1In France, the first postal roads were created in the fifteenth century and then largely extended by the French monarchy from the late seventeenth century, especially after the ambitious program of road construction planned in the "Orry Ordinance" (1738).
2This text gives the main principles of the road policy construction of the XVIIIth century (Lepetit, 1984), fixing the priority of construction and the hierarchy of the itineraries. Main roads ("grandes routes") go from Paris to main ports and borders, simple roads ("routes") link Paris to other regional capitals, main paths ("grands chemins") connect regional capitals, royal paths ("chemins royaux") connect non regional capitals, and traverse path ("chemins de traverse") go from city to city, without postal service.
3The nomination of Daniel-Charles Trudaine to the direction of Ponts et Chaussées in 1743 gave a vigorous impulse to this program, which from this point had the backing of a corps of engineers (Arbellot, 1973). Moreover, notable advances were made with regard to construction and maintenance of paved roads, thanks to the Mémoire presented in 1775 by the chief engineer of the généralité of Limoges, Pierre Trésaguet. In 1776, all the "main roads" from Paris and a part of other roads planned by the Orry Ordinance are finished, which represent a total of 14 000 km. At the beginning of the 1780’s, more than 26 000 km are officially opened in the kingdom, some are still in construction and the last part in maintenance. In 1790, almost all of the roads between regional capitals and main economic centers are finished, even if their maintenance raises major issues, as testified by a contemporary traveler (Young, 1792) (figure 1).
Figure 1: The evolution of the postal road network, digitized from 1632 to 1833
Source: Mimeur et al., forthcoming
4The horse postal service used these royal roads to transport people who were wealthy enough to pay for such prohibitively expensive transport, as well as, in some cases, the staff of the letter post office, who transported bundles of mail. Relays are set out on the roads at an average distance of 10 to 15 km. Part of the mail concerned the administration; another part was transported by the postal letter service to the nearest office, to be collected by users (Belloc, 1886). The relays were managed by post-masters ("Maîtres de Postes"), generally inn-keepers or substantial farmers, who offered a certain number of horses to be used in the system in return for remuneration. Post-horse riders brought the rented horses as far as the next relay, then led them back (figure 2).
Figure 2: The mail coach in front of a horse post relay
Source: A 1805 model, engraving in 1817 by V.J. Adam and L.P.A. Bichebois from a painting by J.A. Duclos. Copyright Photo L’Adresse Musée de la Poste, Paris/La Poste
5The postal route played an essential role in the diffusion of information: for the merchants or bankers established in cities, it allowed for the transmission of commercial instructions, means of payment, account documents, information on the level of prices or credit, on the size of expected harvests or arrivals of merchandises (Cauvin et al., 1987); for the monarchical state, it was a way of quickly sending policies to its administrators to be implemented. Private systems for transporting people and goods existed at the same time, but did not benefit from the privileges of the horse-drawn post office (Conchon, 2008). In addition, coastal navigation was important for exchanging information in regions along the shore. The system of postal routes thus assured only a part of these exchanges, but a part of increasing importance throughout the eighteenth century, as measured by the amount of territory covered (Bretagnolle, Franc, 2021).
6The dataset was created from 2002 but evolved through different versions. The last corrections were added in 2024. The method used to capture stations and segments of the routes that made up the network at a particular date was developed in a three-step process: the display of the group of towns in a specific region in the GIS, the manual identification of the stations, and the sketching of the road segments that connected each of these nuclei.
7In order to produce a list of post houses as exhaustive as possible over the period we were interested in, we began by exploiting the post books as they exist in the Bibliothèque du Musée de la Poste in Paris. This library holds the most comprehensive collection of post books, which were published annually. Post books are usually composed by route, indicating where the route starts and ends, for instance from Paris to Lyon (figure 3). From distance to distance, relays are added to keep the route running. The importance of these places in the transport system is confirmed by the fact that the distance between each relay is indicated as "post", "post and a half", and that these distances correspond to fares. In the above example, to go from Paris to Lyon, you would pay for 64 and a half posts.
