p. M. when it
exceeds 1,700 amperes, while an hour and
a half earlier, or at 4 p. M., it was 800
amperes, and an hour and a half later, or at
7 P. M., it was 930 amperes. At 8.30 p. M.
the load has increased, perhaps, owing to
the lighting of some theatre, after which
the load steadily falls until 2.30 A. M. The
average load during the twenty -four hours
from this diagram is 620 amperes, and since
the maximum load is 1,710, the ratio of the
average to the maximum is 0.363 or 36.3
per cent. This is called the load factor.
If the load represented in Fig. 118, were
supplied without the aid of a storage bat-
tery, it would be necessary to install boil-
STORAGE BATTERIES. 349
ers, dynamos and engines to the extent
necessary to supply a current of 1,700
amperes, although the average load is only
620. By the use of a storage battery,
capable of supplying 800 amperes for,
say four hours ; or, a battery having
a storage capacity of 4 x 800 = 3,200
ampere-hours, then the maximum load
which the boilers, engines and dynamos
would have to supply would be 900
amperes \ or, only about half as much
as in the preceding case, while the load
during the daytime would be increased
by the amount necessary to charge
the storage batteries. Good practice re-
quires that the charging be done during the
time of the day when the load is the least ;
or, in this case, between the hours of 1 and 7
A. M. The smaller the load factor the greater
the probability of obtaining economy in
the installation of a storage battery.
350 ELECTRIC INCANDESCENT LIGHTING.
Another case arises in which an advan-
tage is derived from the use of a storage
battery ; namely, where the engines and
boilers employed to drive the dynamo are
used during the hours of darkness only,
and are stopped during the day, while it is
desired to maintain a few lamps during
the daytime. Under these circumstances
it is often more economical to establish a
storage battery plant, which can be charged
during the time of running at night, thus
permitting the engine and dynamo to be
stopped during the daytime and the stor-
age battery to supply the few lights
required.
Before proceeding to the general descrip-
tion of a storage battery installation, it
may be well to describe briefly the princi-
ples of its operation. A storage cell differs
in no respect from an ordinary voltaic cell ;
»T *$
PHCFLRTY CF
351
for, like a^iftaieHgytj^fiafi^^ essentially
of two plates or elements, called respectively
the positive and the negative plate, plunged
in an acid liquid or electrolyte capable of
acting on one of the plates. As the result
of the chemical action that occurs under
these circumstances, an electric current is
produced which passes in a definite direc-
tion through the electrolyte and issues from
the cell at one of its terminals or poles
and returns to it, after having passed
through the circuit in which the cell is con-
nected, by the other terminal or pole. In
both the ordinary voltaic and the storage
cell, exhaustion takes place when a certain
output of electricity is yielded. In the
case of the voltaic cell both the liquid, and
at least one of the plates, must be renewed,
while in the case of the storage cell, all
that is necessary for renewal is to connect
the terminals of the cell with an independ-
352 ELECTRIC INCANDESCENT LIGHTING.
ent electric source, and send a current
through it in the opposite direction to that
of the current it yields. This current is
called the charging current, and the cell re-
ceiving it is said to be charged. Since a
storage cell thus derives its energy second-
arily from some other electric source, it is
sometimes called a secondary cell, in contra-
distinction to a primary or voltaic cell.
A great number of storage cells have
been devised. Practically all that have
been placed in commercial use, consist of
perforated lead plates or grids, correspond-
ing to the positive and negative plates of
the voltaic cell, with the perforations filled,
respectively, with peroxide of lead and
finely divided metallic lead. These plates
are associated, or placed side by side, in a
solution of sulphuric acid and water. In
the original form given to these cells, they
STORAGE BATTERIES. 353
consisted of a very large positive and nega-
tive plate, suitably supported at a short
distance from each other, and then rolled
up together in a close spiral. It was
found, however, in practice, that when
such a plate became damaged in one part,
the entire cell had to be rejected, so that
for the purpose of convenience, as well as
for the ready inspection of the different
parts of the plate, they are now generally
made in a number of smaller plane plates,
placed parallel to one another, which,
when connected in parallel, are equivalent
to two large plates of the same total sur-
face.
The simplest form of such a cell would
consist of a single positive and a single
negative plate, placed side by side ; but
since the positive plate would only have a
negative plate on one side of it, the other
354 ELECTRIC INCANDESCENT LIGHTING.
side being uncovered, it is preferable to
associate two negative plates with one posi-
tive plate, so as to utilize the entire surface
of the positive plate. Such a cell is shown
in Fig. 119, where two negative plates, JV",
are placed one on each side of a single posi-
tive plate P, inside a jar e7W, filled to
a convenient height with sulphuric acid
and water. The positive plate is shown
separately at A. It consists of a grid or
frame of antimonious lead ; i. e., lead
alloyed with a small amount of antimony,
so as to keep it from being acted upon by
the acid liquid. In this frame are shown
eight small circular buttons a, filling circu-
lar holes in the grid. These, when the cell
is charged, consist of peroxide of lead, and
constitute the active material of the cell.
The negative grids have apertures filled
with square buttons, which when in the
charged condition are filled with porous
STORAGE BATTERIES.
355
FIG. 119. — SMALL STORAGE CELL.
356 ELECTRIC INCANDESCENT LIGHTING.
lead. In the cell shown, the plates are
three inches square, and the weight of the
entire cell, filled with solution, is four
pounds. The capacity of this cell is about
6 ampere-hours when discharged at normal
rates.
In a fully charged storage cell, the ma-
terials filling the apertures in the grids are
dissimilar ; namely, peroxide of lead and
metallic lead. During discharge chemical
actions occur, which result in reducing
these substances to the same substance ;
namely, monoxide of lead. When now
the charging current is sent through the
exhausted cell, a dissimilarity is again
produced, the monoxide being converted to
peroxide of lead on the positive plate, and
into porous lead on the negative plate. It
is evident, therefore, that it is not electric-
ity which is stored, but energy in the form
STORAGE BATTERIES. 357
of chemical energy, and in a form capable,
on discharge, of being released under suit-
able conditions as electric energy.
Fig. 120, shows a larger size of the same
type of storage cell. Here five positive
plates are associated alternately with six
negative plates, so that both external plates
are negative ; these plates are provided
with rectangular apertures. The plates
are bound together by an insulating frame
F, F, F, and are individually separated or
maintained at the proper distance apart- by
sheets of asbestos. In the cell shown, the
plates are 10.5" square, and the entire cell,
with solution, weighs 170 pounds. Its
normal capacity is 500 ampere-hours, or,
approximately, 3 ampere-hours per pound
of total weight. All the positive plates
are connected together to form a single
positive plate with terminal P, and all the
358 ELECTKIC INCANDESCENT LIGHTING.
FIG. 120. — STORAGE CELL.
STORAGE BATTERIES.
359
negatives are similarly connected to form
a single negative terminal N.
When storage cells are employed in large
central stations for the supply of powerful
FIG. 121. — LARGE STORAGE CELL.
currents, it is customary to associate large
plates in single cells, rather than to employ
a number of smaller cells in parallel. Fig.
360 ELECTRIC INCANDESCENT LIGHTING.
121, represents a form of large cell for such
an apparatus. Instead of employ ing a glass
jar, the containing vessel is of wood, with
an interior leaden lining. Here 21 nega-
tive plates are associated with 20 positive
plates, each plate is 15.5" square, and the
total weight of the cell, when filled with
solution, is 800 pounds. The normal
capacity of this cell is 5,000 ampere-hours,
or about 6.4 ampere-hours per pound of
total weight.
It is evident that the storage capacity
increases with the weight of the cells ;
being over 6 ampere-hours per pound,
with larsre cells, and with the smallest cells
O 7
about 1 1/2 ampere-hours per pound of
total weight. The E. M. F. produced by
the average storage cell is about 2 volts.
When fully charged, while discharging, it
is about 2.2 volts, and falls during dis-
charge
energy wcia^ ex-
pressed in watt-hours is, therefore, approx-
imately, 2 multiplied by the capacity of
the cell in ampere-hours. Thus the cell
last mentioned has an energy storage
capacity of 2 X 5,000 = 10,000 watt-
hours, or 10 KW-hours, under normal
circumstances.
Since incandescent electric lamps gener-
ally require a pressure of about 115 volts,
and since a single storage cell has only 2
volts E. M. R, it is necessary to couple at
least 57 cells in series, and, in order to al-
low for drop of pressure in the feeders,
mains and house wires, the usual number
required for this purpose is from 60 to 65
cells. For three-wire installations, twice
this number is necessary, or about 120
cells. Such a three-wire storage battery
STORAGE BATTERIES. 363
installation is represented in Fig. 122.
The cells are arranged in series, the posi-
tive terminal of one cell being connected to
the negative terminal of the next. Each
cell has 11 plates, 5 positive and 6 nega-
tive, and the normal capacity of the cell is
1,000 ampere-hours, giving a 10-hour dis-
charge at the rate of 100 amperes.
