1. Blackpool/Fleetwood: Fluorescents
   A collection of Group-A and Group-B fluorescent lanterns taken in the 1960s.

2. Cambridge: Old Installations
   Before the Candles, before REVO and before BTH, Cambridge was lit by gas. Pete Rivet tracks down some archival photographs.

3. Cheam: ESLAs
   From my original site, here's the only documentation of a huge ESLA installation from South London.

4. York: REVO Festivals
   For years, a small collection of these lanterns lit the approach to York Minster.

5. Dreadful
   Some pictures of some rather dubious street lights from the past.

1. Barnstaple: GEC/REVO
   An attempt to recreate the 'original' installation - which partly succeeds.

2. Beckenham: ESLA
   A unique installation surviving on a private street in Beckenham, South London.

3. Bolton: REVO
   Pictures taken in 1990 of the old REVO installation still lighting the streets.

4. Cambridge: Richardson Candles
   The unique, bespoke, architect designed Richardson Candles of Cambridge.

5. Cambridge: Lanterns
   But there's more in Cambridge than the Candles.

6. Canterbury: Lanterns
   Pictures of surviving ESLA Bi-Multis in Canterbury.

7. Cannock: REVO
   Three REVO Lodestars still survive in Cannock.

8. Folkestone: REVO
   Until recently, the side-streets of Folkestone were lit with antique REVO street lights.

9. Haywards Heath: GEC
   One of the last Group-A fluorescent installations in the country.

10. Kings Lynn: ESLA
    I really, really could not believe this! Sodium ESLAs still in service.

11. Merton Park/Epsom/Ewell: ESLA/REVO
    Lee Gale has discovered several 'rescued' lanterns and columns around Merton Park in South London.

12. REVO City: REVO/ELECO/GEC/Thorn
    A 1930s REVO installation can still be picked out in the back streets of REVO City

13. Tooting: Gas lanterns
    It's not difficult to find interesting lanterns in Tooting; just walk out of the tube station.

14. Wimbledon: ESLA/REVO
    Tucked away in the back streets of Wimbledon, treasures can still be found.

Instead of grouping old streetlights together by location, here they're grouped by category.

1. Concrete Columns And Brackets
   Concrete streetlights are becoming rarer and rarer. Here's a picture of some forgotten ones.

1. Cast Iron Columns And Brackets
   Similarly unspoilt cast iron columns, brackets and lanterns are also becoming depressingly rare. Here's some examples from around the country.

These two installations are unique. The first is a refurbishment of a pole-mounted 35W SOX scheme, proving that with a bit of time and effort, old lanterns can be restored to as good-as-new.

The second is a trial installation in a village for a Parish Council. By taking old 1950s refractor lanterns, and installing state-of-the-art metal halide technology, the scheme is an interesting mesh of old and new.

1. 35W SOX Reburbishment Scheme
   
2. 70W MBI Replacement Scheme
   

There are several museums and/or collections of older lanterns, brackets and columns. In time, I'll document all those known.

1. Concrete Utilities' Museum Of Street Lighting
2. Stratford-On-Avon Collection
3. Dublin collection

I'm a keen urban explorer. My adventures (and misadventures) can be read on my urban exploration website. However, whilst exploring I often come across various street lights which are of interest.

1. Stewartby Brickworks
2. Croxley Green To Watford
3. Location #1
4. Location #2
5. RAF Wyton Bomb Store

Sometimes street lighting columns are used (or disguised) for other uses.

1. BT Mobile Phone Masts

After this scheme, the Department of the Environment agreed to a further scheme of catenary lighting in conjunction with the dualling of a length of the A30 Trunk Road (Great South West Road) within the Borough. The scheme is, at present, under design and it's proposed to install a running wire system rather than a fixed wire system.

Part of the promenade has been lighted very effectively by 27 double-arm decorative concrete columns equipped with two 300W lamps at a mounting height of 20'. In addition 10" diameter opal glass spheres each equipped with 100W lamps are provided at 10'6" for all-night lighting. Metal work is finished in stone colour to match the columns. The low maintenance costs of concrete has influenced the choice, while the lanterns are totally enclosed, and are both weatherproof and dustproof. (A picture of this installation, which includes GEC lanterns, was included in the 1939 journal).

These are Concrete Utilities Blackpool New Type columns and brackets, designed especially for the town.

The main roads are lit by electricity. The lanterns in the most important parts use a 400W MA lamp with two 150W GLS lamps; in less important parts, the lanterns contain a 250W MA lamp and two 75W GLS lamps. Mounting height is 25', spacing is 114'. On other traffic routes, the lanterns are fitted with one 150W MA lamp with two 40W GLS lamps, in an asymmetric prismatic refractor bowl lantern, mounted at 22', and spaced at 113'-150'. The length in use is 8 miles but this will be extended to 50 miles in the next 8 years. The colour modification by separate GLS lamps gives a very pleasing white light.

