Aberdeen to Ballater by BMU:
Notes on the Battery Railcar Experiment
Although battery-powered trains have been making headlines in the railway technical press in recent years it is often forgotten that there were earlier applications of battery technology to rail vehicles. Experiments with 'accumulator railcars' as they were called at the time, took place around 1890 in various countries including Belgium, France, Germany and Italy. Between 1955 and 1995 a fleet of Class ETA 150 railcars (designated Class 515 from 1968) was introduced in Germany using lead-acid batteries. There have also been important battery rail vehicle developments in Japan. Within the UK, the most significant application in the past involved the Scottish Region of British Railways and the North of Scotland Hydro Electric Board (NSHEB) in a joint development project and it is appropriate that we should remember this in 2018, sixty years after it came to fruition.
The story starts late in 1954 when a report of the Electricity Consultative Council for the North of Scotland (ECCNS) included a recommendation that experiments should be made with battery railcars on one or more branch lines within that geographical area which were then under threat of closure. However, the Scottish Area Board of the British Transport Commission (BTC) quickly concluded that battery railcars would not have economic advantages and that experiments were therefore not appropriate. Sir Robert Boothby MP then raised the matter in Parliament and the Joint Parliamentary Secretary to the Ministry of Transport responded with a statement to the effect that battery vehicles would be too costly because the weight of the batteries would necessitate use the use of expensive light-alloy materials for the vehicle body in order to keep the total weight within acceptable limits. The cost of electricity was another controversial topic since it was stated that the NSHEB might be prepared to participate in a project of this type if the batteries could be charged at 0.75d per unit. Pressure was applied to get the issue reconsidered and early in 1956 the Chairman of the ECCNS announced that the BTC had agreed to carry out experiments with a battery railcar on a Scottish branch line.
The formal announcement of the project came at a press conference in Glasgow on 15th October 1957, when Sir Ian Bolton, who chaired the Scottish Area Board of the BTC, announced that a battery-electric multiple unit (BMU) was being developed at the British Railways Carriage & Wagon Workshops at Cowlairs Works and would be introduced the following summer on the Deeside line between Aberdeen and Ballater. Sir Ian was accompanied at the press conference by other members of the Scottish Area Board of the BTC and also by the Rt Hon Thomas Johnston who was Chairman of the NSHEB at that time. Mr James Ness, who was General Manager of the Scottish Region of British Railways, was also present.
As announced, the plans would involve a 117-seat two-car unit to be used within a revised timetable in which the existing four daily return journeys would be increased to six, with three return journeys by the battery railcar and the other three continuing to use steam haulage. If enough additional passenger traffic was attracted to the Deeside branch as a result of the BMU experiment consideration might then be given to the introduction of similar services on other branch lines in Scotland.
The battery-electric multiple unit [1, 5]
The project was a collaborative one involving the NSHEB, along with the Scottish Area Board of the British Transport Commission and the Scottish Region of British Railways, Bruce Peebles Ltd
and Chloride Batteries Ltd. The agreement involved an undertaking by the NSHEB to supply energy at a maximum of 0.75d per unit over the first two years of the project, while Chloride Batteries Ltd would provide suitable batteries.
British Railways was to provide the vehicle and this was based on a two-car diesel multiple unit (DMU) of the 'lightweight' type being built at Derby Works
in 1956. The underframes of the DMU were modified to accommodate the 17 tons of batteries, supplied by Chloride Batteries Ltd under favourable terms. Design and installation of the charging plants at Aberdeen
involved the NSHEB and Bruce Peebles Ltd
. Traction and control equipment was sourced from Germany where there was previous experience of the design and use of battery rail vehicles. Traction motors were provided by Siemens-Schuckert and electrical control equipment by Schaltbau.
The motor coach and control trailer together provided a total of 12 first and 105 second class seats. The motor coach had two 100kW nose-suspended traction motors on one bogie, with power transmitted via single spur gearing. The batteries, with 216 lead acid cells in total gave an output voltage of 440 and a capacity of 1,070 Amp.hr. These were split between the two vehicles, with cradle arrangements that allowed the batteries to be partially withdrawn for inspection and maintenance (see Figure 1). The tare weight of the motor brake second vehicle (SC79998) was 37.5 tons (compared with 27 tons for the equivalent DMU version built at Derby6) and that of the driving trailer composite (SC79999) was 32.5 tons (compared with 21 tons for the DMU
). The design speed was 60mph maximum.