A storage battery plant requires for its
convenient operation a switchboard with its
accessories, such, for example, as is shown
in Fig. 123. S, &, are two double-pole
switches, one for controlling the charging,
and the other the discharging circuit. The
charging here being performed, as is usual
in such cases, by a dynamo whose pressure
is controlled by the field rheostat operated
by a handle at F. A, A, are ammeters in
the charging and discharging circuits ; V, is
a voltmeter which by means of ihepressure
364 ELECTRIC INCANDESCENT LIGHTING.
B
e>
#
FIG. 123. — STORAGE BATTERY SWITCHBOARD.
gwitch P, can be connected either to the
dynamo terminals, or to the battery termi-
nals, to show the pressure before the switch
•is thrown. T, is an automatic cut-out which
STORAGE BATTERIES. 365
disconnects the charging circuit of the
dynamo, as soon as the C. E. M. F. of
the cells exceeds the E. M. F. of the dynamo,
and 0, is an overload switch in the dis-
charging circuit, arranged automatically to
break the circuit of the battery should the
discharge become excessive. It, is a switch
which may either be used to throw in and
out reserve cells, in order to maintain the
discharging pressure constant, or to throw
in and oqt C. E. M. F. cells, that are
employed in place of the rheostat in the
lamp circuit during charge.
The overload switch is shown in greater
detail in Fig. 124. Here a coil of heavy
wire (7, — pivoted upon its axis, has its
ends dipping in mercury cups. The coil
is placed in the main circuit of discharge
in such a manner that when the discharg-
ing current becomes excessive, the electro-
366 ELECTRIC INCANDESCENT LIGHTING.
magnetic force developed by the coil rotates
tlie helix and lifts the contact points P, P,
FIG. 124.— OVERLOAD SWITCH.
out of the mercury cups. H, is the handle
for restoring the circuit when desired.
Fig. 125, shows & form of overload switch
suited for heavier currents. The spring
jaws, J, J, are in contact with the plates
7J T, when the switch is closed, and the
discharging circuit established. S, S9 are
STORAGE BATTERIES.
FIG. 125.— OVERLOAD SWITCH, CLOSED.
spiral springs which tend to disengage the
plates and open the switch, but whose ac-
tion is prevented by the armature J., which
368 ELECTRIC INCANDESCENT LIGHTING.
normally holds the switch plates in posi-
tion. The coil C, of heavy conductor, is
placed in the discharging circuit, and, as
soon as the current strength exceeds a cer-
tain safe limit, it attracts the armature A,
to the iron pofe piece P, thus allowing the
springs &, S to act, and the plates T, T, to
be thrown out of the jaws J, J, as shown
in Fig. 126.
In order to maintain a storage battery
in proper working order, it is necessary to
test the individual cells from time to time,
for E. M. F. This is readily effected in
practice, by means of any suitable voltmeter
connected with electrodes for application
to each separate cell. A convenient form
of apparatus for this purpose is shown in
Fig. 127. A simpler form of testing ap-
paratus, not so accurate as the preceding,
is shown in Fig. 128. It consists of a pair
STORAGE BATTERIES.
FIG. 126.— OVERLOAD SWITCH,
RELEASED.
370 ELECTRIC INCANDESCENT LIGHTING.
of handles with sharp metallic points, to
make connection with the leaden electrodes
of the cells, and connected by a loop of
FIG. 127. — STORAGE CELL VOLTMETER AND ELECTRODES.
wire. In the end of one of the handles is
a lamp socket in which a low volt lamp
is inserted. If the cell tested is in good
STORAGE BATTERIES. 371
FIG. 128.— SIMPLE FORM OF STORAGE CELL TESTER.
372 ELECTRIC INCANDESCENT LIGHTING.
order its E. M. F. will be sufficient to bring
the lamp up to candle-power.
The efficiency of a storage cell or battery
is the ratio of the output to the intake.
An ideally perfect battery would lose no
energy, and would, therefore, have the
same intake and output, representing an
efficiency of unity, or one hundred per cent.
In practice, the ampere-hour efficiency of a
storage cell may be over ninety-five per
cent, so far as regards electric quantity
or ampere-hours, that is to say, if 100
ampere-hours be supplied to a storage
battery during charge, it may yield more
than 95 ampere-hours during discharge.
On the other hand, since the pressure at
the terminals of a cell during charge is
over 2 volts, rising toward full charge to
2 1/2 volts, while the pressure during dis-
charge is from 2.2 to 1.8 volts, the energy
STORAGE BATTERIES. 373
efficiency, as reckoned from the watt-hour
output, is considerably less, and usually
varies from seventy-five to eighty -five per
cent, according to the conditions of the
charge and discharge. A battery dis-
charged at a rapid rate will not have so
high an efficiency, either in quantity or in
energy, as if slowly discharged. In relation,
therefore, to coal consumed, the efficiency of
a storage battery is about eighty per cent.,
but in relation to ampere-hours the effi-
ciency may be over ninety-five per cent.
CHAPTER XVIII.
SERIES INCANDESCENT LIGHTING.
WE have have hitherto described mul-
tiple-connected lamps ; or, in the case of
the three- wire system, lamps in series-mul-
tiple. It sometimes happens, that a dis-
trict to be lighted is scattered and wide
spread, so that scattered lamps have to
be supplied at considerable distances from
the central station. As we have already
shown, under these circumstances a very
great amount of copper would require to
be employed in their multiple-distribution.
In order to avoid this, a series distribu-
tion is sometimes adopted. Here, as we
have already described, a number of sep-
arate lamps are connected in series in one
374
SEKIES INCANDESCENT LIGHTING. 375
circuit. Since in such a system of distri-
bution the extinguishment of a single lamp
would open the entire circuit, a simple au-
tomatic safety device is required, which
will establish a short circuit about the
faulty lamp, in case it breaks, thus pre-
serving the continuity of the circuit.
A system of series-distributed incan-
descent lamps may be operated from a
constant high-potential dynamo. Thus, a
dynamo of 1,000 volts E. M. F. may operate
a circuit of 30 incandescent lamps, each
having a pressure of 33 1/3 volts, includ-
ing drop in connecting wires. Several
such 1,000- volt circuits may be arranged in
parallel, thus producing a multiple-series
system of distribution, as shown in Fig.
129.
Such a series-connected system is some-
376 ELECTKIC INCANDESCENT LIGHTING.
times used for incandescent lamps in street
lighting. It is not suitable for house
lighting, since it is essential that the num-
ber of lamps in each series circuit should
a b c d
129. — MULTIPLE-SERIES SYSTEM.
be as nearly constant as possible, and this
precludes the possibility of cutting lamps
out of a circuit when no longer required.
Various forms of automatic circuit-pro-
tecting devices have been produced. One
of the simplest of these, called the film
cut-out, consists of a thin strip of paper
placed between two spring contacts con-
SERIES INCANDESCENT LIGHTING. 377
nected with two terminals of each lamp.
While the lamp remains lighted, the elec-
tric pressure across the thickness of the
strip of paper is only a few volts ; i. e., the
pressure at the lamp terminals, but if the
filament should break, the pressure at the
terminals is the full pressure of the circuit
which may be 1,000 volts. This pressure
is capable of piercing the thin paper sheet,
thus establishing an arc, which instantly
welds the two metal strips together, thus
short-circuiting the lamp and re-establish-
ing the circuit.
As already stated, series incandescent
lamps are generally employed for out-door
lighting and, therefore, have to be pro-
tected from the weather. The compara-
tively high electric pressure employed on
these circuits, not being safe to handle
under all conditions, requires certain pre-
378 ELECTRIC INCANDESCENT LIGHTING.
cautions in insulation when introduced
into buildings. Fig. 130, shows a form of
lamp and fixtures suitable for out-door
series-incandescent lighting. The lamp Z,
FIG. 130. — STREET FIXTURE, WITH SERIES INCANDESCENT
LAMP.
is placed in its socket beneath tne porce-
lain deflector D, which not only serves to
scatter and reflect the light, but also, in
conjunction with the cover (7, to protect the
SERIES INCANDESCENT LIGHTING. 379
lamp socket from rain. The circuit wires
Cj c, are brought to the socket of the lamp
after being secured to the insulators I, I.
Fig. 131, shows the same fixture with the
FIG. 131. — STREET FIXTURE, WITH LAMP REMOVED AND
SOCKET EXPOSED.
deflector D, removed, showing the socket
>9, placed in the interior. Fig. 132 shows
a form of lamp post suitable for street
lighting with such lamps. Another form
380 ELECTKIC INCANDESCENT LIGHTING.
FIG. 132.— LAMP POST.
SERIES INCANDESCENT LIGHTING. 381
of fixtures is shown in Fig. 133; here the
incandescent lamp is provided with an
external globe.
Where a series circuit has already been
installed for operating arc lamps, it is
FIG. 133. — STREET LAMP FIXTURE.
sometimes desirable to insert incandescent
lamps in the same circuit. This is done
by employing series incandescent lamps of
special manufacture. Since the current
strength, in a series-arc circuit, is usually
382 ELECTRIC INCANDESCENT LIGHTING.
about 10 amperes, and is the same in all
parts of the circuit, it is necessary that the
incandescent lamps be made for this cur-
rent strength. A fifty-watt 16-candle
power lamp, taking 10 amperes, must be a
5-volt lamp, since 5 volts X 10 amperes =
50 watts. Consequently, the resistance
of the lamp must be only 1/2 ohm, since 10
amperes X 1 /2 ohm = 5 volts, and the
filament must be comparatively short and
thick in order to possess this relatively
low resistance. It is evident that lamps
of different candle power can be obtained
from the use of such series circuits, by
suitably adjusting the resistance of their
filaments. The connections for such a
series arc and incandescent circuit are
shown in Fig. 134.
Since series-arc circuits frequently em-
ploy dangerously high pressures, it be-
SERIES INCANDESCENT LIGHTING. 383
comes unsafe to come in contact with the
conductors of such circuits when standing
on wet ground ; for, such circuits com-
FIG. 134. — SERIES ARC AND INCANDESCENT CIRCUIT.
moiily have marked leakage and thus a
dangerously high current might be sent
through the body. Consequently, it is
384 ELECTRIC INCANDESCENT LIGHTING.
necessary carefully to insulate the keys
and connections of incandescent lamps
operated on arc circuits.
