History of Mapping Hong Kong Part 8 – HMS Pegasus

Tymon Mellor: Under a clear blue sky on the 3^rd November, 1924 HMS Pegasus, an aircraft carrier, steamed into Victoria Harbour. On board were nine seaplanes fitted with state-of-the-art camera equipment for aerial photography, along with a band of enthusiastic pilots. Large scale aerial photography was in its infancy and this was one of the most ambitious applications of aerial photography to date. It would change the future approach to map making.

Hong Kong November 1924

The Need for a New Map

The 1904 Newland map of Hong Kong, the first map of the Hong Kong Island and the New Territories, was more of a compilation of information from several sources. Consequently, there was a great variability in the information and this reflected in the quality of the map.

By 1908 the Public Works Department (PWD) recognised the map’s limitations, including the fact that Hong Kong Island had not been surveyed since the Collinson survey of 1845^[1] and also a number of inaccuracies and errors which needed addressing. In early 1909 work commenced on establishing a new set of survey stations across the island in advance of a new mapping exercise. However, the exercise stalled due to the shortage of skilled surveyors and sickness amongst the staff.

The PWD had similar concerns with the mapping of the New Territories. They decided that new maps were required for the rural and urban areas at 200 ft to an inch (1:2,400) and 50ft to an inch (1:600) respectively. In 1909 the PWD established a new baseline at Tai Po, measuring 661m and extended the survey through triangulation using 15 stations to Ngau Shi Wan (Kowloon) where a second baseline of 692m had been established. This allowed the accuracy of the survey to be tested, revealing an error of 1:12,980 or 3.6” – a good result. From this baseline, a total of 52 stations were established but progress was delayed by cloudy weather in May and June and rain during the summer. By the end of the year, survey stations had been established over an area of 15 mile long by 4.5 miles wide.

Survey of the NT (1910)

The initial burst of enthusiasm faded with time, due to a shortage of experienced surveyors, sickness and the discovery of errors in the early work, casting a shadow over the exercise. By 1913, further problems also burdened the survey, including resolving land issues associated with the construction of the Kowloon Canton Railway within Colony. Piecemeal maps were prepared but the vision of a new map of the territory was lost due to the practical challenges of the environment and limited resources.

The Geographical Section General Staff (GSGS)

Back in London, the War Office, responsible for all matters military, established a department called General Staff in 1905. Its personnel were experienced officers and civilians who were responsible for military planning, coordination, analysis and information gathering.

The Geographic Section of the General Staff (GSGS), operating under the Military Operations and Intelligence was responsible for the provision of mapping and cartographic information. It would later be known as MI4. In addition to its role to supply maps to the British forces, it collected data on foreign survey networks, provided training, and prepared survey data for the British Expeditionary Force. GSGS was organized into small sections, each of which specialized in maps of a particular region. Its team operated under a Colonel of the Royal Engineers with surveying qualifications, although there were Royal Artillery and infantry officers providing support. The rest of the staff were civilian technical assistants and clerks, together with some other Royal Engineer ranks, with a total staff of around 70 by 1939^[2].

The role of Aerial Reconnaissance and the codename MI4 were later subsumed into JAPIC (UK) which became the Joint Air Reconnaissance Intelligence Centre (JARIC) in 1953. Its designation as MI4 officially ceased on 8 September 1947^[3]. They published their first map of Hong Kong in August 1905, referred to as GSGS 1393, a 3/4 inch to the mile map of the territory based on Admiralty charts, and the King and Tate maps.

Following the first world war, GSGS was under the command of Lieutenant Colonel H S Winterbotham, an active member of Royal Geographic Society and author on mapping issues. He was responsible for driving the British contribution to an international scheme to prepare maps at a scale of 1 to 1 million covering the whole world. By September 1921, GSGS had produced 89 sheets covering most of Europe and the Near East^[4]. Initial discussions about map specifications were undertaken at an international geographical congress in London in 1909, and the project officially began in 1913. The importance of this scale of map became apparent as long-distance air flights became practical and pilots needed small scale maps to navigate.