Figure 3: An extract from the Post book (1733), relays between Paris and Nemours on the road from Paris to Lyon
Source: Musée de la Poste
8At this stage of the research, we started with the first volume held in this library, the 1708 one, and then chose a 25-year increment to go up to 1833, roughly corresponding to the arrival of the railroads in France. Once the horse posts had been entered, we decided to extend the search back to the 17th century, based on the first printed map of the Poste aux chevaux, drawn by Nicolas Sanson in 1632 (figure 4), many copies of which existing in libraries. This map is also known as one of the first thematic maps showing a network (Palsky, 1996, p.21-22). The survey of post houses based on Sanson's map was then based on the 1708 database, which greatly facilitated data entry by retropolating the information.
Figure 4: The Nicolas Sanson d’Abbeville’s map (1632) "Carte géographique des Postes qui traversent la France"1
Source: Bibliothèque nationale de France
9The location of horse posts differs from letter post offices (figure 5). Relays are places where you can change horses regularly. They are not necessarily located in densely populated areas, but in places that regularly punctuate the horses' itinerary. The aim is to have horses that are always able to gallop, which means that they are sometimes located far from densely populated areas. The letter post offices are where the bundles of letters arrive. This is where, before home delivery was introduced, the neighbouring population came to collect their letters, so the geography of the mailboxes was very similar to the demographics of the time. Wherever there were enough letter readers, there was a post office. There were relationships between these two geographies, one corresponding to a network, the other to places of settlement. The horse post routes start and finish in places where the population is concentrated, which are also places of power, for example Paris and Lyon. Relays could be set up in intermediate towns, villages and market towns. But not all populated areas were systematically served by horse posts. Letter post was therefore obliged to provide transport between the relay stations and post offices that were not on the same route as horse posts.
- 2 The letter post offices database has been recently created by Nicolas Verdier in the framework of a (...)
Figure 5: Horse post relays and letter post offices2 in the west of France from 1706 to 1833
Verdier, 2022.
10The particular geography of the horse post relays along a route meant that some relays were in very small localities, which the communal structure resulting from the French Revolution did not take into account. Their names are sometimes difficult to find. In this context, we used the BD NYME database, which is the IGN's former toponymy database now part of the BD TOPO, to find the correspondence between the historical name of the relay station and its current name, and then the name of the current commune in which it is located (when these names are different).
11The different columns in the Relays database are:
-
ID of the relay (a simple number chosen by us during the digitalization process, relays that last several years keep the same ID)
-
Historical name of the relay (the one present in the Post books when different from the current one)
-
Current name of the relay
-
Name of the commune in which the relay is located
-
Current ID of the commune in which the relay is located (it is the INSEE code of the commune, i.e. the number of the department followed by the number of the commune inside the department, 1999)
-
Latitude and Longitude of the relay (city hall coordinates if no current name is set, place named as the current name coordinates otherwise)
12There were 611 relays in 1632 and 1480 in 1833 (table 1).
13Roads have been digitized as the crow flies between two adjacent relays according to the historical sources. Each segment road goes from relay A to relay B. The graph is not oriented.
14A first digitalization was realized in 2002-2004, by Anne Bretagnolle and Nicolas Verdier (see the publications between 2005 and 2009).
15A second one, more systematic, was realized in 2010 by Timothée Giraud in order to extend the spatial and temporal coverage of the database (see Bretagnolle, Giraud, Verdier, 2010). At this time, different attributes were present in the database:
-
ID of the segment road (a simple number chosen by us during the digitalization process, segments that last several years keep the same ID)
-
ID of the two different relays at the beginning and at the end of the segment road, these identifiers match the ones of the relays dataset
-
- 3 See Bretagnolle, Verdier, 2005b.
Length in number of "posts", which was the historical count unit of the time3
-
Length in kilometers (Euclidean distance)
- 4 We used Roche (2010) and Minetti (2004) in order to model the evolution of horse speed.
16Since 2019, we have been working with Christophe Mimeur in order to estimate an average speed on each postal road, thanks to the different methods developed by him in his PhD (Mimeur, 2016). Euclidian distances have been weighted by the intensity of slopes (we tried both maximal and average slope) and an evolving horse speed, called maximal speed4 (due to major improvements in roads and stagecoach construction, horse speed was multiplied by two between 1765 and 1780).
