Studies in History and Philosophy of Science xxx (2018) 1e12

Contents lists available at ScienceDirect

Studies in History and Philosophy of Science

journal homepage: www.elsevier.com/locate/shpsa

Average rainfall and the play of colors:Colonial experience and global

climate dataq
Philipp Lehmann
UC Riverside/Max Planck Institute for the History of Science, Germany

a r t i c l e i n f o a b s t r a c t

Article history: This paper examines the co-construction of global and local views of the weather and climate at the turn
Received 15 January 2017 of the twentieth century through a history of data gathering efforts in the German colonies in Africa.
Received in revised form While both governmental officials and metropolitan practitioners aimed at producing standardized e
30 September 2017
and thus globally comparable and economically useful e data in different environments, these efforts
Available online xxx
often tended to break down in practice. Rather than being able to turn the field into a finely tuned
laboratory, both European and African data gatherers were confronted with complex and challenging
Keywords:
environmental and institutional realities. Faced with these difficulties, colonial practitioners tended to
Climatology
Meteorology
embrace alternative strategies of recording weather conditions, which placed a higher value on indi-
Colonialism vidual sensory perception and qualitative descriptions. Thus, in a seemingly paradoxical dynamic, the
Instruments attempts to gather quantitative colonial data for global maps and models also facilitated the develop-
Senses ment of a particular colonial approach to climatology that highlighted local specificities and direct
embodied experience.
Ó 2018 Published by Elsevier Ltd.

Among the archival documents in the well-stocked folders of
the German Colonial Office are many prime examples of Wilhel-
mine bureaucratic decorum.1 For the historian reading through the
files, it is thus a refreshing change of pace to discover some clear
expression of emotion e albeit packaged to conform to the stan-
dards of diplomatic exchanges at the turn of the twentieth century.
One of these eye-catching documents is a dispatch from the
meteorologist Georg von Neumayer, the director of the Hamburg-
based Deutsche Seewarte, to the German Foreign Office in 1894.2
All the proper salutations and formalities can scarcely contain
Neumayer’s frustration at the state of meteorological observation in
German East Africa: “Seeing the large number of erroneous re-
cordings, the directorate [of the Seewarte] is forced to refrain from
q This paper appears in the SHPS special issue Experiencing the Global Environ-
ment (Volume 70, August 2018).
E-mail address: lehmann@ucr.edu.
1
The Kolonialabteilung, or Colonial Division, was a section of the Auswärtiges
Amt, or Foreign Office, but directly answerable to the German Chancellor. In 1907,
the division was transformed into the independent Kolonialamt, or Colonial Office.
Today, the files are kept in the Bundesarchiv Berlin-Lichterfelde. All translations are
mine, unless otherwise indicated.
2
The Deutsche Seewarte, or naval observatory, was established in 1875 and led
by Georg von Neumayer until 1903. On Neumayer’s work in geomagnetic and
meteorological research, see: Schlegel, Schröder, and Wiederkehr (2010); Schröder
and Wiederkehr (1992).
calling them out one by one,” he wrote in exasperation, before
repeatedly demanding the deployment of a trained meteorologist
to the colony. In the last paragraph of the dispatch, Neumayer even
went so far as to extort the colonial division by announcing that he
would withhold meteorological instruments needed in East Africa
until the Foreign Office had acknowledged and responded to his
critique.3
Aside from the entertainment value of Neumayer’s irritation, the
document is also a suitable starting point for a place-specific his-
tory of colonial meteorology that pays attention to the different
kinds of experience practitioners had in recording, relaying, sorting,
and translating the data that would serve as the basis for ever-more
scaled up visions of climates in the late nineteenth and early
twentieth centuries.4 Neumayer’s missive introduces the tensions
that existed both between different offices of the colonial
3
Bundesarchiv Berlin-Lichterfelde, Abteilung Deutsches Reich, Reich-
skolonialamt (below: BArch R1001), Folder 6136: Neumayer to Colonial Division,
German Foreign Office, 4 June 1894. All translations are mind, unless otherwise
indicated.
4
On the role of overseas colonial meteorology and climatology in the develop-
ment of the atmospheric sciences, see: Mahony (2016); White (2015); on the
importance of place in the production of scientific knowledge, see: Shapin (1998);
Naylor (2005); Livingstone (2003); on the role of negotiations within transnational
e and I would add, colonial e scientific networks, see: Turchetti, Herran, and
Boudia (2012).

https://doi.org/10.1016/j.shpsa.2018.05.007
0039-3681/Ó 2018 Published by Elsevier Ltd.

Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
2 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
bureaucracy, and between colonial observers and metropolitan
practitioners in questions of meteorological and climatological data
collection. Neumayer, who was tasked with assembling, analyzing,
and publishing recordings from the German colonies in Africa,
called attention to one of the most persistent meteorological con-
cerns throughout the thirty years of the German Empire in Africa:
inaccurate, incomplete, and unverifiable data e an issue that was
not the exclusive domain of colonial practice, but came into stark
relief in the overseas territories.5 It featured as the centerpiece of
long negotiations and disputes between Berlin, Hamburg, colonial
capitals like Dar es Salaam, and the various locations of meteoro-
logical stations in the colonial hinterland, which often consisted of
no more than a simple rain gauge and maybe a recording ther-
mometer. The question of the representativeness of data, however,
was not just a question of degree. Rather, it reflected arguments
between practitioners in the colonies and metropolitan scientists
and officials over the accuracy, and thus also the value, of different
kinds of data e from quantitative tables of instrument-recorded
numbers to the qualitative descriptions of sensory perceptions
and embodied experiences of the data gatherers.
Far from the authoritative neatness of the published meteoro-
logical diagrams and climatological maps that made use of colonial
data, the actual collection of the data was thus a deeply contested
practice. Colonial climatology was uneasily situated between the
rather nebulous realm of practical “colonial science,” expected to
provide quick and economically useful information; the rapidly
developing field of the atmospheric sciences, which had begun to
shift to a focus on quantitative data and ever-larger units of anal-
ysis; and the data collection efforts in underfunded meteorological
stations and makeshift observatories, in which an army of mostly
anonymous and untrained government employees, soldiers, and
volunteers recorded information on atmospheric conditions of the
new overseas territories of the German Empire. Traveling on
horseback, oxcart, or train, and then across the ocean by steamship
or telegraph, the data collected in the colonial stations provided
information for both economic development and scientific inves-
tigation, if not always in the format requested.
From handwritten reports and lists, practitioners in the metro-
pole filtered out the data that appeared most accurate, comparable,
or useful, and assembled these data into comparative tables, or
translated them into maps depicting the territory of new colonial
possessions, continents, or even the entire globe. These repositories
of information then aided the emerging research into trans-
regional climate systems and early global atmospheric models.6 Far
from the places and techniques of data processing, however, the
day-to-day practice of reading and recording colonial skies also
developed its own identity, which sometimes diverged from
practices and conceptions in Europe.7
The formation of a distinctly colonial approach to the atmo-
spheric sciences was connected to both material and personal cir-
cumstances, ranging from the difficulties of transporting fragile
instruments to and within the overseas colonies to the lack of
professional training of the observers. Similar to science in the
British Empire analyzed by Helen Tilley, German efforts at including
the colonies in the scientific projects of the metropole were often
subverted by both practical and scientific concerns with the local
and the vernacular.8 These logistical issues, however, were
frequently accompanied by intimations that not only the
5
For discussions of the role of standardization, precision, and exactitude in
nineteenth-century science, see: Wise (1995); Porter (1986).
6
See: Edwards (2010), pp.22e59.
7
Cf.: Osborne (2005).
8
Tilley (2011).
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
circumstances, but also the phenomena were incongruous and
incommensurable between the different environments of the col-
ony and the metropole and could not be represented accurately by
quantitative methods alone.9 Guided by both circumstance and
design, colonial practitioners thus tended to place a higher value on
narrative reports based on sensory experience.
In a seemingly paradoxical dynamic, colonial meteorology
provided not only quantitative and globally comparable data for a
scaling-up of the climatic vision to encompass continents or even
the whole earth, but also developed place-specific practices that
highlighted the qualitative and sensory experience of the observer
and emphasized local specificities on the scale of the colony, the
colonial district, or even the immediate surroundings of a single
meteorological station.
