ChapterPDF Available

Flow Mapping through the Times: The Transition from Harness to Nazi Propaganda



One of the most commonly used types of maps today are flow maps, which simultaneously depict movement in time, place, and volume on a geographical map, as seen in GPS navigation devices. This type of map-making was invented independently during the 1830-1840s by three railway engineers from the United Kingdom, Belgium, and France. However, as this chapter argues, the growing popularity of the genre had little to do with the intent of the three pioneers. By looking at the context, in which flow maps appeared, rather than the technique used to design them, the chapter shows the importance of culture, politics, and ideology in understanding the changing meanings of flow maps during the nineteenth and twentieth centuries.
Flow Mapping through the Times
The Transition from Harness to Nazi Propaganda
Zef Segal
One of the most commonly used types of maps today are flow maps, which
simultaneously depict movement in time, place and volume on a geographical map, as
seen in GPS navigation devices. Flow maps were first introduced in 1837 but only
became popularized during the second quarter of the twentieth century. Based on the
collection of nineteenth- and twentieth-century European and American atlases in the
Library of Congress, this chapter examines the developing popularity of flow maps as
a graphical method since the mid-nineteenth century.
Historiography of cartography, like historiographies of other sciences and arts,
focuses on the innovators and pioneers of the field. However, despite the obvious
importance of these trailblazers, they are rarely the cause of dissemination and
popularization of new techniques. I will start by describing the first cartographers,
who invented flow maps independently in the United Kingdom, Belgium and France.
Yet, flow maps became something completely different from what they initially
intended to be. The introduction of flow maps in commercial atlases appeared much
later, in places that had very little to do with those original innovators, and reflected
themes that were radically different from those present in the pioneering flow maps.
The history discussed here is not a linear one of scientific progress, but rather an
erratic history of changing paradigms, with no single scientific milestone to mark the
point of change. Furthermore, I will claim that this development was not an
immediate outcome of a new technology, a new set of symbols, or even a conceptual
and cultural change, but rather of a growing western colonialism and nationalism that
required new means to depict western and national global dominance.
Flow maps are cartographic depictions of movement, with an emphasis on its
quantitative values. Flow line symbolization is used when the cartographer wants to
show the type, volume and density of movement between two or more places.
symbolization of qualitative data is most often done by varying the direction, colour,
or shape of the lines in question in order to reflect differences in values. For the
quantitative variety, the widths of the flow lines connecting the places are usually
drawn in proportion to the quantity of movement represented. Flow maps are
important because of their ability to simultaneously represent multiple variables in
such a way that they are ‘integrated gently and unobtrusively that viewers are hardly
aware that they are looking into a world of four or five dimensions’.
The invention of flow maps
Although '[f]lows of people, products, or information often seem to beg for
cartographic portrayal,' as stated by Mark Monmonier, this was only realized in the
early nineteenth century.
The invention of flow maps independently in the United
Kingdom, Belgium and France was part of a larger development in the history of
statistical graphing, described as the beginning of modern graphics.
However, a more
relevant context was the fact that in 1840, the three states had the longest operating
Dent, Cartography, p. 188.
Tufte, The Visual Display, p. 40.
Monmonier, Mapping it Out, p. 189.
Friendly, ‘Milestones in the History of Data Visualization’, p. 37.
railway systems in Europe.
Consequently, the three innovators of flow maps were
engineers looking for new ways to visualize the effects of modern transportation and
individual movement patterns on their societies, and consequently improve the public
transportation infrastructures of their states. The earliest flow maps were created by
Henry Drury Harness (1804-1883) to augment the second report to the railway
commissioners of Ireland in 1837.
Harness was an engineer in the British army, who
had no previous experience or education in cartography. His flow maps show the
relative number of passengers traveling in different directions throughout Ireland
(Figure 18). The lines on the maps are shaded and the varying widths are proportional
to the average number of weekly commuters along that route.
Mitchel, International Historical Statistics, p. 655-656.
Robinson, ‘The 1837 Maps of Henry Drury Harness’.
Figure 18: The first flow map ever made. Map IV ‘Shewing the relative number of passengers in
different directions’, in: Henry Drury Harness, Atlas to accompany 2d report of the Railway
Commissioners Ireland 1838 (1838). Published with the permission of the University College Library.
Within ten years, Alphonse Belpaire (1807-1857) from Belgium and Charles Minard
(1781- 1870) from France began publishing flow maps as well. According to all
records, each of the three had no knowledge of the other two. Belpaire, an engineer
with the Belgian Railways, included two flow maps in his 1847 treatise on railway
expenses. These maps, much like that of Harness, showed transportation movement in
Belgium in 1834, 1835 and 1844. He used these maps to make the point that railroads
did not take traffic away from canals and that in the vicinity of towns there was an
intense use of road carriage.
Unlike the other two mapmakers, who never again published maps, Minard was a
French civil engineer who became a cartographer at age 64 and published 51 maps,
most of which were flow maps. In 1845 he published his first flow map, in which he
depicted the number of travellers from Dijon to Mulhouse. His maps covered a large
array of topics ranging from human flows, and rail traffic to coal and commodities, up
to Napoleon's Russian campaign of 1812. With the exception of a few maps that were
published in his final years, Minard’s work tended to display the movement of people
and commodities within France.
