PART TWO: POWER AND ENGINEERING
362
‘by eye’. The detailed design of the field magnets and coils was studied by John
Hopkinson, Professor of Electrical Engineering at King’s College London, and
a consultant to the American electrical inventor Thomas Alva Edison. Seeking
to introduce a complete electric lighting system, with lamps, generators and
other equipment all of his own design, Edison had been making generators
with very long field poles and coils. Hopkinson made a number of small
models of field systems of different shapes, and measured the magnetic field
produced at the armature. As a result he concluded that Edison’s poles were
much too long, and he designed a fairly squat machine whose general
proportions were followed by many manufacturers.
The machines described above were for direct current. A different pattern
was adopted for the early alternating current generators. These normally had
the armature coils arranged around the edge of a fairly thin disc and moving
between the poles of a multi-polar field system. This design gave a machine
whose reactance was low—important on an alternating current system—and
allowed a sufficient number of poles to be used. It was essential to use
multipolar machines if the generator were to be coupled directly to a steam
engine. Even the fastest reciprocating engines ran at only about 500rpm. A
twelve-pole generator running at that speed would give a 50Hz output. In
practice supply frequencies varied from 16.6 to 100Hz. The disadvantage of
the disc generator was that it was impossible to make such a machine for three-
phase operation. However, before three-phase supplies came into general use
the turbine had replaced the reciprocating steam engine, and generators were
being designed for the higher running speed of the turbine.
ARC LIGHTING
Although the possibility of electric arc lighting had been demonstrated very early
in the nineteenth century, it could not be a practical proposition until a supply of
electricity was readily available. The development of satisfactory generators in
the 1870s stimulated fresh interest in the possibility of electric lighting.

In an electric arc lamp two carbon rods are connected to the opposite
poles of the supply. The rods are briefly touched together and then drawn a
few millimetres apart. This draws a spark, or ‘arc’, which continues as long
as the electricity supply is maintained. The current in the arc produces
considerable heat, and the contact points on the carbons quickly become
white hot. These white hot places on the carbons are the source of the light.
White hot carbon burns in air, and so some arrangement is necessary to feed
the carbons closer together so that the gap is kept constant. Without any
adjustment the gap widens and within a minute or two the electricity supply
will be unable to maintain the arc across the wider gap and the lamp will be
extinguished.
ELECTRICITY
363
The first arc lamps were manually adjusted, and for applications such as
theatre spotlights there was no problem about having a man constantly in
attendance. For general lighting, however, the arc would only be acceptable if
some reliable method could be found for adjusting the carbons automatically.
Much inventive ingenuity went into devising suitable mechanisms for the
purpose.
The first arc lamps to be used in quantity for general lighting were known
as Jablochkoff candles. Paul Jablochkoff was a Russian telegraph engineer
who set out from Russia in 1875 intending to visit the United States
centennial exhibition at Philadelphia in 1876. He only got as far as Paris,
where he became interested in electric lighting, and it was in Paris that he
invented his ‘candle’. Jablochkoff s candle consisted of two parallel carbon
rods placed side by side but separated by a thin layer of plaster of Paris. At
one end the rods terminated in brass tubes which secured the candle in a
holder and made the electrical connections, at the other end they were joined
by a thin piece of graphite. When the candle was in its holder and the
current was switched on, the graphite fused, starting an arc between the ends

of the carbons. As the carbons burned the plaster crumbled away in the heat,
exposing fresh carbon. Provided the candle had been well made the carbon
and the plaster were consumed at the same rate and the result was a steady
light. However, once the light was extinguished, for whatever reason, it could
not be restarted. For street lighting this did not necessarily matter: the candle
would last for an evening, and during the next day a man could go round
putting in fresh candles. Automatic mechanisms were made which brought a
new candle into the circuit when the first was extinguished, but the candle
itself soon became obsolete as regulating mechanisms were devised which
could be mass produced.
The Jablochkoff candle was only used for a few years, but it was used
prominently. It was first installed in Paris, attracting much attention. In July
1878 the London technical journal The Electrician complained that ‘The
application of the electric light is in Paris daily extending, yet in London there
is not one such light to be seen.’ In October the same year the Metropolitan
Board of Works arranged a trial of electric lighting, using Jablochkoff candles,
on the Victoria Embankment. About the same time the City of London
authorities arranged some trials in front of the Mansion House, on Holborn
Viaduct and in Billingsgate fish market. All these installations were working by
Christmas 1878.
Arc lamp regulators have to perform two distinct functions. First the carbon
rods must be brought together and then drawn apart when the current is
turned on, and secondly the spacing of the rods must be maintained. The first
function was quite easy to achieve: the upper carbon was allowed to fall under
gravity and make contact with the lower. An electromagnet connected in series
with the lamp then pulled the lower carbon down a few millimetres to start the
PART TWO: POWER AND ENGINEERING
364
arc. The rate of fall of the upper carbon was controlled by a brake, and the
second function of the mechanism was to control the brake. The smooth