1. Meteorology in the African colonies of the German Empire
The first years of the formal German overseas empire were
inauspicious. In 1884, the government in Berlin declared Southwest
Africa e today Namibia e a German protectorate. In the following
years, the empire grew to encompass further territories in Africa e
German East Africa, Togo, and Cameroon e and cities and islands in
East Asia and the Pacific Ocean. Initially, the colonies were colonies
in name only. With just a handful of colonial officials and soldiers in
each territory and no budget to speak of; Germany exerted little
direct control and limited its activities to support traders, travelers,
and the odd scientific expedition. This situation began to change
gradually in the 1890s. Southwest Africa developed into a settler
colony and remained the only colony of the German Empire with a
sizeable European population until the dissolution of the overseas
empire during and after the First World War. The other African
territories of the German crown were dotted with scattered mili-
tary outposts and lined by a few railroad tracks. While support from
Berlin remained at a modest scale, the funds of the colonial gov-
ernments increased steadily into the twentieth century.10
Similar to the other branches and undertakings of the colonial
governments, colonial science e even if as ostensibly important as
geography or ethnography for control and economic development
e received only scarce funding in the first few years. Private pro-
colonial associations like the Deutscher Kolonialverein provided
some money, but they only supported one-time projects or expe-
ditions, rather than the day-to-day collection of data. The fields of
colonial meteorology and climatology were no exception. While the
creation of government-run meteorological stations had already
been considered as early as the mid-1880s, most of the plans
remained unrealized. The colonial administrations relied on vol-
unteers from among the few officials, settlers, and missionaries for
both time and instruments to gather data e particularly in regions
far away from the centers of colonial power.
While the first simple meteorological stations had been estab-
lished by 1890, there was still no centralized program in place to
collect and analyze the data. The governor’s offices in the colonies
focused on the seizure of political and military control and priori-
tized the creation of topographical maps of the unknown territories
in their early scientific endeavors. They were usually satisfied to see
the locally collected meteorological material used only locally, as
well. For the collection and assembly of data from across different
9
Cf.: Coen (2016); Coen (2006).
10
For recent overviews of the history of German overseas colonialism, see:
Conrad (2008); Speitkamp (2005); Pogge von Strandmann (2009); for a short
overview in English, see: Conrad (2012); the situation of funding for the German
colonies was similar to that of the French Empire, in which government funds were
supplemented significantly by private donations and money from corporations,
learned societies, and the military; see: Osborne (2005).
 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
Fig. 1. Meteorological station in Barombi (German Cameroon), c. 1900.
(Image Archive of the Deutsche Kolonialgesellschaft, University Library, Frankfurt am Main).
parts of a particular colony or from different colonies of the empire,
the colonial administration in Berlin once again depended on the
time and labor of volunteers, such as the Viennese meteorologist
Wilhelm Trabert, who offered his services to the German Foreign
Ministry in 1890.11 But even such small-scale efforts led to a back-
and-forth within the German bureaucracy, until the relatively mi-
nor sum of 90 Reichsmark was finally approved for Trabert’s work
on data from Cameroon and East Africa.12 The situation was similar
in the other colonies of the German Empire, where meteorological
observation and the publication of data were largely contingent on
the efforts of individuals and non-governmental associations like
colonial and scientific societies.13
The organization of meteorological observation in the colonies
began to take a more definite shape in the last decade of the
nineteenth century. The network of meteorological stations grew
gradually but steadily (Fig. 1). By 1905, Southwest Africa already
counted 71 stations. Although, this number may seem more
impressive than it was in reality, many of the stations were
equipped with only a simple rain gauge, and very few of the sta-
tions had supplied complete multi-annual data series.14 By 1914,
the network had grown further to comprise 20 more fully-
equipped stations with at least a rain gauge and a thermometer
in Southwest Africa, 50 in East Africa, 14 in Togo, and about 20 in
Cameroon.15 This meant a wealth of new information on meteo-
rological conditions of different parts of the colonies and, thus,
more insight into economic possibilities. The mode for processing
meteorological data from overseas also became more streamlined.
11
BArch R1001/6132, n.p.: Danckelman to Auswärtiges Amt (AA), 12 March 1890.
12
The research was published as: Trabert (1894).
13
See: Fitzner (1907), pp.iii; the book contains a valuable bibliography of publi-
cations on meteorological observations in the German colonies.
14
Fitzner (1907).
15
Schnee (1920), (vol. III), pp.552e553.
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
3
The observations now travelled in handwritten tables or pre-
printed forms from the field stations to special sections within
the colonial governments, where they were collated. From there,
the data were sent by steamship to the Foreign Office in Berlin, and
finally to the Seewarte in Hamburg for analysis and publication
(Fig. 2). Yet, among both meteorologists and some bureaucrats the
consensus persisted that there was still much room for improve-
ment, particularly in the day-to-day practice of observation and in
the timeliness of transmission of the data to Germany.
2. Colonial meteorology in conflict and crisis
Alexander von Danckelman, who was both an experienced
meteorologist and a scientific advisor to the Colonial Division in the
Foreign Office, alerted his fellow bureaucrats in 1897 that the public
would soon become outraged that “nothing is happening for
[colonial] meteorology.”16 This warning came after Danckelman
had learned about the potential recall of the chief meteorologist
Hans Maurer from East Africa. Maurer’s position had only been
established a few years prior, when Danckelman himself had rec-
ommended a more direct oversight of meteorological observations
in the colonies to the Foreign Office.17 The looming recall of Maurer
would not have simply been a standard bureaucratic reshuffle but
would have meant the elimination of the position itself and, thus,
the dismantling of the only centralized meteorological service
among all of the German colonies. The urgency in Danckelman’s
dispatches speaks to his constant anxiety that not only was there
no tangible progress in the development of colonial meteorology,
16
BArch R1001/6132, n.p.: Danckelman to Schmidt (Colonial Divison), 9 January
1898; see also: BArch R1001/6137, n.p.: Danckelman to Colonial Division, 3 January
1898.
17
BArch R1001/6136, n.p.: Neumayer (Seewarte) to Colonial Division, 15 July 1895.
4 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
Fig. 2. Precipitation tables from German Southwest Africa, 1898e1899.
(BArch R1001/6139: “Regenmeb-Beobachtungen in Deutsch-Südwest-Afrika,” Windhoek, 1900).
but also the very real danger of undoing the modest advances that
had been achieved in the last decade of the nineteenth century.
The sentiment of chronic crisis that Danckelman exhibited in his
missives mirrored the tenor of a large conference for the directors
of German meteorological central offices that had taken place in
October of 1897. Although most of the participants represented
stations in metropolitan Germany, the item on the agenda dealing
with “improvements of German overseas observations and their
publications” was discussed first and garnered the most attention.18
The participants drew a bleak image of the situation of meteorology
in the colonies, criticizing the lack of standardized practices during
the recording of observations, the irregularity of data submissions,
and the problems of finding eligible and trained observers for
service in the colonies. The last of these points was the most
pressing of complaints. Wilhelm von Bezold, the director of the
meteorological stations in Prussia and an early proponent of ther-
modynamic approaches in the atmospheric sciences, described the
situation of official travelers setting out for the colonies with little
or no training: “[t]hey often do not know from where they can
obtain instruments and how those instruments are designed, so
that they regularly take instruments [on their travels], with which
they have no practical experience.”19 The resolution of the confer-
ence called for a more organized teaching and training program for
colonial travelers and officials, although it failed to describe any
concrete steps towards that goal.20
18
BArch R1001/6132: Verhandlungen der Konferenz der Vorstände deutscher
meteorologischer Centralstellen zu Berlin vom 13. bis 17. Oktober 1897.
19
BArch R1001/6132: Verhandlungen der Konferenz der Vorstände deutscher
meteorologischer Centralstellen zu Berlin vom 13. bis 17. Oktober 1879, p.8.
20
BArch R1001/6132: Verhandlungen der Konferenz der Vorstände deutscher
meteorologischer Centralstellen zu Berlin vom 13. bis 17. Oktober 1897, p.31.
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
Bezold’s criticism revealed the full extent of his annoyance at the
state of colonial meteorology, but it also spoke of his underlying
conviction that the issues could eventually be solved and that the data
from the colonies could serve a useful purpose to gain an overview
over the climatic conditions in the colonies or even the continent of
Africa as a whole. This conviction of the feasibility of their project
united the meteorologists at the 1897 conference, who focused their
discussions on approaches to ensure more accurate, standardized,
and comparable data collection both in the metropole and in the
colonies. Implicitly, this also meant the elimination of the individual
observer and his embodied experiences as a factor in the recordings.
The perfect field meteorologist, according to this model, would be
someone trained according to a standard protocol to read, calibrate,
and possibly repair the newest and most accurate instruments used to
record atmospheric conditions. Yet, both the standardized observers
and standardized instruments were often absent from colonial
stations.
