
































Microsoft Word - 8)Williamson & Wilkinson .docx


History of Meteorology 8 (2017)

 
 

 

159 

 

 

 

Asian Extremes:  
Experience, Exchange and Meteorological Knowledge in Hong Kong and Singapore 

c.1840-1939 
 
 

Fiona Williamson  
ariwfc@nus.edu.sg  

National University of Singapore & Climatic Research Unit, University of East Anglia 
 
 

Clive Wilkinson 
Climatic Research Unit, University of East Anglia 

 
 
 
 
 
 
 

Introduction 

 

On 29 July 1939, the aircraft carrier HMS Eagle was off the northern entrance of the Formosa 
Strait, approximately 25⁰N, 121⁰E. The ship’s meteorological officer was formulating the 
current synoptic weather situation, which included a typhoon to the south or south-east of 
Formosa with a second typhoon much further east in about 144⁰E. It might be expected that in 
1939, the existence and position of a typhoon could be corroborated easily by contemporary 
‘experts’ situated nearby. However ‘The utmost confusion prevailed’ noted the officer ‘among 
the experts at Zikawei, Manila & Hong Kong today …I think there is no doubt that a typhoon 
reached Formosa … this was also confirmed by Zikawei’s signals, but Manila gave it a position 
much further East while Hong Kong stoutly maintained that there were no typhoons on the 
map at all’.1 The level of bewilderment over such a significant event may seem surprising to 
our modern eyes but this would not be considered unusual to anyone who had studied the 
correspondence of the above mentioned observatories for the early twentieth century. 
Confusion, mediocre communication channels and, on occasion, outright antipathy, limited 
what might otherwise have been a profitable and progressive relationship between 
meteorological services. This is explained by the history of the development of meteorology in 
the Asia-Pacific region. 

                                                        
1 UK National Meteorological Archive (hereafter NMA), E17, Meteorological Logbook of the HMS Eagle, 29 
July 1939.  



History of Meteorology 8 (2017)

 
 

 

160 

Pertinent to understanding this story has been the reframing of the science within the 
historiography to reveal its regional and global heritage, with emphasis placed on how the 
circulation of knowledge across time and space led to scientific evolution and discovery.2 This 
should be allied in this instance with the concept of an ‘imperial meteorology’.3 The imperial 
framework creates a permeable space for the transfer of knowledge at the same time as 
providing the resources and contexts for its development.4 In the imperial colonies (not just of 
Britain but the German, French, Dutch and Japanese empires especially), concern over 
desiccation – localised climate change induced by man’s impact on the environment – was a 
significant contributing factor toward increasing investment in meteorological research.5 
Colonial officers engaged in fields as seemingly diverse as forestry, medicine, and engineering 
believed in the importance of improving and increasing observations of rainfall and 
temperature to better understand the phenomena. This investment was made possible by 
technological innovation. Key here were the widespread installation of underwater telegraph 
cable from the 1850s,6 improvements in the accuracy and precision of observational 
instruments, and standardisation in observational practice and reporting.7 Crucial here were the 

                                                        
2 See for example: Brett M. Bennett & Joseph M. Hodge, eds, Science and Empire: Knowledge and Networks of 
Science across the British Empire, 1800-1970 (London: Palgrave MacMillan, 2011); D. N. Livingstone and C. 
W. J. Withers, ‘Thinking Geographically about Nineteenth-Century Science’ in D. N. Livingstone and C. W. J. 
Withers, Geographies of Nineteenth-Century Science (Chicago: Chicago University Press, 2011), 1-20; Lissa 
Roberts, ‘Situating Science in Global History: Local Exchanges and Networks of Circulation’, Itinerario 33 
(2009): 19-30; Simon Naylor, ‘Introduction: historical geographies of science – places, contexts, cartographies’, 
British Journal for the History of Science 38:1 (2005): 1-12; D. N Livingstone, Putting Science in its Place: 
Geographies of Scientific Knowledge (Chicago: Chicago University Press, 2003). 
3 M. Mahony, ‘For an Empire of ‘all types of climate’: meteorology as an imperial science’, Journal of Historical 
Geography 51 (2016): 29-39; M. Zaiki and T. Tsukahara, ‘Meteorology on the Southern Frontier of Japan’s 
Empire: Ogasawara Kazuo at Taihoku Imperial University’, East Asian Science, Technology and Society: and 
International Journal 1:2 (2007): 183-203 
4 J. McAleer, “Stargazers at the World’s End: Telescopes, Observatories, and ‘Views’ of Empire in the Nineteenth 
Century British Empire”, British Journal for the History of Science 46:3 (2011): 389-413; David Aubin, Charlotte 
Bigg and H. Otto Sibum, eds, The Heavens on Earth: Observatories and Astronomy in Nineteenth-Century 
Science and Culture (Durham, NC: Duke University Press, 2010); G. A. Good, ‘A Shift of View: Meteorology in 
John Herschel’s Terrestrial Physics’ in J. R. Fleming, V. Jankovic & D. R. Coen, eds, Intimate Universality: Local 
and Global Themes in the History of Weather and Climate (Mass.,: Science History Publications, 2006), pp. 35-
67; S. M. Razaullah Ansari, ‘Early Modern Observatories in India, 1792-1900’ in Chattopadhyaya, ed., History 
of Science, pp. 349-380: Brian Warner, Royal Observatory, Cape Town, 1820–1831: The Founding of a Colonial 
Observatory (Dordrecht: Kluwer Academic Publishers, 1995). 
5 For more on desiccation see: Gregory Barton, ‘Empire Forestry and the Origins of Environmentalism’, Journal 
of Historical Geography 27, no. 4 (2001): 529-552 or Richard Grove, Green Imperialism: Colonial Expansion, 
tropical island Edens and the origins of environmentalism, 1600-1860 (Cambridge: Cambridge University Press, 
1995). 
6 Katherine Anderson, Predicting the Weather: Victorians and the Science of Meteorology (Chicago: The 
University of Chicago Press, 2005), p. 1. 
7 See M. F. Maury, Explanations and Sailing Directions to Accompany the Wind and Current Charts, 7th Ed. 
(Philadelphia: E. C. & J. Biddle, 1855), Chap. XX ‘Maritime Conference Held at Brussels for Devising a Uniform 
System of Meteorological Observations at Sea; 1853’; V. Janković, Reading the Skies: A Cultural History of 
English Weather, 1650-1820 (Manchester: Manchester University Press, 2000), p. 157. 



History of Meteorology 8 (2017)

 
 

 

161 

adoption of Francis Beaufort’s wind force scale, Luke Howard’s classification system for 
clouds, and Henry Piddington’s standardised vocabulary for storms across the British Empire.8  

Whilst attention has been drawn to the story of meteorology in the China Seas region,9 
and British meteorology and British imperial meteorology in this period,10 meteorology in 
Britain’s overseas colonies in Hong Kong and Singapore has been less considered. Notable 
exceptions include Kevin MacKeown, Ho Pui-Yin and Wai Man-Kui’s work on the Hong 
Kong Observatory and typhoon forecasting respectively.11 All three scholars drew attention to 
the role of typhoons in advancing meteorological research, especially the critical necessity of 
improving the storm warning system and, later, to produce more accurate weather forecasts. 
For the cities of Hong Kong and Singapore a typhoon (Hong Kong) or a flood (Singapore), 
could result in disaster, destroying the lives of local inhabitants and/or the livelihoods of 
maritime traders, shipping companies and fishing industry upon which the ports depended 
economically. The main task of the early weather watchers then was to study and improve their 
knowledge of extreme weather systems and to provide advance storm warnings. This was 
achieved through a combination of techniques, including collating barometric observations 
across as wide an area as possible (terrestrial and marine) and by communicating news of 
approaching storms from regional registering stations, observatories and nearby ships. Key to 
the success of the latter was the effectiveness of communications. The laying of electric 
telegraph across the mid-to-late nineteenth century played a major role here, making possible 
the rapid transmission of news across the meteorological hubs of the Philippines, Japan, China, 