Control involved conventional series and parallel connections with three positions of field weakening. Facilities were included to allow either of the motors to be cut out if necessary. Smooth acceleration was achieved through the use of cam-operated contractors driven by an electric motor the speed of which depended on the current being supplied to the traction motors. Cooling-air for the traction motors came from inlet grills on the sides of the motor coach and this reached the motors through flexible ducts. Oerlikon-type air brakes were used and an electrically-driven air compressor was located under the control trailer, providing air not only for the brakes but also for the horn and the window wipers.
Heating of the two coaches was based on an oil-burning air heater with electrical control. The supply for this and also for lighting came from a 2.6kW motor generator set carried below the floor of the motor coach making these supplies independent of the main battery. An automatic voltage regulator was included to ensure a more or less constant electrical output for these heater and lighting supplies.
The driving compartments at each end of the two-coach unit contained a driver's desk with a master controller, situated on the driver's left. This had two handles one of which was used to choose the direction of movement while the other controlled the speed. In front of the driver there was a raised sloping desk on which were mounted a voltmeter, ammeter, speedometer, switches and indicating lamps for the control equipment. The brake valve was positioned on the driver's right, together with a separate panel with two brake gauges (Figure 2). To the driver’s left and mounted to the side, at about eye level, was a 24V switch panel for instrument lighting, cab lighting, destination lights and de-misting fan.
Charging facilities at Aberdeen Joint station were located adjacent to Platform No 1 and involved a 6600V, 3-phase, 50 Hz supply and glass bulb mercury arc rectifier (which was provided by Bruce Peebles Ltd
). At Ballater
station the equipment was similar but was based on a supply at 11,000V. The charging process was stopped automatically when the battery was fully charged and the systems could be used to top-up the batteries during lie-over periods at Aberdeen
. Safety interlocking systems ensured that cables could not be connected or disconnected while they were live and that the train could also not be moved while the charging cables were connected. In order to protect the battery cells during the charging process, cooling air was supplied to the battery compartments using electrically-operated fans. Interlock systems were also used to ensure that charging could not start if the air pipes were not properly connected and the fans switched on.
Figure 1. The BMU departing Aberdeen on the first day of public service (21st April 1958). The battery compartments are clearly visible under both vehicles. Note also the destination boards and brackets visible on the side of the leading coach above the windows; although a common feature on main-line coaches at the time this was an unusual feature on a multiple unit train. (Photograph by D. Murray-Smith).
Figure 2. View from rear of BMU approaching Aberdeen station, having just passed Ferryhill Junction signal box (7th May 1960). The right-hand side of the cab can be seen with the brake valve and brake gauges. (Photograph by D. Murray-Smith).
Although it is understood that several different routes were considered for the experiment when the project was at the planning stage, the choice of the Aberdeen
branch is interesting. The length of the route was about 43 miles and this meant that charging equipment was needed at both terminal stations. Ballater
is 213 m above sea level while the start of the line at Aberdeen
station, being close to the harbour, is only about two or three metres above the sea. The gradient profile for the branch
shows that, as far as Banchory, there were no very significant inclines, apart from a section to the west of Culter
where the rising gradient was one in eighty for a little over a mile. Just after Banchory station there was a short section at one in seventy, increasing to one in 68 for about two miles. Further sections at 1 in 70 started east of Torphins
station and continued for three miles until a summit was reached about 1500 yards east of Lumphanan
station. This was the highest point on the line and the descent towards Dess
stations involved 1 in 70 for almost three miles followed by an undulating profile. After a section of about one mile at 1 in 71 just after Aboyne
station, the line followed the course of the River Dee, with relatively gentle gradients all the way to Ballater
During the later years of steam operation, passenger services on the Ballater
branch were handled mostly by the BR Standard Class 4MT 2-6-4 tank engines based at Kittybrewster Shed
. Even in the early to mid-1950s, occasional use was still made the ex-LNER Class B12s which were also allocated to Kittybrewster. Freight services in the 1950s were handled by a variety of different types from the Kittybrewster allocation, including ex-LMS Class 2P 4-4-0s, ex GNSR Class D40 4-4-0s, ex-NBR Class D34 4-4-0s as well as Class K2 2-6-0s, the BR Standard 4MT 2-6-4 tank engines and BR Standard 4MT 2-6-0s. The occasional royal train in summer months was usually handled by a pair of Kittybrewster Class B1 4-6-0s.