The calls for better instruments and better training were not
new around the turn of the century, but they had rarely led to any
large, structural changes e mostly for financial reasons. The
repeated criticism of meteorological observations in the colonies
did, however, lead to some heated exchanges between the colonies,
Berlin, and Hamburg e such as Neumayer’s dispatch about the low
quality of observations introducing this article. In response, the
government in German East Africa rebutted Neumayer’s critique by
calling his prescriptions for improvement naïve and even “entirely
unfeasible [gänzlich unausführbar].”21 Together with Danckelman,
Neumayer had suggested the regular control of field stations by a
21
BArch R1001/6136: Office of the Governor of German East Africa to the Colonial
Division, 4 July 1894.
 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
trained meteorologist in the colony who could ensure at least some
level of standardization. But e as the governor was quick to point
out e even brief visits by one individual to each station in the large
and mostly roadless colony would take up more than year.22
The rebuttal from East Africa to Neumayer’s missive suggested
that meteorologists in Germany did not have any insight into the
conditions and difficulties of work in the colonies. In a letter a
month later, the governor of East Africa repeated his disagreement
with the directives from Germany. In light of his belief that nothing
could be done about the lack of observers in the colony that would
satisfy the demands of Neumayer, he proposed flippantly to halt
meteorological observation altogether. And in an even huffier tone,
he offered to return all meteorological instruments to the See-
warte.23 While the threat may not have been entirely serious, it
targeted a sore spot by portending to take away any possibility for a
standardized collection of quantitative data. In this particular case,
however, Berlin and Hamburg triumphed over Dar es Salaam. As
seen in Danckelman’s comments above, the Foreign Office ulti-
mately dispatched a meteorologist to East Africa to supervise the
observations of lay practitioners, although the position remained
precarious even into the twentieth century and the model was
never implemented in any of the other colonies.
These bureaucratic exchanges reveal that there seems to have
been a progressive breakdown of communication between the
colonies and the metropole, as the colonial governments developed
their own institutions and approaches to deal with local circum-
stances and difficulties, which officials in Germany neither knew
nor understood fully. The reply of the colonial government to
Neumayer’s critique about the low quality of observations is telling:
step-by-step, the untrained observers would learn on the job and
eventually produce increasingly more accurate observations.24 This
was sufficient for the purposes of the small bureaucracy in Dar es
Salaam, which tried to manage a vast territory and several political
crises at once, and for whom accuracy and comprehensiveness in
meteorological observations was not one of the most pressing
concerns. It was, however, clearly not good enough for the neat
collection of uninterrupted data series that Neumayer aspired to
and that had become the standard of scientific work in publications
such as the Meteorologische Zeitschrift. The reply to Neumayer also
seemed to assume that colonial observers could learn by gathering
more direct experience with their local environments, while the
data series the Seewarte aspired to required learning and experi-
ence that transcended the particular locale. The demands of the
colonial administration e and thus the scale of the colony e clashed
with the demands and scale of metropolitan climatology, which
was seeking to establish firm quantitative, and globally comparable,
data series.25
Colonial budgets increased over the next decade, not least as a
response to two costly wars in East and Southwest Africa.26 While
the wars themselves disrupted meteorological observation even
further, the new level of military, political, and socio-economic
control in the post-war years also increased the chances for more
centralized oversight of observations in the overseas territories.27
Yet, once the dust had settled, the situation of meteorological
22
BArch R1001/6136: Office of the Governor of German East Africa to the Colonial
Division, 4 July 1894.
23
BArch R1001/6136: Office of the Governor of German East Africa to the Colonial
Division, 29 August 1894.
24
BArch R1001/6136: Colonial Division to Seewarte, 7 September 1894.
25
On the development of global approaches and the value of precision in
nineteenth-century meteorology, see: Anderson (2005); see also: Nebeker (1995).
26
Zimmerer and Zeller (2003); Zimmerer (2008); Beez (2003); Giblin and
Monson (2010); Schaller (2010).
27
See, for example: Hann (1906).
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
5
practice on the ground did not change markedly. Not only did the
complaints about an incomplete coverage of the colonies with
meteorological stations continue unabated.28 The collected data
and the practices and technologies of collection also continued to
be the target of critique from Berlin and Hamburg. And while the
Seewarte and meteorological observers in Europe were still hoping
for comprehensive data sets from the colonies that could both
provide material for comparative climatology and further advance
the coloring-in of climatological world maps, the government in
Southwest Africa made it quite clear that other matters were more
pressing than gathering information about the weather: “Meteo-
rology is necessary,” a missive read, “but because of the associated
costs, it has to be limited to what is absolutely essential.”29
3. Instruments, training, and colonial realities
Addressing the difficulties of the day-to-day practice of mete-
orology, Theodore Feldman has written that “[o]bservers were
scarcely more reliable than their instruments. The discipline of
recording daily temperature, pressure, humidity, winds, and cloud
cover over a period of years did not come easily.”30 In this assess-
ment, Feldman is referring to practices in Enlightenment meteo-
rology, but the issues in colonial meteorology more than a hundred
years later would also fit the description. Far from working in a
“laboratory of modernity,” in tune with the disciplinary de-
velopments in the metropole, the colonial observers in Africa had
to grapple with the local circumstances of their small, makeshift
stations and the everyday challenges of recording the weather.31
In 1908, the editor of the Deutsche überseeische meteorologische
Beobachtungen [German Overseas Meteorological Observations]
Paul Heidke used a short review article of his own journal to vent
his frustration and to warn his fellow meteorologists that the
conditions in the colonies were not conducive to good work and
that the meteorological data collected overseas could not be
compared to data gathered in Germany:
Unfortunately, most stations [in the colonies] are not able to
provide complete data series, because most observers do not
have a second-in-command and are sometimes prevented [from
doing their work] through illness, official trips, or other un-
foreseen events. This also explains why many stations had to
close down after the departure of the observer. Unfortunately,
the employment of deputies cannot be demanded, because
there are no funds to provide the observers with even a small
salary, from which they could pay their deputies.32
In fact, the only compensation offered to observers in the col-
onies were copies of Heidke’s journal publishing their observations.
Requests by the colonial governments for even just symbolic pay-
ments to select groups of meteorological observers were cut short
immediately by the Colonial Division, which feared it would set a
bad precedent and could lead to bad blood among the observers
who would not be among the chosen few.33
In his review, Heidke continued by complaining that most
meteorological observers were soldiers in the colonial army, who
28
See, for example: Fitzner (1907), p.iii; Maurer (1911), p.18; Dove (1913), p.94.
29
BArch R1001/6140: Office of the Governor of German Southwest Africa to the
Colonial Office, October 1907.
30
Feldman (1990), p.149.
31
Laak (2004).
32
Heidke (1908), p.238.
33
BArch R1001/6139: Colonial Division to Office of the Governor of German
Southwest Africa, May 1903.
6 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
were frequently detached to new places and who often fell ill due to
the foreign climate (Heidke showed no appreciation of the subtle
irony of the situation). Whether the reported cases of sickness were
all true to the facts was a different question. Some of the observers
seem to have been rather unwilling participants in colonial mete-
orology. Far from the enthusiasm of earthquake observers around
the same time, weather observers complained about the tedium of
regular instrument readings at remote stations early in the morn-
ing or late at night.34 While observers often happily pointed out
particularly extreme conditions or atmospheric phenomena, the
scheduled readings of instruments multiple times per day were less
exciting and often interfered with other obligations. On the back-
end of the process, things did not look any better; Heidke’s re-
view referred to a sizeable lag of up to twelve years between the
recording and the analysis of colonial data, which prevented any
potential inquiries about suspicious figures after the fact.35 It also
meant that the practitioners in the colonies rarely saw the fruits of
their own labor while still employed as observers.
While being particularly forceful in its critique, Heidke’s article
was no exception. Along with the slow transfer of data from the
colonies to Germany, the lack of training among the observers in
the colonies was one of the most intensely debated and most
enduring issues.36 In 1894, Danckelman described the “frequent
change of observers” and the “employment of medical assistants
[Lazarettgehülfen]” as the “cancer, under which the whole system
suffers.”37 It prevented any observer to fully acquaint himself with
the instruments and the local circumstances and thus made the
collection of standardized and comparable data all but impossible.