                                                        
8 H. Piddington, The Sailor’s Horn-Book for the Law of Storms (1848); W. Reid, An Attempt to Develop the Law 
of Storms by Means of Facts (1838). 
9 Robert Bickers, “‘Throwing Light on Natural Laws”: Meteorology on the China Coast, 1869-1912’ in Robert 
Bickers and Isabella Jackson, eds, Treaty Ports in Modern China: Law, Land and Power (Routledge, 2016), pp. 
179-200; Gregory T Cushman, ‘The Imperial Politics of Hurricane Prediction: From Calcutta to Manila and 
Galveston, 1839-1900’ in Erika Marie Bsumek, David Kinkela, and Mark Atwood Lawrence, Nation-States and 
the Global Environment: New Approaches to International Environmental History (Oxford Scholarhsip Online, 
2013); James F. Warren, ‘Weather, History and Empire: The Typhoon Factor and the Manila Galleon Trade, 
1565-1815’ in Geoff Wade and Li Tana, eds, Anthony Reid and the Study of the Southeast Asian Past (Singapore: 
ISEAS, 2012), pp. 183-220; James F. Warren, ‘Scientific Superman: Father Jose Algue, Jesuit meteorology, and 
the Philippines under American Rule, 1897-1924’ in A. W. McCoy and F. A. Scarano, eds, Colonial Crucible: 
Empire in the Making of the Modern American State (Wisconsin: The University of Wisconsin Press, 2009), pp. 
508-519; A. Udías, Searching the Heavens and the Earth: The History of Jesuit Observatories 
(Dordrecht/Boston/London: Kluwer Academic Publishers, 2003); Lewis Pyenson, Civilizing Mission: Exact 
Sciences and Overseas French Expansion, 1830-1940 (Baltimore and London: John Hopkins University Press, 
1993). 
10 See, for example, Mahony, “For an Empire of ‘all types of climate’; P. Moore, The Weather Experiment: The 
Pioneers who Sought to See the Future (London: Vintage, 2015); McAleer, “Stargazers at the World’s End; 
Aubin, Bigg and Sibum, eds, The Heavens on Earth; J. Golinski, British Weather and the Climate of 
Enlightenment (Chicago: Chicago University Press, 2007): Anderson, Predicting the Weather; R. Hamblyn, The 
Invention of Clouds: How an Amateur Meteorologist Forged the Language of the Skies (London: Picador, 2001); 
Janković, Reading the Skies.  
11 See P. Kevin MacKeown, Early China Coast Meteorology: The Role of Hong Kong (Hong Kong: Hong Kong 
University Press, 2011); Wai Man-Kui, ‘The Early Tropical Cyclone Warning Systems in Hong Kong, 1841-
1899’, Hong Kong Meteorological Society Bulletin»14:1:2 (2004): 49-81; Pui-yin Ho, Weathering the Storm: 
Hong Kong Observatory and Social Development (Hong Kong University Press, 2003). 



History of Meteorology 8 (2017)

 
 

 

162 

Hong Kong, and Malaya. After 1915, radio telegraphy was used to communicate between 
shipping and the mainland in Hong Kong.  

Nevertheless, as Jan Golinski reminds us, the history of meteorology was not a linear 
or clear-cut progression; caution should be applied in conceptualising it within a developmental 
teleological framework.12 Certainly, the proliferation of conflicting scientific theories would 
dog would-be meteorologists throughout the nineteenth century. The many different 
explanations for cyclonic storms for example – especially pertinent to the China Seas – caused 
decades of debate.13 The collation of a global observational record – one of the main means of 
studying the weather before 1900 – was problematized by a lack of consistency. This was due 
to a scarcity of registering stations, especially outside of urban areas; short operating periods 
of only a few years in some cases before they were abandoned, and problems of communicating 
the observations themselves. The revolutionary telegraph, for example, was plagued with 
problems, malfunctioning or being abandoned due to political unrest, limited or poor 
maintenance, cost, and with some irony, even bad weather. As MacKeown also rightly points 
out, the contentious personalities of the observatory directors – especially the long-standing 
William Doberck in Hong Kong – also limited the possibilities of progressive research between 
countries.  

This paper seeks therefore to accomplish four interconnected aims. First, to consider 
the development of meteorological services in British Hong Kong, and to a lesser extent 
Singapore, to highlight a lesser known aspect of the story of nineteenth and early twentieth 
century colonial meteorological science. Second, to draw attention to the region’s extreme 
weather as a unique factor which differentiated imperial science in the colonies from 
meteorology in Britain. The weather provided a catalyst for investment in regional research 
ultimately critical to understanding global drivers in the evolution of meteorological science. 
Third, to consider the history of making observations in the region. This history joined the 
maritime community with the terrestrial meteorological organisations. As scholars have noted, 
ships could rightly be considered ‘itinerant observatories’ generating knowledge through the 
intimate experience of ship’s officers with oceanic weather.14 Finally, it will explore something 
of the knowledge network that linked the colonial port city of Hong Kong, and to a lesser extent 
Singapore, via the shipping and communications channels that criss-crossed the China Seas 
and connected East and Southeast Asia.  

   

 

                                                        
12 J. Golinski, British Weather and the Climate of Enlightenment (London and Chicago: Chicago University Press, 
2007), p. 8.  
13 G. Kutzbach, The Thermal Theory of Cyclones: A History of Meteorological Thought in the Nineteenth Century 
(American Meteorological Society, 1979). 
14 John Herschel coined the phrase and it has been used to good affect by historians including: Simon Naylor, 
‘Weather Instruments all at Sea: Meteorology and the Royal Navy in the Nineteenth Century’ in F. MacDonald 
and C. W. J. Withers, eds, Geography, Technology and Instruments of Exploration (Oxon: Routledge, 2016), pp. 
77-96. 



History of Meteorology 8 (2017)

 
 

 

163 

Making observations: imperial storm science across land and sea 

 

In the sixteenth and seventeenth centuries, knowledge of winds, currents, and weather along 
the trade routes that crossed the China Seas stretching from Singapore north to Macau and 
Hong Kong and then east along the China coast to Taiwan, Korea, Japan, and south-east to 
Manila, was considered sensitive commercial intelligence to be kept secret, if possible, from 
potential rivals. That situation could not last. By the nineteenth century, it had become a matter 
of the greatest urgency to collaborate across borders – both geographic and political – to collect 
meteorological observations, especially to facilitate early storm warnings. As James Warren 
has shown for Manila, the monsoon and typhoon seasons had the ability to change the course 
of history for the regional shipping industry.15 With a typhoon season in the China Seas that 
affected mainland Hong Kong between May and November, and a north-east monsoon 
affecting Singapore from November to January, both port entrepôt were affected by extreme 
weather. The typhoons had the force to severely disrupt regional trade for both commercial 
ports, as well as wreaking havoc on Hong Kong through a combination of wind, tidal surge 
and flood. In Singapore, there were no typhoon but the monsoon brought heavy rains and 
storms which had the potential to inundate the low-lying marshy town. An improved 
understanding of the weather was thus critical to all aspects of life across land and sea in this 
region and there should be little distinction made between maritime and terrestrial scientific 
developments. 

At sea, the process of studying the weather began from time immemorial but 
commercially, written sailing instructions were some of the first accounts to detail marine 
weather that were made available publicly. In Britain, the seventeenth century records of the 
English East India Company (EEIC) were invaluable resources. The information in these 
sailing instructions came from accounts and experiences of sailing masters, often accompanied 
by a generalized explanation of the winds, currents and storms that might be encountered on a 
voyage.  Sailing directions were a practical tool to assist the mariner however, and seldom 
incorporated the latest scientific thinking. The reason for this is easily explained. The simple 
descriptive model of the wind system, and in particular the monsoon, as illustrated in the sailing 
directions, accurately reflected what the mariner actually experienced. Scientific refinement 
was thus of academic interest but deemed to have little practical application. Hurricanes, 
typhoons and cyclonic storms were however a different matter, and the study of countless 
observations and experiences of what came to be called revolving storms, promoted strategies 
of avoidance, mitigation and even exploitation.  