17Four different attributes were added in the database:
-
- 5 Slopes were computed from a Digital Elevation Model given by the Alti database, based on a minimal (...)
The maximal slope (in degrees) of each road segment between two adjacent relays5
-
The average slope (in degrees) of each road segment between two adjacent relays
-
The average speed (in km/h) estimated by the model on each road segment
18Between two post houses k and l, this average speed is given by (equation 1):
- 6 Mimeur et al., forthcoming.
where alpha is the average slope between k and l, distance is the length of the segment between k and l and beta is the coefficient of reduction. When the result is less than 0.5 km/h, we retained the minimum value of 0.5 km/h6.
-
The estimated time (in mn) used to travel along each road segment
19The length of the digitized network went from 7 350 km in 1632 to 20 505 in 1833 (table 1).
20The boundaries of mainland France were reconstructed by comparing four publications. The first is Daniel Nordman's Frontières de France, published in 1998 (Nordman, 1998), which provides details of certain borders at very specific times. The second is André Burguière and Jacques Revel's book L'Espace français, published in 1989 (Burguière, Revel, 1989). The third is the Atlas de la Révolution française vol. 4 and 5, for the revolutionary period (Nordman et al., 1989). Finally, the boundary maps have been reworked from the Atlas de l'histoire de France XVIème XIXème siècle (Boutier, 2006).
21We used different historical databases in order to enrich the relay database by including the historical population of the communes in which the relays are located.
22Between 1700 and 1780, urban populations were collected by Paul Bairoch and collaborators (Bairoch et al., 1988, see 1700 and 1780) and partly corrected or completed by Bernard Lepetit (1988, see 1700, 1750 and 1780). These data must be used with caution as historical sources only concern the largest cities.
23We also integrated the urban population counts given for 1809 (Le Mée, 1971) and for 1831 (from the Bulletin des Lois, one of the first census in France, with population collected not at the scale of the whole commune but at the scale of the urban area, as in the paper of Roger Le Mée, see Bretagnolle, 1999, 2009); on the difficulties with these censuses, see Verdier (2011).
Table 1: Total number of postal relays and total length of the network in the database
Date
|
Number of relays
|
Total length (in km)
|
1632
|
611
|
6 998
|
1708
|
832
|
10 638
|
1733
|
887
|
11 756
|
1758
|
1117
|
16 528
|
1783
|
1448
|
23 855
|
1810
|
1323
|
20 899
|
1833
|
1480
|
27 207
|
24The database can be downloaded in different formats. For a GIS or Spatial usage, a single geopackage file contains most of the database. Several GIS layers are available within the Post_Roads.gpkg file.
25Postal relays layers are the following: postal relays can be found relays_1632, relays_1708, relays_1733, relays_1758, relays_1783, relays_1810 and relays_1833.
26Postal roads are in the following layers: roads_1632, roads_1708, roads_1733, rroads_1758, roads_1783, roads_1810 and roads_1833.
27Historical boundaries of France are in the following layers: France_1632, France_1708, France_1733, France_1758, France_1783, France_1810 and France_1833.
28Alternatively relays and roads can be downloaded as csv files. This format may be easier to manipulate in non-spatial context (network analysis for example). These csv files use the prefixes "roads" or "relays" and dates as suffixes.
29The population database can be found in the population.csv file.
Table 2: Description of the files
Files names
|
Layers
|
Type of data
|
Fields description
|
Post_Roads.gpkg
|
relays_1632 relays_1708 relays_1733 relays_1758 relays_1783 relays_1810 relays_1833
|
vector data, POINTS
|
Postal relays at each date with: - IDREL: ID of the relay - HISTORIC_NAME: Historical name of the relay - CURRENT_NAME: Current name of the relay - COMMUNE_NAME: Name of the commune in which the relay is located - COMMUNE_ID: Current INSEE code of the commune - COORDINATES_2154_X & COORDINATES_2154_Y: Coordinates RGF93 / Lambert 93 (EPSG:2154) - COORDINATES_4326_X & COORDINATES_4326_Y: Coordinates WGS 1984 (EPSG:4326)
|
Post_Roads.gpkg
|
roads_1632 roads_1708 roads_1733 roads_1758 roads_1783 roads_1810 roads_1833
|
vector data, LINES
|
Road segment at each date between two adjacent relays with: - IDROAD: ID of the segment road between a relay A and a relay B - IDREL_A: ID of the relay A - IDREL_B: ID of the relay B - N_POST: Number of posts between relay A and relay B at each date except 1632 - LENGTH: distance between relay A and relay B, in kilometers - MAX_SLOPE: Maximal slope of the segment road in degrees - AVG_SLOPE: Average slope of the segment road, in degrees - SPEED: Average speed of the horses in the segment road, in kilometers per hour - TIME: Average time to go from relay A to relay B, in minutes
|
Post_Roads.gpkg
|
France_1632 France_1708 France_1733 France_1758 France_1783 France_1810 France_1833
|
vector data, POLYGONS
|
Historical boundaries of France in 1632, 1708, 1733, 1758 (same as in 1733), 1783, 1810 and 1833.