Winterbotham had been active in mapping the areas of France that became the battlefields of the first world war, and understood the importance of accurate mapping for the soldier^[5]. He probably joined the GSGS in 1919/1920 following the end of active service during the first world war.

Military Needs

The military, like the PWD had been concerned for many years about the quality of the mapping available. The King map at a scale of 8 inches to the mile was considered of high quality, but it only covered Kowloon and the territory up to Tai Po. The militarily important area north of Tai Po and along the border and north of the boarder, were partially covered by the 2 inch Newland map, but this was recognised as being of poor quality. As Winterbotham would later note, “the connection between the Island and mainland is not accurate”.

Under direction from the War Office several surveys were undertaken by the military, and like the civilian effort, these also failed to deliver. This included the preparation of a 2 inch to the mile map from the frontier to the East River, under the command of Captains Johnston and Turner. Five military surveyors completed around 50% of the work during the 1909/10 season, but surveying did not resume in subsequent seasons due to other priorities.

During 1911-1912, a survey was undertaken for a half inch map of the Canton District, consisting of six sheets, but only sheet 5, covering Wai Chau Fu got as far as the plotting of a proof sheet. The map production was abandoned due to the lack of survey information but was resurrected as part of the international goal of a worldwide set of maps at 1:1,000,000, but given the map did not include hill features, it was of little practical use.

Royal Engineering Drawing Team (1909)

In 1920/1921, the GSGS acquired a set of 250 maps sheet covering the Chinese mainland immediately to the north of the frontier. The maps at a scale of 1:25,000 were identified as being surveyed in 1914, but the mapping detail suggested that the survey was carried out in 1910. The maps did not compare favourably with the Johnston survey^[6].

Winterbotham became aware about the concern over the state of the mapping of the territory and recognised the need for improved mapping. In December 1921 a new survey for the colony was proposed, the scope of which included ^[7]:

• a triangulation for Hong Kong Island, the un-surveyed strip from the north of the King map to the frontier and connection to China to tie in the Chinese maps to the Hong Kong survey;
• a topographic survey of the area not covered by the King and Chinese maps;
• checking of the topography on the King and Collinson maps;
• translation of the Chinese names on the maps; and
• preparation of a new map of the colony at a scale of 1 to 25,000.

The survey work was estimated to require one officer and three topographers for a period of seven months at a cost of £2,500.

For the translation of the names, it was recognised that the translator would need to understand Cantonese rather than Mandarin. There were few officers with that skill, but one was identified, a Captain Wahl RE who had recently returned from the far east and was now based in the War Office in London.

The plan for the map production was:

“Drawing could be begun at the W.O. probably six months after field work had commenced, and after that would progress continuously. Assuming that each sheet of the new 1/25,000 map takes six months to draw, and that six men can be employed on the work, the rate of progress would then be 12 sheets a year. The sheets would take about two months after that to be preproduced and published.”

The total cost of the works was estimated at £3,000.

There was no funding and little will for such a massive survey exercise, so the project once more stalled.

In the summer of 1922, a four-man survey team from the Ordinance Survey was sent to Hong Kong under the command of Sargent Holbrook to undertake the necessary survey work associated with the construction of new military barracks. Progress on the new barracks had stalled and the surveyors were employed to update the battery record plans^[8]. However, London requested that they be reassigned to survey the border area to tie the Chinese maps to the King map:

“(a) A small trigonometrical survey along the gap with intersected points well within the 8” to 1 mile to the south, and the Chinese hinterland to the north.

(b) A plane-table survey at some such scale as 1/20,000 (or (say) 3 inches to the mile,) based on (a) and contoured at 20 ft.V.I.”

Colonel Davy, the Chief RE in Hong Kong was not fully on board with the programme as he thought that the existing survey was in good condition and that they could take advantage of the existing maps and the work undertaken by the PWD. He also considered the topographic work to be problematic:

“A strong party would be required and a complete organization for camps etc which is quite impossible to manage with the very limited R.E available.

