Copyright
International Engineering Congress (1901 : Glasgow.

Report of the proceedings and abstracts of the papers read online

. (page 3 of 37)
Online LibraryInternational Engineering Congress (1901 : GlasgowReport of the proceedings and abstracts of the papers read → online text (page 3 of 37)
Font size
QR-code for this ebook


Glasgow Iron and Steel Co., Wishaw.
Glebe Sugar Refining Co., Greenock.
Hyde Park Foundry Co., 54 Finnieston Street.
King, David, & Sons, Manufacturers of Electrical Castings and

Sanitary Appliances, Keppoch Iron Works, Possilpark.
Lang, John, & Sons, Machine-tool Makers, Johnstone.
Lindsay, Burnet & Co., Moore Park Boiler Works, Helen Street,

Govan.

Lloyds Proving House, 82 St. James Street, Kinning Park.
Lobnitz & Co., Ltd., Engineers and Shipbuilders, Renfrew.



MINUTES. 27



London & Glasgow Engineering & Shipbuilding Co., Ltd.,

Govan.
M'Dowall, John, & Son, Saw Mill Engineers, Walldnshaw

Foundry, Johnstone.
M'Farlane, Strang & Co., Iron Pipe Founders, Lochburn Iron

Works.

M'Onie, Harvey & Co., Engineers, 224 West Street, South Side.
M'Millan, Archd., & Son, Ltd., Shipbuilders, Dumbarton.
M'Xeil, John & Co., Engineers, Helen Street, Govan.
Mackie & Thomson, Shipbuilders, Govan.
Martin, Hugh, & Sons, Coatbridge.
Martin & Miller, Tanners, 847 Duke Street.
Mechan & Sons, Engineers, Scotstoun.
Miller, A. & T., Globe Iron Works, Motherwell.
Milne, Jas., & Son, Engineers, Milton House Works, Edinburgh.
Mirrlees, Watson & Co., Scotland Street Iron Works (Afternoons

only).

Muir & Houston, Ltd., Engineers, Kinning Park.
Napier & Miller, Ltd., Shipbuilders, Yoker.

Napier Bros., Windlass Engine Works, 100 Hyde Park Street,
Outfall Sewer and Pumping Station, Dumbarton Road Bridge,

Partick.
Penman & Co., Boilermakers, Caledonian Iron Works, Strath -

clyde.

Ross & Duncan, Engineers, Govan.
Rowan, David & Co., Engineers, 231 Elliott Street.
Russell & Co., Shipbuilders, Port-Glasgow.
Scott & Co., Shipbuilders, Greenock.
Scottish Cold Storage Co., 219 George Street.
Scottish Co-operative Wholesale Society, Ltd., Works, Shield-
hall, Govan.
Shanks & Co., Ltd., Manufacturers of Sanitary Appliances,

Tubal Works, Barrhead.
Simons & Co., Shipbuilders, Renfrew.
Smith, "A. & W., & Co., Eglinton Engine Works, 57 Cook

Street.

Smith, Hugh & Co., Possil Engine W T orks, off Possil Road.
Spencer, John, Ltd., Phoenix Iron Works, Coatbridge.
Stephen, Alexander, & Co., Shipbuilders, Linthouse.
Steven & Struthers, Brassfounders and Engineers, Kelvinhaugh.
Stewart & Menzies, A. & J., Clydesdale Steel Works, Mossend.
Stewart, Duncan & Co., London Road Iron Works, Bridgeton.
Sterne, L., & Co., Engineers, Crown Iron Works, 156 North

Woodside Road.

Summerlee and Mossend Iron and Steel Co., Coatbridge.
Summerlee and Mossend Iron and Steel Co., Mossend.
Tullis, John, & Son, St. Anne's Leather Belt Manufactory,

Bridgeton.
Thornliebank Co., Ltd. (The Calico Printers' Association, Ltd.),

Thornliebank (closed 2-3 daily).

Ure, John, & Son, Regent Flour Mills, Sandyford.
Wemyss Bay Railway Widening, D. A. Matheson, Engineer in

Chief, Caledonian Railway, Buchanan Street Station.
Woodside Steel and Iron Co., Coatbridge.




PROCEEDINGS OF THE SECTIONS.

SUMMARY OF PROCEEDINGS

OF

Section I. Railways.*



TUESDAY, 3rd SEPTEMBER, 19O1.



Sir BENJAMIN BAKER, K.C.M.G., D.Sc., LL.D., F.R.S., in the Chair.



" UGANDA RAILWAY."

Paper by Sir GUILFORD MOLESWORTH, K.C.I. E.
Abstract.