Although the second half of the nineteenth century is defined as the golden age of
statistical graphics, the innovations of Harness, Belpaire and Minard had very little
effect on commercial cartographic production. Harness and Belpaire published their
maps within professional engineering publications and were only rediscovered
decades later by historians of cartography.
Minard, on the other hand, was known to
French officials and French statisticians but had very little contact with contemporary
geographers, cartographers and commercial editors of atlases and maps.
From pioneers to commercial publishing
A survey of 414 different commercial atlases, published between 1837 and 1939 in
four of the most central production centers of cartographic material (USA, UK,
Germany and France), shows that the popularization of flow maps was not directly
Only fourteen of Minard’s flow maps described foreign trade or travel, of which eleven were
produced between 1861 and 1869.
Robinson, ‘The 1837 Maps of Henry Drury Harness’.
Robinson, ‘The Thematic Maps of Charles Joseph Minard’, p. 96; Rendgen, The Minard System, pp.
related to the early pioneers.
Diagram 1 shows the percentage of commercial atlases
published between 1837 and 1939 that contained flow maps. The graphs distinguish
between four historical periods, and three places of atlas production; Germany, the
UK and the USA.
France is not included in this graph since none of the French atlases at the Library of
Congress had any flow maps. Minard’s influence on French cartography was limited
to the realm of French public services and did not spread further to commercial
cartography. In 1879, the newly established French Bureau of Statistical Graphics,
which was established under the Ministry of Public Works, began publishing an
annual series of statistical atlases that expressed graphically the flow of passenger
travel as well as freight. The Album de Statistique Graphique was published between
1879 and 1897 and included many flow maps, with the proposed intention of
supporting the planning, development and administration of public works. In 1897,
the French Bureau of Statistical Graphics was dissolved and the series was
discontinued due to the high costs of production;
its maps and their relative themes
were not reflected in contemporary commercial atlases. While the French Bureau of
Statistical Graphics was interested in visualizing aggregated individual statistics on a
national scale, commercial atlases were much more focused on global movement, as
will be described further on.
The survey included 156 American atlases, 121 British atlases, 106 German atlases and 31 French
atlases. All are kept in the Geography and Map division at the Library of Congress. New editions of
each atlas were surveyed as well but are not added to the atlas total.
Faure,'France', p. 295.
Diagram 1: A graph showing the percentage of commercial atlases published between 1837 and 1939
that contain flow maps. Source: Atlas collection at the Library of Congress.
The first flow maps in commercial atlases appeared in the third quarter of the
nineteenth century, but only in German atlases. Between 1851 and 1880, eight per
cent of German atlases included in the survey displayed a flow map. Its appearance in
German atlases was a continuation of the development of thematic mapping in
Germany. Since the publication of Heinrich Berghaus Physikalischer Atlas in 1845,
which did not include flow maps, German cartographers attempted to visualize
cartographically various physical and human aspects of the earth that had not been
represented previously. The inclusion of flow maps that depict global movement was
a direct outcome of the ethos of German thematic mapping. Accordingly, the first
flow maps did not refer to any of the three pioneers of flow mapping; instead, they all
referred to previous work done on physical, meteorological, and commercial thematic
cartography in German atlases.
British atlas publishers introduced flow maps from 1891 onwards, yet the low number
of atlases that included flow maps (five per cent) reflects a limited popularity among
publishers. This applies even more in the case of American atlases that in general
0 0
0 0 0
1839-1850 1850-1880 1881-1919 1920-1939
refrained from representing motion in their maps, and preferred to depict static places
and locations. Between the two world wars, these percentages grew to over 30 per
cent of all new British and German atlases as well as fourteen per cent of all
American atlases. This growth seems ill placed. It did not reflect the innovation itself,
which had happened almost 80 years earlier, nor was it connected to any new print
technology or graphical innovation. In fact, the first half of the twentieth century is
even referred to as the ‘dark ages’ in terms of introductions of new methods in
statistical graphing and thematic cartography.
However, as Michael Friendly claims,
this was a time of dormancy, application and popularization, rather than new
inventions. Statistical graphics during this period became both mainstream and
standard use in public and private sectors of society; commercial atlases were affected
by this change. The next sections investigate the themes and the visual appearances of
published flow maps in order to have a clearer understanding of the changes in
Early flow maps (1850-1880) and physical geography
The earliest examples of flow lines in German atlases were very different in content
and style from the pioneering flow maps. They were not even flow maps per se since
the depicted flows were not the sole focus of the maps. Unlike the originals, these
maps rarely showed inland movement, nor did they focus on a specific national
territory. Flow lines were not a tool to solve state-centric problems, as they were for
Harness, Belpaire and Minard, but a graphical way to depict growing global trade and
its various modes of trans-oceanic transportation as well as understanding the
emerging globalization. The mapping of flow maps in German atlases followed the
Friendly and Denis, ‘The Early Origins’.
steps of Von Humboldt, who saw geography as a natural science addressing itself to
the whole globe.
One of the first examples appears in the second map of Meyer's Hand-Atlas (1867).