"Lower Intensity Lighting" (or bye-road lighting) is carried out by 6000 two- and three-light gas lamps. A scheme is in progress where they will be converted to three-light suspension lamps fitted with asymmetric glass dish refractors and mounted at 14' with a spacing of 100' to 150'. This will ensure the lighting conforms with the "Final Report" as regards Group B lighting.

The Bournemouth Gas And Water Company are pioneering the Aeration Test Burner (ATB) so the whole of the supply will be under ATB control. All gas burners supplied with this gas will give a regular and consistent performance without any need to interfere with the first setting of the air regulator; it may be found that the regulator is unnecessary. Flame size, under ATB control, will be kept within a very small tolerance and will not vary much in its constancy. The improved bye-road gas lighting will be supplied under ATB control with each lamp individually governed and fitted with a nipple with a calibrated orifice.

It is suggested 5-ft. fluorescent tubes might be adaptable and the GEC when approached make up four sample reflectors, so arranged that they could be carried on bracket arms mounted on the existing trolley poles above the spanwires. The tests with these lanterns are so satisfactory that it is decided immediately to complete the roundabout and also put two additional units in the immediate vicinity of the roundabout in the approach roads to cater for crossing pedestrians. Twenty-five units were made in all.

The installation is lit on the 1st June. The standard of lighting is extraordinarily high for the current consumption of 2¼k.w. including choke losses, whilst the evenness of brightness is most remarkable. At a subsequent date test point and brightness readings were taken on the roundabout and on the kerb edges between 2 and 3 f.c. were obtained in nearly all positions.

Many visiting engineers expressed the opinion that it is probably the finest roundabout lighting that has been seen for years and they were particularly impressed by the absence of glare, due to the cut-off distribution and the low brightness of the light sources.

The GEC put this prototype lantern into production as the One-Eighty.

There are a few roads in Braintree, mostly "unadopted," with pole mounted lights; many of the brackets and control boxes are AC Ford, which makes me wonder if all the Group B lights were once AC Ford HPMV. Although a neighbour of my parents who was once a street lighting engineer has a couple of GEC lanterns that were once HPMV.

Braintree and Bocking probably have more Stanton 9s left than anywhere else, albeit sleeved. There has recently been a replacement programme, with a lot of the Philips being replaced by Urbis Schreder LED lantenrs, on some of which the business part of the luminaire is only about 4 inches square.

Square type gas lanterns have now been fitted with electric lamps and to obtain 0.02 f.c. the upper halves of the four glass panels were painted black whilst the lower halves were "haved" with transparent varnish paint thinned down. For Group A roads where 250W mercury discharge lamps were fitted, they have now been replaced with Starlight fittings with the bottoms removed and 25W lamps inserted.

The old arc lamp pillars were made of cast-iron and broke very easily when hit. It was agreed that the slender modern steel tube, bereft of any unnecessary trimmings, gave a smarter appearance. When hit, steel columns would be bent, and it was easy to bend them straight.

In the war damaged areas, most of the lamps were suspended arcs. Under "Moonlight" conditions, 25' mounting pillars had been erected, and more of the main thoroughfares through this damaged, district gave a more cheerful aspect after dark.

187 Mercra 400W and 250W in BTH Dilen lanterns with BTH control gear were installed along the High Street, Moulsham Street, London Road, Duke Street, Tindal Square, Springfield Road, Baddow Road, Rainsford Road, Broomfield Road, New Street, Rectory Road and at principal road junctions. The poles are arranged in a staggered formation, ranging from 120' to 150' while the lanterns are mounted at a height to conform with modern standards. The centre of the town is lit to Class "D" while Duke Street and parts of London Road are Class "E" and the remainder in Class "F."

637 BTH County Junior lanterns with 150W Mazda gasfilled lamps are used for the lighting of the side streets, the majority being fitted at existing positions. For the remainder, additional columns have been provided in some of the more important residential areas to increase the standard of lighting. 19 of these lanterns are fixed to special brackets fitted to company service poles carrying overhead transmission lines and 8 are used in conjunction with wall brackets.

The lamps are controlled by Solar Dial Time Switches and are in commission daily throughout the day - those in important positions and road junctions being lighted from dusk until dawn and the remainder from dusk until 1AM. The County Council is repsonsible for the general maintenance. This new installation affords increases in illumination of 100% in the main thoroughfares, 50% in the principle side streets and 33% in the remainder while generous lighting is afforded at the more important road junctions.

The thoroughfares in which 1000W lamps are used at a mounting height of 25' have a roadway which varies from 35' to 79' in width. The average width is 16'6". The distances between standards along kerblines in these sections varies from 94' to 124', the average being 108'. These measurements necessitated the projection of beams of maximum intensity from the lanterns at the following angles: (a) vertical angle, or "angle of incidence", min 64°, max 71° average 67°; (b) horizontal angle, or "azimuth angle", min 16°, max 40°, average 26°. In general, the lanterns used had to be capable of producing their greatest intensity of light at an angle of 67° in the vertical plane and 26° from the kerblines in the horizontal plane.

