ARRS - Association of Road Racing Statisticians

The origins of the marathon

To many people the marathon race began in 1896, but the event has a much longer history than that. Putting aside the myth of the Pheidippides connection, the distance of the original Olympic marathon was not just an accident of geography, the distance that needed to be covered between Marathon and Athens.

The current accepted distance is 42.195 km/26 miles 385 yards, but for the first thirty years of its Olympic history, and during the build up to, and following from, the 1896 inaugural modern Olympics, the 25 miles/40 km event was widely accepted as the standard distance. [With 25 miles being 40.233 km, the approximate distance 40 km was equally acceptable to both the metric and non-metric worlds, with many countries having races at 40.2 km, just 33 metres short of 25 miles.]

Professional pedestrians had been contesting the 25 mile distance for many years in Britain. As early as December 1824, the pedestrian John Townsend challenged Robert Skipper who had assumed the title "Champion of England" to a match over "25 miles of ground".

In November 1843 George Bradshaw beat the champion pedestrian Robert Fuller over 25 miles in 3:59 and less than a year later there was another "Great Footrace" between Bradshaw and a B Butler, who were described as "the two greatest pedestrians of this or of any antecedent period." This race was also 25 miles.

When amateurs began contesting long distances, the first 10 mile track mark was set in 1873 by William Fuller, (the first to run the distance within the hour). Later that year Sydney Weall failed in an attempt to cover 20 miles in 2 Hours on the road. {Road racing in this period, being a 'free show' was unusual.} However a 25 mile track walk, in May 1876 was successfully completed and marked the beginning of the amateur events at this distance.

The first go-as-you-please track event for amateurs first took place in 1879 1. By 1881 amateur distance running had progressed so far that the British runner George Dunning set the absolute world best time ever over the 25 mile distance, 2:33:44.0, surpassing even marks made by professionals.

Elsewhere in Europe races of approximately 40km were held. In 1885 Frenchman Louis Saussus won a 38 km (22.5 mile) race between Paris to Versailles in 2:36:30. In 1892 another Frenchman, Auguste Marchais ran 36.7 km in Paris in three hours.

From 1892 onwards in Sweden, 40 km races between Stockholm and Sudertalje and Norrkoping and Linkoping, were held with the former becoming a regular event, varying in length between 33, 35 and 37 km. In 1894, Julius Olsson recorded a time for the course of 2:27:17.6. In 1895, there had been a 36 km race held in Frankfurt in Germany.

In the book The Pedestrian Record by James Lupton published in 1890, which listed the records for every distance and time frame, the 25 mile record of Dunning is shown as the absolute best for the distance. This book codified and reflected what was published in the lengthy, detailed newspaper accounts and reports of the period.

During an era when the leading distance runners were overwhelmingly professional, for an amateur to hold the absolute best time for an event was very unusual, and would have had a particular importance within the Official Amateur hierarchy. This hierarchy was anxious to maintain and promote its position in relation to the professional sport.

When Pierre Coubertin was commissioned by the French Government to undertake a study of modern physical culture methods in 1889, and he subsequently visited Britain, the book, The Pedestrian Record, with its records would undoubtedly have been shown to him. The Olympics that developed from his study were intended to be a celebration of amateur sport - professionals were not allowed to compete.

The 25 mile, the only distance event at which an amateur was demonstratively superior to the professionals, would have had a particular attraction for the new International Olympic Committee when pondering the possible length of the distance race from Marathon to Athens proposed by Michel Breal.

The shortest distance between the village of Marathon and the Panathenic Stadium in Athens was probably around 34 km. Roger Gynn and David Martin actually make the point in their book The Olympic Marathon that the longer coastal route between the village and the capital in 1896 was perhaps chosen because the 40 km/25 mile distance was more familiar to those accustomed to the imperial measurements. (In the 1890s in the first athletic championships held in Berlin in Germany, races were held over the 100 yards and one mile.)

The British runner Len Hurst was one of the early great "marathon" runners. He was successful in the most competitive professional 'marathon' events in the French Paris-Conflans races which were over 40 km during the 1896-1904 period. Hurst's career also shows the importance of the 40 km/25 mile distance at the beginning of the twentieth century.