|
relays_1632.csv relays_1708.csv relays_1733.csv relays_1758.csv relays_1783.csv relays_1810.csv relays_1833.csv
|
|
tabular data
|
Postal relays at each date with: - IDREL: ID of the relay - HISTORIC_NAME: Historical name of the relay - CURRENT_NAME: Current name of the relay - COMMUNE_NAME: Name of the commune in which the relay is located - COMMUNE_ID: Current INSEE code of the commune - COORDINATES_2154_X & COORDINATES_2154_Y: Coordinates RGF93 / Lambert 93 (EPSG:2154) - COORDINATES_4326_X & COORDINATES_4326_Y: Coordinates WGS 1984 (EPSG:4326)
|
roads_1632.csv roads_1708.csv roads_1733.csv roads_1758.csv roads_1783.csv roads_1810.csv roads_1833.csv
|
|
tabular data
|
Road segment at each date between two adjacent relays with: - IDROAD: ID of the segment road between a relay A and a relay B - IDREL_A: ID of the relay A - IDREL_B: ID of the relay B - N_POST: Number of posts between relay A and relay B at each date except 1632 - LENGTH: distance between relay A and relay B, in kilometers - MAX_SLOPE: Maximal slope of the segment road in degrees - AVG_SLOPE: Average slope of the segment road, in degrees - SPEED: Average speed of the horses in the segment road, in kilometers per hour - TIME: Average time to go from relay A to relay B, in minutes
|
population.csv
|
|
|
- IDREL: ID of the relay - COMMUNE_ID: Current INSEE code of the commune POP_1600, POP_1700, POP_1750, POP_1780, ¨POP¨_1809, POP_1831: Population of the commune of the relay in 1600, 1700, 1750, 1780, 1809 and 1831 (when the information is available)
|
30Anne Bretagnolle and Nicolas Verdier analyzed the extension of the postal horse network (length, density by surface, regional differences, etc.) (Bretagnolle, Verdier 2005a, 2007, in French and in English), the evolution and interpretation of the Post measure of length (Bretagnolle, Verdier, 2005b) and the evolution of the location of the gravity center of the network (Bretagnolle, 2009); as well as the question of changing relationships to space (Verdier, 2008a, 2009a), the relationship between the hierarchies of networks and cities (Verdier, 2008b) and finally the question of instabilities at the heart of the network (Verdier, 2009b).
31Anne Bretagnolle, Timothée Giraud and Nicolas Verdier published a modeling of a general accessibility index (using a 900 points grid, we calculated the accessibility "at every place in France"), with a Shimbel index. They considered a secondary road network, characterized by a slower transportation speed, and weighted the accessibility values by geometrical accessibility. They generalized the results by interpolation methods (Bretagnolle, Giraud, Verdier, 2010).
32Anne Bretagnolle and Alain Franc (INRA, UMR Biogeco) published different papers in the framework of the Transmondyn Project (directed by Lena Sander). They worked on a resistance measure (a summary of the facility to go from one relay to another) and computed a betweenness centrality measure and anamorphoses of the postal network. They also modeled regional urban systems based on a Reilly model and postal horse distances (Bretagnolle, Franc, 2017a; Bretagnolle, Franc, 2017b, in French and 2021 in English; Bretagnolle, Franc, 2020). Nicolas Verdier has continued to work on the relationship between network hierarchies and urban hierarchies (Verdier, 2016).