The thoroughfares in which 1500W lamps are used at a mounting height of 30' have a roadway which varies from 101' to 121' in width. The average width is 147'6". The distances between standards along kerblines in these sections varies from 66' to 138', the average being 116'. These measurements necessitated the projection of beams of maximum intensity from the lanterns at the following angles: (a) vertical angle, or "angle of incidence", min 65½°, max 73½° average 69½°; (b) horizontal angle, or "azimuth angle", min 40°, max 67°, average 46½°. In general, the lanterns used had to be capable of producing their greatest intensity of light at an angle of 69½° in the vertical plane and 46½° from the kerblines in the horizontal plane.

22 arc-lamps were installed in 1882. Eastbourne was the first town on the South Coast to supply electricity for public lighting from a central station. The lamps were connected in series across 1200 VDC through special mains from a machine known as a Brush Company's Arc Lighter. Each lamp had an automatic cut-out which short circuited the lamp if the carbons failed. The machine frequently gave trouble and on one occasion the armature burst causing considerable damage.

In 1898 Welsbach incandescent burners with mica chimneys were introduced for the gas lamps and the conversion took four years to complete. The burner cost 7s. 6d. and each mantle cost 2s. 3d. This burner was operated by means of chains dropped through a hole in the glass of the lantern, and in a wind it was a job to capture the small rings at the ends.

In 1902 the original arc-lamps were replaced by Gilbert arc-lamps which were run in series on 2,200V supplied from a constant current transformer. Later the number of Gilbert arc lamps was extended, the later ones being connected on 100V through a transformer and choke. The annual charge per "arc column" prior to 1902 was £30, in 1902 it was £25 and in 1904 it was £22.

In 1904 an Incandescent Gas Company burner was substituted in the gas lamps and was a considerable improvement. Lamp-lighters were paid 22s. per week and had one day off in three weeks.

The first short column electric lighting was commenced in 1905 as new roads were built. The fitting used was known as the Reason G type and carried two 30W lamps with saucer-shaped glass reflectors. Also around this time Nernst 16cp lamps; they were very efficient but gave constant trouble with the heater cut-outs, so they were abandoned in favour of metal filament lamps.

In 1908 further improvements took place with the adoption of inverted fittings and mantles for the gas lamps.

In 1913 the Gilbert arc-lamps were replaced. Trials took place with Oliver Magazine and Excello Flame arc-lamps but in the end enclosed German Sunshine arc-lamps were used.

All street lighting was extinguished during the First World War.

After the Armistice, the arc-lamps interiors were replaced by 300W GLS lamps. At this point, there were 1625 gas lamps and 243 electric lamps in the town.

In 1918 it was decided to convert gradually all gas lighting to electric lighting. This was carried out by fitting a 60W lamp with, first, a straight bracket and then a swan-neck bracket. These replaced the old square gas lanterns.

In 1921 the old converted arc-lamp fittings were placed by fittings carrying a glass refractor and outer globes, which was one of the first lanterns designed for GLS. The use of this fitting with tall cast iron columns was extended through all the main streets between 1921 and 1926.

In 1928 nearly all the short columns with square glazed gas lanterns had been converted to electric lighting. All new lamps had a short swan neck bracket and enamelled reflector.

In 1930 all side street wattages had been increased to 100W and all corners to 200W.

In 1932, the authority decided that full control of the street lighting department was transferred to the Borough Electrical Engineer from the Borough Surveyor.

All the old square glazed gas lanterns remaining (687) on short cast iron columns were removed and replaced by modern brackets giving a mounting height of 15'. (By coincidence this matched the forthcoming Interim and Final MOT Reports). The bracket was made in one length of seamless 1" steam barrel. In the past there had been rusting of the tubes near the mid-bracket foliation and the spigot cap, so it was arranged that these would made of iron and cast onto the barrel. An unusual and novel feature of the bracket was that it was designed to unobtrusively house a selenium bridge housing (see Radiovisor) for a light-sensitive switching housing if required. The bridge is housed in a special housing just above the lantern. However, in the vast majority of cases, they were not using light sensitive control apparatus and therefore the housing was fitted with a lead cap.

At one time the intensive use of light sensitive apparatus was considered and this led to the special design of the bracket as shown above, but:

• About this time that the superimposed high-frequency remote control systems were becoming known.
• It was difficult to get a group of individual light operated apparatus to operate together.
• It could only be used for all-night lighting.
• It prevented subsequent alteration of burning hours.

Light sensitive control is used in the indication of daylight values and operation of signals in connection with a centralised control system.

Some of the lighting is by time switches, either by large capacity time switches, energising switch wires from sub-stations, or by individual time switches in lamp posts. The large capacity time switches gave little trouble. However, in the case of small individual post time-switches, considerable trouble was met:

• The spindle of the high speed type wasn't lubricated in the jewelled bearings sufficiently.
• Some spindles were too thin.
• The spindles were not sufficiently hardened.
• Jewels were insufficiently polished; polishing powder left in jewels; cracked jewels.
• The switches were too weak.
• The contacts were too small. (They did not appear to take into account that filament street lighting has a high initial current).