With steam traction, the summer timetable for 1951 shows the fastest down timing as 81 minutes from Aberdeen to Ballater, which was the 14:00 Saturdays only Aberdeen to Ballater service with five intermediate stops. The fastest up service at that time was the 17:45 departure from Ballater which reached Aberdeen in 74 minutes, also with five intermediate stops. Stopping services calling at up to 13 intermediate stations, at that time, took up to 96 minutes.
Before being moved to the North-East the BMU was tested in the Glasgow area and made a number of night-time trial runs on the former NBR Kelvin Valley line which ran from Kilsyth through Torrance
to a junction near Maryhill Park
station on the route of the West Highland Line
. That line had closed to passenger traffic in 1951 but was still in use for freight traffic until 1961. In 1958 the route was also used for driver training on DMUs.
The formal first run of the BMU on the Deeside line took place in, rather wintry conditions, on Wednesday 26th March 1958
. A special steam-hauled train was run from Aberdeen to Ballater for local representatives and the press, together with Sir Ian Bolton, the Rt. Hon Thomas Johnston and other members of the North of Scotland Hydro-Electric Board, and also Mr James Ness, General Manager of the Scottish Region of British Railways. The party then departed back to Aberdeen at 4.21pm, travelling on the BMU.
Regular working did not start for several weeks to allow a period of training for drivers and other staff. The first day of operation was Monday 21st April 1958, which was a local spring holiday, and the first service left Aberdeen at 9.40am filled to capacity (Figure 1). Initially the BMU shared responsibility for branch services with steam-hauled trains and the final day of scheduled steam operation was Saturday 5th July 1958 (Figure 3). From the following Monday (7th July 1958) diesel multiple units were responsible for three services in each direction while the BMU handled the other three (Figure 4).
Figure 3. Standard Class 4MT 2-6-4T No. 80029 departs Aberdeen on the last scheduled steam service to Ballater on 5th July 1958. (Photograph by D. Murray-Smith).
Figure 4. Metropolitan-Cammmell Class 100 DMU at Platform 1 in Aberdeen station on the first day of diesel services to Ballater (7th July 1958). The battery charging equipment for the BMU was located just beyond the buffer stops on this platform. (Photograph by D. Murray-Smith).
Although the timetable when the BMU operations started was very similar to that for the steam services which it replaced, improvements were gradually introduced but without dramatic reductions in journey times. By 1960, using diesel multiple units and the BMU, times were still not significantly different overall from those when steam had been used. Stopping services with 12 intermediate stations took from 77 to 85 minutes. The fastest service in the Ballater to Aberdeen direction was scheduled for the BMU and took 71 minutes with six stops. An extract
from the summer timetable for 1960 is shown in Table 1. Note that a distinction is made between the 'Battery Car Service' (denoted by the letter 'A') and DMU services denoted by the letter 'D' above the relevant columns. It can be seen that the schedules gave the opportunity to recharge, partially at least, in the lie-over times at the terminal stations. The BMU was allocated to Kittybrewster MPD (shed code 61A). Although initial arrangements in 1958 involved the BMU spending nights at Ballater, it is clear from the timetable in Table 1 that by the summer of 1960 overnight stabling was at Aberdeen.
The debut was highly successful and services over the first few months were marred only by minor problems and a failure at Dinnet on the Monday of the Aberdeen Autumn Holiday in September 1958. However, the unit disappeared to Inverurie Works for attention at the end of October 1958 and remained there for at least two weeks.
The BMU was notable for its smooth and silent running, and it appears to have had operating costs similar to those of two-coach DMUs. The maximum acceleration on straight and level track was officially estimated to be 0.75mph per second up to 30mph although, on one occasion, I did record a speed of 30mph being reached in 32 seconds on departure from Banchory on a service to Aberdeen. There was a short level section there, followed by a down gradient of 1 in 228.
Table 1. An extract from the summer timetable for 1960
. Note the footnote 'A' indicating the services which were scheduled for the BMU.