FIG. 135. — LAMP FOR SERIES-ARC CIRCUIT, WITH
CUT-OUT SWITCH.
SERIES INCANDESCENT LIGHTING. 885
Fig. 135, shows a form of lamp suit-
able for series-arc circuits. Like all
series-connected lamps, it contains an au-
tomatic cut-out, and of the film type al-
ready described. It has no key, but is
provided with a ceiling switch whereby
it may be short-circuited and thus extin-
guished.
CHAPTER XIX.
ALTERNATING-CURRENT CIRCUIT INCAN-
DESCENT LIGHTING.
SINCE incandescent lamps are frequently
operated on alternating-current circuits, it
may be well briefly to discuss some of tlie
characteristics of these currents and in-
stances of their commercial application.
In a continuous current the direction of
electric flow does not change. Its strength
may vary periodically, in which case the
current is said to pulsate; or, it may be
unvarying, in which case the current is
said to be steady. An alternating current,
on the contrary, changes direction many
NATING-CUKKENT OIROU
times in a^s^jtrftfJ, f%ep$ being first' a wave
or flow of curreStHlrro^^r'tEe circuit in
one direction, and then a wave or flow
of current in the opposite direction, and
so on. Each of these waves or flows is
called an alternation, and a complete to-
and-fro motion, or double wave, is called
a cycle. The frequency of alternation is
the number of alternations or of cycles
executed in a second ; thus ordinary com-
mercial alternating circuits have a fre-
quency of from 25 cycles, or 50 alternations
per second, to 140 cycles, or 280 alterna-
tions per second. The number of cycles,
or complete periods, is sometimes symbol-
ized by the sign ~, so that a frequency of
140 complete periods, or cycles, per second
would be written 140 ~.
Between each pair of successive waves,
at the time when the current is chang-
388 ELECTKIC INCANDESCENT LIGHTING.
ing direction, there is no current flowing
in the circuit. Consequently, it might be
supposed, when an alternating current,
of say 120 ~, is sent through a lamp, since
there would be 240 moments in each
second when no current is flowing, that
the light furnished by the lamp would
pulsate, being extinguished and relighted
240 times in a second. In point of fact,
this tendency to pulsate does exist; but
when the frequency is sufficiently high,
before the filament loses enough of its heat
to cease glowing, it receives a fresh acces-
sion of heat from the succeeding wave
of current. Moreover, the retina of the
eye tends to retain its luminous impres-
sions for a sufficiently great fraction of a
second to aid in the apparent uniformity
of the light. The result is, when the fre-
quency is above say 30 ~, or 60 alterna-
tions per second, i. e., 3,600 alternations
ALTERNATING-CURRENT CIRCUIT. 389
per minute, that the light of an incandes-
cent lamp is practically steady. As the
frequency is reduced, the higher economy
lamps ; i. e., those which have a greater
efficiency, or a greater number of candles-
per-watt, are the first to suffer in apparent
steadiness. First, because they are more
brilliant for the same candle-power, or are
operated at a high temperature, and the
eye is much more sensitive to changes
of brilliancy than to changes of candle-
power, or, in other words, to changes in
candle-power per square inch of bright
surface, than to total candle-power; and
second, because such lamps have usually
thinner filaments, and hence chill more
rapidly.
The frequency employed in alternating-
current circuits varies between 25 ~ and
140 ~ per second.
390 ELECTRIC INCANDESCENT LIGHTING.
If an incandescent lamp be supplied by
a continuous pressure, of say 110 volts,
and gives a candle-power of 16 candles at
this pressure, then if it be removed and
connected with an alternating-current pres-
sure of 110 volts, it will shine with equal
brightness and give 16 candles as before,
no matter what the frequency, provided
only that the frequency be sufficiently
great to keep the lamp from flickering.
Similarly, the current strength, which the
lamp will take, on an alternating current
circuit, is the same as that which it takes
on a continuous-current circuit. This is
for the reason that although the strength
of an alternating current is constantly
fluctuating, between the apex of the suc-
cessive waves and zero at the moments of
reversal, yet the current is defined or
measured by its heating effect, so that if
half an ampere of continuous current
ALTERNATING-CURRENT CIRCUIT. 391
brings a given lamp to candle-power, then
the alternating-current strength which also
brings the lamp to candle-power, will be
half an ampere, no matter what the outline
of the current wave may be. Thus, it
would be possible for each wave to have a
current strength of 2 amperes at the apex,
and yet to produce only the average heat
effect of a half an ampere. Such a current
would have an effective current strength
of 1/2 ampere. In general, the maximum
current strength is about forty per cent, in
excess of the effective current strength, so
that when an alternating-current ammeter
shows that a current of one effective
ampere is passing in a circuit, the apex of
the successive current waves will, probably,
attain a strength of 1 — amperes. In
the same way, if the pressure in an alter-
nating-current circuit, as shown by lamps
392 ELECTRIC INCANDESCENT LIGHTING.
or by instruments, be 100 volts, the instan-
taneous maximum value in the successive
cycles will, probably, be about 140 volts.
It is evident, therefore, that it would
be possible to employ alternating-current
generators instead of continuous-current
generators, in a central station, for the dis-
tribution of electric light. The drop of
pressure, however, in the supply feeders
and mains, would, in such a case, for reasons
that it is not necessary here to consider, be
greater than with the same strength of
continuous current, and it is generally con-
sidered, that it is disad vantageous to supply
low-tension systems of incandescent light-
ing from alternating-current generators.
When, however, the lighting has to be dis-
tributed at great distances, and, therefore,
at a high electric pressure, in order to avoid
expense in conductors, the alternating-cur-
ALTERNATING-CURRENT CIRCUIT. 393
rent system possesses decided advantages,
since it readily lends itself to transforma-
tion of pressure; that is to say, it is easy
to take an alternating-current generator of
say 2,000 volts E. M. F., and to transmit
this pressure to considerable distances over
comparatively small conductors, and then
to reduce the pressure locally, within the
precincts of buildings, to say 100 volts, by
means of an apparatus, called an alternat-
ing-current transformer.
A form of alternating-current trans-
former is shown in Fig. 136. Such an
apparatus consists essentially of two coils
of insulated wire, wound around a common
laminated iron core. These coils are con-
nected respectively with a high-pressure
and a low-pressure circuit. In this case,
the high-pressure circuit is the source of
energy, or i\\z primary circuit, and the low-
394 ELECTRIC INCANDESCENT LIGHTING.
pressure circuit is the circuit of delivery, or
the secondary circuit. The effect of send-
ing an alternating current through the
primary coil is to induce alternations of
L,. . ^_
FIG. 136.— ALTERNATING-CURRENT TRANSFORMER.
CAPACITY 1 KW, OR 20 50- WATT LAMPS.
the same frequency, but different E. M. R,
in the secondary coil. If the secondary
coil contains more turns than the primary,
then the secondary E. M. F. is the greater.
If, on the contrary, as in this case, the
ALTERNATING-CURRENT CIRCUIT. 395
secondary turns are fewer, then the second-
ary E. M. F. will be lower. Transformers
are, therefore, of two kinds ; namely, step-up
transformers, or those which raise the pres-
sure, and step-down transformers, or those
which lower it.
Like all machines for effecting the trans-
formation of energy, alternating-current
transformers waste some portion of what
they receive. In large transformers this
loss at full load is relatively very small, only
about two per cent., so that the efficiency
of a large transformer, at full load, is
approximately ninety-eight per cent. On
the other hand, the relative loss at very light
load is necessarily much greater. During
the day time, when comparatively few
lamps are lighted over the distribution
system, the power supplied to magnetize
the transformers may be considerably
396 ELECTRIC INCANDESCENT LIGHTING.
greater than the power usefully expended.
This is the only objection to the action
of alternating-current transformers, and is
outweighed when the distance to which
the current has to be carried is sufficiently
great, since, otherwise, a large amount of
copper would have to be employed to
transmit the necessary energy in any other
way. A continuous current cannot be
transformed without the aid of rotating
mechanism.
In Fig. 136, P, P, are the primary wires
leading to the high-pressure mains, which
are usually supported on poles overhead.
S, 8, are the secondary wires leading to the
interior of the building and acting as ser-
vice wires. The apparatus represented in
the figure has a capacity of 1 KW, and,
therefore, is capable of supplying about
20 fifty-watt incandescent lamps. If the
ALTERNATING-CURRENT CIRCUIT. 397
secondary pressure be 100 volts, the cur-
rent strength at full load would be 10 am-
peres approximately, since 100 volts X 10
amperes = 1,000 watts. In the primary
circuit the current strength will be about
1 ampere at full load, if the primary
pressure be 1,000 volts, since 1,000 volts
X 1 ampere = 1,000 watts. In reality,
owing to some loss of power in the trans-
former, say 50 watts, the current strength
would be somewhat in excess of 1 ampere.
Moreover, in alternating-current circuits,
the current strength is in excess of the
number of amperes which, multiplied by
the pressure in volts, gives the actual
activity in watts.