Overhead switch wires have not been used in the town area, but is used in rural areas.

A considerable part of the area is controlled by hand-switching methods, but this is being reduced by the extension of automatic control. The switching staff consists of one Inspector, one Assistant Inspector, 13 full-time lamp-lighters and 3 half-time lamp-lighters. The Inspectors patrol all the lamp-lighters' rounds each evening and arrange for reliefs when necessary. Each man has a starting and finishing lamp, and reports lamps that have failed at 8AM each morning. The full-time lamp-lighters are employed during the morning as cleaners, fitters and erectors. The half-time lamp-lighters are also employed by the Highways Department as road sweepers (but their role will be dispensed with as automatic control expands). Each round is restricted to an average of 100 lamps-per-man, for which he is allowed one-hour actual switching, finishing up at lighting-up time. The cost of switching is 16s. 0d. per column per annum. The introduction of automatic control was delayed as switch wires, or pilot cables, were not available in many streets. Small time switches were considered but owing to the troubles already experienced, and partly due to the high cost of a time switch, this was not proceeded with. Secondly, a large number of light-sensitive units was considered but rejected (although the standard bracket was designed for this purpose). Therefore an improved form of control using relays operated by a high-frequency impulse superimposed on the existing distribution system is being used. It was installed in a restricted area, and due to its success, has been considerably extended, and is replacing the hand-switched rounds.

It was decided to standardise on a suitable lantern for side-street lighting (to replace the old square glazed gas lamps). Various types of 100-200W GLS fittings using either reflectors or refractors were erected under identical conditions in a trial street. After inspection and completion of photometric and other tests, it was decided to standardise on a fitting with a non-axial asymmetric refractor. No outer globes were used as these absorb 10% of the light output and the difference in cleaning time is neglible. Reflector fittings, particularly those of the multi-facet type, are efficient and extensively used in rural areas, but as many of the fittings in the town were refractors, then this factor had a considerable influence on the final decision. When the fitting was decided upon, alternative combinations of cast-iron, copper and steel were considered for the body, but on the grounds of durability, it was decided to use copper. The lower reflector carries the refractor on special bronze springs, so designed as to facilitate easy removal for cleaning. The refractor is in one piece in which are formed lenticular prisms in two main directions, with vertical and annular prisms to give directional control to light emitted in the upper hemisphere. The asymmetric disposal of the main beams is each 12½ from the axis, and the beam width is broad enough to spread completely across the road. The fittings are mounted at 15', standard side street spacing of 120', and give 0.82f.c. under the lamp and 0.17f.c. a B.S.S test point i.e. almost to Class G standard. Illumination under the lamp is reduced to 0.66f.c. after 1000 hours burning but this does not affect appearance and the lumen output per 100' under average conditions is 717.

For road junctions, where lamp failure may be source of danger, a two-light fitting, which is two of the refractor lanterns in one, is fitted. The refractors are each the standard non-axial asymmetric, or symmetric, depending on the type of junction. It is realised that a certain amount of efficiency is lost due to the "blanketing" effect of two refractors in close proximity, but it was felt that the twin fitting should match closely in appearance the ordinary single fitting.

The majority of the Group B roads in Eastbourne are lighted as described with the exception that on all bus routes, which are not lit as Group A roads, 150W lamps (giving 0.25f.c. at the test point) are fixed instead of 100W. Also at all important junctions, Group B bus routes are illuminated by a 250W MA/V and three 100W GLS lamps mounted at 25', enclosed in a lantern of the dispersive type.

One of the most important annual features is the issuing of specification and obtaining tenders in connection with the supply of materials, especially lamps. Advertisements are issued in December, tenders asked for by January, allowing time for investigations and recommendations before April. Lamp tenders are required to supply a list of undertakings to whom they have supplied lamps in quantity in the past three years, and a certificate of test by the National Physical Laboratory or similar. Samples of lamps are also obtained from various manufacturers and are tested in the Test Room. Illumination values are recorded when new and after a certain length of life. Voltages are regulated and the life of each lamp recorded. These test results are invaluable when chosing a contract.

A number of columns and bollards are knocked down or damaged each year. The average is 20 per annum and the total value of the resultant claims is £60. However, insurance companies only offer to pay 50% of the damage, but pay the full amount in the end. This is influenced by the fact that if there is doubt as to responsibility, then no claim is made.