33Anne Bretagnolle, Christophe Mimeur and Nicolas Verdier modeled the postal speed in order to construct anamorphoses. They modeled historical distances by considering a multigraph (pedestrian paths and postal roads) with edges weighted by the intensity of slopes and an evolving horse speed. The multigraph has a continuous coverage: it allows to estimate historical distances from every commune, with or without relays to another, by using postal or simple roads. The resulting anamorphosis maps show a more and more distorted France, with a shrinking in the north of a Rennes-Genève line despite low relief, and a massive dilation of higher mountains and fragmentation around a multiplication of speed archipelagos in the southern. They interpreted this evolution as the replacement of the pedestrian mobility regime by a new horse mobility regime (Mimeur et al., forthcoming). Nicolas Verdier, after working on the other post, that of letters, has begun to work on the relationship between these two institutions (Verdier, forthcoming).
Bretagnolle A., 2009, "Analyse morphodynamique du réseau des routes de poste en France (XVIe-XIXe siècles)", in Bleton-Ruget A., Commerçon N., Vanier M. (eds), Réseaux en question : utopies, pratiques et prospective, Macon, Institut de recherche du Val de Saône-Maconnais, 117-129, [online]. URL: http://halshs.archives-ouvertes.fr/halshs-00259074/fr/.
Bretagnolle A., Franc A., 2017a, "Emergence of an integrated city-system in France (XVIIth–XIXth centuries): evidence from toolset in graph theory", Historical Methods: A Journal of Quantitative and Interdisciplinary History, Vol.50, No.1, 49-65, [online]. URL: https://hal.science/hal-01603847
Bretagnolle A., Franc A., 2017b, "Vers des systèmes de villes intégrés (France, XVIIIe-XIXe siècles)", in Sanders L. (ed.), Peupler la terre. De la préhistoire à l’ère des métropoles, Presses universitaires François Rabelais, Collection Villes et Territoires, 333-361, [online]. URL: https://hal.science/hal-03566354
Bretagnolle A., Franc A., 2020, "Les routes de la poste à cheval et les petites villes en France (1632-1833)", in Perol C. (ed.), Routes et petites villes de l'Antiquité à l'Époque moderne, Clermont-Ferrand, Presses Universitaires Blaise Pascal, collection Histoires Croisées, 19-41, [online]. URL: https://shs.hal.science/halshs-03823788
Bretagnolle A., Franc A., 2021, "Towards integrated systems of cities (France, 18th-19th centuries)", in Sanders L. (ed.), Settling the world. From Prehistoric to the Metropolis Era, Tours, Presses Universitaires François Rabelais, [online]. URL: https://hal.science/hal-03814678
Bretagnolle A., Franc A., Verdier N., 2020, "Les routes de la poste à cheval en Auvergne", in Gomis S. (dir.), Atlas historique Auvergne, Bourbonnais, Velay, Centre Histoire et Culture, Université Clermont-Auvergne, [online]. URL: https://atlas-historique-auvergne.msh.uca.fr/.
Bretagnolle A., Giraud T., Verdier N., 2010, "Modéliser l’efficacité d’un réseau : le cas des routes de poste en France (1632, 1833)", L’Espace Géographique, Vol.2, 117-131, [online]. URL: https://shs.hal.science/halshs-00504320v1/file/Bretagnolle_Giraud_Verdier.pdf
Bretagnolle A., Verdier N., 2005a, "Images d’un réseau en évolution : les routes de poste dans la France pré-industrielle", Mappemonde, Vol.79, [online]. URL: http://mappemonde.mgm.fr/num7/articles/art05301.html
Bretagnolle A., Verdier N., 2005b, "La mesure postale, reflet des évolutions dans le rapport au temps et à la distance au siècle des Lumières", Cahiers de Métrologie, Vol.22-23, 63-74, [online]. URL: http://hal.archives-ouvertes.fr/hal-00116944/en/.
Bretagnolle A., Verdier N., 2007, "Expanding the network of Postal routes in France (1708-1833)", in Le Roux (ed.), Post Offices of Europe 18th-21th Century. A Comparative History, Paris, Comité pour l’Histoire de la Poste, 155-173, [online]. URL: http://halshs.archives-ouvertes.fr/halshs-00144693/en/.