Negotiating monsoon winds in both the Indian Ocean and China seas was also a matter 
of the practical application of information disseminated through the handing on of personal 
experience, and more formally through printed sailing directions. Where a scientific 
explanation of the monsoon was attempted, it usually involved the accepted understanding of 
the time. This was the seasonal heating and cooling of the Asian continent that caused the 
monsoon winds to blow from one direction and then later in the year to reverse that direction.  It 

                                                        
15 Warren, ‘Weather, History and Empire’.  



History of Meteorology 8 (2017)

 
 

 

164 

was the same phenomenon as the local diurnal change from land breeze to sea breezes, familiar 
to all mariners, but on a continental scale.16 The modern explanation of the passage of the 
monsoon trough, or Inter-Tropical Convergence Zone (ITCZ) would have made little impact 
on the practical application of observed experience. 

This practical application is illustrated well by the advice given in the 1843 edition of 
Horsburgh’s India Directory.  It is worth quoting at length. 

 

Ships bound to China, which depart from Singapore or Banca Straits in 
February, March and part of April, may expect a tedious beating passage: in 
March, April or May, they may proceed by the Inner Passage, along the coast 
of Cochin-China, which is generally the most expeditious route in these months; 
but when June approaches the South-west monsoon is set regularly in, the track 
by the Macclesfield Bank seems preferable, the winds being more steady in the 
open sea than near the coast.  Even so early as April …. A ship may sometimes 
get a westerly breeze blowing out of the Gulf of Siam, to carry her to the 
Macclesfield Bank, and afterwards easterly winds, to run to the Grand Ladrone; 
but if she proceed by the Inner Passage, easterly winds may retard her progress 
around the south-east coast of Hainan, and thence to the entrance of the Canton 
River.17 

 

This is the closest thing to a weather report, or rather weather advice that a ship’s master 
could expect at that time. As such it was based on (as the title page from Horsburgh clearly 
states) ‘original journals of the Honourable Company’s ships, and from the observations and 
remarks resulting from the experience of twenty-one years in the navigation of those 
seas.’18 Sailing directions such as Horsburgh and those that both preceded and came after, were 
some of the earliest (Western) disseminations of weather and climatic knowledge in the 
monsoon seas, based entirely on local observation and the careful accumulation and analysis 
of data collected over many years. It was an extraordinary achievement made purely for 
commercial, rather than scientific, purposes. 

Typhoons were the other major weather phenomena encountered by ships on the 
passage across the China Seas. Typhoons were a danger that could result in the total loss of 
ship, crew and cargo. As with the regime of monsoon winds, sailing directions carefully 
documented the most dangerous season and the areas most prone to these devastating events. 
However, this was still not sufficient, as Horsburgh lamented: 

 

                                                        
16 James Horsburgh, The India Directory or Directions for Sailing to and from the East Indies, China, Australia 
and the Interjacent Ports of Africa and South America. (1852), vol. 1, iii-iv. 
17 James Horsburgh, The India Directory or Directions for Sailing to and from the East Indies, China, Australia 
and the Interjacent Ports of Africa and South America. (London, 1843), vol. 2, p. 292. 
18 Ibid. 



History of Meteorology 8 (2017)

 
 

 

165 

To be able to prognosticate the coming of these tempests would be very useful 
to navigators, but this cannot be done with certainty, for they frequently 
commence without giving much indication of their approach.  ….  Marine 
barometers, if well constructed seem to afford the best means of anticipating 
these tempests; for on the south coast of China, there is a greater fall of the 
mercury than might be expected within the tropics.19 

 

It was in this respect that science needed to be applied to the vast amounts of locally 
gathered observations to be found in ships’ logs and journals. This was not easily achieved 
despite the obvious need and clear benefits. As late as 1875, Labrosse’s sailing directions for 
the China Seas cast doubt on some of the earlier work that had been done in trying to understand 
the behaviour of typhoons. 

 

…[The] track of typhoons …. obeys neither the general law governing cyclones 
of the northern hemisphere, nor any fixed rule ….  After a perusal of 
Piddington’s researches, it must be admitted that the track of typhoons lies 
between NNW and SSW; but it is best not to rely too confidently on this … 

 

Labrosse continues, stating how some typhoons have behaved differently, then; ‘…. It 
is clear that no fixed rule can be given to avoid typhoons; … we shall confine ourselves to 
giving general advice.’ After giving this advice, largely drawn from Horsburgh, he concludes 
that typhoons are to be feared and that almost all measures taken to avoid them are of little 
use.20 Labrosse was clearly aware of Piddington’s work but maybe not the work of others such 
as William Birt.  

Previous decades had produced volumes of work describing hurricanes, cyclones and 
typhoons. There were many scientists and others interested in researching this topic, chief 
among them William Redfield, William Reid, Heinrich William Dove, Charles Meldrum, 
Bernhardus Varenius, Henry Piddington, Benito Viñes and later in the nineteenth century John 
Eliot, to name just a few. Piddington, (referred to by Labosse above) a former mariner turned 
scientist and based in Calcutta, having been inspired by Reid’s Law of Storms, wrote twenty-
three storm memoirs, as well as many other articles on storms, most of them published in the 
Journal of the Asiatic Society of Bengal. In the Fourth Memoir on the Law of Storms, 
Piddington was requested by the secretary of the Marine Board to investigate the loss of the 
Golconda, bound from Singapore to Macao in 1840, and caught by a typhoon in the China Sea. 
Piddington drew upon the logs and journals of other vessels traversing the China Sea at that 
time and his analysis concluded that there were in fact two storms and that those vessels that 

                                                        
19 Ibid., p. 289. This and other sets of sailing directions by others such as Findlay and Rosser contain a profusion 
of accounts of typhoons and hurricanes, littered with barometric pressure observations, gleaned from vessels 
whose logbooks are likely no longer extant.  
20 F. Labrosse (trans J. W. Miller), The Navigation of the Pacific Ocean and China Seas (1875), pp. 27, 29. 



History of Meteorology 8 (2017)

 
 

 

166 

survived had heeded the advice given in the Law of Storms. His analysis also confirmed that 
the storms behaved in accordance with current theory and that the typhoon was a rotary storm.21 

The type of research performed by Piddington was essential, but difficult to carry out 
and to communicate to those most in need of the results.  At a meeting of the Asiatic Society 
in 1840, Piddington submitted a memorandum outlining the difficulties he encountered in 
obtaining extracts from logbooks to carry out his researches. Commanders were unwilling to 
hand over extracts, and Piddington asked the Society to apply to the Government to issue an 
order making it compulsory to hand over information on storms. As commanders were 
answerable only to the ship owners for their logs, it was considered unlikely that the 
Government would consider such a measure, but the Society’s secretary stated that he would 
at least write to the Government asking them to make an appeal to ships’ officers.22 

Piddington’s contemporary in Mauritius, Charles Meldrum, had better fortune in 
gathering data, as he did not rely on goodwill and the submission of logbooks, but sent his 
assistants to transcribe parts of the logbooks of ships in port.  Meldrum was working to a plan, 
outlined and partly implemented by Alexander Thom, which would have ultimately resulted in 
a network of meteorological observations from stations collecting data from shipping. The 
collection points included Mauritius, Bombay, Calcutta, Ceylon, Swan River and Melbourne, 
the intention being to produce published data and results across the Indian Ocean. The 
extension of this plan to the China Sea would have been a next logical step, but the overall plan 
lacked both support, co-operation and funding to publish the results, apart from an initial 
volume covering a single month in 1853.23 The precious and irreplaceable raw data gathered 
by Thom and Meldrum, along with its results still sits (unused) in an archive in Mauritius in 
need of urgent conservation. 

From this it can be seen that early work on storms and typhoons and the benefit of the 
resulting analysis was difficult both to gather and, more critically, difficult to communicate. In 
1848, after reconstructing yet another severe storm, Piddington issued a strong statement in 
conclusion.   