The country may look all right on the map, but it is the most awful country to get about in, except for a couple months in the year when the paddy is comparatively dry.”

In the autumn of 1922 Sargent Holbrook and his team were sent north to survey key points in China to allow the Chinese maps to be coordinated with the Hong Kong maps. The work was problematic as they could not cross the border and had to rely on survey the distant prominent peaks to coordinate. As Davy noted,

“it is out of the question, in the present chaotic state of Kwangtung, to erect beacons on the Chinese triangulation point, and as the peaks are seldom really sharp points, there is plenty of room for errors in the observation.”^[9]

Starting in November, 1922 the survey team started work on the topographic survey using the plane table. It became quickly apparent that the existing survey and PWD work was poor and needed to be redone. By February 1923 Holbrook’s team had completed around half of the required area before they returned to England.

Area Surveyed by Holbrook 1922/1923

The quality of the previous work was also a concern (not to mention poor spelling),

“Tait’s triangulation covering the rest of the leased territories — apparently hap-hazard work with no attempt at marking of points. Tait’s points gradually disappeared and a fresh triangulation was carried out by the PWD. All points are well and permanently marked in concrete. Col. Russell-Brown has, however, considerable doubts as to the accuracy of their values”^[10]

As for the Newland map:

“Longbottom confirms that the 8″ survey is thoroughly good; the 2 inch is good as regards detail but hopelessly inaccurate as regards hill features and bears distinct signs of being done by somebody who had little or no idea of contouring”^[11].

It was clear a new comprehensive map was required but where would they find the money and resources. Winterbotham took a personal interest in the preparation of the new map and looked for an alternative approach. By late 1923 he had developed a plan to utilise the services of HMS Pegasus, an aircraft carrier that was heading to the Far East and planned to call at Hong Kong. The ship was to call in at Singapore and Malaysia to undertake some experimental aerial photography, and this would be an ideal opportunity to use aerial photographs to develop the new map of Hong Kong.

“if the planes arrive with you, to photograph the town and the docks, etc, in order that we may bring up the survey to date without embarking on the difficult job of mapping in a town. We should be quite able to deal with the photographs here, providing that we had these fixed positions and heights so that no specialist knowledge would be wanted on the part of anybody but there”^[12].

The Hong Kong RE military survey team was commanded by a man named Taylor. He was tasked with fixing the points necessary to allow the aerial images to be coordinated. The proposal was to dig crossed trenches at known locations that could be picked up on the images. An alternative approach was to create large crosses with linen pegged to the ground, but this approach had a flaw:

“I suspect in this latter case that the Chinese lasses would acquire pinafores etc at the expense of the Government.”^[13]

Unfortunately, Taylor was busy with the relocation of the Royal Engineers from Hong Kong Island to Lai Chi Kok and Gun Club Hill. Thus, the PWD were asked to assist under the direction of a survey team sent by Winterbotham, led by a Subaltern lieutenant Henry Wace and three other ranks, comprising of a sergeant, a lance corporal and a sapper from the 2nd Colonial Survey Section of the Royal Engineers. The duration of the survey was to be for six months sufficient to assist with the air-survey and address any issues arising. The assignment would actually last nearly three years.

HMS Pegasus

The first world war saw the advance in the use of aircraft for military purposes and the development of the aircraft carriers. Unlike the modern versions, these were ships that carried several sea planes in a hanger while the aircraft took off and landed on the water. It was not until the 2^nd August 1917 that Squadron Commander E H Dunning succeeded in landing on the deck of a moving ship. He attempted to repeat the landing a few days later and drowned when his plane was caught in an updraft^[14]. Landing on a moving object was a high-risk activity. The sea was a safer bet.