Tables 2 and 3 provide details of four journeys on the Deeside line. Two are for trips from Aberdeen to Ballater and two give details of journeys in the other direction. The first journey shown in Table 2 is particularly interesting in that it shows something of the potential of the BMU to improve on the schedule. Departure from Aberdeen was delayed for some reason which was not recorded and the crew made considerable effort to regain time. It is likely that the initial delay was at least ten minutes, since the recorded times show time being recovered throughout the whole journey and an overall time about 9½ minutes less than the schedule (notwithstanding a request stop at Cambus o' May). The second run was on an occasion when the BMU was temporarily unavailable but the reason was not recorded. It may have been due to overnight re-charging problems rather than a fault on the BMU since the unit was back in service later the same day. With a light load of three coaches the Standard 4MT 2-6-0 had little difficulty in maintaining the BMU schedule overall and, indeed, improved on it quite significantly over some sections of the line. One feature of both runs is that station dwell times were very short, often around 20-30 seconds but in some cases (with the BMU) less than 10s.
Table 3 gives details of one run in the up direction with the BMU on the fastest schedule of the day and another when a DMU was being used on the same service. The performance was broadly similar on these two runs but, unlike the runs in Table 2, I was seated at the rear cab interior window and could thus observe speeds throughout. On both occasions speeds of more than 60mph were recorded (a maximum of 62 mph near milepost 18 in the case of the BMU) and a maximum of 65 mph near milepost 12 in the case of the DMU). The DMU also displayed a speed of 62mph between Cults and Ferryhill Junction as the driver was clearly trying to make up time lost at Park while waiting to cross the westbound service from Aberdeen.
|Approximate distance (miles) with corresponding timing points||BMU|
Train: 09:40 Aberdeen-Ballater (delayed departure from Aberdeen)
Date: August 1958
Time (minutes. seconds)
|Standard 4MT 2-6-0 76107|
Train: 08:12 Aberdeen-Ballater (departure 12 minutes late from Aberdeen) with 3 coaches: 97 tons tare
Date: 14th April 1960
Time (minutes and seconds)
|0.0: Aberdeen||Dep.||0min 00s||0min 00s|
|0.9: Ferryhill Junction||Pass||2min 37s (after 30s signal stop)||2min 12s|
7min 43s (schedule 8min)
|8min 27s (after 10mph pwr)|
8min 58s (schedule 8 min)
13min 15s (schedule 15min)
14min 57s (schedule 17min)
|9.7: Drum||Pass||16min 34s||19min 33s|
18min 23s (schedule 22min)
21min 30s (schedule 25min)
23min 05s (schedule 27min)
26min 37s (schedule 30min)
27min 44s (schedule 33 mins)
32min 25s (schedule 36min)
|18.0: Milepost 18||Pass||35min 33s|
|19.0: Milepost 19||Pass||32min 30s||37min 46s|
35min 03s (schedule 41min)
42min 10s (schedule 44min)
39min 22s (schedule 46min)
47min 08s (schedule 49min)
45min 06s (schedule 53min)
55min 13s (schedule 56min)
48min 57s (schedule 57min)
59min 49s (schedule 60min)
52min 47s (schedule 62min)
65min 22s (schedule 65min)
61min 00s (schedule 70min)
74min 02s (schedule 73min)
|39.4: Cambus O’May||Arr.|
66min 07s (schedule 76min)
|78 min 48s (pass: request stop) (schedule 79min)|
|43.3: Ballater||Arr.||72min 35s (schedule 82min)||83 min 38s (schedule 85min)|
Table 2. Timings recorded by the author for two services from Aberdeen to Ballater scheduled for the BMU, one of which involved a delayed start and a journey time almost 10 minutes better than the schedule. The other record involves a replacement steam service on which the BMU schedule was bettered overall.
Although there were occasional problems with the BMU, there were also problems from time to time with the charging equipment and it has been reported that train crews had more confidence in the charging system at Ballater than at Aberdeen. Around October 1960, the BMU was taken out of service for a considerable period, being replaced by DMUs of Class 100 and Class 105. Timings on the line were relaxed slightly in 1962 to allow for the opening of the Dee Street Halt in Banchory. In June 1962 the BMU was again taken out of service for a period and equipped with a new form of battery and new publicity material for summer services on the line that year again gave emphasis to the use of the BMU. However, subsequent breakdowns and fires saw the unit being reported as being withdrawn in August 1962 and the public timetable for the period from September 1962 showed diesel operation for all services. The BMU was apparently sent to Hyndland depot5 in Glasgow for a decision on its future but there are accounts of it being back in service on the Aberdeen-Ballater line as late as 1966. The exact extent of its use after 1962 remains in doubt. Figures 5, 6 and 7 show the unit at three different locations.