 

If warnings like these are not listened to, it is difficult to say what will be 
required.  Nothing short of a whole fleet would seem sufficient to rouse the 
attention of those who it behoves to insist upon the laws of our science being as 
duly attended to as the lead and the chart, and upon every Commander intrusted 
(sic) with public property noting in his log his reasons for standing on or heaving 
to on the approach of bad weather; and this will, in case of his return to port in 

                                                        
21 Henry Piddington, ‘A Fourth Memoir of the Law of Storms in India, being remarks and documents relative to 
the loss of the Ship Golconda, in the Tyfoons of 22nd to 24th September 1840, in the China Sea’, Journal of the 
Asiatic Society of Bengal, vol. 10 (1841): 895-906. 
22 Journal of the Asiatic Society of Bengal, vol. IX (1840), ‘Proceedings of the Society’, 552-4. 
23 Charles Meldrum, Contributions to the Meteorology and Hydrography of the Indian Ocean (part 1) a 
Meteorological Journal of the Indian Ocean for the month of March 1853, (1854), i-v. 



History of Meteorology 8 (2017)

 
 

 

167 

a disabled state, at once show if he understood his position or not.  If he did not, 
he is unfit for the command of a vessel till he does.24 

 

The slow and piecemeal accretion of observational data, with little in the way of a 
formal system of analysis or communication, except by individual publication, did enable some 
progress in understanding typhoons and cyclones. In 1858, M. F. Maury remarked on the 
recurvature of typhoons and their unusual tracks, but admitted that the reason for this behaviour 
was a puzzle to both meteorologists and navigators.25 By the 1860s W. H Rosser, in his sailing 
directions included very detailed ‘meteorological signs preceding a hurricane, which included 
portents observed by sight, sound and feeling. This was drawn largely from Birt’s Handbook 
of the Law of Storms, published in 1853 and re-issued in 1879.26 Rosser himself published his 
own law of storms, joining a pantheon of similarly titled volumes. Most of these volumes were 
derivative, borrowing heavily from previous publications, yet slowly producing a 
comprehensive list of advice for mariners to observe, or ignore if they chose. Rosser’s 
publication was unique however, and its chief interest, for the historian at least, is not the advice 
given to mariners but the extraordinarily detailed historiography of the development of the 
science surrounding cyclones and cyclone theory from the time of Columbus up until the time 
Rosser was writing.27 

At this point the somewhat troubled progress of the early communication of knowledge 
and understanding of hurricanes and typhoons had reached a logical conclusion in the advice 
issued through sailing directions and the innumerable variations of the Law of Storms. Effective 
typhoon forecasting, as opposed to simply recognizing the portents, required a network of 
observers and a means (such as telegraph) to issue warnings. Nevertheless, the behaviour of 
typhoons was now well documented if imperfectly understood. This was in part due to 
governmental intervention in requesting that information contained in logs be handed over to 
port or meteorological authorities on land, as had been hoped back in the 1840s. In Hong Kong 
for instance, Colonial Secretary Frederick Stewart intervened to make sure that the observatory 
received observations taken on board ships that docked at port from the 1880s.28 These 
observations were collated by the observatory’s first female member of staff Anna Doberck 
(sister to director William) from 1892; from which she personally constructed pilot charts. 

Thus in Hong Kong, as in most parts of the world, the story of meteorological science 
began in earnest with the making and collation of instrumental weather observations, a 
trajectory that closely mirrored and connected with maritime developments. In line with early 

                                                        
24 Henry Piddington, ‘Fifteenth Memoir of the Law of Storms’, Journal of the Asiatic Society of Bengal, vol XVII 
(1848): 27-56. 
25 M. F. Maury, Explanations and Sailing Directions to Accompany the Wind and Current Charts, (1858), vol. 1, 
p. 262. 
26 W. H. Rosser & J. F. Imray, The Seaman’s Guide to the Navigation of the Indian Ocean and China Sea (1867), 
pp. 148-9. 
27 W. H. Rosser, The Law of Storms Considered Practically (1876). 
28 Public Records Office Hong Kong (hereafter PRO), Hong Kong Record Series (hereafter HKRS) 356 1-1-2, f. 
94. Frederick Stewart to William Doberck, 5 November 1888.   



History of Meteorology 8 (2017)

 
 

 

168 

nineteenth century thinking to advance understanding of the weather, the recording of 
observations was considered the essential step in building a body of data from which to 
extrapolate patterns and to understand the working of storms.29 From the 1840s the British 
Association for the Advancement of Science (BAAS) and the Royal Society had worked 
toward a grand plan of establishing a global network of imperial observatories from which 
could be gathered systematic, daily terrestrial observations, to augment information from the 
naval and East India Company ships log books. In line with this, the number of registering 
stations was gradually expected to increase. 

 

Ideal versus reality? The situation in Hong Kong and Singapore 

 

The reality was a far less structured and less co-ordinated affair. The British colonies at 
Singapore and Hong Kong are excellent examples of the degrees of success and variances from 
country-to-country. Between the 1820s and 1860s registering stations were limited and 
experiments in producing consistent meteorological series ad hoc. A temperature and rainfall 
series survives for Singapore for 1820-23 made by then Governor William Farquhar at Fort 
Canning, who also made observations in Malacca (1809) and in Georgetown, Penang (1815-
16, 1820-21, 1823).30 Between 1839 and 1841 Joseph S. Travelli – a missionary – made 
observations at Ryan's Hill at the American Board of Foreign Missions where Travelli was 
then based.31 Then from 1835 to 1836 rainfall and temperature observations were published in 
The Singapore Free Press and Mercantile Advertiser but thereafter only sporadic records 
survive until the late 1860s. There was, for example, an observatory operating in Singapore 
between 1841-5. Established as part of Edward Sabine and the BAAS ‘Magnetic Crusade’,32 
the Singapore Observatory was closed due to a lack of funds and political will.33 Although 
meteorological observations were only made as a by-product of the magnetic investigation, the 
observatory did produce a consistent four year temperature series for Singapore and, for a few 
months in Sarawak, Borneo during 1842.34 Likewise, a series of meteorological and tidal 
observations were started in 1844 at Penang Island but ceased in 1846 by order of the 

                                                        
29 Anderson, Predicting the Weather, pp. 6, 132-140; Mahony, ‘For an Empire’: 30-1; Golinski, British Weather, 
p. 210; F. Williamson, ‘Weathering the British Empire: meteorological research in the early nineteenth-century 
Straits Settlements’, British Journal for the History of Science 48:3 (2015): 475-492, 6-7; Good, ‘A Shift of View’, 
p. 35. 
30 Thermometrical and Barometrical Tables communicated by Lieut. Col. William Farquhar, Transactions of the 
Royal Asiatic Society of Great Britain and Ireland, Vol. 1, Appendix 1: 585-599. 
31 Joseph S. Travelli, “Abstract of a Meteorological Register kept at Singapore”, The American Journal of Science 
and Arts XLIV (1843): 150-8.  
32 J. Cawood, ‘The Magnetic Crusade: Science and Politics in Early Victorian Britain’, ISIS, 70:4 (1979): 493-
518.  
33 Williamson, ‘Weathering the British Empire’: 486, 488.  
34 The observations made by Lieutenant Charles Elliot are available online at the British Geological Survey: 
http://www.geomag.bgs.ac.uk/data_service/data/yearbooks/sin.html 



History of Meteorology 8 (2017)

 
 

 

169 

government.35 As the century progressed and the call for extending observations became more 
mainstream, the Straits Settlements rather fell by the wayside. A small observing facility was 
located at Mount Faber from the 1880s36 until 1934 when it moved to the new Kallang airfield, 
but it was 1928 before it was equipped with the latest wireless receiving equipment and a 
traveling micrometer for time determination.37  

Weather observation throughout most of the nineteenth century in Singapore was not 
the job of a formal meteorological service but a combination of the colonial medical branch, 
public works departments, or private plantation or estate owners.38 In the 1860s for example, 
observations were made at Killeney Estate on River Valley Road, by Jonas Vaughan, ex-
seaman, who published the rainfall, temperature and wind readings in the Government 
Gazette.39 Other stations were gradually added over the 1870s and 1880s, including 
Perseverance Estate at Geylang owned by the Alsagoff family.40 Registering stations were 
often located at hospitals too in order to study the relationship of rain and temperature with 
health. In Singapore, this included the Convict Hospital at Bras Basah and the Pauper Hospital, 
likely part of the large Tan Tock Seng hospital facility. There were also stations at the P&O 
Depot at New Harbour; Water-works Reservoir, Thompson Rd (made by James McRitchie 
Municipal Engineer); and of course the Botanic Gardens (made by Superintendent Nathanial 
Cantley). 