THE Uganda Railway is instructive

ist. In showing the inferences that may be deduced from the
study or maps and books of travel.

2nd, As an example of an excellent reconnaissance based on
astronomical and barometrical observations.

3rd, As an instance of the combination of difficulties different
from those ordinarily encountered by the engineer.

In 1891 the author had to advise the I.B.E.A. Co. on the question
of railway communication with Lake Victoia. He had never been
in the country, which before 1888 was practically a terra incognita,
the only European who had succeeded in penetrating the country
being Mr. Joseph Thomson in his rapid and necessarily superficial
expedition through Masailand. What was known of the rest of
the region was the result of conjecture, or native reports, gathered
by missionaries. Stanley visited Lake Victoria via Congo, and
Fischer had in 1883 passed through German territory to the
Dogilani Plain and Navasha, In 1888 Jackson and Gedges
expedition passed via Machakos to Navasha, and thence via Stotik
to Lake Victoria. From these sources Ravenstein's map was
compiled; and from it, and from the records of Thomson's and

* The full proceedings of Section I. are published by Messrs. Wm.
Clowes & Sons, Ltd., Duke Street, Stamford Street, London, S.E. Price
55. 6d. post free.



UGANDA RAILWAY. 29

Jackson's expeditions published by the Royal Geographical Society,
the author gleaned the information on which his advice was based.
A map thus compiled must necessarily be sketchy and in points
inaccurate; but, notwithstanding these defects, it afforded valuable
information. Some idea of its inaccuracy may be inferred by the
results of recent surveys near the mouth of the Nyando.

Little information was given about the escarpments which
bounded the great rift that traversed the country. There were
no records of any European having visited either the Man Plateau
or the Valley of the Nyando.

After careful study of the sources of information, he submitted
to the I.B.E.A. Company a sketch map, on which he had marked
the line of reconnaisance which he recommended for first trial,
giving also the reasons for his advice, which may be summarised
as follows :

1. A typical section in a straight line from coast to lake was
assumed.

2. A great volcanic rift existed, at least 20 miles in breadth, with
escarpments 1500 to 2000 feet high.

3. A chain of lakes indicated that the rift extends throughout
British territory, and therefore cannot be avoided.

4. A longitudinal section of the rift and its escarpments was
assumed.

5. Close to the coast the Rabbai Hills, 700 feet high, had to be
rounded.

6. Voi was an obligatory point for purposes of water supply.

7. From Rabai Hills the land rises steadily to 5000 feet at the
rift.

8. The Tsavo River should be crossed between its confluence
with the Sabaki and the River Mbololo.

9. Mackakos must be avoided either by the Athi Valley or an
alternative route.

10. The ramifications of the Athi River indicated the probability
of a low point in the escarpment, and the best approach to the
rift near Ngongo.

11. The descent of the eastern escarpment should run in the
direction of the rising rift floor.

12. The line should pass along by Lakes Navasha and Elmenteita
to the culminating point at Nakuro.

13. An easy line would be obtained in the rift floor at this part.

14. The best point for ascending Mau escarpment was at Lake
Nakuro.

15. The ascent should run in the direction of the fall of the
escarpment.

1 6. A railway by Jackson's route through Sotik was
impracticable.



30 UGANDA RAILWAY.

17. The only probability of a favourable line descending to Lake
Victoria was by Mau Plateau and the Nyando Valley.

1 8. A line via Nzoia River would involve a considerable detour
and broken ground.

19. Beyond Ngongo, excepting the portion in the rift floor, the
line must be difficult and costly.

Macdonald's expedition in 1891-92 entirely confirmed these
inferences, with one exception, the main point of difference being
that the route via Nzoia was followed instead of the Nyando, which
was considered impracticable. This change involved a detour
of about 100 miles, but when the permanent survey was
made in 1898 it was discovered that the Nyando Valley was quite
practicable, and the railway is now being made through it.

Macdonald's reconnaisance was very ably made by compass,
pedometer, and aneroid barometer. The cross sectional slopes
of the country were taken by Abney's level. Corrections were made
for the diurnal barometic wave, which is very important in the
tropics. Plans and sections were plotted in camp each day, and
linked in by triangulation where feasible ; otherwise by astronomical
observation. The position each day was checked either by
latitude and longitude with chronometer, or by longitude from
occupations. Notes were taken of the dimensions, slopes, flood-
marks, soil in bed and banks, all waterways, and of the general
physical and geological features of the country.