Transformers of small size, weight and
capacity, are more expensive per KW, or
per lamp, both to purchase and to operate
than large transformers, since they require
398 ELECTRIC INCANDESCENT LIGHTING.
relatively more labor and material and have
a lower efficiency. It is customary, there-
fore, when possible, to supply several sets of
secondary conductors, say in several adjacent
buildings from a single large transformer,
rather than have a transformer for each
house. For the same reason, where single
lamps have to be supplied by alternating
currents at considerable distances apart, as, r
for example, in street lighting, instead of
employing a small ^th KW transformer
for each lamp, the method is adopted of
connecting a number of lamps in series,
as in an arc circuit, and shunting each lamp
by a coil called a reactive coil. This reac-
tive coil allows almost the entire current
strength in the circuit to pass through the
lamp and takes only a small portion
through its own circuit. If, however, the
lamp filament breaks, thus interrupting the
ALTERNATING-CURRENT CIRCUIT. 399
circuit of that lamp, the reactive coil per-
mits the full current strength to pass
through it with only a comparatively small
drop in pressure; namely, the pressure
FIG. 137. — SERIES-CONNECTED STREET LAMP FOR ALTER
NATING-CURRENT CIRCUITS.
equal to that of the lamp which has be-
come extinguished. This drop is due to a
C. E. M. F. established by the current in
the circuit in passing through the reactive
coil. Such a series lamp shunted by its
reactive coil is represented in Fig. 137.
400 ELECTRIC INCANDESCENT LIGHTING.
The alternating currents required for
alternating-current incandescent lighting,
FIG. 138. — ALTERNATING-CURRENT GENERATOR.
are obtained from a form of dynamo known
as an alternator. This dynamo does not
ALTERNATING-CURRENT CIRCUIT. 401
differ in principle from an ordinary con-
tinuous-current generator, except that it
does not commute the current in its line
circuit. Fig. 138, shows a form of alter-
nator having fourteen poles. Since the
fourteen field magnet coils require to be
supplied by continuous currents, the ma-
chine is accompanied by a small continu-
ous-current generator, called an exciter.
Incandescent lamps, intended for alter-
nating-current circuits, possess no peculiar-
ity, that is to say, a lamp may be used
indifferently on a continuous or on an alter-
nating-current circuit, when the working
pressures in each case, and the sockets, are
the same.
CHAPTEK XX.
MISCELLANEOUS APPLICATIONS OF INCANDES-
CENT LAMPS.
BESIDES the various uses for the incan-
descent lamp, which we have described,
there are many others which want of space
will prevent our discussing, except very
briefly.
The fact that the incandescent lamp is
capable, within reasonable limits, of being
made of almost any size, and the additional
fact that the glowing filament is entirely
protected by a surrounding glass chamber,
permits the incandescent lamp to be
employed for purposes in which other
artificial illuminants would be impossible..
MISCELLANEOUS APPLICATIONS. 403
We may mention, as one of such purposes,
the employment of small suitably shaped
incandescent electric lamps for exploring
the cavities of the body. Incandescent
lamps are thus employed by physicians
and surgeons. For this purpose a small
lamp, shaped so as to permit of its ready
introduction into the cavity to be exam-
ined, is mounted at the extremity of a
suitable support. In some cases the ex-
ploring lamp is sometimes placed in a
sheath or tube, through the interior of
which the illumined area is directly ob-
served.
An entirely distinct method, however,
from the preceding is sometimes adopted ;
namely, the method of trans-illumination.
Here an attempt is made to illumine the in-
terior cavity so as to permit it to be visible
through the body as a translucent screen.
404 ELECTRIC INCANDESCENT LIGHTING.
The amount of light required to attain
this result, is so great as to necessitate the
liberation of considerable heat activity, and
means have usually to be provided for
keeping the lamp cool. This is done by a
FIG. 139. — MINIATURE INCANDESCENT LAMPS.
jacket, in the exterior glass globe, through
which cold water is kept circulating.
Examples of three small lamps employed
for surgical and dental purposes are seen
in Fig. 139. Such lamps are generally
called battery lamps, because they are cap-
MISCELLANEOUS APPLICATIONS.
405
able of being operated from a primary
or secondary battery. The usual voltage
is from three to six volts. The efficiency
of such lamps, in candle-power per watt, is
3-Candle Lamp. 4-Candle Lamp. 6-Candle Lamp.
FIG. 140. — MINIATURE INCANDESCENT LAMPS.
much less than that of larger lamps, owing
principally to the rapid conduction of heat
from the short filament through the con-
ducting wires. Other forms of battery
lamps, of candle-power as marked, are re-
presented in Fig. 140.
406 ELECTRIC INCANDESCENT LIGHTING.
Sometimes in medical or surgical ex-
aminations, a small incandescent lamp is
mounted before a concave reflector, strapped
FIG. 141. —INCANDESCENT LAMPS, WITH REFLECTOR AND
CONDENSING LENS.
to the head of the observer. In other cases
the lamp and reflector are mounted as
shown in Fig. 141, within a tube mounted
MISCELLANEOUS APPLICATIONS. 407
on a suitable stand and provided with a
condensing lens.
Another use for the incandescent elec-
tric lamp which is due to the fact that the
glowing filament is hermetically sealed, is
the use of safety incandescent lamps in
mines, where the presence of fire-damp is
feared. This form of safety lamp is in-
tended to replace the Davy safety lamp.
Such lamps are supplied by either a pri-
mary or secondary battery placed in the
lamp case. A form of such lamp is shown
in Fig. 142.
Battery or miniature incandescent lamps
are also often employed for use with the
microscope in the illumination of the
object, since they are capable of being
placed conveniently for observation. Va-
rious designs of supports and reflectors are
408 ELECTRIC INCANDESCENT LIGHTING.
employed in connection with lamps for
this purpose.
The incandescent lamp is used exten-
FIG. 142. — SAFETY LAMP FOR MINES.
sively on the modern steamship not only
for purposes of general illumination of the
vessel, but also for the lighting of the
409
stern-lights
case of such
lights, it is a matter of considerable im-
portance that the continuity of service of
the lamps be ensured, since the rupture
of the filament might jeopardize the vessel.
In order to lessen the liability of disabling
the side-light, it is common to employ a
double-filament lamp, sometimes called a
twhi-filament lamp, so arranged that if one
filament should fail, the other may con-
tinue burning. At other times two sep-
arate lamps are employed for the same
purpose. A double-filament lamp is shown
in Fig. 143.
Incandescent lamps on board ship are
operated at pressures between 100 and 120
volts, almost invariably on the two-wire
system. In warships, the pressure is usu-
ally 80 volts, in order to facilitate the use
410 ELECTRIC INCANDESCENT LIGHTING.
FIG. 143. — DOUBLE FILAMENT LAMP FOR SHIP'S
SIDE-LIGHT.
MISCELLANEOUS APPLICATIONS. 411
of arc-light projectors without the addition
of much resistance in the projector circuits.
In the early history of ship lighting, the
method was adopted of employing the
ship's iron sheathing as a common return ;
that is, one pole of all the lamps and also
one pole of the dynamo being connected to
the sheathing of the vessel, the other pole
being connected to insulated conductors.
This method was, however, found objec-
tionable in practice and has been completely
given up. Especial care has to be given
to the insulation of wires and fixtures on
board of ship. The connection boxes,
switches, etc., are often arranged so as to
be rendered completely water-tight. In
Fig. 144, JTj is the switch handle, and H,
the receptacle for the insertion of local
connecting wires, leading to the lamp or
other device to be operated. The cover
4:12 ELECTRIC INCANDESCENT LIGHTING.
C\ is intended to effect a water-tight seal
when the receptacle is removed. In Fig.
FIG. 144. — WATER-TIGHT MARINE SWITCH 'AND RECEP-
TACLE.
145, the switch handle H, is recessed into
the water-tistfit box in such a manner that
O
the cover 6y, can enclose it in a water-tight
MISCELLANEOUS APPLICATIONS. 413
seal. The wires enter the box through the
water-tight openings at O O.
Miniature incandescent lamps are occa-
sionally employed as a variety of electric
FIG. 145. — WATER-TIGHT MARINE SWITCH.
jewelry, as in a scarf pin or ornament for
the hair, such lamps being always operated
from a small battery carried on the person.
Similar lamps are also employed on the
414 ELECTRIC INCANDESCENT LIGHTING.
stage, for stage effects, being worn by the
actors.
Incandescent lamps are colored either
by employing globes of tinted glass, or by
dipping the clear lamp globes into a solu-
tion of suitable dye. In the former case
the coloring is permanent, in the latter
case, it is temporary.
Incandescent lamps, either plain or col-
ored, are extensively employed in illumi-
nated electric signs, where the lamps are
grouped in the shapes of letters. Some-
times, in order to attract attention to the
sign, automatic switch devices are em-
ployed, which at intervals extinguish some
or all of the lamps ; or, blocks of lamps are
arranged so as to be automatically cut out
of circuit and cause letters to successively
appear and disappear and words to be thus
spelled out.
416 ELECTRIC INCANDESCENT LIGHTING.
The in -indescent electric liii'ht lends
o
itself very readily to decorative effect.
This arises not only from the readiness
with which the light is subdivided and
distributed, but also from the fact that the
glowing filament being protected by the
surrounding glass chamber, permits the
light to be partially buried in walls or
ceilings, in a manner which would be im-
possible with any other form of artificial
illuminant.
No description of decorative effect by
incandescent lamps would be complete
without some allusion to the magnificent
spectacle that was afforded by the incan-
descent lighting of the Court of Honor, at
the World's Columbian Exhibition, at Chi-
cago, in 1893. A faint conception only of
the beauty of this scene may be gathered
from the accompanying illustration in Fig.
MISCELLANEOUS APPLICATIONS. 417
146. The building in the background is
the Administration Building, whose exte-
rior is lighted by incandescent lamps. On
the right hand side is the facade of Elec-
tricity Building, and on the left hand is
the facade of Machinery Hall. The entire
bank of the waterway was illumined by
incandescent lamps.