|Approximate distance (miles) with corresponding timing points||BMU|
Train: 17:35 Ballater - Aberdeen
Date: 8th August 1959
Time (minutes. seconds)
|Cravens Class 105 unit Sc51475|
Train: 17:35 Ballater- Aberdeen
Date: 14th April 1960
Time (minutes and seconds)
|0.0: Ballater||Dep.||0min 00s||0min 00s|
|3.9: Cambus O’May||Pass||5min 46s||5min 54s|
10min 47s (schedule 10min)
10min 14s (schedule 10min)
18min 26s (schedule 17min)
17min 35s (schedule 17min)
|7.5: Dess||Pass||23min 15s||23min 38s|
28min 14s (schedule 27min)
27min 26s (schedule 27min)
34min 15s (schedule 34min)
33min 25s (schedule 34min)
|16.9: Glassel||Pass||37min 53s||37min 11s|
44min 47s (schedule 41min)
43min 25s (schedule 44min)
|23.8: Crathes||Pass||49min 05s||47min 34s|
52min 36s (schedule 53min)
|51min 43s (after sig. check to 5mph)|
57min 22s (schedule 53min)
|33.6: Drum||Pass||55min 24s||59min 19s|
59min 57s (schedule 59min)
62min 22s (schedule 59min)
|39.5: Cults||Pass||64min 24s||67min 44s|
|42.4: Ferryhill Junction||Pass||69min 55s (after signal check)||72min 34s (after pwr to 10mph and signal check)|
|43.3: Aberdeen||Arr.||71min 49s (schedule 71min)||74min 21s (schedule 71min)|
Table 3. Timings recorded by the author for two services from Ballater to Aberdeen, one of which involves the BMU and the other a DMU.
Following the closure of the Ballater branch in February 1966 the BMU was believed to have been at Inverurie Works but there are accounts of it being used on overnight runs from Aberdeen to Montrose and back, perhaps in order to satisfy contractual obligations with the battery manufacturer5. It was officially withdrawn in December 19665 and placed in storage at Cowlairs in Glasgow.
Later history of the BMU
The unit passed to the Railway Technical Centre in Derby in 1967 where it was known as 'Laboratory Vehicle 16', latterly carrying the name "Gemini". It was painted in the red and blue colours then used for Departmental stock and was used for testing of the British Rail Automatic Train Operation system ("BRATO") using a combination of track and train-mounted equipment. Some second class seating was removed to house test equipment, but other parts of the unit remained in original condition. It was usually based at Mickleover in Derbyshire and operated on the Mickleover to Eggington Junction test line.
Having been declared redundant early in 1984, the unit was put up for sale and was purchased by the West Yorkshire Transport Museum for use on its proposed Bradford Low Moor "Transperience" scheme. It was put into storage at the redundant BR Hammerton Street depot in Bradford. While there coach 79998 underwent complete asbestos removal. The front saloons, behind the cabs, still retained the original seats, with the original shell-pattern moquette, which was in remarkably good condition. Most of the original brown marble formica surfaces remained and, externally, the bodywork was in excellent condition.
Figure 5. The BMU approaching Ferryhill Junction with the 14.45 Aberdeen to Ballater service (April 1958). The “Aberdeen – Ballater” boards on the sides of both vehicles can be seen clearly in this view. The steam locomotive on the left is one of the Class N15 0-6-2T engines used for station pilot duties at the south end of Aberdeen station at the time. (Photograph by D. Murray-Smith).
Figure 6. The BMU passing Cults station on 13.55 Aberdeen to Ballater service on 18th April 1960. (Photograph by D. Murray-Smith).
Figure 7. The BMU entering Ballater station (probably on the 15:25 service from Aberdeen) (8th August 1959). (Photograph by D. Murray-Smith).
The unit was loaned to the East Lancashire Railway, and No. 79998 went into service there, with second class seating replaced. Portable battery charging equipment was apparently fitted in the guard’s area and the coach was repainted into the original green livery.