There were many reports from the colonies that observers did not
have any instructions how to operate the instruments in their
charge, leading to problems ranging from difficulties to take read-
ings from hygrometers, to the improper installment of rain gauges
in partially covered places or the improper installment of ther-
mometers in places exposed to direct sunlight.38
Even in the last years of the German colonial empire, the criti-
cism still revolved around the same issues: neither could the
available colonial data be fully trusted, nor was enough work being
done to get to the bottom of climatological conditions in the col-
onies, because most of the observers were laymen e most
commonly low-level colonial officials, teachers, low-level officers
in the service of the colonial army, or the medical assistants that
Danckelman had criticized so harshly.39 One group of observers
that almost disappeared entirely in the exchanges and writings
about colonial meteorology were the practitioners drawn from the
local population, although there is evidence to suggest that they
played a non-negligible part in the recording of meteorological
information.40 The lack of attention paid to African observers also
meant, however, that they were not singled out for criticism. On the
contrary, colonial officials occasionally argued for the replacement
of Europeans with less costly local observers. This happened in
German East Africa, when the chief meteorologist sent from
34
Coen (2013).
35
Heidke (1908), p.237.
36
See: Ruppenthal (2007), p.29.
37
BArch R1001/6136: Office of the Governor of German East Africa to the Colonial
Division, 29 August 1894; see also: Jahresbericht über die Tätigkeit der deutschen
Seewarte (1894), p.32.
38
See: Wohltmann (1898), p.9; BArch R1001/6139: Office of the Governor of
German Southwest Africa, July 1902.
39
See, for instance: Dove (1913), p.93; BArch R1001/6138: Booklet “Der Pflanzer,”
(1911); see also: Gräbel (2015), p.181. On amateur science, see: Allen (2009), McCray
(2006); on the constantly negotiated nature of data in amateur and citizen science,
see: Aronova (2017).
40
See: Raj (2016).
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
Germany was allegedly not up to his tasks and the colonial gov-
ernment floated the idea to replace him with “a cheap Indian
worker,” presumably drawn from the large South Asian population
in the coastal cities of the colony.41
Another issue that remained an enduring topic of conversation
between colonial officials, observers, and metropolitan meteorol-
ogists were the problems with meteorological instruments in the
colonies. In his article, Heidke referenced not only widespread is-
sues with the control and adjustment of minimum-maximum-
thermographs, which recorded the lowest and highest tempera-
ture in a given day; he also referred to the fragility of some in-
struments e for instance the “rather expensive” mercury
thermometers, of which more than half would break during
transport overseas.42 The fragility of instruments had already been
a problem in the early years of the German colonial empire. Some of
the inland stations in German East Africa had to wait for several
years for new instruments, as the attempts to keep mercury ba-
rometers whole on their transport from the coast had failed
repeatedly.43 The fragility of barometers and thermometers would
remain a chronic issue both in the colonial stations and on scientific
expeditions, made traceable by the sudden appearance of blank
spaces for one particular atmospheric condition in meteorological
data sets.44
Like his colleagues about ten years earlier, Heidke also warned
that the training of colonial observers was insufficient in many
ways, and in particular with regard to the operation of hygrome-
ters.45 One of the most commonly used instruments was a pair of
wet and dry bulb thermometers, often used in the form of a sling
psychrometer, which had to be whirled around on a string or a
ratchet at constant speed for a few minutes to obtain an accurate
reading.46 Aside from the difficulties of handling these instruments
correctly e particularly for untrained observers e the psychrome-
ters also fared badly in some of the environmental conditions in the
colonies. Extremely arid conditions could dry out the wet bulb
thermometer and sand or other airborne particles could stick to it
and prevent an accurate recording.47 Other instruments were less
complex but necessitated sustained and delicate efforts at keeping
them running accurately. Thermographs, for instance, required the
observer to “gently lift the paper registration strip” every day to
ensure accurate recordings.48 While many of the issues of handling
instruments were not insurmountable in theory, they required a
level of discipline and regularity that many of the volunteers and
conscripted observers did not have, or did not care to have. Thus, as
the precious few instruments that the small colonial stations had at
their disposal were often out of order and observers had little or no
41
BArch R1001/6137: Note, Office of the Governor of German East Africa, 14
August 1904.
42
Heidke (1908), pp.236e237.
43
BArch R1001/6135, n.p.: Medical Division, German East Africa to Colonial Di-
vision, 17 March 1893.
44
See, for instance: BArch R1001/6135, n.p.: Foreign Office to Wissmann,
November 1889; on the difficulties of transporting fragile instruments on expedi-
tions, see also: Gräbel (2015), pp.198e199.
45
Heidke (1908), pp.237e238.
46
“The instrument [.] consists of two thermometers, as nearly as possible
identical, the one marked Dry, the other Wet. The bulb of the Wet thermometer is
covered with thin muslin, and round the neck of the bulb and over the muslin is
twisted loosely, or tied in a loose knot, a conducting-thread of lamp-wick, common
darning-cotton, or floss silk; this passes to an adjacent vessel of water placed at
such a distance as to allow a length of conducting-thread of about three inches.”
This description is taken from: Glaisher (1893), p.iii.
47
See, for instance: Gülland (1907), p.7; psychrometers are also sensitive to cli-
matic conditions, so that particularly cold, hot, or dry conditions require the use of
calculational corrections or specialized tables to obtain reliable data; see: Bloss and
Orloski (1972), p.265.
48
BArch R1001/6140: Note by the Deutsche Seewarte, February 1909.
 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
training to repair them, colonial practitioners often found them-
selves forced to rely on alternative methods to record atmospheric
conditions, placing more emphasis on descriptive and qualitative
approaches that reflected the personal experience of the observers.
This state of affairs, however, did not mesh with the demand from
metropolitan scientists and officials for comparable numerical data,
expressing the collective experience of the discipline rather than
the experiences of individual observers.
With the founding of the Hamburg-based Kolonialinstitut in
1908, there was finally an official institution for the standardized
training of colonial administrators and other individuals interested
in work or research in the colonies.49 The Colonial Institute would
indeed go on to offer courses in practical meteorology, although the
subject was not made mandatory in the curriculum.50 Over the next
e and incidentally also the last e six years of German overseas
colonialism, however, the impact of the new institute was negli-
gible e at least if we can trust the continuing calls from metro-
politan meteorologists for better education of observers in the
colonies and for the establishment of a centralized and more
thoroughly structured organization of colonial meteorology per se.
At the third Colonial Congress (Kolonialkongress) in 1910, the par-
ticipants agreed on a resolution that called for the establishment of
a professional meteorological service for the colonies and for the
employment of a thoroughly trained expert in each colony.51 By the
eve of the First World War, German East Africa was still the only
colony with a centralized meteorological service. In 1914, the
government of Cameroon still felt the need to pressure the Colonial
Office in Berlin for a meteorologist familiar with non-European
environments to be employed at the Colonial Institute, particu-
larly to ensure an improved training of future officials in the
handling of meteorological instruments.52 A few months later, any
possible measures of improvement were cut short by the start of
the war and the swift defeat of most of the German colonial military
forces in Africa and beyond.
The amateurish nature of the German attempts to gather
meteorological data in the colonies may not seem too outlandish in
the context of the continuing complaints about the shortage of
funds for all kinds of colonial ventures. In scientific endeavors, as in
the economic development of the overseas protectorates, the
German government had initially hoped for private enterprise to
take up the reins. When those hopes were not fulfilled, the German
government increased the funding, but remained cautious in light
of growing colonial deficits.53 Yet, the feeble attempts to establish a
working network of weather observation in the colonies stands in
stark contrast to the rhetorical importance given to meteorology
and climatology for practical colonial matters. Both governmental
publications and the pro-colonial press usually classified overseas
meteorology as part of applied geography, and thus as part of the
applied colonial sciences, which remained closely connected to
human concerns and were ultimately supposed to aid in the eco-
nomic exploitation of the territories.54
49
On the history of the institute, see: Paul (2008); Ruppenthal (2007).
50
The mandatory subjects were “colonial policy, colonial law, administration and
jurisprudence, applied geography, tropical hygiene, ethnography and agricultural
knowledge,” and e for those students interested in colonies with a Muslim pop-
ulation, “lectures about the culture of Islam”; see: Ruppenthal (2007), p.212.
51
BArch R1001/6132, n.p.: Deutscher Kolonialkongreß (1911); see also: Ver-
handlungen des deutschen Kolonialkongresses 1910 zu Berlin (1910).
52
BArch R1001/6132: Colonial Government of German Cameroon to Colonial
Office, 7 March 1914.
53
See: Pflanze (1990), (vol. 3), pp.113e184; Fitzpatrick (2008), pp.116e159.
54
The geographer and colonial politician Hans Meyer drew some of the most
direct connections between the practice of colonial climatology and the economic
valorization of the overseas territories; see: Meyer (1902), p.75; Meyer (1910),
p.722.