A map entitled Erdkarte (Map of the Earth) emphasizes postal ship routes and the
colonial division of the world and uses a unique graphical technique in order to
describe the routes of ships, as well as the duration of travel time and frequency of
journeys. Much like other maps of the mid-nineteenth century, the map's prominent
symbol is that of natural motion, dotted white lines, which signifies ocean currents,
and sporadic black arrows that describe their direction. However, only ship routes are
drawn as multivariate lines. Two types of black lines are used to distinguish between
routes leading from Europe to routes leading to Europe, dots above the lines define
the frequency of the ships along this course (monthly, fortnightly or weekly), and
numbers describe the length of travel in days. Although the graphic is simple and still
relies on verbal notation, it provides information about the direction, frequency,
duration in travel time as well as length in distance traveled. The purpose of the map
is to reflect the multiplicity and complexity of trans-oceanic traffic.
A similar, but more advanced map entitled Allgemeine Welt-Karte (General Chart of
the World), was drawn by Hermann Berghaus in 1863 and published by Perthes
Publishing in German and in English (Figure 19).
This map, in its various editions,
depicts telegraphs, submarine cables and railway lines, all of which are drawn as lines
with no other quantitative value attached. However, oceanic currents and steam ship
routes are drawn as multivariate lines indicating direction, route and the magnitude of
the movement. Four different types of currents are depicted: Equatorial currents,
Wardenga, ‘German Geographic Thought’, p. 137.
Meyer, Meyers Hand-Atlas.
Berghaus, Allgemeine Welt-Karte.
Equatorial counter currents, periodical currents and Polar currents. The currents
include arrows indicating their direction and numbers indicating their mean velocity.
Steam ship routes are identified by their national flags (sixteen different nationalities
indicated by colour), principal companies (56 different companies indicated by
initials), as well as their frequency (monthly, fortnightly, weekly and bi-weekly), and
direction, visualized by shape. Numbers above the depicted lines describe distances in
nautical miles and days of travel. The resulting image is that of a flow map since the
multiplicity of lines and the abundance of data reflects the amount of transport along
oceanic routes. Despite the tremendous achievement of Berghaus’ map, its influence
on atlas production was limited, as evinced by the miniscule number of atlases that
included similar types of maps.
Figure 19: The South Atlantic Ocean segment of the Chart of the world’ by Heinrich Berghaus,
published by Perthes Publishing (1879). Published with the permission of the Library of Congress.
The turn of the century (1880-1914) and economic geography
In contrast to the German school of cartography, which related global trade to the
physical world and natural sciences, the British interest in global trade was more
nationalistic and economically oriented. British flow maps were introduced as part of
the late nineteenth-century emergence of economic geography in the academy as well
as in school education, which was a consequence of imperial British commerce, a
global trading system and the system of free trade.
British economic geographers
believed that trade was above all a geographical phenomenon, and that Britain was at
its centre.
Accordingly, the first British atlases to simultaneously depict time, space
and movement were the Atlas of Commercial Geography (1889) by John
Bartholomew and the Atlas of Commercial Geography (1892) by W.A.K. Johnston,
both published almost at the same time as the first text book on economic geography,
George Chisholm’s A Handbook of Commercial Geography (1889). Bartholomew and
Johnston were two of the most prominent atlas publishers in late nineteenth-century
Britain, but their only atlases that included these types of maps were atlases of
commercial geography. These atlases had no flow maps, but utilized a different
graphical style, developed a few years earlier, called isochrone maps.
Unlike flow
maps, which focus on specific routes, an isochrone map depicts areas of equal travel
time from a specific point. The thirteenth map of the Bartholomew atlas and the first
map of the Johnston atlas depict the earth's surface as coloured regions, identifying
areas of equal travel from London. The map in Johnston's atlas shows railways,
caravan routes, steamer tracks, sailing ship tracks and telegraphs, but only those
related to routes leading to or from London. While German flow-maps were interested
in the nature of global trade in general, and as such, usually centred its maps on the
Atlantic Ocean, British flow-maps were far more interested in the characteristics of
British trade.
By the first decade of the twentieth century, the interest in commercial cartography
made trade flows into a significant element in specialized atlases in Germany as well
as in Britain. The first map of the Handels Atlas zur Verkehrs und
Barnes, ‘In the Beginning was Economic Geography’.
Barnes, ‘Inventing Anglo-American Economic Geography’, p. 15
Galton, ‘On the Construction of Isochronic Passage-Charts’.
Wirtschaftsgeographie (1902) is titled Weltverkehr, Kolonien und Handelsflotten
(World Traffic, Colonies, and Merchant Fleets).
It depicts the earth as divided
among eleven colonial powers and connected through roads, railway lines, rivers and
ship routes, in that order of significance. All routes are defined as lines of various
kinds, but the important routes of steam ships are depicted as stripes whose widths
indicate the frequency of traffic along that route. The distance in travel time is written
inside the relevant stripe. Unlike mid-nineteenth-century German maps, physical
features, such as ocean currents and winds, are ignored. The shift from physical
geography to economic geography transformed global trade into a central force
without any relation to natural phenomena.