One advantage of hand switching is that the lamp-lighter is able to note any lamp failures on his round. These are presented on a form to the Inspector when reporting for his half-day's ordinary duties at 8AM. Lamps out in the automatically controlled areas are noted either by the Inspectors or by the Police. Each morning the Inspector visits the Police Station and receives a list of lamps reported as being out. In the case of the more important lamps any failures are replaced at once the sme evening by a man on duty for the purpose until 11PM - the Inspector arranges in the mornings the replacement of the other lamps. The identificaiton of posts is facilitated by the numbering which is carried out by means of a neat, thin zinc plate with stencilled numbering. This system is better than painting as they are more discernible at height and not obliterated when the column is repainted. For short columns, box tricycles are used, the lamps and cleaning materials being carried in the box and a short ladder fitted to carry along the box. For tall column work, three hand tower wagons and a motor tower wagon are available.

A form covering lamp failures is completed each morning by the Inspector. On this the life of each lamp is calculated. No difficulty is experienced in obtaining free replacements for lamps which have definitely failed prematurely and the average life of lamps is:

• Sodium discharge: 3000 hours.
• Mercury discharge: 2000 hours.
• Tungsten filament: 1000 hours.

Attention has to be given to appearance and preservation given the salt laden air and hot summer sun. One coat of paint is applied every two to three years by brush rather than spray gun. Spray guns were used but discarded because:

• The round surface tends to cause excessive waste of paint.
• The resultant coat is thin.
• Time is wasted cleaning out the gun between columns.
• Spray of paint may fall on a passer-by.

Cleaning of lamps is carried out in the morning by 10 of the light-lighters at an annual cost of £487, or 3s.8d. per column per annum. Each lamp is cleaned on an average of every five weeks, but some are cleaned every two weeks. (The accumulated dirt is such as to reduce the lumen output by 10%).

The records are much the same as those used by most street lighting authorities. One, which is invaluable, is a complete set of ordnance plans of 1/1,250 scale of the whole of the town. On these plans are plotted all lamp positions with the column number adjacent. The colour of the plot or ring indiciates the light source (red for filament; blue for mercury; amber for sodium) and the shape of the point indicates the type and make of fitting and wattage of lamp. All switch wires are indicated by red lines connecting these points and in addition the time switch or other control positions are clearly indicated.

The costs are met entirely from the General Rate, and for 1938, the figure is estimated at £18,310, representing 5.09d. in the pound. The costing of items (shown separately) are facilitated by the use of allocation numbers:

• PL.1.A. : Switching etc.
• PL.1.B. : Cleaning.
• PL.2 : Sundries.
• PL.3 : Transport.
• PL.4.A : Repairs.
• PL.4.B : New lamps.
• PL.5 : Light sensitive apparatus.
• PL.6 : High frequency impulse control.
• PL.7 : Improvements (Class B Roads).
• PL.8 : Parade decoration work.
• PL.9 : Improvemetns (Class A roads).

The estimated consumption of electricity for 1938 was 1,412,000 units at 1.25d per unit, which is an average figure for towns the type and size of Eastbourne. The consumption of electricity is arrived at by meter readings. Meters are fixed in columns with lamps of different wattages, and from the readings of several of these meters, an average is taken. In some cases, where meters are difficult to fix, the reading is calculated from burning hours and the wattages.

In 1934, it was decided to improve the lighting of the main approach roads to the town and also the main shopping streets. This followed a trial of mercury discharge lamps on eight tall steel columns. It was immediately realised that the optical value and efficiency of discharge lighting as a road illuminant outweighed the objections as to colour. Therefore there was no question that discharge lighting in one form or other. Therefore Eastbourne was a pioneer installer of mercury vapour lamps (MA) in 1934 (as mentioned in GEC catalogue from 1935).

In 1937, the expenditure for main road lighting was £2,538 and for improvements on side roads it was £718. For 1938, these figures were £2,209 for main roads, and £617 for side roads. For 1939, it is estimated that these figures will be £4,310 for main roads (this includes the lighting of main roads in an area added to the County Borough) and £750 for side roads.

The layout for main roads could be separated into two types:

• Motor approach roads lined principally by residences.
• Shopping streets lined by fairly high buildings and shops.

The decision was then arrived at to light the main approach roads by sodium discharge lighting and the shopping streets by mercury discharge blended with filament lighting. The success of this decision has been so obvious that the authorities have not had the slightest hesitation in continuing the principle in the following years.

There is much controversy on the question of sodium discharge lighting or mercury discharge lighting and non-cut-off fittings and cut-off fittings. However, when applied to sodium lighting, the question of non-cut-off and cut-off is exaggerated due to the low intrinsic brightness of sodium. Sodium is used as the light source for all approach roads: it is the most efficient light source (at 60 lumens per watt), chromatic aberration is eliminated and visual acuity is greater than with polychromatic sources, the instrinsic brightness is low (10 c.p. per sq. c.m., compared with 600-900 c.p. per sq. c.m. in a filament lamp and 100 c.p. per sq. cm. in a mercury lamp) so the risk of glare is low, the contrast of objects is on a road is very marked and the lamps are least affected as regards lumen output by voltage changes. Therefore the lamp was selected for motor roads where good visibility and clear distance perception are the most important requirements.