On 23 September, 1901 Hurst won the professional 25 Mile world championships at the Tee-To-Tum ground at Stamford Hill in London. Two years later, on the 27th August 1903, Hurst set a professional world record over the 25 mile distance of 2:32:42, just clipping Dunning's 1881 time. He had been paced to break Dunning's mark, which one must suspect, was an affront and a challenge to the pride and status of professional runners.

During the early years of the 20th century, Olympic marathons were being held over approximately the same distance. The 1900 Paris Olympic marathon was 40.260 km, the St Louis Marathon in 1904 was 40 km, [reported as such in the daily program] the Intercalated Athens Marathon in 1906 41.860 km. The actual exact measurement of all these races is uncertain but the key point is that the official nominal distance was around 40 km.

In line with this, across the globe, countries had held their inaugural 'marathons' over the approximate 40 km distance. Gyula Kellner of Hungary had earned his trip to the 1896 Olympic Games by running a 40 km trial run in three hours in Budapest. Norway had their first 'marathon' over 40.2 km in 1896. The first non-European marathon was held in New York in 1896 over 25 miles [40 km]. The first Boston marathon in 1897 was over 24.7 miles/39.75 km. The first Danish race in 1898 was over 40.2 km, and the same year Arthur Techtow won the first German 'marathon', yet again over 40 km. The first Swiss and Austrian races in 1901 in Geneva and Vienna were over 40.2 km and 40 km respectively.

The first Finnish race in 1906, was over 40.2 km. 40km was chosen for the first 'marathon' race in what was to become Czechoslovakia, at Dobris in 1908, and across the Atlantic in Canada the same year, 25 miles.

The first Union championships in South Africa that year was also over a distance of 25 miles. An earlier event held in September 1903 over 25 mile from Durban to Isipingo and back had attracted 148 entrants.

The first marathon run in the Russian Empire in Kiev in 1913 was a race of 38 verstas or 40.2 km, won by Aleksandr Maksimov. 25 miles was the distance of the first Japanese marathon held in 1911. The 'marathon' held as part of the Far Eastern Games Marathon in 1917 in Tokyo was once again 25 miles.

The early Olympic trials held by the USA show the same degree of consistency. The 1900 race was held over 38.752 km, the 1908 over 39.750 km, and the two 1912 trials over 39.750 km and 40.233 km.

The original programme for the 1908 Olympics submitted by the British organisers and approved by the IOC at its meeting at The Hague in the spring of 1907 stated: "The Marathon race of 40 kilometres will be run on a course marked out on public roads by the Amateur Athletic Association and will finish on the running track inside the Stadium, where the last 1/3 of a mile will be run (1 lap = 536 metres)." (Offical Report of the 1908 Olympic Games, page 410) This would have been sent out to the various competing nations. It is more than likely that Dorando Pietri and others came to London expecting a 40 km race.

In 1912, the IAAF was formed and shortly afterwards the first list of official world track record was published. This list is of particular interest and seems to indicate what was viewed as the "correct" distance for the 'marathon' at that time.

In May that year the British runner, Harry Green, had set the world's fastest time for the 1908 Olympic marathon distance of 26 miles 385y/42.195 km on the track. However it was not that mark of 2:38:16.2 that the IAAF chose to ratify and include in their new world record schedule. Instead they selected his 25 mile split which had broken George Dunning's thirty-one year old track record - 2:29:29.4.

Why Green's 25 mile split was chosen in preference to his final marathon time is unknown. However as we have seen, the 40 km/25 mile distance had been widely accepted internationally as the 'marathon' distance. One could speculate that to the new amateur body, perhaps uninformed about distance running, Green's 25 mile time would appear superior to the professional marathon runner Willie Kolehmainen's 2:29:39.2 set over the longer distance in 1912. Possibly the longer distance was seen as being tainted by its adoption as the marathon standard by professional runners.