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
7
The further the German colonial project developed into the
twentieth century, the more directly the colonial sciences became
concerned with matters of practical colonization.55 In his inaugural
lecture as the first professor of colonial geography in Germany, Fritz
Jaeger stressed in 1911 that “colonial geography is [.] an important
basis for a rational colonial economy.”56 This economic importance
was also openly stated by some of the meteorologists themselves,
who realized the significance of information on climatic conditions
for agricultural development or, at least, saw an opportunity for
funding in emphasizing practical utility: Fitzner, for instance, used
the first sentence of his study on the geography of rain in the
German colonies to highlight the preeminent role of precipitation
as the most economically important climatic element.57 Beyond
this, knowledge of the weather, and thus at least the possibility of
control over the skies and landscapes of a particular territory, could
also contribute to the expansion of power of the colonial state over
both landscapes and people.
The rhetoric, however, did little to change the situation.
Throughout the thirty years of the German overseas empire, colo-
nial meteorological stations remained underfunded and colonial
meteorologists remained under-trained and under-equipped to
fulfill the goals of metropolitan scientists and officials. The German
attempts to gain a “place in the sun,” and thus a position of global
reach, had reached new heights with the colonial expansion over-
seas, but what that sun and all the other elements of the atmo-
sphere were actually doing remained difficult to grasp throughout
the period of the empire’s existence. It is telling that the meteo-
rologist Wilhelm Meinardus, who promoted the study of large-
scale systems of circulation, declared that the time for an analysis
of the periodic changes in the North Atlantic circulation had not yet
come as “that would require a deeper insight into the oceanic and
meteorological conditions of the tropics, and today we are unfor-
tunately still very far from getting there.”58 Despite all of these
problems, however, work at the meteorological stations in the
colonies continued, developing far away from the institutions and
publications of science in Europe.
4. Provincializing climates
Faced with the discrepancy between the rhetoric of trans-
continental or even global empires and the local inadequacies of
meteorological work outside of Europe, the well-known geogra-
pher Alfred Hettner drew a distinction between the meteorology
undertaken in “civilized states” and that practiced in the colonies.
The two were, in his view, entirely different kinds of endeavors:
regarding the study of the atmosphere in Europe, Hettner
emphasized the importance of instruments, especially thermom-
eters and barometers, and designated the creation of dense
observation networks as central to the foundation and the pros-
pects of meteorology and climatology as independent scientific
disciplines.59 Describing the study of weather and climate beyond
Europe, and thus beyond dense observation networks, Hettner’s
take had a very different ring to it: “Instrumental observations do
not suffice for the purpose [of scientific investigation],” he wrote:
“there are some weather conditions that can not be grasped with
[instruments], but which can only be perceived through the eye or
the direct impression [Gefühl] and that are not understood
55
See: Gräbel (2015), pp.90e92, pp.103e109.
56
Jaeger (1911), p.405.
57
Fitzner (1907), p.iii.
58
Meinardus (1904), p.362.
59
Hettner (1911), p.426.
8 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12

quantitatively, but qualitatively.”60 Hettner even went so far as to
claim that work that only relied on numbers and ignored qualita-
tive information could not be called “fully scientific” and he criti-
cized “most climatological representations [as] far too statistical
and not sufficiently physiological.”61
Hettner, a student of Georg Gerland in Strasbourg and of Frie-
drich Ratzel in Leizpig, had developed the notion of geography as a
chorological science, or a science of space and spatial relations. In
his own empirical work on South America and Russia, he empha-
sized the scale and particularity of the region as the most important
unit of analysis.62 Thus, the emphasis on regional differences was
part and parcel of Hettner’s geographical approach in general, and
not reserved to his portrayal of meteorological practice in the col-
onies. The focus on the senses and on qualitative description
became particularly pronounced, however, when Hettner referred
to what climatological work e and geographical work in general e
should look like beyond the borders of Europe. His stance was also
reflected in the writings of some of his students and colleagues,
who worked e in one way or another e on meteorological obser-
vation in the colonies. And his emphasis on non-instrumental,
direct sensory, and embodied ways of understanding climate was
at least partly connected to the lack or dearth of instruments and
the generally difficult conditions of scientific work in non-
European environments.63 Hettner, however, did not see colonial
practices as less “scientific” than those done in Europe and thus
came close to ascribing two different kinds of rationality, or even
objectivity, to the different approaches.64
One of Hettner’s students, who collected a wealth of first-hand
experience in colonial meteorology, was the German geographer
Franz Thorbecke. In the introduction to his report on an expedition
in Cameroon from 1911 to 1913, he attempted to explain the gaps
that his report contained. Quite early into his journey he wrote, his
party had been left without any registering thermometers, which
had inadvertently been packed in reserve suitcases and deemed
lost. On top of that, both the psychrometers broke during the trip
and could not be repaired or replaced.65 Even during less eventful
days of the expedition, Thorbecke struggled to make regular
quantitative measurements, citing the time-intensive “manage-
ment of the expedition” and “care for the often large staff of colored
servants and porters.”66 All of the adverse circumstances also
meant that the quantitative parts of Thorbecke’s “climatological-
meteorological diary” remained incomplete, and that he used
qualitative e and sensory e descriptions of the weather to make up
for the gaps.67
Thorbecke, however, also had less pragmatic and more pro-
grammatic and even aesthetic reasons for his qualitative ap-
proaches, as he explained in a particularly lyrical passage of the
expedition report:
60
Hettner (1911), p.426.
61
Hettner (1911). pp.427e429.
62
Wardenga (2000); Wardenga (1995); Plewe and Wardenga (1985); Harvey and
Wardenga (2006), pp.427e428.
63
See, for instance, Jaeger and Waibel’s description of taking altitude measure-
ments in Southwest Africa without the aid of an aneoroid barometer; in: Jaeger and
Waibel (1920), p.4.
64
See: Douglas (2004); Daston and Galison (1992).
65
Thorbecke (1951), vol.4, pt.2, p.7; Thorbecke’s expedition report was published
in four volumes, which appeared between 1914 and 1924, except for the final part
of the fourth volume, which only appeared posthumously in 1951. It was edited by
Franz Thorbecke’s wife Marie Pauline Thorbecke, who accompanied her husband
on his expedition and recorded landscapes, people, and the daily work of the
expedition in photographs, drawings, and paintings; see: Horstmann (2013); Kraus
(1971); Gräbel (2015, 180).
66
Thorbecke (1951), vol.4, pt.2, pp.6e8; c.f.: Hassert (1908), p.627.
67
Thorbecke (1951), vol.4, pt.2, pp.115e285.
I was especially captivated by the play of colors e changing
through all scales of the rainbow e that were visible on the
horizon and on the borders of clouds, the vast, always-
wandering, tropical cumuli, as well as the cloud covers
spanning the entire sky, made up of stratus and cumulostratus
clouds. Later I realized that in the literature on the weather
and climate of the tropics, these color phenomena, which
often determine the appearance of the landscape, and their
sometimes rapidly changing manifestations are barely noticed
or even overlooked; especially to geographers, however, these
atmospheric phenomena e only to be observed and retained
by the senses and not by instruments e should be of
interest.68
Thorbecke thus turned a weakness of colonial meteorology into
a purported strength. The complex and fragile instruments e his
message seemed to say e were not that important in non-European
environments, where the direct sensory perception was key to both
aesthetic experience and scientific inquiry.69
This was not simply an idiosyncratic way of conceptualizing
scientific work in the colonies. The Head of the German Colonial
Office, Bernhard Dernburg, agreed with the sentiment by closing
his description of “colonialism as science and technology” with a
particular emphasis on the senses: sitting in lecture halls would not
suffice for the next generation of colonial officials, who would have
to gain first-hand experience in the colonies and learn to under-
stand the conditions in overseas territories through Augenschein, or
“visual inspection.”70 This focus on sensory perception, rather than
instrument readings and recordings, also managed to break
through the colonizer/colonized distinctions, as in Karl Dove’s
praise of the indigenous populations in Southwest Africa as the
“best experts of the territory,” who alone understood the true
importance of precipitation for the colony. Even their “rain magic,”
the meteorologist and expedition leader Dove argued, was thus
superior knowledge to that of the “armchair theoreticians [Theor-
etiker vom grünen Tisch]” in Germany.71
For the colonial observers themselves, colonial meteorology and
climatology often became subjects in themselves, apart and
different from the disciplines taught and developed at universities
and other institutions in Europe. This was not just a matter of
placing more emphasis on particular climatic conditions in the
colonies e for example on the striking variability of rainfall in arid
territories e but also contained qualitative shifts.72 Far from being
mere “fact gatherers,” colonial practitioners developed their own
approaches and methods.73 Critical of the value of instruments and
standardization, colonial observers tended to place a higher value
on sensory data and, thus, also exhibited a more holistic and
embodied approach to the study of climates. In practice, this shift
could look like Thorbecke’s description of meteorological condi-
tions in Njua in Cameroon: “Occasionally, gusts of wind arrive from
the mountain (NE); when they blow with force, they unsettle the
red dust from the village square; in a more attenuated form, they
have a refreshing and cooling effect.”74
68
Thorbecke (1951). vol.4, pt.2, pp.6e7.