INDEX.
Actinic Effect of Light, 208.
Activity, 62.
and Candle-Power, Relation Between, 183.
, Surface, of Incandescent Filament, 165.
, Surface, of Positive Crater in Arc. 166,
, Unit of, 63.
Adapters for Lamps, 133, 134.
Adjustable Lamp Pendant, 240.
Age Coating of Lamp, 178.
Air Pump, Mechanical, 125.
Alternating-Current Circuit for Incandescent
Lighting, 386 to 401.
Current Generator, 400.
Current Transformer, 393.
Alternation, 386.
, Frequency of, 387.
Alternator, 400.
Ammeters, 306.
419
420 t INDEX.
Ampere, 56.
Hours, 335.
- Efficiency of Storage Cell, 372.
Amyloid, Use of, for Filaments, 87,
Analysis of Light of Glowing Filament, 74.
- of Sunlight, 74.
Antimonious Lead, Use of, in Storage Cells, 354.
Arc, Voltaic, 18.
Armature for Central-Station Generator, Winding
of, 322, 323.
— of Central-Station Generator, 314.
Artificial Illtiminants, Requisites for, C.
Illumination, 1 to 17.
Automatic Cut-Out for Storage-Battery Switch-
board, 365.
Safety Device for Incandescent Lamp, 375.
Switch, 273.
B
Bamboo Filament, 107.
- Filament, Preparation of, 86, 87.
Bases, Lamp, 131, 132.
Batteries, Storage, 345 to 373.
Battery Incandescent Lamps, 407.
-of Boilers, 311.
, Voltaic, 47.
INDEX. 421
Begohrn, 247.
Belt-Driven Generator, 327.
Belting, 327.
.Bipolar Generator for Isolated Plant, 330.
Blackening of Lamp Chamber, 178.
Block, Branch, 274, 275.
Blowing of Fuse, 276.
Boilers, Battery of, 311.
Bougie-Decimate, 200.
Boulyguine, 33.
Boulyguine's Incandescent Lamp, 32, 33.
Box, Carbonizing, 93.
Boxes, Distribution, 280, 282.
, Connecting, 263, 264.
, Coupling, 295, 296.
, Field Regulating, 306.
, Junction, 300 to 303.
Bracket, Lamp, 243.
— Lamp, Movable Arm for, 246.
Branch, 274, 275.
Coupling Boxes, 297.
Cut-Out, 274.
Branches, 251.
Brass- Covered Conduit, 262.
Brilliancy, 181.
of Incandescent Filament, 168.
British Candle, 199.
422 INDEX.
Brushes for Generator, 316.
Bus Bars, Definition of, 223.
c
Candle Power and Activity, Relation Between,
183.
• Power, Effect of Varying Pressure on, 188
to 191.
Power of Luminous Source, 199.
Power, Mean Spherical, 207.
Capacity, Energy of, Storage of Cell, 361.
, Storage, 349.
Carbon Filament, Life of, 16V.
, Suitability of, for Incandescent Filament,
83, 84.
Carbonization, General Processes for, 84, 85.
, Methods Employed for, 91 to 98.
Carbonizing Box, 93.
Frame, 94, 95.
Cell, Charged, 352.
, Counter-Electromotive Force of, 365.
, Secondary, 352.
Celluloid Filaments, 90.
Central-Station Generator, Double Brushes for,
316.
Station Generators, 313.
INDEX. 423
Central Station, Load Diagram of, 347, 348.
Station, Six-Pole Generator for, 317 to 320.
Station Smashing Point of Lamp, 193.
Station, Storage Cell for, 359.
Station Switchboard, 305 to 308.
Central Stations, 304 to 333.
Charged Cell, 352.
Charging Current, 352.
Chlorine and Bromine, Residual Atmospheres of,
in Lamp Chambers, 130.
Circuit, Open, 46.
, Primary, of Transformer, 393.
, Secondary, of Transformer, 394.
, Series Arc and Incandescent, 383.
Circular-Mil, Definition of, 54.
Circular-Mil-Foot, 54.
Cleat Wiring, 254.
Cleats, 255.
, Wooden, 255.
Coil, Reactive, 398.
Cold Light, 10, 78, 79.
Color, Cause of, 71, 72.
Values, Day-Light, 72.
Commutator of Central-Station Generator, 314.
, Sparking at, 317.
Concave Panel Shade and Reflector, 153.
Concealed Work, 260.
424 INDEX.
Conduction of Heat, 75.
Conductor, Neutral, of Three-Wire System, 222.
— , Twisted-Double, 248.
Conductors, Double-Flexible, 249.
, Drop in, 252.
— , Parallel, 248.
— , Silk-Covered, 248.
— , Solid, 247.
— , Supply, 250.
— , Twin, 248.
, Stranded, 247.
Conduit, Brass-Covered, 262.
of Creosoted Wood, 288, 289.
Conduits, 287.
, Interior, 261, 262.
Connecting Boxes, 263, 264.
Connection Boxes for Interior Conduits, 264.
in Series, 47.
Consumers' Smashing Point of Lamp, 194.
Continuous Electric Current, 304.
Convection of Heat, 77.
Cord, Flexible Lamp, 240.
Corrugated Lamp Reflector and Shade, 155.
Cotton Thread, Use of, for Filament, 87.
Coulomb, 56.
Coulomb-per-Second, 56.
Counter-Electromotive Force Cells, 365.
INDEX. 425
Coupling Boxes, 295.
— Boxes, Branch, 297.
Crater, Positive, of Arc, Surface Activity in,
166.
Creosoted Wood Conduit, 288, 289.
Current, Continuous Electric, 304.
, Electric, Pulsating, 386.
, Steady Electric, 386.
Strength, Effective, 391.
Cut-Out, Branch, 274.
Film for Incandescent Electric Lamp, 376,
377.
- Fixture, 278, 279.
- Mains, 277.
Switch for Series Circuit, 384.
Cycle, 387.
D
Day-Light Color Values, 72.
De Changy, 25.
Incandescent Lamp, 26, 27.
De Moleyn, 24.
Diagram of Central-Station Load, 347, 348.
Direct-Driven Generator, 328, 329.
Distributing Point, 280.
Distribution Boxes, 280, 283.
426 INDEX.
Distribution, Centers of, 253.
, Series-Multiple Lamp, 221.
, Three- Wire System of Lamps, 221 to
224.
Double Brushes for Central-Station Generator,
316.
Filament Lamps, 409.
Flexible Conductors, 249.
- Pole Switches, 267.
— Pole Switches, Forms of, 269, 270, 271.
Drop in Conductors, 252.
Dummy Moulding, 260.
Dynamo-Electric Generator, 48.
Dynamos or Generators, 311.
E
E. M. F., 44.
Early History of Incandescent Lighting, 18
to 42.
Early Horse-Shoe Lamp, 104.
Early Illuminants, 1 to 5.
Incandescent Lamps, 19, 20.
Effect of Temperature on Resistivity of Insula-
tors, 55.
Effective Current Strength, 391.
Efficiency, Ampere-Hour, of Storage Cell, 372.
INDEX. 427
Efficiency of Incandescent Lamp, 170 to 233.
' of Lamp, Effect of, on Duration of Life,
184, 185.
of Storage Cell, 372.
of Transformers. 395.
Electric Jewellery, 413.
Lighting, Life Risks of, 14, 15.
Pressure, 45.
Quantity, Unit of, 56.
Resistance, 49.
Electrolier, 243.
Electrolytic Meter, 335 to 339.
Seal, 108.
Electromotive Force, 43.
Elementary Electrical Principles, 43 to 64.
Elements or Plates of Storage Cell, 351.
Emissivity of Incandescing Filament, 80, 81.
of Lamp Filament, Effect of Surface on,
180.
Energy Efficiency of Storage Cell, 372, 373.
Storage Capacity of Secondary Cell, 361.
Storage of Cell, 361.
Equalizer Switch, 235, 236.
Ether, Luminiferous, 67.
, Universal, 67.
Evaporation of Incandescent Filament, 176, 177.
Exciter, 401.
428 INDEX.
F
Factor, Load, 349.
Farmer, 36.
Farmer's Incandescent Lamp, 35, 36.
Feeder Distribution, 229.
Distribution, Three- Wire System of, 231,
232.
Equalizer Resistance, 234.
— Load, Methods of Overcoming Inequali-
ties of, 233, 236.
— Regulators, 233.
- System, 253.
— Tubes, 298, 299.
Feeders for Lamp Distribution, 228.
Feeding Point, 298.
Field Magnet Coils of Generator, 314.
Regulating Boxes, 306.
Filament, Bamboo, Preparation of, 86, 87.
— , Effect of Flashing on Emissivity of, 116,
117.
— , Effect of Surface on Emissivity of, 180.
— , Incandescing, Surface Activity of, 80, 81.
— , Mounting of, 99, 100,
, Sealing-in of, 118.
— , Shadows, 1/9.
— , Spotted, 112.
Filaments, Amyloids,
— , Celluloid, 90.
, Flashing Process
, Methods Employed for
91 to 98.
— , Squirted, 88, 89, 97.
— , Stopper-Mounted, 122, 123.
— , Use of Cotton Thread for, 87.
Fire Fly, Radiation of, 78.
Fittings, Lamp, 135 to 162.
Five- and Four-Wire Systems of Lamp Distribu-
tion, 225.
Fixture Cut-Outs, 278, 279.