The 25 mile mark was to stay on the IAAF record books for years. In October 1934, the Italian Michele Fanelli set a new world record at 2:26:10.8, a time that was still listed when the IAAF finally removed the event from their record list in 1954. Green's British record had been broken by Frank Farmer in 1939. In 1961, the Dane, Thyge Togerson, ran 2:15:30 for a recognised national 40 km record on the track at Zaandam. The British Road Runners Club was to maintain the world records for the 25 mile event after that and the current best is held by Eric Austin with 2:10:48.0 set nearly forty years ago. More recently John Cramer (1974) and Barney Klecker (1981) set US national 25 mile records - 2:28:50 and 2:15:56.9.

Although it is very well known that the Italian, Dorando Pietri, was the catalyst for the change to what with hindsight seems an the present unlikely distance, ironically he was selected to run in the 1908 Olympics because he set a 40 km track record!

To cover familiar territory, for various reasons the 1908 London Olympic Marathon had been extended to well over 26 miles/42.195 km. Dorando Pietri collapsed as he was coming into the stadium. The controversy over his disqualification for being helped over the line created a whole series of lucrative professional marathon races, the Marathon Craze which to sustain the link with the controversy, were exactly the same distance.

Despite the Marathon Craze, the next Olympic Marathon in 1912 was over 40.2 km. The 1920 race was stretched to 42.75 km and it was not until 1924 that the present distance was adopted for the Olympic marathon. The International Olympic Committee decided that this race and all subsequent Olympic Marathons would be held over 42.195 km. Why this decision was made is unknown. Perhaps a compromise between the metric and imperial worlds, between 40 km and 25 miles, choosing a totally different distance? Perhaps pressure from the Anglo-Americans who, as a result of the prolific number of professional events in their countries, had adopted the Dorando distance as the standard?

Thus, the origins of the marathon are focussed at the 25 mile/40 km distance. It was only later that the longer distance was officially recognised - and that was not until 1924. Despite the IOC's decision, in Scandinavia and elsewhere, the 40 km and 40.2 km distances continued to be recognised for the "marathon" event for many decades afterwards, into the late 1920s and beyond. Even as late as 1946 the European marathon championship held in Oslo NOR, was held at 40 km.

If the aberration of the longer Olympic marathon in London in 1908 had been ignored and the distance continued to be standardised at 40 km/25 miles, then the subsequent history of the event would have been changed, probably resulting in different winners of major events. Etienne Gailly might have won the 1948 Olympics, and perhaps Jim Peters would have won the 1954 Vancouver Commonwealth Games marathon, instead of collapsing before the finish.

The Two Hour marathon would now be a reality. The splits by Haile Gebrselassie in his recent marathon world record, show he passed 40 km in 1:58:08, and would have also clocked well under the two hours for the slightly longer 25 mile run.

So the marathon race has a history that extends way before 1896, a history obscured by the IOC's decision to standardise the Olympic Marathon at the 1908 length, thus cutting the event's link with the original Greek race in the first Olympics, and with its deep roots into nineteenth century distance running.

1 In the eighteenth century British distance runners like Woolley Morris and Gruffydd Morgan died after over-extending themselves at ten miles. By the 1840s and 1850s 20 miles had been run just under 2 hours but in general distances beyond that were contested by walkers. Wagers usually related to what another pedestrian had done, and record setting was still an emerging concept. In an era when sudden death was a frequent occurrence, there was no incentive for a runner to extend himself and race at a longer distance. However long distance walks were sometimes go-as-you-please, which meant a mixture of running and walking, running to ease cramp for example, as opposed to 'fair heel and toe'. With 6 day walking races with their extensive prize money becoming go-as-you-please from March 1878, the incentive was then there for professional runners to learn how to pace longer events and to contest the distances formerly the province of walkers. As their knowledge and experience increased, soon professional runners rapidly reduced the best times for 25 miles and beyond. This advance was soon emulated by the amateurs, with the first go-as-you-please distance amateur track event held in April 1879.

This article was written by Andy Milroy. The author would like to thank Peter Lovesey and Jacques Carmelli for their help with this article.