The difficulties encountered in the construction were very great.
A port had to be established, with jetties, moorings, cranes, steam
launch and lighters, and connected with the terminus by
a short railway with a gradient of i in 50. Store sheds
and workshops had to be built, labourers housed, nearly all
the labour had to be imported from India, many subordinates
obtained in India or locally were incapable or inebriates, those
sent from England were satisfactory. The staff was new to the
work, the language, and each other. No supplies were available
in the country; even poles and thatch for coolie sheds had to be
imported. Native raids necessitated military escort for the first
survey parties. The construction involved an organisation
equivalent to the .maintenance of an army of 15,000 men in a
practically waterless country devoid of resources, and of all means
of animal or wheeled transport, with a base of operations to which
everything had to be imported from a distant country. Large
condensing plant was needed to supplement the water supply, and
a corn mill to grind the imported food. The line had to be
constructed telescopically, and it was impossible to maintain
working parties far in advance of railhead. Separate water trains
had to be run, and locomotives had to take a heavy water tank to
supplement the tender. Heavy temporary works were necessary



UGANDA RAILWAY. 31

to expedite the progress of railhead; 34^ miles of temporary
diversions were needed for the first 300 miles ; amongst these being
the Macupa Bridge and the Mazeras Viaduct, built in 91 and
25 working days respectively. The ruling gradients on these
diversions was i in 30, with curves 400 feet radius; these limited
the power of the engines. On one temporary diversion the descent
to the rift was made by four rope inclines with a maximum gradient
of i in 2, making a total descent of 15,000 feet with a length of
6000. The engineering strike in England delayed the supply of
locomotives, rolling stock, and bridges. The first 250 miles were
infested with tsetse fly, fatal to transport animals; nearly all of
those imported died. Jiggers abounded, causing ulcers, which
often necessitated amputation of one or more toes. Man-eating lions
killed 28 of the Indian labourers, and caused a panic. Waves of
fever passed over the country, and at one station 90 per cent, of a
working party were down with it. It was necessary to organise
an agency in India for labour and materials, a postal service with
regular mails, a force of 200 police, complete hospital staff, a
temporary telegraph beyond railhead; and a small steamer
had to be carried piecemeal by porters to the lake. The
viaducts over the deep ravines in the descent into the rift
had to be constructed telescopically. The responsibility for
the whole of this organisation rested on the chief engineer, and
very great credit is due to him and his staff for the able manner
in which these difficulties hav.e been met.



Mr. A. E. Welby and Mr. Wigham Richardson took part in the
Discussion.

The author replied, and on the motion of the Chairman a vote
of thanks was accorded to him.

Mr. A. E. Welby, at the Chairman's request, contributed some
additional notes on the paper.



Mr. ALEXANDER Ross, Vice-Chairman, in. the Chair.

THE ECONOMY OF ELECTRICITY AS A MOTIVE
POWER ON RAILWAYS AT PRESENT DRIVEN BY
STEAM.

Paper by Professor C. A. CARUS-WILSON, M.A.



Abstract.

THE paper deals with the economic considerations which will
probably govern the substitution of electricity for steam as a motive
power on railways.

Branch or cross country lines are the least profitable part
of present railway systems, and in many cases the receipts
barely cover expenses. The competition of electric tram
lines now being built throughout the country will still
further accentuate the unremunerative character of branch
lines. With steam traction it is necessary to make up long
trains, so that on branch lines with little traffic the interval
between trains is large and entails delay in making connections with
main line stations. This infrequency of service causes unpunctuality,
as the limited traffic does not warrant the employment of a staff
adequate to cope rapidly with long trains heavily laden with pas-
sengers and luggage, which come in at infrequent intervals. This
need not necessarily be the case if the traffic were evenly distributed
over the working day, as the existing staff would be able to cope
with a considerable increase of passenger traffic. By breaking up
the train service on branch lines into> smaller units moving more
frequently, cross-country travel would be greatly facilitated.

It is thereore necessary to ascertain upon what the cost of any
increase of train service depends, so as to deduce the minimum
traffic required to pay for it. To do this with steam railways, the
running expenses, such as coal, drivers' and conductors' wages, etc.,
per train-mile, which vary with the number of trains run, must be
separated from the fixed expenses, such as maintenance of way,
traffic expenses, etc., which do not so vary.

The fixed expenses per train-mile, multiplied by the number of
trains per day on any given line, will then give the contribution
of that line- per day-mile to the general fund for maintenance. This
constitutes a fixed sum per day-mile which must be provided for



ECONOMY OF ELECTRICITY AS A MOTIVE POWER 33

under the new conditions, together with the increased running ex-
penses. The traffic per day-mile must exceed this amount, plus
a sum required to pay interest on the electric installation, before
the line can be said to pay.