— Molding, 259.
for Street Incandescent Lamp, 381, 382.
Fixtures and Wiring for Houses, 237 to 283.
Flashing Process for Filaments, 113, 114, 115.
Flexible Lamp Cord, 240.
— Lamp Pendant, 239, 240.
Flush Switches, 272.
Foot-Pound, 59.
Foot-Pound-per-Second, 63.
Four- and Five-Wire Systems of Lamp Distribu-
tion, 225,
Frame, Carbonizing, 94, 95.
French Standard Candle, 200.
Frequency, Effect of, on Steadiness of Light, 388.
430 INDEX.
Frequency of Alternation, 387.
, Luminous, 68.
Full- Wire Guard for Incandescent Lamp, 158.
Fuse, Blowing of, 276.
— , Cut-Out, 273.
— , Safety, 274.
G
Geissler Type of Mercury Pump, 127.
Generator, Alternating-Current, 400
, Dynamo-Electric, 48.
, Field Magnet Coils of, 314.
Unit, 308.
Generators, Multipolar, 316.
Generators or Dynamos, 313.
Glass Lamp Shades, 156.
Glow Worm and Fire Fly, Radiation of, 78.
Glowing Filament, Analysis of Light Produced
by, 74.
Grids of Storage Cell, 351.
H
Half-Shade for Incandescent Lamp, 151.
Half Wire-Guard for Incandescent Lamp, 157.
Heat, Conduction of, 75.
, Convection of, 77.
INDEX. 431
Heat, Molecular Transfer of, 76, 77.
•, Radiation of, 75.
High-Economy Lamps, 389.
Horizontal Intensity, Maximum, 207.
Horse-Power, Definition of, 63.
Hours, Ampere, 335.
, Watt, 335.
House Fixtures and Wiring, 237 to 283.
Illuminants, Early, 1 to 5.
Illuminated Electric Signs, 414.
Illumination, Actual Values of, 206.
, Artificial, 1 to 17.
, Law of, 203, 204, 205.
, Significance of, 198.
, Unit of, 202.
Incandescent Lamp, 163.
Lamp, Automatic Safety Device for, 375.
Lamp, Efficiency of, 170 to 233.
- Lamp, Half Wire-Guard for, 157.
Lamp, Multiple-Series Distribution of,
376, 377.
Lamp, Street Fixture for, 378.
- Electric Lamp, Film Cut-Out for, 376, 377.
Electric Lamp, Physics of, 65 to 82.
432 INDEX.
Incandescent Filament, Brilliancy of, 168.
Filaments, Evaporation of, 176, 177.
- Filament, Total Candle-Power of, 168, 169.
- Filaments, Surface Activity of, 165.
Lamp, Full Wire-Guard for, 158.
- Head-Lights for Ships, 409.
Lamps, Early, 19, 20.
- Lamps, Miniature, 404, 405.
Lamps, Miscellaneous Applications of,
402 to 417.
Lamp, Varying Candle-Powers of, 209.
Lighting, Early History of, 18 to 42.
— Lighting, Fire Risks of, 13, 14.
— — Lighting, Alternating-Current Circuit for
386 to 401.
Lighting, Decorative Effects in, 416, 417.
• • Side-Lights for Ships, 409.
Signal Lights for Ships, 409.
Stern-Lights for Ships, 409.
Filament, Emissivity of,-80, 81.
Filament, Temperature of, 82, 174.
Intake Wires, 280.
Intensity, Luminous, 69.
, Maximum Horizontal, 207.
Interior Conduit Joints, 263.
Conduit Junction Boxes, 264.
Conduit, Junction Boxes for, 265.
INDEX. 433
Interior Conduits, 261, 262.
Isolated Lighting Plants, 324 to 333.
— Plant, Smashing Point of Lamps, 194, 195.
- Plants, 324 to 333.
Plants, Quadripolar Generator for, 330, 332.
Jewellery, Electric, 413.
Joints of Filament with Leading-in- Wires, Bolt
and Nut Type, 108.
, Butt Joint Type, 111.
— , Socket Type, 108.
— , Interior Conduit, 263.
Joule, 60.
Joule-per-Second, 63.
Junction Boxes, 300 to 303.
Interior Conduit, 264.
K
Keyless Wall-Socket, 138.
Key-Socket Push Button, 145.
— Wall-Socket, 139.
Keys for Sockets, 140 to 142. .
King, 27.
Kosloff, 29.
Konn, 29.
Komi's Incandescent Lamp, 30, 31.
434 INDEX.
Lamp Adapter, 133.
, Age-Coating of, 178.
- Bases, 131, 132.
- Bracket, 243.
Chamber, Blackening of, 178.
, Filament Shadows on, 179.
— , Machine Sealing of, 121.
— , Sealing Off of, 127, 128.
, Steam-Tight, 161, 162.
Cord, Flexible, 240.
— Cords, Silk, 248.
— Distribution, Series-Multiple System of, 221.
— , Distribution Systems of, 209 to 236.
— Filament, Relation Between Efficiency,
Candle-Power, and Surface Activity of,
180.
— Fittings, 135 to 162.
— Guard, Portable, 160.
— , Incandescent, 163.
— , Leading-in- Wires of, 100.
— , Mean Spherical Candle-Power of, 207.
— Pendant, Adjustable, 240.
— , Flexible, 239, 240.
— , Portable Incandescent, 238.
— Post for Incandescent Lamps, 380, 381.
INDEX. 435
Lamp Reflector and Shade, Corrugated, 155.
Renewals, Rules for Best Commercial
Results, 196, 197, 198.
Shades, Glass, 156.
Shadows, 149 to 156.
, Smashing Point of, 192.
, Semi-Incandescent, 37, 38, 39,
Socket, Temporary, 147.
Socket, Weather-Proof, 148.
Sockets, 135 to 140.
, Stopper, 121, 122, 123.
Switch, 140.
Switches, 267.
, Twin-Filament, 409.
Lamps, Battery-Incandescent, 407.
— , Connection in Parallel, 210.
, Diagram of Multiple Connection of, 212.
, Distribution of by Four- and Five-Wire
Systems, 225.
, Double-Filament, 409.
, Incandescent, Use of, for Surgical Explo-
ration, 403.
, Miniature Incandescent, 404.
, Safety Incandescent, 407.
, Series Connected, Diagram of, 211.
— , Series Connections of, 210.
, Spring Socket for, 146.
436 INDEX.
Lamps, Tree Distribution of, 230.
Law of Illumination, 203.
Leading-in Wires of Lamp, 100.
Life of Filament, Circumstances Governing, 167.
of Incandescent Filament, 167.
- of Lamp and Efficiency, Relation Between,
183.
Light, Actinic Effect of, 208.
, Monochromatic, 73.
— , Objective, Significance of, 66.
, Significance of term, 198.
, Subjective, Significance of term, 66.
— , Two-Fold Use of Word, 66.
— , Unit of Total Quantity of, 201.
Lighting Plant, Isolated, 325.
, Series Incandescent, 374 to 385.
Load Diagram of Central Station, 347, 348.
— Factor, 348.
Lodyguine, 28.
Long Life vs. Low Efficiency, 185, 186.
Lumen, 201.
Luminiferous Ether, 67.
Luminous Frequency, 68.
Frequency, Effect of Temperature on,
68, 69.
Intensity, 69.
Intensity, Standards of, 200.
INDEX. 437
Luminous Intensity, Unit of, 199.
— Source, Candle-Power of, 199.
Lux, 202.
Lux-Second, 208.
M
Machine Seal of Lamp Chamber, 121.
Main Conductors, Overhead, 284.
- Cut-Outs, 277.
- Switch, 277.
Tubes, 293.
Mains, 250.
— , Street, 284 to 303.
— , Three- Wire, 276.
— , Two- Wire, 276.
Man Holes, 287.
Marine Switch for Lamps, 412.
Maximum Horizontal Intensity, 207.
Mean Spherical Candle-Power, 207.
Mechanical Air Pump, 125.
Mercury Pump, 125.
— Pump, Geissler Type, 127.
Mercury Pump, Sprengel Type, 127.
Metallic Half-Shade for Incandescent Lamp, 151,
• Lamp Shades, 149 to 153.
Meter, Electrolytic, 335 to 339.
Meters, Electric, 334, 344.
438 INDEX.
Methods for Overcoming Inequality of Feeder
Loads, 233, 236.
Mil-Foot, Circular, Definition of, 54.
Miscellaneous Applications of Incandescent Lamps,
402 to 417.
Molecular Transfer of Heat, 76, 77.
Monochromatic Light, 73.
Molding, Dummy, 260.
, Picture or Ornamental, 259.
— , Section of, 258.
, Three- Wire, 257.
Mounting of Filament, 99, 100.
Movable Arm for Bracket Lamp, 246.
Multiple and Series Systems of Lamp Distribution,
Relative Advantages of, 212 to 220.
Connected Lamps, Diagram of, 212.
Series System of Distribution of Incan-
descent Lamps, 376, 377.
Multipolar Generators, 316.
N
Negative Plate of Storage Cell, 351.
Pole, 44.
Terminal, 44.
Neutral Conductor of Three- Wire System, 222.
Non-Luminous Heat, 10.
INDEX. 439
o
Occluded Gas Process, 128, 129.
Ohin, 57.
, Definition of, 50.
Ohm's Law, 57.
Open-Circuit, 46.
Ornamental Moulding, 259.
Outlet Boxes, 266.
Output Wires, 282.
Overhead Main Conductors, 284.
Wires, 284.
Overload Switch for Storage Battery Switch-
board, 366 to 370.