Bretagnolle A., Verdier N., 2007, "L’extension du réseau des routes de poste en France (1708-1833)", in Le Roux (ed.), Postes d’Europe 18ème-21ème siècles. Jalons d’une histoire comparée, Paris, Comité pour l’Histoire de la Poste, 155-173, [online]. URL: https://shs.hal.science/halshs-00144693
Bretagnolle A., Verdier N., 2014, "Les routes de la poste à cheval, de 1632 à 1833", in Zadora-Rio E. (ed.), Atlas Archéologique de Touraine, [online]. URL: http://a2t.univ-tours.fr/notice.php?id=12
Mimeur C., Verdier N., Bretagnolle A., forthcoming, Modeling the transformation of the French postal exchange space through multipolar anamorphoses (1632-1833).
Verdier N., 2008a, "A Méltanyossag logikajatol az elvont mértékegységig", Korall Tarsadalomtörténeti Folyoirat, No.32, 31-59, [online]. URL: https://shs.hal.science/halshs-00331428/document
Verdier N., 2008b, "Gerarchie urbane e città in rete : l’urbanizzazione della Francia tra 1700 e 1836", in Iachello E., Militello P. (dir.), L’insediamento nella Sicillia d’étà moderna et contemporanea, Bari, Epuglia. 46-58, [online]. URL: https://shs.hal.science/halshs-00357347/document
Verdier N., 2009a, "Dal territorio alla carta : posta a cavallo e acculturazione dei rapporti col spazio tra la fine del settecento e l'inizio dell'ottocento", Quaderni Storici, Vol.131, 579-606, [online]. URL: https://shs.hal.science/halshs-00429232/document
Verdier N., 2009b, "La labilité du réseau aux XVIIIe et XIXe siècles. Le cas de la Poste aux Chevaux dans le quart nord-Ouest de la France", in, Robert S., Verdier N. (dir.), "Du sentier à la route, une archéologie des réseaux viaires" Nouvelles de l'Archéologie, No.115, 13-17, 53-56, [online]. URL: https://journals.openedition.org/nda/645?lang=en
Verdier N., 2016, "Avant le chemin de fer, quelles relations entre réseau de transport et urbanité ?" in Térade A. (dir.), Infrastructures de transport et transformations des villes et du territoire - XVIIIe-XXIe siècles, Vol.2, Paris, Les Cahiers de l’IPRAUS, 17-26.
Verdier N., forthcoming, "La poste aux chevaux et la poste aux lettres en Bretagne aux XVIIIe et XIXe siècles", in Colleoni F., Watteaux M. (dir.), Chemins et mobilités bretonnes. Regards croisés sur la protohistoire et l'histoire des circulations en Bretagne, Rennes, Presses Universitaires de Rennes.
34The first limit concerns the time period covered by the database. The final date that we chose is 1833, a bit far away from the period of the largest expansion of the network, which is around the beginning of the 1850s. Due to the huge time spent in the digitalization of the relays, we never achieved our initial goal, which was completing the GIS with another final date (around 1855). The difference may be important, for instance in Brittany, which was the last region in France to be equipped in postal relays.
35Another limit concerns the degree of confidence related to the location of relays, especially in mountains (Alpes for instance). Sometimes, even with the help of the database BD NYME, it was not possible to find the correspondence between the historical name of the relay and the present one. For this reason, the database must be used with caution in very low density places.
36A third limit is related to the road digitalization method. Roads have been digitized as the crow flies between two adjacent relays according to the historical sources. As such, they don’t follow the real trajectory of the historical roads. It would have been impossible to find these real trajectories, if only because itineraries were fluctuant, from one year to another, and even from one season to another one. When analyzing the results at a national or regional scale it does not introduce huge differences, but at a local level, the geometrical aspect of the modeled roads strongly appears, as for instance in the map built for the Atlas de la Touraine (Bretagnolle, Verdier, 2014) (figure 6).