Despite the change in content, German maps were not nationalistic as were the British
maps. The 1871 unification of Germany did not alter, at first, the decentralized nature
of German politics, nor did it hinder the independence of German publishing houses,
which operated from many small towns and cities. As a result, its influence on
German atlas production was underwhelming in the first few decades. Accordingly,
the centre of the maps is the Atlantic Ocean and the flows have no national colours;
they are all coloured with a similar tone of pink. In contrast, British flow maps
remained nationalistic in style and content. The Atlas of the World's Commerce
(1907) includes two flow maps, and an isochrone map centered on London.
Although all three maps show the entire world, their primary focus is Britain. The
editor, John Bartholomew pronounces his intention in the preface of his atlas:
Commerce leads the way, and in this new age, it has come to be realized that
commerce is the real basis of our modern material civilization, and that the
nations, which maintain commercial supremacy, will also be assured of
Scobel, Handels-Atlas, pp. 2-3.
Bartholomew, Atlas of the World's Commerce.
political supremacy. National competition for the world's trade must every
year become keener, and in such competition, a thorough appreciation of the
whole economic situation will be of primary importance.
The first flow map (Figure 20) is entitled 'Commercial Highways of the World' and
depicts railway, telegraph, rivers, and steamship routes. Of these principal highways
on land and sea, only steamship routes are depicted in varying widths. However,
unlike Drury, Belpaire and Harness, it wasn’t the intention to draw the width to scale.
The map key states that 'the routes with the greatest traffic are indicated by thicker
lines', but Bartholomew never clarifies the meaning of 'greater traffic' in his thorough
introduction or in a footnote. In this map, wider and narrower lines are not exact
representations of quantitative differences but rather visual means to emphasize
Bartholomew, Atlas of the World's Commerce, Preface,
Figure 20: 'The commercial highways of the world', in: J.G. Bartholomew, Atlas of the World's
Commerce (London: G. Newnes, 1907). Published with the permission of the Library of Congress.
The second flow map in the Atlas of the World’s Commerce depicts telegraphic
communication. Unlike the previous map that uses width to indicate 'greater traffic',
this map indicates traffic by the number of cables connecting two points on the map.
For example, five British cables and three foreign cables connect south Britain to
Nova Scotia, while only one foreign cable connects south Britain to the Azores.
Graphically, the map itself is extremely simple. There are only five map symbols in
the map key: three types of lines signifying British cables, foreign cables and land
telegraphs, and two colours of the earth's territories, 'countries with telegraphic
communication' and 'regions without telegraphic communication'. Both distinctions,
created by the map symbols, reflect the imperialistic agenda of Bartholomew. The
distinction between the two primary types of cables follows a nationalistic tendency
of the mapmaker, which distinguishes between us (Britain) and them (foreign
nations). The territorial distinction between developed 'countries' and undeveloped
'regions' divides the world between countries that could not be colonialized and
regions that are available for conquest. In short, it was a visual manifestation of
Bartholomew's preface.
Interwar maps: geopolitical cartography
Atlas production slowed to a halt during World War I and was only revived in the
mid-1920s. In British and German atlases published subsequently, flow maps
appeared more frequently than before, and their colonial agenda was much more
apparent than in previous years. In addition, in the first quarter of the twentieth
century there was a transition in the geographical paradigm: commercial geography
changed the emphasis from race and climate to prioritizing trade and commerce.
Consequently, the discourse of civilization described the world as divided according
to an evolutionary logic of stages of human development.
Accordingly, many of the
flow maps mutually depicted flows (primarily European) and territorial hierarchy.
Movement and flows became the signifiers of cultural and scientific progress, while
lack of movement was a sign of retardation.
For example, Putnam's Economic Atlas (1925) includes a flow map entitled 'Means of
transport and communication' (Figure 21).
In this map, the colours of the regions of
the world are based on the primary means of transport in that region. The colours
range from yellow, used for motorized transport, to pink, indicating dog or reindeer
sledges and canoe transport. Three different orange tones mark methods such as
horses and oxen. The gradation in the colour scheme (yellow-orange-pink) establishes
a hierarchy of civilizations. Simultaneously, the map depicts primary transport routes,
which include railway lines, inland waterways and principal channels of maritime
trade. The latter are symbolized with blue stripes whose widths indicate the relative
value of marine trade. Unlike the indistinct widths in Bartholomew's 1907 map,
George Philip, the editor of the Putnam's Economic Atlas, clearly defines the widths.
However, most stripes are identical in width and the variation that exists is an overt
statement of his European and British bias. The stripes are widest around Europe and
become narrower as the route moves further away. In addition, Britain is depicted as
the only hub connecting Europe and North America.
Domosh, ‘Geoeconomic Imaginations’, pp. 947-948.
Philip, Putnam's Economic Atlas, pp. 3-4.
Figure 21: 'Means of transport and communication' in George Philip, Putnam's Economic Atlas
(London : G. Philip, 1925). Published with the permission of the Library of Congress.
Similar maps appear in other British atlases, and although the phraseology is different,
the style and content of the various maps remain the same. For example, the fifth map
of Cassell's New Atlas (1932) entitled The world: commercial development follows
a similar style as it depicts principal steamship routes with varying widths together
with territories whose varying colours are based on their commercial development.