In the early twentieth century, formal responsibility for meteorology was passed from 
the medical department to the Museums Department in Singapore under the direction of Mr 
Herbert Robinson. Robinson, Director of Museums between 1908 and 1926, had an active 
interest in the science. He appointed a Mr Kelliher as a Meteorological Officer in the direct 
pay of the Museums Department and it was under the ward-ship of these two men that 
meteorological study experienced something of a revival in Malaya. Robinson was critical of 
the quality of observations taken by the medical branch previously, suggesting that staff were 
ill-trained, instruments were not standardised, and readings were temporally erratic.41 Whether 
or not his criticisms can be credited with any veracity, it was he, along with George Maxwell 
– then Chief Secretary for the Federated Malay States (FMS) – who argued the urgent need for 

                                                        
35 British Library (hereafter BL), India Office Records (hereafter IOR), F/4/2170/104987: Extract of a dispatch 
from the Government of India to the Honourable Court of Directors, 23 May 1846. 
36 The observatory is marked on a map of 1882: NAS, C.O. 129/205. Media - Image Number: D2014040098, 
Survey of Mount Faber Range and country surrounding Singapore, 1882.  
37 The National Archives UK (hereafter TNA), CO273/552/12: Apparatus for installation at Mount Faber 
Observatory, 1928. 
38 TNA: CO273/541, G. Maxwell and H. C. Robinson, ‘A Meteorological Service for Malaya’, p. 1 and The 
Singapore Free Press and Mercantile Advertiser, 24 January 1927, p. 7. 
39 Observations available in the Straits Settlements Government Gazette for 1864-6, available at National 
University of Singapore Law Library (Singapore-Malaysia Collection). 
40 Observations available in the Straits Settlements Government Gazette TNA, CO273 (5), 1875.  
41 TNA: CO273/541/4: G. Maxwell and H. C. Robinson, ‘A Meteorological Service for Malaya’, p. 1. 



History of Meteorology 8 (2017)

 
 

 

170 

consistency and improvement in the meteorological services in the 1920s. This was made all-
the-more pressing by the expansion of aviation services across Singapore and the peninsula.42  

Responsibility for meteorology was passed over again on 1 January 1927, this time to 
the Survey Department (part of the Public Works Department).43 This was an interim measure, 
as plans were afoot to create a dedicated Meteorological Branch. In 1927 it was vaunted to 
establish 

 

groups of the second order stations down each coast and along the centre of the 
Peninsula … In addition it is hoped to establish a large number of stations at 
which temperature and rainfall data will be collected with the co-operation of 
the planting community.44 

 

The service began formal operations in Kuala Lumpur in 1929. Under the new branch, 
seventeen new fully-equipped stations were built across the peninsula, in addition to those 
already established.45 The headquarters moved to Singapore in 1931. 

In Hong Kong, the situation was a little different. Before 1860, meteorological research 
was – like Malaya – small-scale. Observations were made at government departments, 
hospitals, police stations and at the Harbour Master’s Office.46 There were also sporadic 
weather records made by visiting scientists including Frank Julius Ferdinand Meyer in the early 
1830s and – like Singapore – some interested individuals made meteorological diaries. Diaries 
were kept by meteorology enthusiast and Dutch consul Mr Blettermann and an English 
merchant Thomas Beale, for instance.47 The early collection of observations along the China 
coast by the Chinese Maritime Customs Service (CMCS) were far more significant. These were 
published in the Canton Register, an English-language newspaper founded by two Scotsmen 
in 1827.48 Robert Hart’s revival of the CMCS in the 1860s was a significant staging point in 
the modernisation of meteorology and the establishment of a regional observational network. 
Hart was the mainland service’s Inspector-General from 1863 and it was he who requested that 
each customs station in China and Hong Kong be outfitted with meteorological instruments; 
                                                        
42 NAS, M712/14-15, f. 79. Extract from the Malaya Tribune, 14 June 1926: Air Liners: Big Scheme for Malaya; 
M712/14-15 ff.84-5. Report submitted by Samuel G. Ebling, Vice Consul in charge 14 Sept 1927. Proposed 
service by Air Survey Company Ltd. 
43 TNA, CO273/541/4 Proposal to establish a Meteorological Department for Malaya: Memorandum on a 
Pamphlet entitled ‘A Meteorological Department for Malaya’ by Sir George Maxwell and Herbert C. Robinson, 
written by Victor A. Lowinger, Surveyor General (Federated Malay States and Straits Settlements), 24 October 
1927, pp.1, 3.  
44 R. L. Jarman, Annual Reports of the Straits Settlements, 1855-1941, Vol. 9., 1927-31 (Slough: Archive Editions: 
1998) Report for 1927, p. 304. 
45 TNA, CO273/541/4 Memorandum on the general scheme for the systematic study of the climate of Malaya, 14 
October 1927, pp. 1-5.  
46 Ho, Weathering the Storm, p. xii.  
47 Ibid. 
48 A. Le Pichon, ed., China Trade and Empire: Jardine, Matheson & Co. and the Origins of British Rule in Hong 
Kong 1827-1843 (Oxford: Oxford University Press, 2006), p. 67. 



History of Meteorology 8 (2017)

 
 

 

171 

that observations made at lighthouses be made use of, and that meteorological registering 
stations be established along the coast and linked by telegraph.49 Although his scheme was met 
with a certain amount of ambivalence, especially when it came to Hong Kong soil, Hart did a 
great deal to standardise and reform meteorological services along the China coastline. 
Observations of temperature, rainfall, wind, cloud, ozone, and pressure made at Hong Kong 
were published in the China Mail from at least the 1860s.50  

In the 1880s the trajectories of meteorology in Hong Kong and Singapore began to 
diverge. In 1883, the British government paid for an observatory to be opened in Hong Kong. 
Its staff were expected to take charge of meteorological research and, from 1884, the institution 
provided the main source of observational data for terrestrial Hong Kong. It was envisaged that 
the observatory would provide an important link in a regional observational network, filling a 
gap essential for improving early storm warnings for Hong Kong, Macau and Manila.51 The 
early emphasis of the Hong Kong Observatory was on protecting the interests of the shipping 
industry. The provision of a time service for ships, the making of weather observations and 
issuance of typhoon warnings were considered vital. Despite some suggestion that wider 
meteorological research lagged initially,52 the letters, reports and correspondence of the 
successive directors – including the first director William Doberck – show that impetus to catch 
up came from Colonial Secretary Frederick Stewart who continually pushed Doberck to spend 
more effort on making observations and issuing daily reports and storm warnings for the 
public.53 Doberck – trained as an astronomer in Europe – may arguably have lacked experience 
in the field.54 After arriving in the Far East from his previous position in Ireland, Doberck was 
thus asked to spend two months visiting Swatow (Shantou), Amoy (Xiamen), Shanghai and 
other cities on the Chinese coast, as well as travelling on the Imperial Customs Revenue Cruiser 
to lighthouses including those on Macau and Penghu Island in the Taiwan Strait.55 Whilst in 
Shanghai and Amoy he inspected, adjusted and verified extant meteorological equipment, and 
took his own observations during the expedition.56  

                                                        
49 Fir more on Hart and the Service see: Bickers, “‘Throwing Light on Natural Laws”. 
50 See for example, The China Mail, 11 January 1868, available online from Hong Kong Public Libraries: 
https://mmis.hkpl.gov.hk/ 
51 Anthony Dyson, From Time Ball to Atomic Clock (Hong Kong: Hong Kong Government Printer, 1983), p. 19.  
52 Correspondence between the Acting Colonial Secretary Frederick Stewart and Doberck in 1886 critiqueS the 
latter’s 1885 report for its lack of effort in producing daily weather reports or storm warnings for the public, and 
neglect of any analysis of the China Coast observations: HKRS 356 1-1-2, ff. 4r-5v Colonial Secretary’s Office 
to the Government Astronomer, 2-4 March 1886. 
53 HKRS356 1-1-2, ff. 4r-5v Colonial Secretary’s Office to the Government Astronomer, 2-4 March 1886; f. 60 
Invitation from Frederick Stewart to William Doberck to sit on a specially convening storm signalling board, 13 
March 1886.  
54 MacKeown, China Coast Meteorology, p. 62-3. 
55 HKRS356 1-1-1, 79. Despatch 1351 from the Colonial Secretary’s Office, Hong Kong, 23 August, 1883; 
HKRS356 1-1-1, Inspectorate General of Customs Circular No. 235, 7 September 1883. 
56 Hong Kong Government Reports Online (hereafter HKGRO); Administrative Reports (hereafter AR), 1883: 
Observatory Reports. Report from the Government Astronomer, together with Instructions for making 
Meteorological Observations, 17 November 1883, p. 1. 