69
For a similar argument, see: Schultze (1910), p.145; Schultze argued that the
precautions that meteorologists had to take to get accurate instrument readings
prevented them from appreciating and recording the often-rapid changes of tem-
peratures in Southwest Africa.
70
Dernburg (1907), p.7.
71
Dove (1896).
72
See: Maurer (1911), p.19; Maurer (1902), pp.547e48.
73
Cf.: Harrison (2005), p.56.
74
Thorbecke (1951), vol.4, pt.2, p.9.

Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
Fig. 3. “Climates of the Earth.”
(Wladimir Köppen, “Klassification der Klimate nach Temperatur, Niederschlag and Jahreslauf,” Petermanns Mitteilungen 64 (1918), pp.193e203).
With their shift away from regularized and quantified instru-
ment readings, colonial practitioners broke with Julius Hann’s
highly influential definition of climates as the “mean condition of
the atmosphere in a particular place at the surface of the earth.”75
The arguments for the particular importance of sensory impres-
sions seemed to fit much better into the climate definition of
another influential practitioner, who had himself done most of his
work in colonial environments, if long before the establishment of
German colonies or even the German nation-state itself: in his 1845
treatise Kosmos, Alexander von Humboldt had described climates
as changes in atmospheric phenomena that “markedly affect the
human organs.”76 And this definition was quite close to Hettner’s
and Thorbecke’s descriptions of what they deemed important
features of meteorology and climatology in the colonies.
5. The coproduction of climatological scales
Humboldt’s definition of climate had generally receded into the
background with the emergence of more quantitative and statis-
tical definitions in the late nineteenth century. His visual in-
novations, on the other hand, enjoyed a more lasting effect.77
Among the most important visual tools from Humboldt’s reper-
toire were isolines, which connect points on a map sharing a
common value of a particular quantifiable feature.78 Climatologists
around the turn of the twentieth century used this visual tech-
nology widely, particularly in the growing number of representa-
tions of atmospheric phenomena on a trans-regional, continental,
75
Hann (1883), p.1; for the English version, see: Hann (1903).
76
Humboldt (1845), (vol. 1), p.340; see also: Heymann (2010), p.587; Bernhardt
(1984); Bernhardt (2003); for the embodied practices of Humboldt’s climatology,
see: Cushman (2011), pp.24e25.
77
Güttler (2014).
78
See: Godlewska (1999); Grevsmühl (2014a); Grevsmühl (2014b).
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
9
or even global scale. Vladimir Köppen’s work on world climatic
zones may be one of the most famous examples (see Fig. 3).
The data that Köppen used to construct his maps came from all
around the globe, including the German colonies. And in that sense,
the colonies were highly important to permit and facilitate a
“scaling-up” of the climatological vision to encompass the entire
earth.79
Were Köppen’s maps a truly global vision of the world? This is, in
the last instance, a matter of definition or emphasis, depending on
whether the isolines are read horizontally, and thus as connecting
points around the globe, or vertically, as dividing areas with distinct
atmospheric characteristics. There is, however, a strong argument for
the climatological world maps representing an “appearance of
globality,” just as the spread of European science gave off the
“appearance of universality,” rather than embodying a truly univer-
sal and ahistorical phenomenon.80 Like most representations of
global scale in the colonial period, Köppen’s maps were built on
highly local and contested data, which in turn were based on highly
local and contested practices of data gathering. The conditions of
data production remained in the pages of governmental memo-
randa. They only rarely made it into published articles in meteoro-
logical journals, which often used the colonial data quite literally as
“a given.” Meteorology in the nineteenth century was thus certainly
one of the prime examples of a “data-driven science,” but it was
driven by data that were generally cleared of their own history.81
79
Katharine Anderson has made the point that isolines could also be thought of as
the clearest expression of the connection between local and global scale in
nineteenth-century meteorology: “Meteorologists were particularly aware that
their science was based on connecting local experiences and global phenomena. At
its most vivid level, these shifts in perspectives were the basic premise of
nineteenth-century mapping techniques, the isobars and isotherms, lining up
places where barometers and thermometers gave the same numbers.” From:
Anderson (2005), p.290.
80
This is discussed in: Chambers and Gillespie (2000), p.239.
81
The term is borrowed from: Strasser (2012).
10 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
The case of data from the German colonies is also a prime
example of what Paul Edwards has described as the process of
“making data global,” that is “building complete, coherent, and
consistent data sets from incomplete, inconsistent, and heteroge-
neous data sources.”82 The practitioners who were working in the
colonies had little to do with the global vision that began to emerge
on the basis of the data they produced. They worked, usually un-
paid and often unwillingly, in the context of their particular station
and with all the immediate problems of what the day-to-day work
of reading and fixing instruments brought with them. To some
extent, this distinction between the extremely local e found in
diplomatic exchanges e and the global e found in climatological
articles of the time, was a consequence of the stark geographic and
temporal separation between the collection of data in the colonies
and the analysis of the data in Hamburg or other places in Germany
and Europe. Unless they received some critique of their methods
from Berlin or Hamburg, the observers who recorded meteoro-
logical information in the colonies had no knowledge of what
happened to their data once they sent it to their superiors in the
colonial capitals, who then relayed it to Germany by steamship. And
because of the isolation of stations in the colonies, the observers
also had no, or at least far fewer, opportunities to see and interact
with the meteorological publications and maps than their col-
leagues in Europe.
However, even among the group of geographers who combined
the collection with the analysis of colonial data and who traveled
between Europe and Africa the global scale remained in the deep
background. The scientific travelers who toured through the over-
seas regions of the German colonial empire tended to frame their
work along similar lines, and described similar issues, as the ob-
servers in colonial stations. While the gaze of the scientific travelers
was usually not bound to one particular station, it also often fell
short of a global vision e and thus a vision that would have been
congruent with approaches and questions in large-scale clima-
tology, built on practices of standardization and quantification
across borders. The scale that travelers and colonial geographers
emphasized was that of the colonial region or that of the colony,
and usually of the particular region or colony.83 Similar to Austrian
approaches of Kleinklimatologie, or small-scale climatology, that
Deborah Coen has described, German colonial practitioners
emphasized the heterogeneous nature of climatic conditions
overseas.84 Jaeger emphasized the “infinite diversity [Man-
nigfaltigkeit] of landscapes” around the earth, Hettner used the
same term to emphasize the inability of meteorological stations to
capture the entirety of atmospheric conditions in extra-European
territories, and Thorbecke introduced his study of the Abyssinian
highlands with a paragraph on their unique, and thus incompa-
rable, climatic characteristics.85 In the instance of colonial meteo-
rology, the experience of far-away lands often engendered an
emphasis on local specificity and difference, rather than a more
inclusive or even global vision.
This, however, was not just an issue of scale, nor merely an issue
of the problem of induction, which bothered many practitioners
who questioned the construction of a global understanding of
meteorological phenomena on the basis of particular, local data.86
The vision of the contingency of colonial landscapes e not least
82
See: Edwards (2010), pp.251e286.
83
On the importance of the scale of the region in processes of scaling in the field
sciences, see: Vetter (2010); Coen (2011).
84
Coen (2006).
85
Jaeger (1911), p.401; Hettner (1911), p.427; c.f.: Hettner (1897); Thorbecke
(1933).
86
See: Anderson (2005), pp.131e170.
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
founded on the very material difficulties of meteorological obser-
vation in colonial environments e also led practitioners to seek
alternative routes for observational practice. And as instruments
were often uncalibrated, broken, or misplaced, and observers had
little or no training to handle or fully rely on them, some practi-
tioners developed alternative strategies, which placed a higher
value on individual embodied experience and sensory
perception.87
In the context of the development of the global atmospheric
sciences, the episode of German colonial meteorology is certainly
not the central strand, nor can it claim to stand for the development
of colonial climatology in different times and places. It is none-
theless an illuminating example of the different scales on which
meteorology was practiced and envisioned, which were neither
only representations of the mere extension in space of the obser-
vational gaze and of the observable phenomena themselves; nor
did they follow any inherent logic of disciplinary development.88
Rather, they were entangled in the particular circumstances of
colonial politics and the realities of data collection on the ground.
The development of the atmospheric sciences around the turn of
the century was thus not a simple, albeit slow, move from local to
global visions of the earth. This move was happening, to be sure,
but it was not uncontested and it went along with other lines of
development.