The lighter the road surface the better the visibility e.g. a near matt concrete surface is better than a rough stoned asphalte surface. However once the asphalte had developed a high degree of polish then the road surface showed a very high degree of specular reflection. Therefore it would appear that concrete roads present the best type of road surface for lighting, but concrete road surfaces are not considered to be permanent and are usually covered with a bituminous material. And asphalted roads with a dressing of stone chippings are the only surface considered best from an economy, non-skid, durability point of view. Therefore the Street Lighting Engineer, from a colour point of view, must not look for any assistance from the Road Engineer.

The type of fitting to be used with sodium lighting was the next thing to be decided. The three main types were:

• Reflector fittings with controlled cut-off
• Refractor fittings with controlled cut-off
• Fittings with a definite cut-off at 75° from the vertical. However glare not considered a problem so fitting not considered.

The output of the lanterns is 6,600 lumens, or 4,400 per 100' at 150' staggered spacing which conforms to the MOT Recommendation. The B.S.S. test point illumination is 0.15f.c. which is equivalent to B.S. Class E. Where 120' staggered spacing is adopted the lumens per 100' is 5,500 giving a B.S.S. test point illumination of 0.31f.c. which is Class D. After 2,800 hours, the initial output in lumens of 9,600 has fallen to 6,900 and the foot-candle test under the lantern from 1.3 to 0.97. The lighting shows little worsening in appearance and does not fall below Class E.

The main shopping streets required different treatment from the approach roads. The illumination should be the same as the approach roads but other considerations were taken into account: where most of the "life" of the town concentrates, the appearance by night must be considered important; the buildings and foot walks lining the street must have a good aspect; the colour of the light must be considered, principally becuase the ratio or pedestrian traffic to road traffic is high; and the shop window displays are lit by the street lighting when the shop window lighting is extinguished but before pedestrian traffic has finished. The apperance of the installation must be in keeping with the importance of the thoroughfare and therefore highly scientific fittings may be out-of-place and unsightly. The light must be as close to daylight as possible; therefore both filament and mercury discharge lamps were used. It was decided in the main thoroughfares to house one 400W mercury discharge lamp with three 100W filament lamps in each fitting, giving an efficiency of 30.9 lumens per watt; and in the secondary thoroughfares one 250W mercury discharge with three 75W filament lamps, giving 24.1 lumens per watt. The use of this blended mercury lamp restricts the choice of lantern. A lantern was found in which the light sources are entirely enclosed. The distribution of light is symmetrical, substantially in the lower hemisphere, and the panels are constructed of a special low absorption diffusing glass which mixes the radiations from the two types of sources. The fittings are mounted at 25' and spaced at 120'. The output of the larger wattage lanterns is 10,240 lumens or 8,530 lumens per 100' at 120' staggered spacing, giving a test point illumination of Class D. The output of the smaller wattage lanterns is 5620 lumens, or 4683 lumens per 100' at 120' staggered spacing, giving test point illumination Class E.

The whole principle of the lighting of a shopping street, where congestion of road and pedestrian traffic must be taken care of, is completely different from that of a motor road where road and kerb illumination is almost the only consideration. The generally accepted standard of high road brightness and visibility by silhouette, which is apparently correct for traffic routes, does not apply so much to shopping streets where it is important that people see, or are seen, by direct vision. Silhouette vision does apply to a certain extent, of course, where the shopping street is not congested, but is of secondary importance. Therefore the only way to illuminate them satisfactory is by direct illumination. This is the case of a point duty policeman: silhouette vision would be useless and dangerous, whilst direct illumination is hardly sufficient. After experiements, it was found necessary, completely independent of the street lighting, to have a special cut-off fitting, of a distribution giving no spread directly above the policeman.

To obtain public opinion on discharge lighting, a questionaire was distributed to local bus drivers, policemen and a number of residents. It could be seen that sodium discharge illumination was preferred for general road visibility. Of the residents, 100% considered the discharge lighting a considerable improvement on filament lighting, and 56% preferred the blended mercury to the sodium lighting.

In the case of shopping streets, before the improvements, there were already cast iron columns with heavy round arm ornamental brackets supporting high wattage filament lamps at a height of 19'. When the first improvement took place, it was decided to remove the heavy round arm and substitute a steel extension piece with new round arm bracket so that the mounting height was 25'. The steel extension was spigoted into the top and a steel tube ran from the bracket through the centre of the cast iron column where a steel "spider" was threaded on the tube and, on being tightened up into the top of the base, held the extension piece in vertical compression on the top of the column. This was carried out in one street only but there were disadvantages: the erratic positioning of the old columns was not well suited to modern standards; the cast iron bases would not withstand an impact to the same extent as a steel column; the steel spider required periodic tightening; and the cost was no less than the alternative of scrapping the columns and setting out a properly planned installation of modern steel columns.

The general use of steel columns of modern slender design was then decided upon for both shopping streets and the main approach roads.