Panel Reflectors, 153, 154.
Parallel Conductors, 248.
Parchmentizing Process, 85, 86.
Pendant Lamp, 244.
Petrie, 24.
Phot, 208.
Physics of the Incandescent Lamp, 65 to 82.
Plant for Isolated Lighting, 325.
Plants, Isolated, 324 to 333.
Plates or Elements of Storage Cell, 351.
440 INDEX.
Platinum- Wire Incandescent Lamps, Requisites
for, 21, 22.
— Wire, Use of, for Sealing-in Lamp Chamber,
105.
Plug, Cut-Outs, 277.
Point, Distributing, 280.
, Feeding, 298.
Pole, Negative, 44.
, Positive, 44.
Portable Electric Incandescent Lamp, 238.
— Lamp Guard, 160.
Positive Plate of Storage Cell, 351.
- Pole, 44.
— Terminal, 44.
Pressure, Electric, 45.
— , Unit of, 46.
— Switch for Storage Battery Switchboard,
364.
Wires, 298.
Primary Circuit of Transformer, 393.
Pulsating Electric Current, 386.
Pump, Mercury, 125.
Push-Button Key Socket, 145.
Q
Quadripolar Generator for Isolated Plants, 330, 332.
INDEX. 441
R
Radiation of Glow- Worm and Fire-Fly, 78.
- of Heat, 75.
Rate-of- Doing- Work, 62.
Radiation, Selective, 79.
Rays, Ultra-Violet, 70.
Reactive Coil, 398.
Recording Wattmeter, 341 to 349.
Reflector Shade for Incandescent Lamp, 152.
Regulators, Feeder, 233.
Requisites for Artificial Illuminants, 6.
Residual Atmosphere, Lamp Chambers Inten-
tionally Provided with, 130.
Resistance, Electric, 49.
- Electric, Unit of, 50.
— for Feeder Equalizer, 234.
, Specific, 53.
Resistivity, 53.
of Conductors, Effect of Temperature on,
55.
of Insulators, Effect of Temperature on,
55.
Reynier, 38.
Reynier's Semi-Incandescent Lamp, 38, 39.
Reynier- Werdermann's Incandescent Lamp, 41, 42.
Risers, 250.
442 INDEX.
s
Safety Incandescent Lamps, 407.
- Fuse, 274.
Sawyer, 33.
Sawyer's Incandescent Lamp, 34.
Safety Device, Automatic, for Incandescent Lamp,
375.
Screw Cleats, 256.
Seal, Machine, of Lamp Chamber, 121.
Sealiug-in of Filament, 118.
Sealing-off of Lamp Chamber, 127, 128.
Secondary Cell, 352.
Circuit of Transformer, 394.
Selective Absorption of Light Radiations, 71, 72.
Radiation, 79.
Semi-Incandescent Lamp, 37, 38, 39.
Series and Multiple Distribution, Relative Advan-
tages of, 212 to 220.
Arc and Incandescent Lamp Circuit, 383.
Connected Lamps, Diagram of, 210.
Connection, 47.
Connections of Lamps, 210.
Incandescent Lighting, 374 to 385.
Multiple System of Lamp Distribution,
221.
Service Wires, 250.
INDEX. 443
Shades for Incandescent Lamps, 149 to 156.
Shadows, Filament, 179.
Ship Lighting by Incandescent Lamps, 412.
Ships, Incandescent Head-Lights for, 409.
, Incandescent Side-Lights for, 409.
, Incandescent Stern-Lights for, 409.
Short-Circuit, 272.
Short Life vs. High Efficiency, 185, 186.
Signal Lights for Ships, Incandescent, 409.
Signs, Illuminated Electric, 414.
Silk-Covered Conductors, 248.
Lamp Cords, 248.
Single-Pole Switch, Simple Form of, 268.
Pole Switches, 267.
Six-Pole Generator for Central Station, 217 to
320.
Smashing Point of Lamp, 192.
Point of Lamp from Central Station Stand-
point, 193.
Point of Lamp from Consumers' Stand-
point, 194.
Point of Lamp from Isolated-Plant Stand-
point, 194, 195.
Socket Keys, 140 to 142.
Sockets, Simple Form of, 137.
Solid Conductors, 247.
Wires, 246.
444 INDEX.
Sparking at Commutator, 317.
Specific Resistance, 53.
Spotted Filament, 112.
Sprengel Type of Mercury Pump, 127.
Spring Socket for Lamps, 146.
Squirted Filaments, 88, 89, 97.
Starr, 27.
King Incandescent Lamp, 27, 28.
Standard Candle, French, 200.
of Luminous Intensity, 200.
Storage Cell Tester, Simple Form of, 371, 372.
Stations, Central, 304.
Steadiness of Light, Effect of Frequency on, 388,
Steady Electric Current, 386.
Steam-Tight Lamp Chamber, 161, 162.
Step-Down Transformer, 395.
Step-Up Transformer, 395.
Stern-Lights for Ships, Incandescent, 409.
Stopper Lamp, 121, 122, 123.
Stopper-Mounted Filaments, 122, 123.
Storage-Battery Switchboard, 363 to 369.
Batteries, 345 to 373.
Capacity, 349.
Cell, Efficiency of, 2, 372.
- Cell, Energy Storage Capacity of, 361.
Cell, for Central-Station Work, 359.
Cell, Energy Efficiency of, 372, 373.
INDEX. 445
Storage Cell, Plates or Elements of, 351.
— Cell, Negative Plate of, 351.
Cell, Positive Plate of, 351.
Cell, Voltmeter and Electrodes, 370.
Stranded Wires, 246.
Conductors, 247.
Street Incandescent Lamp Fixture, 381, 382.
Fixture, for Series-Incandescent Lamp, 378.
Lamps, Series Connected for Use on Alter-
nating-Current Circuits, 399.
Mains, 284 to 303.
Sub Mains, 250.
Subways, 285, 286.
Sunlight, Analysis of, 74.
Color Values of Artificial Illuminants, 8, 9,
12.
, Frequencies Present in, 70.
Supply Conductors, 250.
Surface Activity of Incandescing Filament, 80, 81,
165.
Activity of Lamp Filament, 180.
Activity of Positive Crater of Arc, 166.
Surgery, Use of Incandescent Lamps in, 403.
Suspended Lamps, Wire Guards for, 159.
Switch, Automatic, 273.
— , Flush, 272.
— for Lamps, 140.
446 INDEX.
Switch, Main, 277.
, Marine, for Lamps, 412.
, Pressure, for Storage-Battery Switch-
board, 364.
, Overload, for Storage-Battery Switch-
board, 365, 366 to 370.
Switchboard for Central Station, 305 to 308.
Switches, Double-Pole, 267.
for Lamps, 267.
, Single-Pole, 267.
System, Feeder, 253.
, Three-Wire, of Lamp Distribution, 221 to
224.
Systems of Lamp Distribution, 210.
T
Taps, 251.
Temperature, Effect of, on Luminous Frequencies,
68, 69.
, Effect of, on Resistivity of Insulators, 55.
of Incandescing Filament, 82, 174.
Temporary Lamp Socket, 147.
Terminal, Negative, 44.
, Positive, 44.
Tester for Storage Cells, Simple Form of, 371,
372.
INDEX. 447
Thermostat, 339.
Three-Wire Mains, 276.
Moulding, 257.
System, Neutral Conductor of, 222.
System of Feeder Distribution, 231, 232.
System of Distribution, 221 to 224.
Time Illumination, Unit of, 208.
Total Candle-Power of Incandescent Filament,
168, 169.
Transformer, Alternating- Current, 393.
•, Primary Circuit of, 393.
, Step-Up, 395.
, Secondary Circuit of, 394.
Transformers, Effect of Size and Weight on Cost
and Efficiency of, 397, 398.
Efficiency of, 395.
Step-Down, 395.
Trans-Illumination, 403.
Tree Distribution of Lamps, 230.
Twin Conductors, 248.
Filament Lamp, 409.
Twisted Double Conductor, 248.
Two- and Three-Wire Systems of Lamp Distribu-
tion, Relative Economy of, 223, 224.
Two-Wire Mains, 276.
Tube, Underground, 292.
Tubes, Feeder, 298, 299.
448 INDEX.
u
Ultra-Violet Rays, 70.
Underground Tube, 292.
Unit Generator, 308.
of Activity, 63.
• of Electric Activity, 63.
- of Electric Flow, 56.
• of Electric Power, 63.
of Electric Pressure, 46.
— of Electric Quantity, 56.
— of Electric Resistance, 50.
of Illumination, 202.
of Illumination, Intensity of, 199.
of Time Illumination, 208.
— of Total Quantity of Light, 201.
— of Work, 59.
Universal Ether, 67.
V
Violle, 200.
Volt, 46.
Ampere, 63.
Coulomb, 61.
Coulomb-per-Second, 63.
Voltaic Arc, 18.
Battery, 47.
INDEX. 449
Voltmeters, 306.
and Electrodes for Storage Cell, 370.
w
Wall Socket, Key, 139.
Socket, Keyless, 138.
Watt-Hours, 335.
Wattmeter, Recording, 341 to 349.
Weather-Proof Lamp Socket, 148.
Werdermann, 40.
Werdermann's Semi-Incandescent Lamp, 41.
Wire-Guards for Suspended Lamps, 159.
Wires, In -Take, 280.
, Overhead, 284.
, Pressure, 298.
, Service, 250.
— , Solid, 246.
— , Stranded, 246.
Wiring and Fixtures for Houses, 237 to 283.
Cleat, 254.
