Road Numbers:
How it happened

A 1920s road numbering diagram
A 1920s road numbering diagram

It's all very well talking about zones, boundaries, numbering rules, conventions, oddities and all the rest of it. But all that doesn't answer the often-asked question of how we came to have the road numbers that we have in the first place. Why were roads ever numbered? Why hubs and spokes? Who was responsible for it and how did they go about doing it?

The story of classification in the UK is not quite as straightforward as it might be - disrupted by war, complicated by Frenchmen and sabotaged by catrographers. This is the story of how the Men from the Ministry came up with an awful lot of numbers.

In the beginning...

...there was nothing. Well, there were roads, but they all looked the same and it was quite difficult to tell them apart. Local authorities found it difficult to keep track of maintenance work. Central government had no accurate measure of how far apart towns were, let alone what state any of its roads were in. Travellers had trouble finding their way around because one road was much like every other, and road signs were generally poor. So in 1910, with the motor car making road journeys competitive and spurring improvement to the roads, action was taken to sort out the mess. A government body called the Roads Board was set up under William Rees Jeffries, instructed to upgrade existing roads and build new ones using money from the new road and petrol tax.

Sample of a road itinerary for Kent. Click to enlarge
Sample of a road itinerary for Kent. Click to enlarge

The trouble Rees Jeffries and his colleagues faced was in working out which roads should be funded, upgraded or replaced. There was nothing to tell them apart and no data available to say which roads were busy. In 1913, work began under Sir Henry Maybury, one of the Board's senior engineers. Its principal concern wasclassification - categorising each road depending on how busy it was - and it followed that they could be assigned numbers for reference. A document from about 1914 explains:

"The object of numbering roads under the scheme for the classification of roads is obviously for the purpose of easy reference between the Central Department and the Local Authorities and others.

"The proposal is to give a road one reference number from its commencement to termination; e.g. the Great North Road from London to Inverness will be given one number throughout its entire length."

Memorandum on Numbering Roads

The scheme evolved so that each road was split into numbered sections, with a new section starting at every road junction. A request for maintenance work could then be identified easily as being on "5-315", that being section 315 of road 5. A circular was sent to County Councils in April 1914 asking them to commence traffic surveys so that classification on this basis could begin, and a sample scheme for Kent was drawn up (a sample from which is above right - click to enlarge it).

The initial scheme, even then, allowed for Class I and Class II roads. The original plan was revised, and numbering was restricted to individual counties with the numbers changing at county boundaries.

A Classification Form, circa 1914. Click to enlarge
A Classification Form, circa 1914. Click to enlarge

The outbreak of the Great War in 1914 brought work to an abrupt stop. It was only in September 1919, as home affairs were returning to normal, that Sir Henry was invited to resume his classification work, now under the new Ministry of Transport.

Building a system

Setting up the Ministry took time. Once there were enough people working there to make a go of it, Sir Henry and his colleague Colonel Richmond wrote again to County Councils asking them to perform traffic surveys on all their roads and return the data so that classification could begin. The work only really got under way in 1920. The system of Class I and Class II roads was resurrected, but now with a new focus of providing a means of navigation, the idea of duplicating numbers in each county was scrapped.

It's hard to say how influential the Michelin company was in the numbering system that was eventually adopted, but André Michelin himself (founder of the Michelin company) took a great interest and wrote a series of four short papers on the subject of French and British road numbering for the Ministry's attention. He explained the French system, and its disadvantages, and went on to propose a system for Britain:

"A first numbering would include the main roads leaving London and the main transversal roads, such as:-

  • Glasgow - Edinburgh
  • Stranraer - Newcastle
  • Liverpool - Scarborough
  • Birkenhead - Newcastle

"We think these might well be lettered "N" (corresponding to National) which would afford an interesting similarity to the classification of the French roads.

  1. The first numbers would be given to the roads leaving London in a northerly direction and then turning from left to right in the sense of the hands of a watch.
  2. Transversal roads would be numbered from north to south.

"For the other roads, in order to avoid a change of number for purely artificial reasons (administrative boundary, etc.), we would suggest to divide the country into sectors bounded by the National roads. These sectors would reach a maximum of 24 and would radiate from London in accordance with the scale map attached herewith.

"Each sector would be distinguished by a letter of the alphabet (except the letter "N").

"All the first class roads in a sector would be given a capital letter as allocated to the sector, followed by a number (Example: A49, H54). All the second class roads in the sector would be given a serial number followed by a lower case letter as allocated to the sector (Example: 59a, 54h)."

André Michelin, 26 April 1921

Sir Henry Maybury thanked Michelin for his trouble and the whole lot was forgotten.

As the surveys came back in, maps were marked up to show which roads carried enough traffic or connected sufficiently important places to be regarded as Class I, and which were Class II. The initial proposal to use the designations T and L for first and second class roads - standing for "Trunk" and "Link" - was soon replaced by the idea of simple A and B designations*. The letters A and B don't stand for anything in particular. It's more like a school report where the best roads are awarded grade A, and the smaller connecting roads dread parents' evening because they only got a B.