History of Meteorology 8 (2017)

 
 

 

172 

His trip was vital in establishing links with the Imperial Maritime Customs Service – 
the CMCS – which was to prove the richest and most important source of regional observations, 
and the main network for communicating them, for many years to come. During his first few 
months in Hong Kong, Doberck (usually at the insistence of the Colonial Secretary) was to 
highlight the absolute necessity of expanding the range and diversity of meteorological 
observations and registering stations, a mission that was to consume a large part of his time at 
Hong Kong Observatory. The extension, he argued, could be done only with the cooperation 
of Hart and the CMCS and it was largely a result of the Hart’s efforts along with Father Marc 
Dechevrens, Director of the Zikawei Observatory at Shanghai, that the wide-ranging network 
of coastal and riverine registering stations were established and their observations standardised. 
Doberck requested that registers from the old and new stations mainland China and Macau 
(including Kiungchow [Hoihow], Pakhoi [Beihei], Canton [Guangzhou] Swatow [Shantou], 
and Ningpo [Ningbo] amongst others), be sent to the Observatory monthly and that the network 
of Chinese stations making observations be extended to thirty-five (the latter request was 
ignored).57 After receiving the observations – via the customs services at Amoy and Shanghai 
– Doberck intended that his observatory staff would make corrections and revisions as 
necessary and construct daily weather maps from the data contained therein.58 Although 
Doberck’s prime function was that of astronomer, his early work on the law of storms in the 
China Seas published in 1886 – without governmental financial support or impetus – shows a 
commitment to meteorological research, especially that which directly affected Hong Kong.59 
It is also likely that he was trying to make a name for himself as his pamphlet drew directly 
from the earlier studies and styles of Reid, Piddington, Rosser and Birt. It did his reputation 
little good however, especially as the maritime community threw doubt on Doberck’s findings, 
suggesting the text had little applicability in the real experiential world of seafaring. 

On joining Hong Kong Observatory Doberck had become part of a major 
communications and knowledge network that incorporated some of the most significant sites 
of learning and important scholars of meteorology in Asia. Because of the regular progression 
of typhoon tracks across the China Seas, excellent communications between China and Hong 
Kong were essential but so too was a good relationship and rigorous communication with 
Manila, Taiwan, Korea and Japan. In 1883, an arrangement was made with the Great Northern 
and the Eastern Extension Telegraph Companies to send – at no cost – observations between 
Hong Kong, Manila, Xiamen, Fuzhou, Shanghai, Vladivostock and Nagasaki.60 By 1895, the 
Eastern Extension Telegraphic Company (EETC) exchanged daily observations with 

                                                        
57 MacKeown notes that Doberck’s proposal for 35 stations was not taken up, possibly because of a lack of 
enthusiasm on the part of Hart and Dechevrens to take direction from the Colonial Office: MacKeown, China 
Coast Meteorology, pp. 77-8. 
58 HKGRO, AR, 1883: Observatory Reports. Report from the Government Astronomer, together with Instructions 
for making Meteorological Observations, 17 November 1883, p. 2. 
59 HKRS 356 1-1-2, f. 18. Government Order from the Colonial Secretary to the Director of the Observatory in 
regard to the publishing of a pamphlet, 11 September 1886. Here Stewart tells Doberck that there will be no 
financial or resource support for the publishing of his book though Doberck can publish by ‘any private 
arrangement he likes’. 
60 HKGRO, AR, 1883: Observatory Reports. Report from the Government Astronomer, together with Instructions 
for making Meteorological Observations, 17 November 1883, p. 3. 



History of Meteorology 8 (2017)

 
 

 

173 

Singapore, Haiphong, Xiamen, Shanghai, Bolinas, Nagasaki, Vladivostock, and Macau.61 
Observations were also received at Hong Kong and Manila from stations in Taiwan via 
Taihoku (which included readings made at stations at Taihoku, Taichu, Tainan, Koshun 
[Hengchun] and the Penghu Islands).62 In 1898 the board of Directors of the EETC had 
authorised observations to be taken by their own officers stationed at Capiz and Tuburan in the 
Visayas (Philippines) and promised that these would be communicated to Hong Kong.63 By 
the early 1900s thirty-one stations were included in the China Coast Meteorological Register 
(CCMR) observational network.64 By 1904, telegrams were received at Hong Kong twice daily 
from Cap St Jacques (Vũng Tàu, Vietnam), Cap Padaran (Mui Dinh, Vietnam), and Labuan in 
the Straits Settlements (Sabah, Malaysia).65   

The extension of the observational network directly resulted from the need to address 
the extremes of weather by improving storm forecasting. The weather was the catalyst that led 
to many investments in the science. Certainly, the establishment of the Hong Kong Observatory 
and the Malayan Meteorological Department can be cited as good examples of this. In Hong 
Kong, there had been calls from at least 1861 for an institution to provide a time service and 
storm warnings for shipping without success. The major typhoon that devastated Hong Kong 
in 1874 alerted the public to the benefits of a local observatory and heightened pressure on the 
government to enact their earlier plans. Drawn-out discussions were moved forward and the 
Observatory opened in 1883.66 In similar vein, it could be argued that the restructuring of the 
Malayan meteorological service from 1927-1929 could have been inspired by some of the most 
serious floods to have been known in that region. Although discussions had been underway 
from as early as 1919 that the colony needed an improved service, 1925 and 1926 saw major 
flooding in Singapore and in rural Malaya respectively. The floods crippled large swathes of 
agricultural land and caused massive destruction of infrastructure.67 A significant component 
of the planned new service was to increase the number of manned meteorological registering 
stations across the peninsula to improve the collection of rainfall data, especially in rural areas. 
This was intended as ‘a preliminary to scientific investigation and a systematic study of the 
climate ... [and a] … foundation on which can be built a Meteorological Department capable 
of dealing with the needs of aviation, upper air research work, the distribution and exchange of 
information and weather predictions’.68 

                                                        
61 HKRS8421/2 General Correspondence 1895-7, 28/4-5, Communication from L. Webster RE Meteorological 
Telegrams 17 October 1895. 
62 HKRS8421/2, 7, Letter to Doberck from H. Kondo, Taihoku Meteorological Station, 6 November 1896.    
63 HKRS842/3 1898-1903. Letter to the Colonial Secretary at Hong Kong from Manager-in-China, EETC, 17 
March 1898.  
64 An example of Acting Observatory Director Frederick Figg’s copy of the CCMR can be found at: HKRS842/4 
General Correspondence 1903-1913, CCMR, 10 September 1903. 
65 HKRS842/4 General Correspondence 1903-13, Letter from Director of Meteorological Services Indo-China to 
HKO, 2 May 1902 & Letter from Eastern Extension Telegraph Company 20 August 1904.  
66MacKeown, China Coast Meteorology, pp. 27-8. 
67 For more on the 1926 floods, see F. Williamson, ‘The Great Flood of 1926: Environmental Change and Disaster 
Governance in British Malaya’, Journal Ecosystem Health and Sustainability, Environmental Impact of Disasters 
- special issue, 2:11 (2016).  
68 Jarman, Annual Reports: Report for 1927, p. 304. 