In one section of an approach road there were some tramway columns with mounting height of the existing filament lamps of 19'. Since these are steel it was not as conclusive as the cast iron, although they had disadvantages as no base for gear accommodation and a too low mounting ehight. As an experiment, one of the tramway brackets has been altered to give the new mounting height of 25' (with an upward sloping bracket).

The new columns are constructed of fluted tubular steel of high tensile strength and pleasing appearance. The base portion is a large diameter tube, part of which is cut away to house discharge lighting gear, and is provided with a hinged door. This removes the necessity for an old boxed out base or the unslightly alternative of fixing the gear externally to the tube high up the column. Steel columns are much lighter in weight and can be more roughly handled than cast iron or concrete, and this shows in transport and erection costs. A foreman and five labourers can erect an average of three columns per day at 26s.3d. per column. After four days of "setting", one fitter and two other men erect the arms, scrollwork and lantern by means of the motor tower wagon. They further fit the equipment in the base and carry out the wiring. Two columns per day are compelted by this fitter and the cost is 26s.1d.: therefore total labour cost, apart from servicing, per column is £3 6s. 1d. One further advantage of steel is that when underground obstructions are met, special roots such as cranked offset or flanged root can be used.

With brackets, the modern tendency for straight lines is preferred instead of the former round arm ornamental brackts and the overhang is regulated to suit the width of the road and the light distribution of the fitting. To avoid trees and obstructions, 10' overhang from the kerb are necessary. This is considered preferable to suspension lighting. This long overhang does not interfere with the kerb and footwalk illumination, this being due to a fitting with a broad beam spread in the case of sodium lighting, and the use of a lantern with symmetrical distribution in the case of the mercury lighting.

The disadvantages of a centrally suspended systems are obvious: low kerb visibility and the tendency of drivers to drive along the lane of light on the crown of this road (although this does not necessarily apply in the case of specially designed short spaced cut-off installations).

Also considered is the daylight apperance of the installation: overhead wires are not things of beauty, particularly those for central suspension systems. Span wires do not find favour with the Fire Brigade Services either.

The initial cost per column of sodium installations, including erection, jointing and reinstatement averages £25 10s, and at 150' staggered spacing, this averages £897 per mile. The running cost per column per annum for lighting and maintaining one 150W sodium discharge lamp on all-night lighting (3905 hours) amounts to £6 6s, which is £223 per mile. The practice in Eastbourne is to reduce the lighting by approximately half at midnight on main roads and so the running cost is £162 per annum per mile.

The initial cost per column of blended-mercury, including erection, jointing and reinstatement, averages £28 and at 150' staggered spacing averages £985 per mile, and for 120' staggered spacing, £1231 per mile. The running cost per annum of lighting and maintaining one 400W mercury and 300W filament lamps on all-night lighting amounts to £18 8s. The practice is to extinguish the filament lamps at midnight and therefore the running cost is £14 13s per column per annum.

The initial cost per column for the 15' Group B installation is £8 16s, complete with service, reinstatement and one 100W filament lamp. The running cost of this per annum assuming all-night lighting is £2 5s 8d per column.

In 1934 it was seen that the necessity would arise from some system of automatic control to replace hand switching. Around this time a form of remote control system which did not require the existance of pilot wires was introduced, and it was decided to install an initial installation confined to the equivalent of one hand-switched round i.e. approximately 100 lamps. This installation was one of the first in the country, and after a certain amount of teething trouble, has operated since very satisfactorily. This installation consisted of a high frequency transmitter situated in a sub-station and from thsi the impulse signal was superimposed on the distribution network. The impulse operates the relays fitted in the column bases, which will either turn on or extingush the lamps. The local transmitter, each of which operates up to 200 relays, is a spark gap high-frequency generator with an output frequency of 4000 cycles. For each switching operation this high frequency impulse is superimposed for a period of four sections between the neutral bus-bar and the earth connection in the substation and the relays in the individual lamp columns are operated by this impulse. The local transmitter is set to emit three impulses per twenty-four hours, and relays react accordingly (at dusk, midnight and dawn). For two years, the initial installation was operated without extension in order to give the system a proper trial. There were teething troubles (a) paused by some old services not being lead covered and having poor earth return and (b) the occasional erratic operation of the spark gap which operated some relays and not others (this disappeared after adjustment and improvement of the spark gap design). After two years trial, it was decided to put down two further installations, and these gave a similar experience to the initial trial area. Until very recently, each of the three transmitters was operated by an astronomical dial time switch with three arms set for the three impulse operations. In April 1938, this was replaced by a central control: this consists of a master transmitter of the valve type which superimposes a high frequency impulse of 1900 cycles on the nearest low tension network via blocking condensers. This impulse potential is stepped up to the high voltage network via the existing supply transformers, and is available if required for use at every sub-station on the high voltage network. This master impulse is received at the local transmitters by means of sensitive maisn relays employing a Thermionic value in duplicated for reliability. This receiving relay closes the contactor circuit of the local transmitter for the four seconds required. The master impulse was controlled by a time switchm but as an eningeer was always on duty at the power station it was decided he should operate the transmitter depending on weather conditions. To give the engineer the necessary indication, light senstivie bridges are fitted on the roof of the control room, and these opearte relays which operate signal bells. (A time switch operates another signal and the difference in times between the time switch and actual switch on was noted). The local transmitters cost £50, the master transmitter cost £100 and the relays were 50s each.