operation of the lamp depended on the brake, and any sudden movement of
the upper carbon would cause the light to flicker. Most of the earlier lamps
used the series electromagnet to control the brake. This was easily arranged: as
the gap widened and the arc lengthened the current would fall, and the
electromagnet, which had been holding the brake on, would weaken, releasing
the brake until the gap was restored (see Figure 6.7).
The disadvantage of using the series electromagnet to control the arc lamp was
that only one lamp could be connected to the supply. If two lamps were connected
in series to one generator, then a fall in current due to one gap widening would
affect both regulators, and the brake on one lamp would be released too soon. (Arc
lamps will not operate in parallel, either, because the arc is a negative resistance
and one lamp would take all the current while the other went out.)
More satisfactory arc lamps were made by introducing another
electromagnet, connected in parallel with the arc. As the arc widens the current
falls, but the voltage across it increases. The parallel electromagnet therefore
became stronger as the gap widened, and it was used to release the brake
which was normally held on by a spring. Most arc lamps from the mid-1880s
onwards worked in this way. Many designs were made, with the object of
producing a cheap, reliable, yet sensitive control mechanism.
Arc lighting was widely adopted in places where it was suitable—in large
buildings like markets and railway stations and for street lighting. King’s Cross
Figure 6.7: Compton arc lamp mechanism of about 1878.
ELECTRICITY
365
Station in London, for example, was lit by twelve lamps in 1882. The lamps, rated
at 4000 candle power, were hung ten metres above the platforms and supplied
from four Crompton-Burgin generators all driven by a single steam engine.
Two improvements in arc lighting were made during the 1890s. One was
the enclosed arc, which had the arc contained within a small glass tube that
restricted the air flow. The effect of this was to reduce the rate of burning of

the carbons. The second improvement was the addition of cores of flame-
producing salts, mainly fluorides of magnesium, calcium, barium and
strontium, to the carbon rods. They increased the light output, and also gave
some control over the colour of the light.
In 1890 there were reported to be 700 arc lamps in use for street lighting in
Britain, and probably a similar number in private use. About 20,000 were
installed in the following twenty years, but by then the filament lamp had been
developed to an efficiency at least equal to that of the arc. Although few, if any,
further arc lamps were installed, those that were already in place continued in
use. London retained some arc street lighting into the 1950s.
THE FILAMENT LAMP
There was no call for a public electricity supply until the invention of a
satisfactory filament lamp. Electric arc lighting was proving its worth for streets
and public buildings, but it was quite unsuited to domestic use. The individual
lamps were far too bright, too complex, and too large physically for use in the
home, and they would probably also have been a fire hazard.
The idea of the filament lamp was almost as old as the arc lamp. Many
people had tried to produce light by heating a fine wire electrically so that it
glowed, but they all faced a series of seemingly insuperable problems. They
had to find a material that would stand being heated repeatedly to white heat
and then cooled, then it had to be sealed into a glass vessel in such a way that
the glass did not crack when the wire was hot, and finally the air had to be
pumped out so that the filament did not oxidize.
The early attempts at making a practical incandescent filament lamp all
failed, mainly because of the difficulty of obtaining an adequate vacuum. After
the invention of the Sprengel mercury pump in the mid-1870s several inventors
succeeded in making viable lamps, the best known being Edison and Swan. At
the first International Electrical Exhibition, held in Paris in 1881, four
manufacturers had filament lamps on display: Swan and Lane-Fox from Britain
and Edison and Maxim from the USA. There was little to choose between the