Figure 6: The postal road network in the Indre-et-Loire department, in 1833
Source: Bretagnolle, Verdier, 2014
37Another limitation of the database is the fact that the graph is not oriented, whereas some of the postal roads were used only in one direction. For instance, some water roads were used instead of land roads when going down rivers, as for instance, the use of barges on the Saône between Lyon and Chalon. Likewise, when modeling an average speed, we did not consider the difference between up and down senses in mountain regions, which in fact introduced huge differences in the rhythm of horses.
38Several Historical GIS exist in the world. The first ones were national ones, like in China (Bol, Ge, 2005) or Great-Britain (Gregory, 2005), mainly targeting the distribution of population in the past. The Postal horse relays and roads in France from the 17th to the 19th centuries database is one of the first initiatives for reconstructing geohistorical networks in France. Other initiatives focus on infrastructures, like the one dedicated to French railroads called FRANcE (Thevenin et al., 2016), including stations and railways. FRANcE also deals with the evolution of speed between 1830 and today. Another database focuses on canals and navigable rivers (Litvine et al., 2023), and it relies on infrastructure since 1790: locks, barrages, bypass. Other geohistorical databases have been built in France, based on automatic algorithms to deal with digitized maps, like the Cassini network (Perret et al., 2015).
39In archaeology and geo-history, a lot of researchers are now interested in building and sharing a geo-historical referential that would allow people from different institutions and disciplines for working together through a common geographical space (see the Projet Time Machine, a consortium of the Infrastructure of the Humanities Huma-Num: https://ptm.huma-num.fr/). Thanks to the Zenodo open access and present data paper, the Postal horse relays and roads in France from the 17th to the 19th centuries database will be indexed in this consortium and integrated in the interface of the Huma-Num website, among the 198 actual databases currently shared through the Projet Time Machine/Humanum website7.
40This database could also be specifically used in economic history, among other historical databases, for econometric estimations. Indeed, geohistorical data are appropriate in order to limit endogeneity between several variables (Duranton, Turner, 2012), especially in a long time period. Many historical variables could be used as instrumental variables, by assuming they are highly correlated with explicative variables but not related with the dependent one. This database would also have a huge significance for researchers, geographers, historians, economists, archaeologists, who model multimodal accessibility over centuries.
41Another advantage of this database is that it provides precise information on the best routes available for rapid transport between the beginning of the 18th century and the first third of the 19th century. This network can be used in conjunction with the data from the Geohistoricaldata project (https://geohistoricaldata.org/) (Dumenieu et al., 2019). Numerous analysis are possible, such as those carried out by Davide Gherdevich for the Friuli Venezia Giulia region of Italy in the medieval period (Ghedevich, 2017). We should also consider the serialisation of databases of the fastest routes, such as that developed by Thomas Thevenin for the following period (Mimeur, Thevenin, 2020).
42Beyond reuses in academic research, this database could also benefit to a large community of potential users, for instance simple citizens trying to find historical information about some places in France (we received a lot of emails asking about the existence of an old postal relay in one area or another), but also associations or people involved in web-mapping who want to find open-access historical data related to their domain.
43Repository location: https://doi.org/10.5281/zenodo.11196161
44Authors: Nicolas Verdier, Timothée Giraud, Christophe Mimeur, Anne Bretagnolle
45Publication date: February 2025
46Langage: English
47Time period covered: 1632-1831
48Data type: vector data, tabular data
49Data sources: 1) For the historical relays, 1733-1831, Post books as they exist in the Bibliothèque du Musée de la Poste in Paris; 1632, The Nicolas Sanson d’Abbeville’s map (1632) Carte géographique des Postes qui traversent la France. 2) For the French historical boundaries, Nordman (1998), Burguière, Revel (1989), Nordman et al. (1989), Boutier (2006). 3) For the historical populations of communes including the relays, Bairoch et al. (1988), Lepetit, 1988, Le Mée, 1971, Bulletin des Lois, 1831.
50Geographical coverage: Current French boundaries
51Geographical units: Historical relays (points), Road segment at each date between two adjacent relays (lines).
52Type of article: Data paper
53License: ODC Open database license V1.0
54Acknowledgments: the authors would like to express their gratitude to the reviewers for their valuable feedback and insights, which greatly contributed to improving this work, and to the journal Cybergeo for their support.