There is still a great deal to be done, before the whole town can be said to be well and efficienty lighted to the standards suggested by the MOT.

Main Road Lighting: There will be no deviation of the policy of sodium discharge lamps for the approach roads and blended mercury for shopping streets and town centres. In Germany and Holland, exactly the same policy is followed. Blending filament lamps are extinguished after midnight and this is also the custom in Eastbourne. At first the mercury lamps were extinguished at midnight, leaving the filament lamps for all-night lighting. But it was found at B.S.S. test points:

• Both illuminated: Class E: 7880 lumens per 100'.
• Filament only: Class G: 1330 lumens per 100'.
• Mercury discharge only: Class F: 6550 lumens per 100'.

Experiments were tried with filament-and-sodium-discharge and sodium-and-mercury-discharge lamps in the same lantern, but this did not result in the blending of the light, but the superimposition of one on the other.

The luminescent mercury discharge lamp was tried but the initial cost of the lamp is rather high compared with its rather limited colour correction value and reduced efficiency. There were also general doubts as whether the colour corrections will continue at the same value through the lamp's life.

Side Road Lighting: Considerable work is required before these could be considered as illuminated to MOT standards. The standard of lighting of the side roads generally is higher than most towns, and headlights are not necessary in most roads. Most of the new roards and roads on housing estates are already to Group B standards. In some of the older districts, erratic spacing, varied mounting heights and shading by trees are common.

Graded Lighting: Experiments are being carried out with the grading of certain secondary roads at the point where they meet main approach roads (sodium) or shopping streets (blended mercury). The MOT Report does advise grading. The policy is of erecting low wattage sodium lamps between a sodium lit road and a filament lit side road; and low wattage mercury lamps are erected between a mercury lit road and a filament lit side road. Two 65W sodium lamps used for grading but 50W and 100W soidum will be tried as the auxiliary apparatus for these three sizes is the same. In the case of mercury, one 125W and one 80W mercury intervene.

• Sodium grading: 6600 lumens (main road), 2835 lumens (grading), 860 lumens (side road).
• Mercury grading: 10240 lumens (main road), 3240 lumens (grading), 860 lumens (side road).

Statistics show that 38% of the accidents occur in side roads i.e. almost a proportion of two to every three on main roads.

Type of Road: Residential
Fittings: Silvered Mirror Reflector
Lamps: Sodium Vapour
Light Source: Central Suspension
Mounting Height: 23'
Spacing: 120'
Width of Road: 35'
Illumination: Class D (nearly Class C)
Average Illumination: 0.56 foot candles
Diversity Ratio: 2.24 to 1
Consumption: 150W per point

A motorist travelling along this roadway can see an obstruction 300 yards distance. No fatal accident has occurred on the roadway since the installation. Mr Greenhalgh states: "I think I am right in saying that the averaging of foot candles on a given length of roadway is retrograde in application since one could have a shocking example of street lighting which, measured under this system, could have excellent results on paper. I think our Association of Public Lighting Engineers should insist on the diversity factor in all computation and comparisons."

Class B: Central traffic crossings and circuses

Class C: Main roads and shopping thoroughfares within the city centre

Class D: Main roads from the city boundary to the city centre

Class E: Main arterial roads and suburban roads with heavy through traffic.

Class F: Residential roads with through traffic.

The capital cost is a little less than £6 per foot and the maintenance cost no more than for the conventional method of lighting with tall columns and lanterns.

• Observe curb visibility - how many lamps away can one follow the distinct edge of the curb.
• Observe the road surface - is it evenly lit, are there pools of darkness.
• How far distant can one see a small object such as a child, dog or cat?

Classifications are:

• Class D: Main through traffic routes and principle shopping streets.
• Class E: Principle district routes and susidiary shopping streets and potential Class D routes.
• Class F: Principle local residential routes and potential Class E routes and major estate roads.
• Class G: Internal residential roads and estate roads.

Outside sources provide illumination: the Romford Gas Company and the Country of London Electricity Supply Company. If costs are to be comparable and competitive it is (1) essential to extend the standard specification to local needs and provide a tender form of comprehensive layout and (2) provide a means of checking the resulting illumination both initially and in service.

Council's specification is based on B.S.S. 307:1931 with modifications to suit local conditions and advances in technique as embodied in the Intermin Report of the Departmental Committee on Street Lighting such as Gradation Units and Glare. The spacing of units on bends or curves taken from Wilson, Waldram, Damant paper [1]. Lighting up times taken from Wilson paper [2] adjusted for latitude. 25% of the lamps in the district are put into extended night lighting until sunrise.