four lamps at the exhibition, though Swan and Edison soon captured the
market while the others disappeared from the scene.
Swan and Edison were very different men, with different approaches to
their common objective of developing an electric light suitable for domestic
PART TWO: POWER AND ENGINEERING
366
use. As Sir James Swinburne later remarked, ‘Edison and Swan were hardly
racing, as they were on different roads.’
Joseph Wilson Swan, later Sir Joseph, was born in Sunderland and
apprenticed to a firm of druggists there. Subsequently he set up in business
with his friend John Mawson as chemists and druggists in Newcastle upon
Tyne. A man of wide scientific interests, he was intrigued by some of the early
experiments towards an incandescent filament lamp, and made several
experimental lamps himself in the 1840s and 1850s. He made carbon filaments
from strips of paper treated with sulphuric acid to give a very smooth material
called ‘parchmentized paper’ because of its resemblance to parchment. He
carbonized the paper and mounted it in a glass vessel closed with a rubber
stopper and then evacuated. However, he could not obtain a sufficient
vacuum, and left the experiments for twenty years during which time he
worked on other matters, especially photography: the most important of his
inventions in that field was the silver bromide photographic paper still used for
black-and-white prints. In 1867, John Mawson, who was by then his brother-in-
law as well as close friend and colleague, was killed in an accident. Swan found
himself responsible for the families and for a large chemical business.
About 1877, Swan was able to resume his interest in electric lighting, and the new
Sprengel airpump gave him fresh impetus. He first spoke publicly about it on 19
December 1878, at an informal meeting of the Newcastle Chemical Society when
several members gave brief talks. It seems probable that he did not have a working
lamp to demonstrate then, but he certainly did so at several public meetings in the
area in January to March 1879. During 1879 he worked to improve his lamp. The

main problem, which was discussed at another Newcastle Chemical Society meeting
on 18 December 1879, was that residual gas was occluded in the carbon and came
out when the filament became hot. This gas carried particles of carbon which were
deposited on the cooler glass, causing blackening. The solution was to pump the
lamps while the filament was hot, and Swan applied for a patent for that process on 2
January 1880. He never tried to patent the basic idea of a carbon filament lamp,
since he considered that there was nothing novel in that. During 1880 he worked on
the filament material. While studying the problems of evacuating and sealing the
lamp he had used mainly thin arc-lamp carbons for his filament—they were available
only about one millimetre in diameter, and were relatively strong. The carbonized
paper and carbonized parchmentized paper were not entirely satisfactory. He tried
other substances and found a suitable material in parchmentized cotton, which was
ordinary cotton thread treated with sulphuric acid, as the parchmentized paper had
been. This gave a compact and uniform material which could be carbonized in a
furnace to give satisfactory filaments. Swan applied for a patent for this filament on
27 November 1880, and went into commercial production in 1881 (see Figure 6.8).
He never sought to manufacture other components for an electric lighting system,
although he worked closely with Crompton who was making generating equipment
and arranging electrical installations.
ELECTRICITY
367
Edison’s approach was quite different. He had already made a name for
himself as an electrical inventor, and had built up a research organization. He
became interested in electric lighting late in 1877, after visiting William
Wallace’s electrical factory in Connecticut. Wallace made arc lighting
equipment. Edison thought that a viable electric lighting system should have
lamps of about the same power as the gas jets then in use and that electricity
should be distributed in a similar way to gas, with each light being
independently controlled. He wanted to produce both the lamps and the
electricity supply system to feed them, and all the resources of his laboratory

and staff were turned to the subject.
Edison’s friend, the lawyer Grosvenor P.Lowrey, put up $300,000 to
establish the Edison Electric Light Company in October 1878. It was in
Figure 6.8: Advertising drawing published by the Ediswan company in the late
1930s, showing Swan’s first successful filament lamp of 1881.
PART TWO: POWER AND ENGINEERING
368
that month that Edison announced publicly: ‘I have just solved the
problem of the subdivision of the electric light.’ The search for a viable
filament lamp was often called the problem of ‘subdividing’ the electric
light because the perceived need was for a much smaller lighting unit
than the arc light. Edison’s announcement, which received wide publicity
and caused an immediate slump in gas shares, was based on a lamp with
a platinum filament. This lamp contained a thermostat which
momentarily cut off the current when the filament was in danger of
overheating and melting. It was not until late 1879 that Edison turned
again to carbon as his filament material. When he began commercial
production of filament lamps in 1880 he used filaments made from
Bristol-board—a thick paper with a very uniform texture. Continuing the
search for better materials, however, he found that fibres from a
particular variety of bamboo gave the best results, and used that from
mid-1881 to about 1894.
Several other people made workable filament lamps, and at least two of them
went into commercial production. Hiram S.Maxim, who is better known now
for his work on guns and on aerial navigation, was an American by birth, but
later became a naturalized Briton and was knighted. The other was an
Englishman, St George Lane-Fox, who also designed a complete distribution
system. His patents were acquired by the Anglo-American Brush Electric Light
Corporation.
The manufacture of Swan’s first commercial lamps was a complex