But enough letters: it was time to start pencilling in some numbers.

Taking shape

hub-and-spoke system, loosely modelled on the one used by France, was the intention from the outset. At first the numbering encompassed only England and Wales, with Scotland numbering its own roads shortly afterwards. But exactly what form the system would take, and where the spokes would lie, was more problematic. One document lists the London radials and numbers them from one to five - missing out the London-Dover route and giving England and Wales just five zones. It's not clear whether this is just a mistake in an early draft, though, as there are references on the same page to six zones.

The wheel of France and its many spokes, from a Michelin map posessed by the Ministry in the 1920s

The wheel of France and its many spokes, from a Michelin map posessed by the Ministry in the 1920s

A document from about May 1921 (referring only to England and Wales) refers to the A6 as running to Glasgow, and states that "number 6 sector is bounded on the North by the Carlisle - Berwick-on-Tweed Road to which No. 7 is allotted" (Col. Richmond, "Suggested System of Route Numbering"). Aside from the minor gripe that there is no direct or obvious road along the border from Carlisle to Berwick, this splits the system into two very clear parts, with zone 6 prevented from entering Scotland.

Letter codes in use around Kelso
Letter codes in use around Kelso

While numbers were still being pondered upon, there seems to have been an intriguing system of lettered codes indicating first and second class roads, and perhaps what sort of number it was hoped to apply. The example to the left shows Kelso. It appears that Class I roads in a given area were given a single letter and Class II roads a pair of letters (among the lines shown here are A698 and A699). It's not clear what relevance these had, if any; it may only have been to distinguish one route from another.

By August 1921, much of this work seems to have been finished, and correspondence was being sent north for Scotland's highway engineers to begin considering their own numbers. There was still great uncertainty about what would happen north of (and along) the border.

WL Campbell at the Roads Department in Edinburgh was instructed that "the routes you should number in Scotland are 7, 70, 71...79; 8, 80, 81...89, and 9, 90, 91, 92...99." The letter continued:

"There is an objection to route 6 terminating at Edinburgh. All roads in each English sector can be numbered with 3 numerals - the extension of Sector 1 into Scotland would necessitate some of the roads in the Scottish portion of zone 6 bearing four numerals.

"I, personally, see no objection to number 7 being given to the border road and the length Hawick to Edinburgh being 7a."

Letter from the Ministry of Transport to WL Campbell dated 8th August 1921

This letter goes to show that - despite English numbers having been allocated, and one- and two-digit numbers already appearing on signs - the system was still in its infancy. Four digit numbers started to appear rather rapidly south of the border, many of them in London (and, of course, B-roads had them from the outset). The idea of having "A7a" is also rather nonsensical, especially considering it would only exist to contrive a change of zone along the border.

A plan for Scotland in two zones. Click to enlarge
A plan for Scotland in two zones. Click to enlarge

The letter went on to suggest that, north of the recommended route for A8, which was Leith to Gourock, there would probably be enough zone 8 numbers to dispense with zone 9 completely. This was a serious proposal for a while and diagrams were drawn up to show how it might work (such as the one on the right - click to enlarge it)**. This interesting scheme had the rest of the zone 8 roads allocated as follows:

  • A80 Edinburgh - Inverness via ferry across Forth estuary
  • A81 Rosyth - Aberdeen via ferry across Tay and Dundee
  • A82 Aberdeen - Inverness
  • A83 Oban - Perth - Dundee
  • A84 Glasgow - Perth - Stonehaven
  • A85 Edinburgh - Stirling
  • A86 Glasgow - Fort William - Inverness
  • A87 Fort Augustus - Kyle of Lochalsh (for Skye)
  • A88 Inverness - Thurso

Why this exercise was carried out is a little obscure: the idea that there were enough numbers to go round doesn't seem a good enough reason to eliminate one whole zone. It is, of course, possible that - with Irish independence still a few months away and by no means certain - there was an idea of reserving zone 9 for Ireland.You probably only glanced at it before reading on, but the image at the very top of this page alongside the introduction shows the road numbers proposed for Aberdeen had it been in zone 8 instead of zone 9.

All the same, documentary evidence turns a little patchy, but there's general agreement that classification was done by 1922 (zone 9 included) and the numbers were cleared to go on signs. The small matter of telling the public about it was all that remained.

Going public

The various map publishers had long since known that road numbers were coming, and most of them were very eager to get hold of these exciting new codes that identified each link in the network. Bartholomew was reportedly so eager to get a head start on its competitors that it published a map with leaked information on the routes of A1 to A99. The company was left looking a little foolish when the final scheme was revealed with several noticeable differences the following year.

Michelin were certainly very keen. When the first batch were released in July 1921 they were invited to send representatives to the Ministry in order to inspect the new scheme. In order that there would be no doubt whatsoever, no matter how minor the question, a series of obsessively large-scale maps were drawn up for the handful of places where the running of two route numbers concurrently could not be avoided.