HMS Pegasus in Tolo Harbour (1924)

HMS Pegasus was commissioned in 1917, with a displacement of 3,300 tons and a crew of 258 including 100 aviation personnel along with nine Fairey IIID seaplanes of the RAF housed in a rear hanger. The ship was fitted with two photographic dark-rooms allowing for film to be processed and printed once removed from the aircraft.

Preparing for Launching (1924)

The process of launching the planes for the aircraft carrier was described by one of the pilots on the 1924 survey, G E Livock^[15]:

“Each morning the duty seaplane was wheeled out of the hanger and prepared for take-off. One of the cranes on each side of the hangar doors was swung inwards, and the patent slip hooked on to the wire slings on top of the centre section of the fuselage. If the weather conditions appeared favourable the order was given to hoist out. The engine was started and the pilot indicated that he was ready, and the RAF duty officer, standing on deck with a small flag in each hand, signalled to the crane driver for the plane to be hoisted a few feet off its trolley. The crane was then swung outboard until the aircraft was clear of the ship’s side. When the signal was given, the seaplane was slowly lowered until floats were touching the water. The duty officer then signalled ‘slip’ to the observer, who was standing on top of the fuselage. The observer jerked the quick release. The seaplane dropped gently into the water and taxied away for take-off. In two or three minutes the aircraft was in the air and climbing towards the area chosen for the day’s photograph”.

Sea Plane on HMS Pegasus (1924)

The ship had been ordered to carry out an aerial survey of Singapore Island and southern Johore, before proceeding to Hong Kong for a re-fit in the dock. The use of photography in large scale air surveys was in its infancy requiring time for preparation and practice of the flight crews. Under the command of Squadron Leader E L Tomkinson, the seven pilots gathered at RAF Mount Batten, in Plymouth in early 1924 for training. It was not easy flying as instrumentation was rather rudimentary, and the technology temperamental but after several months of practice the pilots were ready and HMS Pegasus departed in the middle of March 1924.

 

The development of aerial photography accelerated during the Great War with improvements in techniques, material and equipment. Two of the major developments were the adoption of large format roll film and the use of a fast focal plane shutter. Brigadier Martin Hotine, of the Royal Engineers and a Research Officer of the Air Survey Committee developed a method of plotting mapping details from strips of aerial images, eliminating errors of scale caused by tilt and variations in ground levels. The techniques were perfected during trials over the town of Arundel, in south England ahead of the Hong Kong survey. The technique would be known as the Arundel method of triangulation from air photographs. Hotline would be a consultant for the Hong Kong survey, providing guidance on technical details associated with image interpretation.

The Fairey IIID planes had been fitted with the new 450 hp Napier Lion engine along with the F8 camera (manufactured by Williamson Kinematograph Company Limited) for taking vertical and oblique images. The camera had a 10-inch lens with an f/4.5 aperture and, reportedly, an 80-exposure magazine producing 7-inch square prints.

F8 Cameral and Controls & Oblique Mounting

The camera system was new and had several teething problems while the pilots had to learn how to maintain a suitable flight path. As Livock recorded:

“A minute or so before passing over the starting point the observer said: ‘Steady as you go. Starting camera in one minute.’ The pilot concentrated hard on maintain his height at exactly ten thousand feet, and adjusting the trim of the aircraft so that it kept absolutely level and on a steady course.”

Pegasus arrived in Singapore by the end of April before proceeding to Johore. Flying in the humid Asian air introduced new problems, and it was found that after completing a flight it was necessary to descend slowly to prevent condensation forming in the camera and on the films.

Livock View of Hong Kong (1924)

HMS Pegasus arrived in Hong Kong on the 4^th November, 1924 to find Victoria Harbour crowded with liners, cargo ships, war ships and Chinese junks. Work on the aerial photography commenced immediately to take advantage of the good visibility and flying conditions. It was soon apparent that with so much marine traffic, taking off and landing within the harbour was not practical, so the ship moved and anchored in the Tolo Harbour, an ideal spot and more convenient for the photographs of the New Territories.