An analysis of the Hoard of Trade returns* of the working
expenses of the principal English railways for 1900 shows that the
fixed expenses increase when the proportion of passenger traffic
to goods traffic is increased. Thus, on the Midland Railway, where
the passenger train-mileage is 40 per cent, of the whole, the fixed
expenses are only 22.6 pence per train mile; whereas with the Great
Western and Great Northern Railways, where the goods and pas-
senger train-miles are equal, the fixed expenses vary from 23d. to

2 5 d.

On the other hand the item of running expenses remains fairly
constant for all the principal lines, despite the difference in the
proportion of passenger and goods traffic. Thus, the Midland
Railway, with 60 per cent, of goods train-miles and 1.45 tons per
train-mile, has the same running expenses as the Lancashire and
Yorkshire Railway with 35 per cent, of goods train-miles and 3.33
tons per mile. An exception occurs in the case of the London and
Brighton and South Eastern Railways owing to the high price they
had to pay for coal last year. The analysis demonstrates that the
running expenses do not rise above the average unless there is a
very large proportion of heavy goods traffic.

In comparing steam with electric traction, we may assume the
case of a branch line with six steam trains each way per day.
Taking the fixed and running expenses of a normal line like the
Great Northern for the purpose of illustration, the running expenses
will be 12 x n.85d. = 1428., and the fixed expenses will be
12 x 2i.38d. = 256d., per day-mile. If the line is to contribute to
the general revenues a sum proportional to the trains run arid to
the average cost per train mile for the whole of the line, the
receipts per day-mile must equal 398d.

Instead of the steam train running every two hours we may sub-
stitute an electric train, composed of motor-driven cars with ordinary
carriages trailing, running every half-hour, but with a quarter of
the seating accommodation. About 20 per cent, dead weight is
saved by dispensing with the locomotive ; and, as the weight of the
carriages is only a quarter, the new trains will weigh one-fifth of
the old ones. This reduces the coal item in the running expenses
to o.68d.,t and water, oil, etc.. to o. i.^d., as against 3-36d.t and
0.7 yd. respectively for steam trains. The experience of the City
and South London Railway shows that the cost of wages and

* This Table is given in the Paper.
{The price of coal is taken at 8s. per ton.



34 ECONOMY OF ELECTRICITY AS A MOTIVE POWER.

materials for repairs is halved, bringing these items down to 0.6 yd.
and o.52d. for electric railways. The simplicity of the electric
equipment makes it possible to substitute one motor-man for the
highly-paid driver and fireman; so that the item of wages on the
locomotive is also halved. The electric motor car is ready to start at
any time, and a larger proportion of actual working hours can be
usefully employed; so that the men can put in about 50 per cent
more train-miles, thus reducing the wages item to 2.2$d. To this
must be added the wages of the men at the generating station,
estimated at 0.6 2 d., or half the motor-man's wages, thus making
the wages per train-mile altogether 2. 8 yd. The total cost per train-
mile for running expenses for electric traction is therefore i.&yd.,
as against n.85d. for steam traction.

With electric traction the fixed expenses would be the same as
with steam traction, but the running expenses would increase with
the frequency of the service. In the case assumed, with trains every
half-hour, or twenty-four each way per day, the running expenses
would be 48 x 4.89d. 24od., and the fixed expenses being as
before, 256d., the total expenses per day-mile would come out at
496d. In order to pay expenses the receipts would have to increase
from 398d. to 496d. per day-mile, or about 25 per cent. This,
however, would not pay the interest on the capital required for
the electrical equipment. The generating station, rolling stock,,
and distributing system for a half-hour's service of 40-ton trains on
a line 15 miles in length, would probably be about ^8000 per
mile, which at 3^ per cent, interest would require additional receipts
of i84d. per day-mile. The total increase of traffic required to-
pay all expenses and interest would therefore be about 70 per cent.
Assuming that a fourfold increase in the number of trains per day
were to double the traffic, the profits per day-mile would be IDS. ;
if the traffic were trebled the profits would be 435.

The profits per day mile on the whole of the Great Northern
system average 1245. ; so that the adoption of electricity on branch
lines is worth considering as a means of making them contribute a
more substantial proportion of the total profits than they do at
present.



Mr. Hurry Riches and Sir Douglas Fox took part in the Discus-
sion. The author replied, and has also replied by correspondence.

On the motion of the Chairman a vote of thanks was accorded
to the author.

The meeting was then adjourned.



WEDNESDAY, 4th SEPTEMBER, 19O1.



Mr. ALEXANDER Ross, Vice-Chairman, in the Chair.