Although the categorization of the various lands is based on density of population, the
title of each category reflects its level of sophistication: 'Highly developed
manufacturing districts', 'Highly developed agricultural and plantation regions', 'Other
industrial and agricultural regions', 'Productive agricultural or pastoral regions, less
highly developed', 'Less productive or under-developed pastoral, forest, or agricultural
regions', and 'Other under-developed regions'. Instead of referring to the actual data,
Philip, Cassell's New Atlas.
density of population, the titles provide a value judgement for each region. The map is
also surrounded by comparative diagrams showing the principal states of the world in
order of population and area. These diagrams emphasize the magnitudes of empires,
most importantly that of Great Britain.
German flow maps followed the British example and began emphasizing German
flows, rather than global flows. The Westermanns Welt Atlas (1926), for example,
includes a map entitled Was die verschiedenen Teile der Erde dem Weltmarkte liefern
(What the different parts of the world deliver to the world market).
This map
shows the distribution of manufactured commodities around the world and in greater
detail the production of major European states. The map, except for the letters
marking the various commodities, is a reproduction of the opening map of the
previously discussed Handels Atlas (1902). However, the reproduction includes a
number of significant changes, reflecting the new agenda of German flow maps. The
centre of the map is shifted eastward in order to focus on Germany rather than the
customary centre of global trade in the Atlantic Ocean. In addition, the map
distinguishes between foreign ship routes, coloured light purple, and German ship
routes, coloured with a dark purple. Despite its lengthy key, this map, unlike the older
map, does not explain the various widths. Much like the British maps, this map is a
representation of European and German dominance, rather than reflecting actual
quantities of trade.
Germany had now become the centre of the frame of German flow maps, and German
flows were the most important element of these maps. In the Lange Diercke
Schulatlas (1932), the map depicting global trade only mentions 'Germany's world
Liebers, Westermanns Welt Atlas, p. 108.
trade and world traffic'.
It marks German ships and cables, centres of production of
imported goods and trade flows between Germany and the rest of the world. The map
itself does not visually depict any cultural hierarchy, but two small inset maps on the
same page provide relevant information. The first map, in the bottom left corner of the
page, shows the world divided by ways of animal transportation: horses, oxen,
camels, yaks, llamas, and reindeers. The only colour without any explanation is that
of the ‘civilized’ world, Europe and the United States. The second map, in the bottom
right corner of the page, complements the first one by dividing the world based on
agricultural advancement. These two inset maps are an integral part of the flow map
above them. The superiority and centrality of Germany and its flows, as visually
detailed in the main map, are the direct outcomes of sophistication and superiority of
Germany's communication and agricultural systems, as portrayed in the inset maps.
Interwar maps: schematic cartography
Gradually, flow maps became a visual rhetoric. Mapmakers started to simplify the
iconography of the map with fewer symbols and colours, exaggerating the widths of
the flows, and most importantly excluding references to the meaning of those widths.
For example, the second and third maps of the Practical Atlas of Modern Geography
(1931) are rainfall and wind maps.
Both maps have only four colours to reflect the
rainfall amounts in different territories and a single symbol, a black arrow, to mark the
wind trails. Although the arrows are differentiated by size and width, the
symbolization is never explained or elaborated upon. Although each map has a long
Original: 'Deutschhands Welthandel und Weltverkehr'. Lange-Diercke Schulatlas Lange-Diercke
Schulatlas Lange and Diercke, Lange-Diercke Schulatlas, pp. 46-47; see also Herders Welt und
Wirtschaftsatlas, pp. 13-14 and Eggers, Deutsches Land, pp. 70-71.
Stamp, The Practical Atlas.
caption, verbally describing its content, the text only explains the seasonal weather
and not the map symbology.
This minimalistic cartography enabled atlas publishers to portray the world as
borderless, thus facilitating the smooth transition of capital around the world, forming
a new type of imperialism, commercial imperialism. The Welthandels Atlas (1927),
for example, is not explicitly a colonial atlas since its maps are extremely minimalistic
and do not show political borders. The atlas only consists of maps that show global
import, export and the production of various commodities related only to Europe
(Figure 22). Three maps are drawn for each commodity, one describing world
production and trade flows, and two smaller inset maps focusing on North America
and Europe. The various maps have only two distinct symbols, wide arrows depicting
flows and circles representing centres of production. The use of colour is also limited.
Land surface is not coloured, while flows and location of production are coloured red.
The maps in this atlas are very scientific in their verbal methodology; each symbol is
described, the sizes of circles and arrows are in direct proportion to their relative part
of global trade of a specific commodity, and the sources from which the information
was gathered is noted. However, flows are now the sole element of the maps. While
maps in the nineteenth century show trade flows alongside winds and ocean currents,
thus relating human movement to the natural world, and maps at the turn of the
twentieth century show flows as another consequence of European colonialism, flows
in these maps are important in themselves. The world was reducible to human traffic,
most importantly European traffic.
Figure 22: 'Gerste', in: Walther Schmidt and Georg Heise, Welthandels-atlas: Produktion, Handel Und
Konsum Der Wichtigsten Welthandelsgter (Berlin: Columbus-Verlag, 1927). Published with the
permission of the Library of Congress.
Similarly, the Deutsches Land, Deutsche Volk und die Welt (1937) includes a map
entitled Völker und Kulture: Staaten, Handel und Verkehr (People and Cultures:
States, Trade and Traffic).