Fairy IIID Landing in Hong Kong Harbour (1924)

The initial plan had been to photograph only key areas of the colony, but in September 1924 prior to the arrival of Pegasus it was agreed that the whole territory would be photographed with an overlap of 50% so that the photographs could be fitted together to form a strip^[16]. Photographs of the outlying islands were particularly important, as they were; “troublesome to visit”.

With both vertical and oblique images, they could use them as stereoscopic pairs to get a 3D image, allowing the topography to be plotted. They did however need to have a number of known points on each image to locate and scale the image. This required support from the PWD to establish the markings, but it did not happen. They would have to locate the images based on existing features, and so these would need to be surveyed. Lieutenant Wace’s job had just got harder.

Central, 13 November 1924 (https://ncap.org.uk/frame/20-1-2-9-4)

In addition to the survey of the colony, the pilots were requested to undertake reconnaissance of both Mirs Bay and Bais Bay in an attempt to locate the hideouts of pirates. Flying over the Mainland was technically illegal, but given the chaos and in-fighting going on there, no one was going to pay attention. The flights identified several walled towns by their “distinctly medieval appearance”.

HMS Pegasus stayed in Hong Kong to see in the new year before departing in early January 1925 to finishing the aerial survey of Singapore. It was now left to Wace to undertake the necessary survey work for the production of the map.

 

1924 Aerial Images are available at National Collection of Aerial Photography www.ncap.org.uk

Source

1. Report on the Land Survey work for the year ending 31^st March 1910, Director of Public Works, 16 April 1910, CO129/366 ↑
2. The Second World War 1939-1945 Army, Maps and Survey, Compiled by Brigadier A B Clough, The War Office, 1952 https://en.wikipedia.org/wiki/MI4#:~:text=The%20role%20of%20Aerial%20Reconnaissance,Reconnaissance%20Intelligence%20Centre%20(JARIC). ↑
3. Nature Geography at the British Association, 29 Sep 1921 ↑
4. British Survey on the Western Front, Lieutenant Colonel, H S L Winterbotham, Jan 1919 ↑
5. Observations regarding Para 3 of letter K/952 from Hong Kong dated 7^th July 1923, WO 181/81 ↑
6. Survey proposal by Lieut-Colonel G.S 1 Dec 1921, WO 181/81 ↑
7. Letter Colonel C W Davy Chief Engineer to Lieut Colonel H Stj L Winterbottom, 26 Sep 1922, WO 181/81 ↑
8. Letter Colonel C W Davy Chief Engineer to Lieut Colonel H Stj L Winterbottom, 5 Feb 1923, WO 181/81 ↑
9. File Note, MI4(Col), 13 May 1924, WO 181/81 ↑
10. File Note, MI4(Col), 13 May 1924, WO 181/81 ↑
11. Letter, 27 Nov 1923 Colonel H S Winterbotham, GSGS to Colonel V W Davy CMG Chief Engineer, The Forces in China, Hong Kong ↑
12. Letter Colonel H S Winterbotham to Russel-Brown, 11 Sep 1924, WO 181/81 ↑
13. Edwin Harris Dunning, https://en.wikipedia.org/wiki/Edwin_Harris_Dunning ↑
14. To the Ends of the Air, Group Captain G E Livock, 1973 ↑
15. Letter Colonel H S Winterbotham to Russel-Brown, 11 Sep 1924, WO 181/81 ↑

This article was first posted on 27th July 2024.

Related Indhhk articles:

1. Mapping Hong Kong Part 1 – Where Are We?
2. Mapping of Hong Kong Part 2 – 1841 The Belcher Map
3. Mapping of Hong Kong Part 3 – 1845 The Collinson Map
4. Mapping Hong Kong Part 4 – 1866 Map of San On District
5. History of Mapping Hong Kong Part 5 – Mapping Kowloon
6. History of Mapping Hong Kong Part 6 – 1901 The Tate Map
7. History of Mapping Hong Kong Part 7 – 1904 The Maps of King and Newland