The transfer of information from the local to the global e which
was also often a spatial transfer of data from the colonial places of
production to the metropolitan places of analysis e went along
with a selection for particular kinds of data e quantitative data that
could be stripped of the individual experiences of their collectors,
portrayed in tables, and made comparable, or in short: data that
served the collective empiricism inherent in climatology of the turn
of the twentieth century.89 It was only in this step, performed by
practitioners in Europe, that the selected and filtered data took on
the appearance and scale of the global. Direct sensory data
collected and described by the observers did not make it very often
from the colonies to the pages of meteorological and climatological
journals, and that despite the importance that some practitioners
placed upon the personal experience of atmospheric conditions.
While observers recording non-European climatic conditions thus
aided in the envisioning of an interconnected, global environment,
they did so at the expense of pushing their own embodied expe-
rience to the colonial margins of climatology.
References
Allen, D. E. (2009). Amateurs and professionals. In P. J. Bowler, & J. Pickstone (Eds.),
The Cambridge history of science: The modern biological and earth sciences (pp.
15-33). Cambridge, UK: Cambridge University Press.
Anderson, K. (2005). Predicting the weather: Victorians and the science of meteo-
rology. Chicago: University of Chicago Press.
Aronova, E. (2017). Citizen seismology, stalinist science, and Vladimir Mannar’s cold
wars. Science, Technology & Human Values, 42(2), 226-256.
Beez, J. (2003). Geschosse zu Wassertropfen: Sozio-religiöse Aspekte des Maji-Maji-
Krieges in Deutsch-Ostafrika (1905-1907). Cologne: Köppe.
Bernhardt, K.-H. (1984). Alexander von Humboldts Auffassung vom Klima und sein
Beitrag zur Einrichtung von meteorologischen Stationsnetzen. Zeitschrift für
Meteorologie, 34, 213-217.
Bernhardt, K.-H. (2003). Alexander von Humboldts Beitrag zu Entwicklung und
Institutionalisierung von Meteorologie und Klimatologie im 19. Jahrhundert.
Algorismus, 195-221.
87
On the value of a history of observation as a bridge between the history of
science and the history of the senses, see: Daston (2008), p.98.
88
See: Coen, Fleming, Jankovic (2006).
89
This is not to suggest that there was ever anything like “raw data,” even before
the processes of transmission to and selection in the metropole; see: Bowker
(2005); Gitelman (2013).
 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12
Bloss, R. L., & Orloski, M. J. (1972). Precision measurement and calibration: Selected
national bureau of standards papers on mechanics. U.S. Government Printing
Office.
Bowker, G. C. (2005). Memory practices in the sciences. Cambridge: MIT Press.
Chambers, D. W., & Gillespie, R. (2000). Locality in the history of science: Colonial
science, technoscience, and indigenous knowledge. Osiris, 15, 221-240.
Coen, D. R. (2006). Scaling down: The ‘Austrian’ climate between empire and re-
public. In J. R. Fleming, V. Jankovic, & D. R. Coen (Eds.), Intimate universality:
Local and global themes in the history of weather and climate (pp. 115-140).
Sagamore Beach, Mass.: Science History Publications.
Coen, D. R. (2011). Imperial climatographies from Tyrol to Turkestan. Osiris, 26, 45-
65.
Coen, D. R. (2013). The earthquake observers: Disaster science from Lisbon to Richter.
Chicago: The University of Chicago Press.
Coen, D. R. (2016). Big is a thing of the past: Climate change and methodology in the
history of ideas. Journal of the History of Ideas, 77, 305-321.
Coen, D. R., Fleming, J. R., & Jankovic, V. (Eds.). (2006). Intimate Universality: Local
and global themes in the history of weather and climate. Sagamore Beach, Mass:
Science History Publications.
Conrad, S. (2008). Deutsche Kolonialgeschichte. München: Beck.
Conrad, S. (2012). German colonialism: A short history (S. O’Hagan, trans.). Cam-
bridge, UK: Cambridge University Press.
Cushman, G. T. (2011). Humboldtian science, creole meteorology, and the discovery
of human-caused climate change in South America. Osiris, 26, 16-44.
Daston, L. (2008). On scientific observation. Isis, 99, 97-110.
Daston, L., & Galison, P. (1992). The image of objectivity. Representations, 40, 81-128.
Dernburg, B. (1907). Koloniale Lehrjahre. In Zielpunkte des deutschen Kolonialwe-
sens: Zwei Vorträge. Berlin: Mittler und Sohn.
Douglas, H. (2004). The irreducible complexity of objectivity. Synthese, 138, 453-
473.
Dove, K. (1896). Deutsch-Südwest-Afrika. Ergebnisse einer wissenschaftlichen Reise im
südlichen Damaralande. Ergänzungsheft zu Petermanns Mitteilungen 120. Gotha:
Justus Perthes.
Dove, K. (1913). Deutsch-Südwestafrika (2nd ed.). Süsserotts Kolonialbibliothek 5.
Berlin: W. Süsserott.
Edwards, P. N. (2010). A vast machine: Computer models, climate data, and the politics
of global warming. Cambridge: MIT Press.
Feldman, T. (1990). Late enlightenment meteorology. In T. Frängsmyr, H. L. Heilborn,
& R. E. Rider (Eds.), The quantifying spirit in the 18th century (pp. 143-178).
Berkeley: University of California Press.
Fitzner, R. (1907). Die Regenverteilung in den deutschen Kolonien. Berlin: H. Paetel.
Fitzpatrick, M. P. (2008). Liberal imperialism in Germany: Expansionism and nation-
alism, 1848-1884. New York: Berghahn Books.
Giblin, J. L., & Monson, J. (2010). Maji Maji: Lifting the fog of war. Leiden: Brill.
Gitelman, L. (2013). Raw data is an oxymoron. MIT Press.
Glaisher, J. (1893). Hygrometrical tables adapted to the use of the dry and wet bulb
thermometer. London: Taylor & Francis.
Godlewska, A. (1999). From enlightenment vision to modern science? Humboldt’s
visual thinking. In D. N. Livingstone, & C. W. J. Withers (Eds.), Geography and
enlightenment (pp. 236-280). Chicago: University of Chicago Press.
Gräbel, C. (2015). Die Erforschung der Kolonien: Expeditionen und koloniale Wissen-
skultur deutscher Geographen, 1884-1919. Bielefeld: Transcript.
Grevsmühl, S. V. (2014a). La terre vue d’en haut: l’invention de l’environnement global.
Paris: Seuil.
Grevsmühl, S. V. (2014b). The creation of global imaginaries: The Antarctic ozone
hole and the isoline tradition in the atmospheric sciences. In B. Schneider, &
T. Nocke (Eds.), Image politics of climate change visualizations, imaginations,
documentations (pp. 29-53). Berlin: Transcript.
Gülland, A. (1907). Das Klima von Swakopmund. Berlin: Friedrich-Wilhelms-
Universität.
Güttler, N. (2014). Das Kosmoskop: Karten und ihre Benutzer in der Pflanzengeogra-
phie des 19. Jahrhunderts. Göttingen: Wallstein Verlag.
Hann, J. v (1883). Handbuch der Klimatologie. Stuttgart: J. Engelhorn.
Hann, J. v (1903). Handbook of climatology (R. DeCourcy ward trans.). London:
Macmillan.
Hann, J. v (1906). Der tägliche Gang der Temperatur in Windhuk. Mitteilungen von
Forschungsreisenden und Gelehrten aus den deutschen Schutzgebieten, 30-35.
Harrison, M. (2005). Science and the British Empire. Isis, 96, 56-63.
Harvey, F., & Wardenga, U. (2006). Richard Hartshorne’s adaptation of Alfred
Hettner’s system of geography. Journal of Historical Geography, 32, 422-440.
Hassert, K. (1908). Vorläufiger Bericht über einige Ergebnisse der Kamerun-
Expedition 1907/8 des Reichs-Kolonialamtes. Geographische Zeitschrift, 14,
625-628.
Heidke, P. (1908). Literaturbericht: Deutsche überseeische meteorologische Beo-
bachtungen. Meteorologische Zeitschrift, 25, 236-238.
Hettner, A. (1897). Venezuela. Nach den Forschungen von W. Sievers. Geographische
Zeitschrift, 3, 401-408.
Hettner, A. (1911). Die Klimate der Erde. Geographische Zeitschrift, 17, 425-435, 482e
502, 618e633, 675e685.
Heymann, M. (2010). The evolution of climate ideas and knowledge. Wiley Inter-
disciplinary Reviews: Climate Change, 1, 581-597.