The map itself is a typical interwar colonial flow map
which simultaneously shows the colonial division of the world, the production centres
and the major trans-oceanic ship routes with changing widths. Although Germany is
in the centre of this map, it is not emphasized in any other way. However, four inset
flow maps on the bottom of the same page, which follow the question Woher kommen
die wichtigsten Handelsgüter, die wir einführen müssen? (From where do the most
important goods, which we must import, come?), construct the centrality of
These maps show the major import routes into Germany of ten different
commodities. The German territory is coloured black in an otherwise uncoloured
Eggers, Deutsches Land, pp. 70-71.
Eggers, Deutsches Land, pp. 70-71.
world, the flows are coloured blue, orange, red, and light brown, and are the only
elements drawn on the map besides for the borders of the continents. The world
constructed by these maps has a singular centre, Germany, which is both the centre of
a borderless static world and the centre of the dynamic world of flows.
The use of flow maps as persuasive cartographies was not an innovation of the 1920s
and 1930s. Harness, Belpaire and most notably Minard repeatedly used such maps to
influence their readers.
However, the original flow maps remained rich with
different types of data, and the pioneers were consistent in their motivation to use the
method as an exact representation of quantitative values. Furthermore, the world seen
through the early flow maps is not meant to be reducible to flows, and unlike flow
maps of the 1930s, which will be discussed consequently, the resulting image was not
intended to be propaganda.
Schematic flow maps with a nationalistic agenda became popular in British and
German atlases from the early 1930s. The Practical Atlas of Modern Geography
(1931), for example, includes a schematic flow map entitled 'The British Empire and
the main ocean routes of the world'.
The map contains very few symbols and has no
key to explain them. The British Empire is coloured red while the rest of the world is
uncoloured and without borders, except for the borders of the continents. Black stripes
of varying widths connect Britain to the world thus marking its centrality in world
flows. Much like the German map of 1937, Britain is both the centre of the static
world, with its dominant red colour, and the centre of the dynamic world, through its
Rendgen, The Minard System, pp.24-25.
Stamp, The Practical Atlas, p. 6.
Interwar maps: Goode and American flow maps
Unlike British and German cartography, American atlases tended to be far less
international, and much more focused on the American continent.
However, the
growing importance of global trade and economic geography in the early twentieth
century forced the USA to be placed in the global framework rather than seeing itself
isolated and superior.
Flow maps were still a rare sight in American cartography as a
result of its focus on political world maps, and on place-names rather than
In 1923, the chief cartographer of Rand McNally John Paul Goode, who
was extremely critical of the American style of cartography, edited an atlas that was
meant to change the landscape of atlas production in America.
His atlas is
innovative in many cartographic aspects, one of which is the inclusion of flow maps
(Figure 23); a pioneering effort in the otherwise static worldview reflected by
American published maps.
The first flow map is a polar projection of the northern hemisphere showing cyclone
tracks circling the Earth. The tracks are sketched as dotted lines and not stripes, but
the multiplicity of these lines manifest the frequency and volume of the cyclones, thus
operating as width. The next page contains a map of ocean cables, operating in a
similar way. Each line signifies a submarine cable and their multiplicity reflects the
volume of traffic. On the bottom of that same page, Goode includes a map entitled
The flow of commerce’. This map is a classic flow map, with red stripes of varying
widths marking the flows, and an explanation defining the width as ‘proportional to
the flow of commerce’. Unlike contemporary European examples, these maps
emphasize global trade and global flows, rather than national flows. State borders are
Schulten, The geographical Imagination, p. 21.
Domosh, ‘Geoeconomic Imaginations’, p. 349.
Dent, Cartography, p. 192.
Goode, Goode's School Atlas, pp. 12-13; Schulten, The geographical Imagination, p. 194.
not included or even mentioned, and the three maps are centred on the North Pole,
Britain, and the Pacific Ocean, respectively. In addition, Goode deliberately avoids
creating a global centre by changing map projections in each map.
However, the
influence of European mapmaking is still prevalent in the schematic nature of these
maps. Each map has a single symbol, notating the relevant type of flow, and a single
colour, marking that specific movement. The land surface is not coloured, except for a
pink tint marking commercially developed lands in the map of ocean cables.
Figure 23: three types of flow maps published in John P. Goode, Goode's School Atlas (New York:
Rand McNally, 1923). Published with the permission of the Library of Congress.
Goode’s maps did not affect other American mass-market atlases, who refrained from
adding flow maps. American atlases were cautious of adding innovative maps since
they were wary of confusing the consumer public. Rand McNally, for instance,
continued to publish the Goode’s World Atlas, but marketed it primarily for schools.
The globalized world-view of Goode’s 1923 flow maps was slightly altered in a
reproduction published in Rand McNally World Atlas (1935), which was also edited
More on Goode’s criticism of the Mercator projection in Schulten, The geographical Imagination, p.
by Goode.
The reproduced map shows Goode’s unique holomographic projection
and the same flow routes. The map is less schematic than Goode’s map but still
contains very few symbols; seven colours distinguish between regions of economic
activities, small black dots mark mines, and red flows mark shipping routes. The main
difference is the centre of the map. The map projection is altered in order to situate
the American continent in the centre, because like British and German flow maps,
American flow maps had also become nationalized.