History of Meteorology 8 (2017)

 
 

 

174 

 

Controversy and confusion: the 1906 typhoon and its consequences 

 

All such systems however – whether national or international – relied heavily on the 
effectiveness of the communications network. This was far from perfect. As the example of the 
HMS Eagle in 1939 showed, observers could not always agree on the trajectory or even the 
existence of a typhoon. This could have disastrous consequences. Waking up on 18 September 
1906, Hong Kong inhabitants were entirely unprepared for what that day would bring. A 
typhoon, short but almost unprecedented in strength, was to wreak utter havoc in the colony. 
At 8am the first warning was given out in the form of the hoisting of a black signal drum (the 
final in a stage of red to black storm warnings), and the typhoon gun was fired at 8.40 signalling 
the immediate arrival of the storm. It was too late for the ships, fishing boats and sampans 
which were already out at sea and could not return to shore in time. It has been suggested that 
fifty per cent of Hong Kong’s Chinese junks and sampans were destroyed that day and a 
number of larger merchant and naval vessels were grounded or sunk.69 These included the 
German steamer Petrarch, the British sloop-of-war Phoenix, the French destroyers Fronde and 
Francisque, and the Canadian Monteagle.70 On land, serious flooding, wind-driven high waves, 
and a massive storm surge of 6.10m at Tai Po – one of the highest ever recorded for that area 
– caused massive destruction.71 Contemporary estimates suggested that between 4000 to 
10,000 people died that day.72 

The Observatory received a large apportionment of blame for the tremendous loss of 
life and damage to property during this typhoon and a public inquiry was initiated. The 
investigation, led by a Royal Navy Officer, a master mariner and the manager of the Eastern 
Extension Telegraph Company, was confidently expected to ‘exonerate the Director and Staff 
of the Observatory’.73 Certainly, Governor Matthew Nathan argued that he saw no possible 
way the organisation could have released an earlier warning. An inquiry was necessary to 
assuage the public outcry – taken up by the contemporary press – that staff should have and 
significantly, could have, given a much earlier warning of the impending storm. This view was 
publicised by French consul to Hong Kong Monsieur Liébert, who contended that French ships 
in the area had been warned by Zikawei Observatory staff long before Hong Kong Observatory 
had raised the Black Drum.74 Hong Kong Observatory contested this, arguing that the last news 
they had received from Zikawei had been the previous evening and that ‘this typhoon gave no 

                                                        
69 Li Woon Yee, The Typhoon of 18 September 1906, Royal Observatory, Hong Kong, Occasional Paper no. 36 
(1976), p. 3. 
70 The Calamitous Typhoon at Hong Kong, 18 September, 1906 (Hong Kong: Hong Kong Daily Press, 1906) p. 
5. http://ebook.lib.hku.hk/HKG/B36228084.pdf. Accessed 05.01.2017.    
71 Li, Typhoon of 18 September 1906, p. 3. 
72 Calamitous Typhoon, p. 1.  
73 HKGRO, Report of a Committee, p. 42 and HKGRO (1842-1941), Hong Kong Hansard, Reports of the 
Meetings of the Legislative Council, Session 1906: Report of the Meeting 20 September 1906, pp. 51-4.  
74 HKGRO, Report of Committee appointed to enquire whether earlier warning of the typhoon of September 
18th, 1906, could have been given to shipping; Supplement to the Hong Kong Gazette, 22 March 1907. 



History of Meteorology 8 (2017)

 
 

 

175 

indication of its existence until close to the Colony’.75 The incident is also revealing of the 
strained relationship between the French Jesuits at Zikawei and staff – Doberck especially – at 
Hong Kong.76  

After the 1906 typhoon, Doberck made it a personal mission to contact comparable 
observatories around the world to step up the collection of meteorological observations to 
better understand the weather patterns of the China Seas. Much of this work was actually 
undertaken by Doberck’s capable meteorological assistants, Anna Doberck and Frederick Figg, 
the latter succeeding Doberck as Director in 1907. The efforts taken to improve the quality, 
quantity and, critically, the frequency of observations can be seen to have paid dividends by 
1912. A letter to then Hong Kong director Thomas Claxton from Frere Louis Froc, director of 
Zikawei Observatory, in 1912 noted how ‘owing to increased certainty in the telegraphic 
transmissions of meteorological data, it is now possible to ascertain the direction in which the 
typhoon is travelling or from which a gale may be expected to a closer degree than hitherto’.77 
Claxton continued the work, attempting to persuade all those observatories in his immediate 
network – namely Tokyo, Zikawei, Manila and Indo-China – of the importance of simultaneous 
observations. Claxton argued that this was the only way to guarantee effective daily weather 
maps for the region.78  

To facilitate research, successive directors of the Far East stations had also sought to 
improve the historical record of observations for the region for some time. In a government 
notice issued from the Hong Kong Observatory on 16 February 1893 for instance, Doberck 
invited ‘Owners, Agents and Masters of Vessels … to forward old logbooks on loan to this 
Observatory’ for the purpose of tabulating ‘all observations made on board ships between 
Singapore and 180º E, Gr. Longtitude and between 0º and 45º latitude…with the view of the 
issue of pilot charts for the China Seas and the North Pacific’.79 Writing to various 
meteorological organisations globally, including those in Washington, Paris, and St 
Petersburg,80 as well as shipping companies like the P&O, he asked if they could furnish him 
with observations made during voyages on the China Seas.81 Doberck also communicated with 
Gilbert T. Walker, Director-General of Indian Observatories who, in 1905, had sent copies of 
‘all charts issued by the various meteorological offices of India’ to Hong Kong along with 

                                                        
75 Calamitous Typhoon, p. 1. 
76 For more on this controversy see, F. Williamson, Uncertain Skies: ‘Forecasting’ Typhoons in Hong Kong c. 
1874-1906, Quaderni Storici (Forthcoming, 2018). 
77 HKRS356 1-1-3, Zikawei to HKO concerning addition to the typhoon warning signal, 1912.  
78 HKRS356 1-1-3, especially 86, 92, 97 HKO 41/12, Claxton to Harbour Master Macau, 12 March 1912; HKO 
19/12 Claxton to Director Central Observatory Indo-China, 17 February 1912; HO 18/12 Claxton to 
Superintendent Eastern Extension Telegraph Company, 21 February 1912. 
79 HKRS8421/1 General Correspondence 1891-4, 107a, Government Notification No. 63, 16 February 1893. 
80 This communication worked two-ways: in 1906 the US Department of Agriculture Weather Bureau also 
requested daily weather information from HKO: HKRS842/4 General Correspondence 1903-1913 f. 20. Letter 
from USDOA Weather Bureau to HKO 8 August 1906.  
81 HKRS8421/1 General Correspondence 1891-4, 107a/c, Letter from Washington to Doberck 17 March 1893; 
Letter from Shipmaster’s Society Fenchurch Street London Letter from P&O Managing Directors to Doberck, 9 
June 1893;  



History of Meteorology 8 (2017)

 
 

 

176 

signal codes and storm telegraphic code for reference.82 An important avenue for furthering 
collaboration on such matters was attendance at the meteorological conferences for the Far 
East, the first of which was held in 1913 in Tokyo. Over succeeding years, a variety of 
conferences brought the various directors – including those of Malaya and Hong Kong – into 
direct contact, including during the Conference of Empire Meteorologists held in 1919, 1929, 
and 1935;83 the meetings of the International Meteorological Committee and the Conferences 
of the Regional Meteorological Commission for the Far East (first held in Hong Kong in 
1937).84 