enterprise. The ladies of the Swan household in Newcastle upon Tyne
prepared the filaments and Swan himself carbonized them. The bulbs were
blown by Fred Topham and all the components were conveyed to Birkenhead
where C.H. Stearn mounted the filament assemblies in the bulbs and
evacuated them.
A catalogue published by the Swan United Electric Light Company in 1883
lists more than a hundred houses and other buildings and twenty-five ships lit
with Swan’s lamps. Probably the most prestigious contract was for lighting the
new Law Courts in London, which opened in December 1882. Crompton
supplied the generators and six arc lamps for the large hall, and Swan supplied
filament lamps for the courts and other rooms.
A number of large private houses were lit electrically, using current
supplied from their own generating plant. Sir William Thomson lit his
house in Glasgow. In a letter to Sir William Preece, who was lighting his
house in Wimbledon, Thomson noted the need for lampshades. ‘The
high incandescence required for good economy is too dazzling and I
believe would be injurious to the eyes if unmitigated. I have found that
very fine silk paper round the globe spreads out the light quite
sufficiently to make it perfectly comfortable to the eye while consuming
but a small percentage of the light and Lady Thomson has accordingly
ELECTRICITY
369
made little silk-paper globes for nearly all our lights.’ He goes on to say
that there were 112 lights in the house.
CENTRAL POWER STATIONS
The first central electricity generating station offering a supply of electricity to
the general public was probably the one that began operating at Godalming,
Surrey, in the autumn of 1881.
Until that year the streets of Godalming were lit by gas, under a contract
that expired at the end of September. In 1881 the town council and the gas

company were unable to agree on the price to be charged for the coming
winter’s street lighting. An influential figure in the town was John Pullman,
of R. & J.Pullman, leather dressers, who had a business based at Westbrook
Mill, on the River Wey. It was probably Pullman who suggested that the
town should have the new electric light rather than gas; he offered the use
of his waterwheel to drive a generator in exchange for free light at the mill.
The apparatus was soon installed. A Siemens generator at the mill supplied
seven arc lamps and about forty Swan filament lamps. A few of each type
of lamp were at the mill, the remaining arc lights were in the main streets of
the town and the filament lamps in the side streets. It was announced that
people who wanted electric lighting could have the wires taken into their
homes. Although the lighting created great local interest and was reported
in the local and national press, very few people took up the opportunity. In
May 1882, Sir William Siemens said there were only ‘eight or ten’ private
customers with a total of 57 lamps between them. The installation was
never a commercial success, though the Siemens company felt that they
learned useful practical lessons from it before, in 1884, the town reverted
to gas lighting.
The River Wey proved to be an insufficiently reliable source of power for
electric lighting, and within a few months the generator was removed from
Pullman’s mill and set up in the town centre, where it was driven by a steam
engine. That move also helped with the solution to another problem, volts-
drop in the wires. It was soon found that the voltage at the end of the supply
cables was less than the voltage at the generator, and that therefore if the
voltage was right close to the generator, then lamps at the other end of the
circuit only glowed dimly. This problem was reduced when the generator was
moved to the town centre, but it lead Sir William Siemens, in evidence to a
House of Commons Select Committee, to express the opinion that no
electricity supply station could be more than about half a mile from its most
distant customer.