The late 1930s and racial flow maps
As we have seen, most flow maps in commercial atlases were devoted to trans-
oceanic world trade. Even the emerging national tendencies of the early twentieth
century did not change this. However, the combination between schematic flow
mapping and the racial agendas of the Nazi party in Germany introduced new types of
flow maps with explicit racial and nationalistic content, which had very little to do
with global trade. This was part of a larger process, in which the German Right
became increasingly aware of the persuasive power of maps.
Arnold Hillen
Ziegfeld, one of the leading figures of the German school of suggestive cartography,
stressed the importance of cartographically depicting dynamic aspects of life and
During the 1930s, this type of cartography was a tool of propaganda and
among its graphical elements arrows as well as flows were used repeatedly to
emphasize the dynamic nature of the German territory. These maps intentionally
‘Economic Activities’, in Goode, Rand McNally World Atlas.
Herb, Under the Map of Germany, pp. 76 and 134. On the insignificance of suggestive cartography
in the German Left, see Herb, Under the Map of Germany, pp. 152-153.
Ziegfeld, ‘Kartengestaltung’.
connect two contradicting worldviews, a world divided into discrete and isolated
racial groupings and a world described by movement.
This contradiction is emphasized in flow maps that show movement of races and
nations. The purpose of these flow maps was not to depict a world of free movement
but rather a world of growing tension. A map entitled Volk ohne Raum (‘A People
without Space’) was published in Deutsches Land, Deutsche Volk und die Welt
The map shows the locations of Germans worldwide by red colours and
their migration routes by red stripes with varying widths. The red flows are not
explained and are used primarily as a rhetorical device. This map does not distinguish
between the static geography of places and the dynamic geography of flows. It is part
of a political argument, mentioned in the map’s title: Germany needs to grow and its
current borders are irrelevant for this growth.
Flows had become a sign of struggle and contest over land. The Werden und Wachsen
atlas (1938), for example, includes a number of flow maps that show racial
The first shows ‘the races of the modern times’ and divides the world
into four groups, European races, other Caucasian races, Black races, and Mongol
races (Figure 24). Large brown arrows emerge from the brown European continent
and spread across the world. These arrows have varying widths but are not explained,
except for a side note stating ‘the passion for travel of the northern people’. This flow
map merges depictions of static territories and dynamic flows into what looks like a
single organism that occupies the world.
Eggers, Deutsches Land; see also Die Erde, Verbreitung der Deutschen in Herders Welt und
Kumsteller, Werden und Wachsen.
Figure 24: The races of the modern times’, published in: Bernhard Kumsteller, Werden und Wachsen,
(Braunschweig: Westermann, 1938). Published with the permission of the Library of Congress.
Two other maps show ‘the northern race and the Germans as culture bearers’ and
The spread of the Jews in Europe’, and provide larger scale maps of movement than
those we have seen previously.
These illustrate the historical routes of Germans
within Europe and Jews into Europe respectively. Different colours and types of
arrows depict the movement in various time periods. The arrows do not have varying
widths, but their depiction as wide stripes resembles that of flows rather than arrows,
which are usually narrow lines with an arrow head. The purpose of these maps is not
to reflect the historical movement of Germans or Jews within Europe, nor is it to
depict their current locations, but rather to illustrate the state of conflict between the
races. The same atlas uses flow mapping and its relation to the static depiction in two
different ways: the first combined flows and territories into a single brown surface,
while the second showed the contradiction between flows (Jewish immigration) that
supposedly infiltrate and harm the integrity of the European territory. On the one
hand, flow maps had returned to the original large scale of the 1830s, focusing on
Original: 'die nordische Rasse und die Germannen als Kulturträger', and 'Ausbreitung des Judentums
in Europa'. In general, cartographers during the Nazi period turned their attention to the homeland, see
Heske, ‘Political Geographers of the Past’, p. 278.
states and on Europe, but on the other hand, they had changed drastically from the
scientific intent of their innovators.
Flow maps were invented by three engineers, who sought for a way to visualize and
solve state-centric problems that involved personal and commercial inland traffic.
However, their source of motivation and themes were not replicated in commercial
atlases. In fact, flow maps only began appearing a few decades later as a response to
the growing world trade and the extending European imperialism. Much like other
types of late nineteenth-century maps, flow maps made the complex imperial project
visible and comprehensible. The scale of these maps is much smaller than the
pioneering flow maps, as they deal with the whole world, and the basic flow is always
trans-oceanic traffic, which connects Europe and the world. On-going changes and
trends in the study of geography, from physical geography to economic geography
and later geopolitical geography, changed the content and context of the flow maps.
Winds and ocean currents, colonial demarcations, and territorial hierarchies were
added and removed as the trends shifted.
Flow maps gradually became a way to emphasize specific types of movement. Instead
of being a sophisticated visual tool, accurately and objectively reflecting a scaled
representation of spatial relations, they had become a simplified visualization used for
the sake of national agendas. This reached its apex in German cartography of the late
1930s, as flow maps became a significant component in Nazi propaganda, showing
the dynamism of the German territory and people.
Denis Cosgrove states that the thematic map reveals the presence of phenomena that
are beyond our normal bodily senses’.