Wooden Cleats, 255.
Work, 59.
— , Concealed, 260.
, Unit of, 59.
PUBLICA
THE W. J.
The Electrical World.
Review of Current Progress in Elect?!
tical Applications. Annual subscription .
The Electric Railway Gazette. An Illustrated
Bi-monthly Record of Electric Railway Practice and
Development. Annual subscription i.oo
Johnston's Electrical and Street Railway
Directory. Containing Lists of Central Electric
Light Stations, Isolated Plants, Electric Mining Plants,
Street Railway Companies — Electric, Horse and Cable —
with detailed information regarding each ; also Lists of
Electrical and Street Railway Manufacturers and Dealers,
Electricians, etc. Published annually 5.00
The Telegraph in America. By Jas. D. Reid.
894 royal octavo pages, handsomely illustrated. Russia, 7.00
Dictionary of Electrical Words, Terms
and Phrases. By Edwin J. Houston, Ph.D.
Third edition. Greatly enlarged. 667 double column
octavo pages, 582 illustrations 5.00
The Electric Motor and Its Applications.
By T. C. Martin and Jos. Wetzler. With an appendix
on the Development of the Electric Motor since 1888, by
Dr, Louis Bell. 315 pages, 353 illustrations 3.00
The Electric Railway in Theory and
Practice. The First Systematic Treatise on the
Electric Railway. By O. T. Crosby and Dr. Louis
Bell. Second edition, revised and enlarged. 416 pages,
183 illustrations 2.50
Alternating Currents. An Analytical and GrapK-
ical Treatment for Students and Engineers. By Frederick
Bedell, Ph.D., and Albert C. Crehore, Ph.D. Second
edition. 325 pages, 1 12 illustrations 2.50
Publications of the IV. J. JOHNSTON COMPANY.
Gerard's Electricity. With chapters by Dr. Louis
Duncan, C. P. Steinmetz, A. E. Kennelly and Dr. Gary
T. Hutchinson. Translated under the direction of Dr.
Louis Duncan , $2.50
The Theory and Calculation of Alternat-
ing-Current Phenomena. By Charles Proteus
Steinmetz 2.50
Central Station Bookkeeping. With Suggested
Forms. By H. A. Foster 2.50
Continuous Current Dynamos and Motors.
An Elementary Treatise for Students. By Frank P.
Cox, B. S. 271 pages, 83 illustrations 2.00
Electricity at the Paris Exposition of
1889. By Carl Hering. 250 pages, 62 illustrations. 2.00
Electric Lighting Specifications for the use of
Engineers and Architects. Second edition, entirely re-
written. By E. A. Merrill. 213 pages 1.50
The Quadruplex. By Wm. Maver, Jr., and Minor
M. Davis. With Chapters on Dynamo -Electric Machines
in Relation to the Quadruplex, Telegraph Repeaters, the
Wheatstone Automatic Telegraph, etc. 126 pages, 63
illustrations . . , 1.50
The Elements of Static Electricity, with Full
Descriptions of the Holtz and Topler Machines. By
Philip Atkinson, Ph.D. Second edition. 228 pages,
64 illustrations 1.50
Lightning Flashes. A Volume of Short, Bright
and Crisp Electrical Stories and Sketches. 160 pages,
copiously illustrated 1.50
A Practical Treatise 011 Lightning Pro-
tection; By H.W. Spang. 1 80 pages, 28 illustrations, 1.50
Publications of the W. J. JOHNSTON COMPANY.
Electricity and Magnetism. Being a Series of
Advanced Primers. By Edwin J. Houston, Ph.D. 306
pages, 1 16 illustrations $1.00
Electrical Measurements and Other Ad-
vanced Primers of Electricity. By Edwin
J. Houston, Ph.D. 429 pages, 169 illustrations i.oo
The Electrical Transmission of Intelli-
gence and Other Advanced Primers of
Electricity. By Edwin J. Houston, Ph.D. 330
pages, 88 illustrations « i.oo
Electricity One Hundred Years Ago and
To-day, By Edwin J. Houston, Ph.D. 179 pages,
illustrated i.oo
Alternating Electric Currents. By E. J.
Houston, Ph.D. and A. E. Kennelly, D.Sc. (Electro-
Technical Series) , , i.oo
Electric Heating. By E. J. Houston, Ph.D. and
A. E. Kennelly, D.Sc. (Electro-Technical Series) i.oo
Electromagnetism. By E. J. Houston, Ph.D. and
A. E. Kennelly, D.Sc. (Electro-Technical Series) i.oo
Electro-Therapeutics. By E. J. Houston, Ph.D.
and A. E. Kennelly, D.Sc. (Electro-Technical Series). . i.oo
Electric Arc Lighting. By E. J. Houston, Ph.D.
and A. E. Kennelly, D.Sc. (Electro-Technical Series). . i.oo
Electric Incandescent Lighting. By E. J.
Houston, Ph.D. and A. E. Kennelly, D.Sc. (Electro-
Technical Series) i.oo
Electric Motors. By E. J. Houston, Ih.D. and A.
E. Kennelly, D.Sc. (Electro-Technical Series) i.oo
Publications of the W. J. JOHNSTON COMPANY
Electric Street Railways. By E. J. Houston,
Ph.D. and A. E. Kennelly, D.Sc. (Electro-Technical
Series) $1.00
Electric Telephony. By E. J. Houston, Ph.D.
and A. E. Kennelly, D.Sc. (Electro-Technical Series). . i.oo
Electric Telegraphy. By E. J. Houston, Ph.D.
and A. E. Kennelly, D.Sc. (Electro-Technical Series). . i.oo
Alternating Currents of Electricity. Their
Generation, Measurement, Distribution and Application.
Authorized American Edition. By Gisbert Kapp. 164
pages, 37 illustrations and two plates I.oo
Recent Progress in Electric Railways.
Being a Summary of Current Advance in Electric Rail-
way Construction, Operation, Systems, Machinery,
Appliances, etc. Compiled by Carl Hering. 386
pages, no illustrations i.oo
Original Papers on Dynamo Machinery
and Allied Subjects, Authorized American
Edition. By John Hopkinson, F.R.S. 249 pages, 90
illustrations i.oo
Davis' Standard Tables for Electric Wire-
men. With Instructions for Wiremen and Linemen,
Rules for Safe Wiring and Useful Formulae and Data.
Fourth edition. Revised by W. D. Weaver i .00
Universal Wiring Computer, for Determining
the Sizes of Wires for Incandescent Electric Lamp Leads,
and for Distribution in General Without Calculation,
with Some Notes on Wiring and a Set of Auxiliary
Tables. By Carl Hering. 44 pages i.oo
Publications of the W. J. JOHNSTON COMPANY.
Experiments With Alternating Currents
of High Potential and High Frequency.
By Nikola Tesla. 146 pages, 30 illustrations $1.00
Lectures on the Electro-Magnet. Authorized
American Edition. By Prof. Silvanus P. Thompson.
287 pages, 75 illustrations i.oo
Dynamo and Motor Building for Amateurs.
With Working Drawings. By Lieutenant C. D. Park-
hurst i.oo
Reference Book of Tables and Formulae
for Electric Street Railway Engineers.
By E. A. Merrill i.oo
Practical Information for Telephonists.
By T. D. Lock wood. 192 pages I.oo
Wheeler's Chart of Wire Gauges i.oo
A Practical Treatise on Lightning Con-
ductors. By H.W. Spang. 48 pages, 10 illustrations. .75
Proceedings of the National Conference of
Electricians. 300 pages, 23 illustrations 75
Wired Love ; A Romance of Dots and Dashes. 256
pages 75
Tables of Equivalents of Units of Measure-
ment. By Carl Hering 50
Copies of any of the above books or of any other electrical book
published, will be sent by mail, POSTAGE PREPAID, to any address
in the world on receipt of price.
THE W. J. JOHNSTON COMPANY,
253 BROADWAY, NEW YORK.
THE PIONEER ELECTRICAL JOURNAL OF AMERICA,
Read Wherever the English Language is Spoken,
The Electrical World
is the largest, most handsomely illustrated, and most widely
circulated electrical journal in the world.
It should be read not only by every ambitious electrician
anxious to rise in his profession, but by every intelligent Ameri-
can.
It is noted for its ability, enterprise, independence and
honesty. For thoroughness, candor and progressive spirit it
stands in the foremost rank of special journalism.
Always abreast of the times, its treatment of everything
relating to the practical and scientific development of electrical
knowledge is comprehensive and authoritative. Among its
many features is a weekly Digest of Current Technical Electri-
cal Literature, which gives a complete r£sum£ of current origi-
nal contributions to electrical literature appearing in other
journals the world over.
Subscription] Ml^SSi?r4S^u-&'}$3 a Year.
May be ordered of any Newsdealer at 10 cents a week.
Cloth Binders for THE ELECTRICAL WORLD postpaid, $1.00.
The W. J. Johnston Company, Publishers,
253 BROADWAY, NEW YORK.
RETURN TO the circulation desk of any
University of California Library
or to the
NORTHERN REGIONAL LIBRARY FACILITY
Bldg. 400, Richmond Field Station
University of California
Richmond, CA 94804-4698
ALL BOOKS MAY BE RECALLED AFTER 7 DAYS
2-month loans may be renewed by calling
(415) 642-6753
1-year loans may be recharged by bringing books
to NRLF
Renewals and recharges may be made 4 days
prior to due date
DUE AS STAMPED BELOW
— MAY ^21991
JUL 81992
M289308
THE UNIVERSITY OF CALIFORNIA LIBRARY
r*