The Select Committee was considering the Bill which became, in August
1882, the world’s first Electric Lighting Act. This Act laid down a general
PART TWO: POWER AND ENGINEERING
370
legislative framework for electricity supply, and was drafted on the assumption
that an electricity supply undertaking would be a fairly local matter, and that
the local authority should, if it so wished, have a large measure of overall
control. At least three other public supply systems were in operation before the
Act came into force. Edison had an experimental steam-powered station in
London, at Holborn Viaduct, which began supply in January 1882 and was
really a trial for his first New York station, which opened in September of the
same year. Crompton extended a street lighting and market lighting installation
in Norwich to supply private houses from about March 1882. Perhaps the
most important of these early schemes, however, was the one at Brighton,
Sussex. The electrical pioneer Robert Hammond had visited Brighton in
December 1881 to give an exhibition of Brush arc lighting. The demonstration
was so successful that Hammond was asked to extend it, and on 27 February
1882 he opened a permanent supply undertaking. Brighton has had an
electricity supply for longer than anywhere else, for all the other very early
undertakings closed within a short period.
A supply undertaking that did not use overhead wires and did not need to
break up the streets could operate outside the provisions of the Electric
Lighting Act, 1882. The largest company to use that loophole was the
Kensington Court Electric Light Company. Kensington Court was a new
housing development just south of Kensington High Street, in west London.
About a hundred houses on the estate were linked by a system of subways, in
which the company laid its mains.
Crompton was the leading figure in the Kensington Court Company,
which was registered in June 1886 and commenced supply in January 1887.
Initially they had only three customers, and by the end of the year there were

still only nine. Requests for a supply soon came from people outside the
estate, and the company obtained a licence from the Board of Trade to
increase their area of supply. The initial generating plant was rated at 35kW
(47hp), but additional plant was soon added, and by 1890 the generating
capacity was 550kW (738hp).
TRANSMISSION: AC v DC
Except for very small local systems, all supply undertakings had to solve the
problems of transmitting electricity at high voltage and then reducing and
stabilizing the voltage at a point near the customer. A great rivalry, the ‘battle
of the systems’, ensued between the proponents of alternating current and
those of direct current. The advantages of AC were that it was easy to change
the voltage up and down by means of transformers, and the voltage could be
adjusted by tap-changing on the transformers. If, however, it was desired to
maintain the supply day and night, then at least one generator had to be kept
ELECTRICITY
371
running all the time. With a DC system batteries could be used to maintain
the supply at times of low demand. Furthermore, DC generators could easily
be operated in parallel: parallel operation of AC machines was always difficult.
Another disadvantage with AC, initially, was the lack of a practical electric
motor, but after induction motors were developed around 1890, AC supplies
became attractive to potential industrial customers.
The leading pioneer of AC electricity supply was Sebastian Ziani de
Ferranti, a Liverpudlian of Italian extraction. At the age of 17 he was working
for Siemens Brothers at Charlton, but he soon branched out on his own. His
first company, Ferranti, Thomson and Ince Ltd, was formed in September
1882, with Alfred Thomson, another engineer, and Francis Ince, a lawyer, to
manufacture generators; the following year the firm was dissolved and
Ferranti, still only 19 years old, set up in business on his own, manufacturing
generators, meters and other equipment in London.

Within a few years he was chief engineer of the Grosvenor Gallery
Company, which had sought his help with technical problems. This
company had been formed by Sir Coutts Lindsay to light his art gallery in
New Bond Street. Neighbours had been impressed and sought a supply, with
the result that the company was soon supplying electricity over a substantial
area of Westminster and surrounding districts. They employed the Gaulard
and Gibbs distribution system in which electricity was distributed at high
voltage to transformers (known as ‘secondary generators’) at or near to each
customer. The primary windings of all the transformers were in series and
the current in the system was maintained constant at 10 amps. Individual
loads were fed from secondary windings on the transformers, and the
transformer design was such that the secondary voltages were roughly
constant whatever the load.
As the Grosvenor Gallery Company’s network expanded, every additional
secondary generator required an increase in the circuit voltage, and the practical
limit was soon reached. When Ferranti took charge he rearranged the system for
parallel working, using distribution at 2400 volts and a transformer on each
consumer’s premises to reduce the voltage to 100. He also replaced the Siemens
generators originally used by machines of his own design, each able to supply
10,000 lamps of 10 candle power, which required about 350kW. The supply was
not metered: each customer paid £1 per 10 candle power lamp per year.
As business expanded still further a new company, the London Electric
Supply Corporation Ltd, was formed in 1887 with a capital of one million
pounds. Ferranti planned a massive power station on the banks of the River
Thames at Deptford. Land there was relatively cheap; there was easy access
for coal barges, and there was ample cooling water. He designed generators of
10,000hp (7460kW) and planned to transmit electricity into London at 10,000
volts. Such a pressure was quite unprecedented, and he had to design
everything himself, including the cables. When the Board of Trade questioned