"MODERN PRACTICE IN RAILWAY SIGNALLING."
Paper by I. A. TIMMIS.



Abstract.

THE changes which steam effected when it came into use as an
aid to more rapid movement of people and material on land and
water, created an ever-increasing desire and want for more perfect
and faster means for effecting that movement. And, now that
another force of nature electricity has come to the aid of steam,
the growth of railways has developed enormously, and the desire
and necessity for intercommunication in all countries has not only
increased, but must go on increasing; as a consequence, the
engineers of railways are obliged to fit new signalling systems in
order to deal with the larger stations, greater number of main lines
and sidings, larger cabins, the increase in number of trains
and higher speeds. It has become necessary to place the points
and danger signals at a greater distance from the cabins. The
result of these altered conditions is that some other power is required
to take the place of manual. Three systems have been tried
hydraulic, pneumatic, and electric.

HYDRAULIC. The experience gained from signal work operated
by this system proves that it cannot compete with the pneumatic
and electric systems, and so the author did not deem it advisable to
take up time in describing it.

PNEUMATIC. There are two systems that use air:
i. The Westinghouse High Pressure. In the first installations
that were fitted in the United States the signals and points were
operated by the air conveyed through a main supply pipe and its
branches to a cylinder on each signal post and at each pair of
points, and the air was admitted to the cylinders by hydraulic
power, which was put into action at the signal cabin by the signal
man moving a small lever. The hydraulic pressure acted on a
valve, which admitted the air into the cylinder, and moved a piston.
But a later development introduced an electric current as the
controlling agent. The levers in the cabin are interlocked; when
a signal lever is pulled over, an electric current is sent to an electro



36 MODERN PRACTICE IN RAILWAY SIGNALLING.

magnet on the signal post, which compresses a spring, closes th?
exhaust port, and opens the high pressure air admission valve.
The piston in the cylinder then lowers the signal. When the
electric current is interrupted the spring closes the valve, opens the
exhaust, pushes back the armature, and the counter weight puts
the signal to " danger." When a point lever is moved in either
direction the operation of the points, in each direction, is practically
as described for the signals. Thus there is a magnet controlling
each end of the point cylinder with one slide valve. But there
is a third magnet to lock the slide valve, and in addition it breaks
and changes the electric circuit and sends an indication current
back to the signal cabin when the points are over and locked, and
this current operates an electro magnet in the cabin, which enables
the signalman to lower the necessary signals.

2. The Low Pressure Pneumatic. This system is altogether on
different lines from the high pressure. The operating is effected
by air at i5lbs. pressure, and the controlling by air at half that
pressure.

To operate points the lever is pulled over half way, and is then
stopped. The controlling current goes to the points and admits
the higher pressure air into a cylinder, when the points are moved
and locked, and a return indication is sent to the cabin, which
releases the lever and completes its throw. The movement of
the points and the locking bar and locking bolt are effected by a
plate or flat bar with grooves and studs in it. There are four
pipes to work each signal main supply, two controllers, and one
return and there are five pipes to a pair of points.

Both the above systems can be fitted to work with a track
circuit, but this involves the use of electricity, and adds consider-
ably to complication of detail.

ELECTRIC. In the United States a system is fitted by the Union
Switch and Signal Co., where the power is supplied from primary
batteries, and each signal is lowered to " line clear " by a small
motor geared 1000 to i. An electro magnet then holds the signal,
and the motor is cut out. When the circuit is broken the signal
goes to " danger " automatically.

Another system, fitted by the Taylor Co., uses secondary
batteries, and the signals are operated practically in the same way
as just described. Points are also worked by motors geared 20
to i to the driving wheel. The first quarter revolution of the
wheel unlocks the points, and the last quarter locks them and
closes the return indication circuit to the cabin, and reverses the
connections for a reverse movement. Interlocking is effected in
the cabins in the lever frame.

In this country the first practical system fitted was on the
Liverpool Overhead Railway. This is an automatic system, and



MODERN PRACTICE IN RAILWAY SIGNALLING. 37

of course only works the signals. A full description is in
"Engineering" of February loth, 1893. As a train leaves a
station it puts the starting signal to danger by means of a striking
bar fitted to the rear vehicle operating a breaking contact; and
when the train is a suitable distance ahead of the signal the same
bar operates a making contact which closes a cuircuit. This
circuit is completed by the signal just passed being at " danger,"
and then the signals in the rear block are lowered automatically



Online LibraryInternational Engineering Congress (1901 : GlasgowReport of the proceedings and abstracts of the papers read → online text (page 3 of 37)