The A15 and A18 multiplex is made plain

The A15 and A18 multiplex is made plain

There was also a great deal of interest from members of the public, and archive documents on the subject contain an incredible number of letters asking how road numbers were allocated or how the system worked. Each one got its own reply, individually written but always along the same lines.

In 1941, the Encyclopædia Britannica expressed an interest and the response from the Ministry contained a very helpful (and, for once, concise) outline of their intentions for the different classes of road.

"Under this scheme the most important roads connecting large centres of population and other roads of outstanding importance from the point of view of through traffic were classified as Class I roads. Roads forming important links between Class I roads and between the smaller centres of population were classified as Class II roads. Roads of purely local importance were not classified."

Letter to Director of Research, Encyclopædia Britannica, 6 March 1941

Local authorities were actively encouraged to put numbers on new signs and to begin appending them to existing ones as soon as the first batch of numbers were made public in summer 1921. The Ministry of Transport covered all such costs, and once the first 99 A-roads were covered they were keen to see the signposting applied to progressively smaller and smaller routes.

B1217, running west of the A1. Shock horror!
B1217, running west of the A1. Shock horror!

It's just as well that it took a while to get the numbers out in public, because errors did creep in to the original system. In September 1924, the Ministry sent a letter to their Divisional Road Engineer in the Northern Division to ask whether any signs had yet been erected on the Hook Moor-Towton Road in West Yorkshire - the reason being that, while it started on the A1 and headed east, thus being entirely in zone 1, it was originally numbered B6379.

There was a sigh of relief when the response dropped onto the Ministerial doormat, because the answer was no, and the number B1217 was applied without having to change any signs. In 1999 the closure of its connection to the A1 meant that it was extended west, into zone 6, and is now once again wrongly numbered (the picture to the left was taken west of the A1). Sadly the B6379 couldn't possibly make a return from its 75 years in exile, because it now lies between Scholes and Low Moor elsewhere in the county.

T for trunk

That would be the end of the story if it wasn't for the fact that, by the 1930s, long-distance motoring was becoming a reality and the Ministry of Transport was concerned that its major long-distance routes were inconsistently maintained by County Councils. The Trunk Roads Act of 1936 aimed to sort things out by giving the Ministry direct control of major roads.

Proposed Trunk Road cartouche with mileage and number
Proposed Trunk Road cartouche with mileage and number

The Act stipulated that the new trunk routes must be numbered for reference, and there was much chin-stroking and head-scratching about what the numbers should be. There was every intention that they would replace the A-road numbers once the system was decided. The system of numbers was, in the main, produced by copying the number of the A-road that the trunk route used for the greatest length (T30 from London to Penzance), but included a few surprises such as the T42 from Birmingham to Birkenhead via Chester.

The trunk routes were at first intended to be conspicuous in every way. A draft policy document states:

"All posts on Trunk Roads carrying directional, mandatory or other signs to be painted red and white chequer instead of black and white as at present. This would ensure immediate recognition of any Trunk Road."

David Service, Roads Classification Section, 2 February 1937

They went a step further, suggesting that not only should the T-number appear on road signs, but a distinctive cartouche should be drawn up for use on road signs and it should be sited regularly on marker posts to show the mileage and other information (like the one above right). The impetus behind the numbers being public was - in all honesty - for maintenance crews to efficiently locate faults. For this reason alone it was expected that motorists between, for example, Doncaster and Grimsby would have to get used to driving on the T22 instead of the A18.

A15 and A18 become Trunk Road T22

A15 and A18 become Trunk Road T22

There was a simple reason for the numbers not being made public. In mid-1937 it was suddenly realised that the Ordnance Survey provided a cunning new way of locating points on their maps by use of the National Grid - so any point could be located to within a few yards by a string of co-ordinates.

It was discovered quite by accident, when a senior cartographer at the Ordnance Survey was enlisted to write a leaflet explaining the system, and sent his draft text to an old friend to get his comments on it. The old friend happened to be a senior Ministry of Transport civil servant working on the new Trunk Roads. He then realised that it was possible to send work crews out with a map and a set of co-ordinates, and to keep the whole messy business of Trunk Road numbers secret.

If it wasn't for the old school tie network, we might even today be driving on T-roads.

Motorway zones

The history of how motorways came to be numbered is every bit as interesting, and slightly more panicked, than this one. The full story is over at Pathetic Motorways.

* In 1922 Ireland declared its independence from the UK, inheriting classification data but no road numbers. Until 1977, its roads carried T and L numbers, not A and B.

** This plan isn't very well thought-out: there is an arc drawn showing how road numbers should increase as they progress clockwise around Edinburgh, but the ones labelled on the map do the opposite, with A81 on the east coast and A87 in the west.

With thanks to Brian Rolph and Nevis Hulme for information on this page.