The West Yorkshire Transport Trust's "Transperience" project went into liquidation in 1997 and most of the assets were disposed of by auction. The Royal Deeside Railway Preservation Society successfully negotiated with the liquidators to buy the BMU vehicles and they arrived back in Scotland in 2001 for storage at Ferryhill in Aberdeen. They were moved on to their present site at the Royal Deeside Railway at Crathes about five years later. Coach No. 79998 has now returned to passenger services but the cost of replacement batteries and lack of charging facilities means that the vehicle cannot, at present, operate under its own power and is used as a locomotive-hauled coach.
Lessons from the BMU experiment?
Traffic did increase on the branch with the introduction of the BMU and the DMUs and it is reported5 that revenue increased by 64.9 percent over the first year, with the following year showing a further increase of 11.5 percent. Day-return fares were kept at a level that made them very similar to those of the competing bus services.
Although it is clear that battery technology was perhaps not at a stage to allow the full potential of battery-powered trains to be achieved in the late 1950s and early 1960s the project should certainly not be forgotten. Although it was successful in many ways it is disappointing that, in the new enthusiasm for battery powered vehicles, little mention is being made about this pioneering project. In comparison with modern experiments with unconventional rail vehicles, including the recent trials of a battery-powered Class 379 “Electrostar” unit between and Harwich and Manningtree (involving Network Rail, Bombardier and Greater Anglia), the Scottish BMU experiment was on a small scale and was a relatively low-key exercise. Although the total capital cost of was estimated at the time as £63,000 (perhaps about £1.5M today) a large part of that total was for the two vehicles themselves and it was, in many ways, a remarkably low-cost project. The BMU was popular with passengers, due to the absence of vibration, the very low noise levels and general comfort compared with the steam-hauled rolling stock that it had replaced. Except in hours of darkness, drivers usually raised the blinds between the cab and the passenger compartment so a view ahead was normally available and those interested could not only admire the scenery of the line but also observe in detail how the unit was driven. During the first few weeks of BMU operation passengers commented very favourably on the novelty of being able to see the open views of the line through the front and rear cabs. The BMU was also popular with drivers, many of whom were said to prefer it to the DMUs in terms of performance.
In retrospect, there is little doubt that the decisions in the 1960s and 1970s to close the Ballater branch and the lines to Fraserburgh and Peterhead, along with many stations on the outskirts of Aberdeen, was a serious mistake. The Aberdeen area has changed very considerably in the years since then due to the oil industry and improved transport facilities are now being recognised as a priority. Re-opening of the Deeside line, at least between Banchory and Aberdeen, is being talked about and, further north, re-opening of the Buchan line between Dyce and Ellon is now a serious proposition which has the support of many at a local level. At a national level organisations, such as the Scottish Association for Public Transport and Transform Scotland, have been maintaining pressure to have rail-based solutions adopted rather than further road improvements in North East Scotland. Following successful re-openings of Dyce and Laurencekirk stations, new projects between Aberdeen and Inverurie should lead to significant improvements in local rail services. Those living on Deeside should think about the fact that a little over fifty years ago a journey by BMU or DMU could have got them from Banchory to Aberdeen (17 miles) in a time of between 26 and 30 minutes. Current bus timetables show a journey time of slightly over one hour.
1. Anonymous, “Battery Multiple Unit – Derby 2-car BMU”, Railcar.co.uk website: http://www.railcar.co.uk/type/battery-multiple-unit/background, (accessed 10th January 2018).
2. Anonymous, “A Scottish battery railcar” Trains Illustrated, December 1957, pp626-627.
3. Anonymous, “First battery railcar on BR”, The Railway Magazine, June 1958, pp419-420.
4. Anonymous, “Scottish battery railcars”, Trains Illustrated, June 1958, p284.
5. Mullay, A. “Batteries Included: the story of Britain’s strangest ‘emu’”, “The Railway Magazine”, August 2004, pp56-58.
6. “ABC British Railways Locomotives, Combined Volume”, Summer 1959 Edition, Ian Allen, London, 1959
7. “Gradients of the British Main-Line Railways”, The Railway Publishing Co. Ltd, London, 1947, p28B.
8. “British Railways, Passenger Services Scotland, 13th June to 11th September inclusive 1960 (or until further notice)”, British Railways (Scottish Region), British Transport Commission, London, 1960, Table 40.