This was the initial intent of Harness,
Belpaire, and Minard. However, a map is also ‘a creative process of inserting our
humanity into the world and seizing the world for ourselves.’
This creative process
was reflected in this chapter through the changing roles of flow maps. These maps are
used to represent colonial, imperialistic, nationalistic and later racial worldviews,
none of which had to do with the initial maps which depicted mainly individual
railway travel.
T.J. Barnes, ‘”In the Beginning was Economic Geography” A Science Studies
Approach to Disciplinary History’, Progress in Human Geography, 25 (2001), pp.
T.J. Barnes, ‘Inventing Anglo-American Economic Geography, 1889–1960’, in A
Companion to Economic Geography, ed. by E.S. Sheppard and T.J. Barnes (Malden:
Blackwell, 2000), pp. 11-26.
J.G. Bartholomew, Atlas of Commercial Geography (Cambridge: University press,
J.G. Bartholomew, Atlas of the World's Commerce (London: G. Newnes, 1907).
H. Berghaus, Allgemeine Welt-Karte (Gotha: Justus Perthes, 1869).
D. Cosgrove, Geography and Vision: Seeing, Imagining and Representing the World
(London: IB Tauris, 2012).
Cosgrove, Geography and Vision, p. 168.
B.D. Dent, Cartography: Thematic Map Design (Boston: WCB/McGraw-Hill, 1999).
M. Domosh, Geoeconomic Imaginations and Economic Geography in the Early
Twentieth Century’, Annals of the Association of American Geographers, 103 (2013),
pp. 944-966.
W. Eggers, Deutsches Land Deutsche Volk und die Welt, (Berlin: Deutsche Buch-
Gemeinschaft, 1937).
F. Faure, 'France', in The History of Statistics, their Development and Progress in
Many Countries, ed. By J. Koren (New York: Macmillan, 1918), pp. 215-330.
M. Friendly, ‘Milestones in the History of Data Visualization: A Case Study in
Statistical Historiography’, in Classification: The Ubiquitous Challenge, ed. by C.
Weihs and W. Gaul (Berlin: Springer, 2005), pp. 34-52.
M. Friendly and D. Denis, ‘The Early Origins and Development of the
Scatterplot’, Journal of the History of the Behavioral Sciences, 41 (2005), pp. 103-
F. Galton, ‘On the Construction of Isochronic Passage-Charts’, Proceedings of the
Royal Geographical Society, 3.11 (1881), pp. 657658.
H. de Beltgens Gibbins, Atlas of Commercial Geography (Edinburgh: W.A.K.
Johnston, 1892).
J.P. Goode, Goode's School Atlas (New York: Rand McNally, 1923).
J.P. Goode, Rand McNally World Atlas (New York: Rand McNally, 1935).
G.H. Herb, Under the Map of Germany: Nationalism and Propaganda 1918-1945
(London: Routledge, 2002).
Herders Welt und Wirtschaftsatlas (Freiburg im Breisgau: Herder, 1932).
H. Heske, ‘Political Geographers of the Past III- German Geographical Research in
the Nazi Period: a Content Analysis of the Major Geography Journals, 1925
1945’, Political Geography Quarterly, 5 (1986), pp. 267-281.
B. Kumsteller, Werden und Wachsen (Braunschweig: Westermann, 1938).
H. Lange and C. Diercke, Lange-Diercke Schulatlas (Braunschweig: Westermann, 1932).
A. Liebers, Westermanns Welt Atlas (Braunschweig: Westermann, 1926).
H.J. Meyer, Meyers Hand-Atlas (Hildburghausen: Bibliographischen Instituts, 1867).
B.R. Mitchel, International Historical Statistics Europe 1750-1988 (New York:
Stockton, 1992)
M.S. Monmonier, Mapping It Out: Expository Cartography for the Humanities and
Social Sciences (Chicago: The university of Chicago Press, 1993).
G. Philip, Cassell's New Atlas (London: Cassell, 1932).
G. Philip, Putnam's Economic Atlas (London: G. Philip, 1925).
S. Rendgen, The Minard System: The Complete Statistical Graphics of Charles-
Joseph Minard, From the Collection of the Ecole Nationale des Ponts et Chaussees
(New York: Princeton Architectural Press, 2018).
A. H. Robinson, ‘The 1837 Maps of Henry Drury Harness’, The Geographical
Journal, 121 (1955), pp. 440-450,
A.H. Robinson, ‘The Thematic Maps of Charles Joseph Minard’, Imago Mundi, 21
(1967), pp. 95-108.
S. Schulten, The Geographical Imagination in America, 1880-1950 (Chicago:
University of Chicago Press, 2001).
A. Scobel, Handels-Atlas zur Verkehrs und Wirtschaftsgeographie (Leipzig:
Velhagen & Klasing, 1902).
E.C. Rea Stamp, The Practical Atlas of Modern Geography (London: G. Gill & Sons,
E. Tufte, The Visual Display of Quantitative Information (Cheshire: Graphics Press,
U. Wardenga, ‘German Geographical Thought and the Development of
Länderkunde’, Inforgeo, 18/19 (2006), pp. 127-147.
A.H. Ziegfeld, ‘Kartengestaltung Sport oder Waffe?‘, Zeitschrift für Geopolitik, 12
(1935), pp. 243-247.
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.