It should not be assumed however that these networks were effective or, that 
communications were trouble-free. The outcome of the meteorological Conference held in 
Tokyo in 1913 for instance was a lack of agreement and discord on the main item under the 
discussion: the standardisation of storm signals, especially the types of code used and the 
manner of signalling them, a problem that was to dominate correspondence between Hong 
Kong, Manila, Tokyo and Shanghai for many years to come.85 The breakdown of telegraphic 
communication was also to dog the meteorological services throughout the nineteenth and early 
twentieth centuries. Cable lines may have existed but they broke down with regularity during 
bad weather. In 1887 Doberck had noted the problem – even within Hong Kong – of telegraphic 
communication. During typhoons, when swift communication was utterly essential, the 
telegraph between the Observatory and the Central Police Station would not work.86 This was 
not just a localised problem. Trans-marine cables regularly broke down and were considered 
unreliable. In 1903, for example, a complaint was made to the Consul General at Shanghai 
regarding the meteorological observations ‘which are supposed to be received twice daily from 
the Imp. Mar. Customs [sic] authorities at Hankow, Hainan. These observations rarely arrive 
in time to be made use of by the Hong Kong Observatory & repeated representations on the 
subject have had no result’.87 On other occasions the outbreak of war brought disruption or the 
total breakdown of services. In 1898 for instance, the outbreak of the hostilities between Spain 
and America meant the loss of the Manila cable between Luzon and Hong Kong: an essential 
node in the regional storm warning network.88 At the same time, the cable between Samsui and 
Amoy had been destroyed during troubles in mainland China necessitating observations from 
that region be communicated via Japan.89 This was also the case during the Chinese Revolution 

                                                        
82 HKRS842/4 General Correspondence 1903-1913, fol. 13. Letter from G. T. Gilbert to HKO, 30 October 1905. 
83 For more on the conferences see Mahony, ‘For an Empire’.  
84 TNA, CO323/1525/1. International Meteorological Committee: Far East Regional Committee, Preliminary 
Report of the Proceedings 13-21 January 1937. 
85 HKRS356 1/2/2, unmarked folio. Letter from the Colonial Secretary Hong Kong to the British Consul, Hanoi, 
nd. 1917; HKRS356 1/2/2, unmarked folio. Letter from the Coast Inspector, Maritime Customs at Shanghai to Mr 
Le Cadet, Director Hanoi Observatory 8 February 1917.    
86 HKRS356 1/1/2, f. 32r. Report of the Director of the Observatory for 1886, 31 December 1886.  
87 HKRS356 1/2/1, unmarked folio. Letter from the Colonial Secretary’s Office at Hong Kong to H.B.M. Consul 
General Shanghai, 23 July 1903.  
88 HKRS842 3 1898-1903, unmarked folio. Letter to EETC from the Colonial Secretary 14 June 1898 & 18 June 
1898.  
89 HKRS842/3, f. 20. Communication between the Hong Kong Chamber of Commerce and Director of the 
Observatory, 17 June 1898.  



History of Meteorology 8 (2017)

 
 

 

177 

of 1911 and, although the service was resumed during 1913, Claxton reported that ‘the morning 
observations are never received in time to be utilised for the Weather Map’.90 Other 
communication problems were caused by finances (many observations were transmitted freely 
on the good will of the telephone company)91 and by personality. In 1899, for example, there 
was a ‘suspension of telegraphic typhoon warnings given at Manila, for any place outside of 
the Philippines’ due to ongoing hostility between Doberck and José Algue, Director of the 
Manila Observatory, peaking after Doberck’s statement that ‘the Observatory in Manila is in 
the hands of men who possess very little scientific education’ and his continued assertion that 
the Manila Central Observatory continually communicated sensational and inaccurate typhoon 
warnings to the Hong Kong press.92 

 

Conclusions 

 

Building on recent work highlighting British meteorology as a lens into the networks of 
knowledge that stretched across an increasingly connected world and Kapil Raj’s call to 
‘relocate’ modern science, this article has sought to offer a corrective to a meteorological 
historiography which privileges institutional and epistemological histories from the metropoles 
of European and North American power.93 The history of meteorological services in Hong 
Kong and Singapore shows how a combination of global and local factors influenced and 
directed the development of meteorological research. Studies of the weather in late nineteenth 
century and early twentieth century Hong Kong and Singapore were based on collating historic 
and contemporary observations over a wide geographic area, in order to perceive climatic 
patterns and to establish the laws governing events such as typhoons. Over the period in 
question the number of registering stations and observatories in Hong Kong and in Singapore 
was increased, through a variety of private and public investment. By the early twentieth-
century, the system of sporadic observations made by interested individuals or under different 
government or military departments had been overtaken by the more formal provision of a 
government sponsored meteorological service. This approach to research reflected global 
meteorological studies over the same period.  

                                                        
90 Report of the Director of the Royal Observatory, Hong Kong, for the year 1913, E.5. 
http://sunzi.lib.hku.hk/hkgro/view/a1913/193.pdf. Accessed 04.01.2017. 
91 HKRS842/3, unmarked folio. Correspondence between the EETC and Hong Kong Observatory, 14 March 1906. 
92 HKRS842/3, General Correspondence 1898-1902, 28, Letter to the Director of the Hong Kong Telegraph from 
José Algue, Director of Manila Central Observatory, 7 March 1899. Doberck’s relationship with, and antagonism 
towards, the Manila Observatory is covered extensively in MacKeown, China Coast Meteorology, esp. Ch. 4.  
93 John McAleer, ‘Stargazers at the World’s End: Telescopes, Observatories, and “Views” of Empire in the 
Nineteenth Century British Empire’, British Journal for the History of Science 46:3 (2011): 389-413; David 
Aubin, Charlotte Bigg and H. Otto Sibum, eds, The Heavens on Earth: Observatories and Astronomy in 
Nineteenth-Century Science and Culture (Durham, NC: Duke University Press, 2010); Kapil Raj, Relocating 
Modern Science: Circulation and the Construction of Knowledge in South Asia and Europe, 1650-1900 
(Basingstoke: Palgrave Mamillan, 2007); Katherine Anderson, Predicting the Weather: Victorians and the 
Science of Meteorology (Chicago: The University of Chicago Press, 2005). 



History of Meteorology 8 (2017)

 
 

 

178 

In both ports however, it is possible to see the influence of factors closer to home. The 
weather, especially the typhoon and monsoon seasons respectively, played a role in directing 
the trajectories of the meteorological services in both colonies. In Hong Kong, the typhoons 
that annually wreaked havoc on the island demanded attention, stimulating discussion and 
research in storm theory, and investment in storm warning systems. In Singapore, the lack of 
such extremes meant that the formal meteorological service developed later, although the need 
to improve rainfall observations was a significant stimulus toward increasing registering 
stations across the country. The mariners who traversed the China Seas were also major players 
in understanding the tropical weather during the nineteenth century. It was they who witnessed 
first-hand storms and typhoons and made breakthroughs based on experience, often in contrast 
to prevailing theory. Thus, the combination of marine and terrestrial weather observers across 
the China Seas was critical to regional meteorological science. 

Personalities and people also played a major role in the different histories of 
meteorology in the two ports. In Singapore, despite the lack of an observatory to direct research 
work, support and interest from the influential George Maxwell and Herbert Robinson helped 
to eventually reframe the meteorological services there. In Hong Kong, the series of colonial 
governors, especially Frederick Stewart in the early years, were key to pushing the successive 
observatory directors to focus more on meteorological work. Likewise, the difficult 
relationships sustained between William Doberck and the Jesuit scholars at Manila and at 
Shanghai was a significant, even dangerous, problem. The personality of key players affected 
the communication of meteorological observations between Manila, Shanghai and Hong Kong 
which, as we saw in 1899 and in 1906, could have serious consequences. This article has not 
focussed on the controversies generated by Doberck in his relationship with these two 
observatories as that topic has been covered in depth elsewhere, especially by Kevin 
MacKeown. Suffice to say that Doberck’s difficult relationship with the Jesuit scholars at both 
observatories added to the problems already inherent in the communicational infrastructure. 
The telegraph was a vital but erratic tool, subject to breakdown and interruption. When 
working, however, it formed the basis on a remarkable network of meteorological scholars and 
organisations crossing Asia from Japan to Hong Kong, from Russia to Singapore. This physical 
network formed the basis of the scholarly network that corresponded on research and 
innovation including the regional storm signal system, and which gathered at the imperial 
meteorological conferences of the early twentieth century. Much more work needs to be 
undertaken on piecing together the personal dynamics and scientific discoveries of this 
incredible network, in addition to that already undertaken for Manila and Shanghai.  

 

Acknowledgements 

This project benefitted from the financial support of a Singapore Ministry of Education 
Academic Research Fund Tier 2 grant entitled “Governing Compound Disasters in Urbanising 
Asia” (MOE2014-T2-1-017). 
 

 