Horstmann, A.-K. (2013). Koloniale Geographie - das Ehepaar Marie Pauline und
Franz Thorbecke. In M. Bechhaus-Gerst, & A.-K. Horstmann (Eds.), Köln und der
deutsche Kolonialismus. Eine Spurensuche (pp. 101-106). Cologne: Böhlau.
Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007
11
Humboldt, A. v (1845). Kosmos: Entwurf einer physischen Weltbeschreibung, 1.5
volsStuttgart: Cotta.
Jaeger, F. (1911). Wesen und Aufgaben der kolonialen Geographie: Antrittsvorle-
sung, gehalten im Geographischen Institut der Universität Berlin am 10. Mai
1911. Zeitschrift der Gesellschaft für Erdkunde zu Berlin, 400-405.
Jaeger, F., & Waibel, L. (1920). Beiträge zur Landeskunde von Südwestafrika. Berlin: E.
S. Mittler und Sohn.
Kraus, T. (1971). Marie Pauline Thorbecke. Geographische Zeitschrift, 59, 301-308.
Laak, D. v (2004). Kolonien als Laboratorien der Moderne? In S. Conrad, &
J. Osterhammel (Eds.), Das Kaiserreich Transnational: Deutschland in der Welt
1871-1914 Göttingen: Vandenhoeck & Ruprecht, 257e27.
Livingstone, D. N. (2003). Putting science in its place: Geographies of scientific
knowledge. Chicago: University of Chicago Press.
Mahony, M. (2016). For an empire of ‘all types of climate’: Meteorology as an im-
perial science. Journal of Historical Geography, 51, 29-39.
Maurer, H. (1902). Das Klima von Deutsch-Ostafrika. Meteorologische Zeitschrift,
543-548.
Maurer, H. (1911). Kurze Charakteristik des Klimas der deutschen Schutzgebiete.
Geographische Zeitschrift, 17, 18-30.
McCray, P. (2006). Amateur scientists, the international geophysical year, and the
ambitions of fred whipple. Isis, 97.4, 634-658.
Meinardus, W. (1904). Über Schwankungen der nordatlantischen Zirkulation und
ihre Folgen. Annalen der Hydrographie und maritimen Meteorologie, 32, 353-362.
Meyer, H. (1902). Die geographischen Grundlagen und Aufgaben in der wirt-
schaftlichen Erforschung unserer Schutzgebiete. Verhandlungen des Deutschen
Kolonialkongresses, 72-83.
Meyer, H. (1910). Die landeskundliche Kommission des Reichskolonialamtes. Kolo-
niale Rundschau, 722-734.
Naylor, S. (2005). Introduction: Historical geographies of science: Places, contexts,
cartographies. The British Journal for the History of Science, 38, 1-12.
Nebeker, F. (1995). Calculating the weather: Meteorology in the 20th century. Inter-
national Geophysics series 60. San Diego: Academic Press.
Osborne, M. A. (2005). Science and the French empire. Isis, 96, 80-87.
Paul, L. (2008). Vom Kolonialinstitut zum Asien-Afrika-Institut: 100 Jahre Asien- und
Afrikawissenschaften in Hamburg. Gossenberg: Ostasien Verlag.
Pflanze, O. (1990). Bismarck and the development of Germany: The period of fortifi-
cation, 1880-1898 (Vol. 3)Princeton: Princeton University Press.
Plewe, E., & Wardenga, U. (1985). Der junge Alfred Hettner: Studien zur Entwicklung
der wissenschaftlichen Persönlichkeit als Geograph, Länderkundler und For-
schungsreisender. Stuttgart: F. Steiner Verlag Wiesbaden.
Pogge von Strandmann, H. (2009). Imperialismus vom grünen Tisch: Deutsche Kolo-
nialpolitik zwischen wirtschaftlicher Ausbeutung und “zivilisatorischen” Bemü-
hungen. Berlin: Links.
Porter, T. M. (1986). The rise of statistical thinking, 1820-1900. Princeton: Princeton
University Press.
Raj, K. (2016). Go-betweens, travelers, and cultural translators. In B. Lightman (Ed.),
A companion to the history of science (pp. 39-57). Chichester: Wiley Blackwell.
Ruppenthal, J. (2007). Kolonialismus als “Wissenschaft und Technik”: Das Hambur-
gische Kolonialinstitut 1908 bis 1919. Stuttgart: Steiner.
Schaller, D. J. (2010). From conquest to Genocide: Colonial rule in German South-
west Africa and German East Africa. In A. D. Moses (Ed.), Empire, colony,
Genocide: Conquest, occupation, and subaltern resistance in world history (1st
ed.). (pp. 296-324). Berghahn Books.
Schlegel, K., Schröder, W., & Wiederkehr, K.-H. (2010). Georg von Neumayer and
Geomagnetic Research. History of Geo- and Space Sciences, 1.2, 77-87.
Schnee, H. (Ed.). (1920). Deutsches Kolonial-Lexikon, 3 volsLeipzig: Quelle & Meyer.
Schröder, W., & Wiederkehr, K.-H. (1992). Georg von Neumayer (1826e1909) und
die internationale Entwicklung der Geophysik. Gesnerus, 49, 45-62.
Schultze, L. (1910). Südwestafrika. In H. Meyer (Ed.), Das Deutsche Kolonialreich: Eine
Länderkunde der deutschen Schutzgebiete, 2 (Vol. 2, pp. 129-298)Leipzig and
Vienna: Bibliographisches Institut.
Shapin, S. (1998). Placing the view from nowhere: Historical and sociological
problems in the location of science. Transactions of the Institute of British Ge-
ographers, 23, 5-12.
Speitkamp, W. (2005). Deutsche Kolonialgeschichte. Stuttgart: P. Reclam.
Strasser, B. J. (2012). Data-driven sciences: From wonder cabinets to electronic
databases. Studies in History and Philosophy of Biological and Biomedical Science,
43, 85-87.
Thorbecke, F. (1933). Das Hochland von Abessinien. Geographische Zeitschrift, 39,
292-302.
Thorbecke, F. (1951). Im Hochland von Mittel-Kamerun (Vol. 4)Hamburg: Cram, de
Gruyter & Co. pt. 2. 4 vols.
Tilley, H. (2011). Africa as a living laboratory: Empire, development, and the problem of
scientific knowledge, 1870-1950. Chicago: University of Chicago Press.
Trabert, W. (1894). Einiges über den täglichen Gang der Lufttemperatur zu Bis-
marckburg und des Luftdruckes zu Kamerun. Mittheilungen von For-
schungsreisenden und Gelehrten aus den deutschen Schutzgebieten, 7, 247.
Turchetti, S., Herran, N., & Boudia, S. (2012). Introduction: Have we ever been
‘transnational’? Towards a history of science across and beyond borders. The
British Journal for the History of Science, 45, 319-336.
Vetter, J. (2010). Introduction. In J. Vetter (Ed.), Knowing global environments new
historical perspectives on the field sciences (pp. 1-16). New Brunswick: Rutgers
University Press.
12 P. Lehmann / Studies in History and Philosophy of Science xxx (2018) 1e12

Wardenga, U. (1995). Geographie als Chorologie: zur Genese und Struktur von Alfred
Hettners Konstrukt der Geographie. Erdkundliches Wissen 100. Stuttgart: Franz
Steiner.
Wardenga, U. (2000). Defining geography: The structure and development of alfred
Hettner’s methodological construct. In M. J. Watts (Ed.), Struggles over geogra-
phy: Violence, freedom and development (pp. 113-126). Heidelberg: University of
Heidelberg.
White, S. (2015). Unpuzzling American climate: New world experience and the
foundations of a new science. Isis, 106, 544-566.
Wise, M. N. (1995). The values of precision. Princeton: Princeton University Press.
Wohltmann, F. (1898). Deutsch-Ostafrika, Bericht über die Ergebnisse seiner Reise
ausgeführt im Auftrage der Kolonial-Abteilung des Auswärtigen Amtes, Winter
1897/98. Schöneberg: F. Telge.
Zimmerer, J. (2008). Colonial genocide: The Herero and nama war (1904-8) in
German Southwest Africa and its significance. In D. Stone (Ed.), The historiog-
raphy of genocide (pp. 323-343). Houndmills, Basingstoke: Palgrave Macmillan.
Zimmerer, J., & Zeller, J. (2003). Völkermord in Deutsch-Südwestafrika: Der Kolo-
nialkrieg (1904-1908) in Namibia und seine Folgen. Berlin: Links.

Please cite this article in press as: Lehmann, P., Average rainfall and the play of colors:Colonial experience and global climate data, Studies in
History and Philosophy of Science (2018), https://doi.org/10.1016/j.shpsa.2018.05.007