Captain Henry William Gordon was employed during the Anglo-Boer War (1899-1902) in the Survey Section of the Royal Engineers under the director of military intelligence. Towards the end of 1902 he was transferred to the primary triangulation of the Orange River Colony and Transvaal Colony. The purpose of the work was to establish geodetic chains in the two territories that would extend the existing primary triangulation of the Cape Colony and Natal. The project was initiated by Sir David Gill*, who became scientific adviser for the surveys to the two governments. Colonel W.G. Morris* was the superintendent of the survey, the results of which were published in 1908 with Morris, Gordon and Gill as authors. Most of the field work was carried out by Gordon, W.B. Robinson*, A. Simms*, and Captain Ley. Gordon first took part in a reconnaissance survey and then carried out observations in the field from July 1903 to August 1905. He surveyed the following chains: From Newcastle through Belfast to the Limpopo River, forming part of the 30 degrees East arc of meridian (656 km, with Simms); from Kimberley through Kroonstad to Newcastle, connecting existing chains (536 km); from Hopetown to Wepener (296 km); from Ottoshoop to Belfast; and a short secondary chain between Johannesburg and Pretoria. The latter connected the surveys of the Witwatersrand with the geodetic survey and its results were published by W.B. Robinson in 1907.
In July 1906 Gordon was transferred from the Transvaal to survey a connecting chain between the Transvaal survey and a geodetic chain surveyed by Simms in Southern Rhodesia (now Zimbabwe). This work would close a gap of about 2 degrees in the 30 degrees East arc of meridian. The observations were difficult owing to irregular refraction, haze, dense bush and the absence of roads, but he managed to complete the work in January 1907. The geodetic survey of South Africa was now completed, and the arc of meridian extended unbroken from the Cape Colony to the southern shores of Lake Tanganyika. The results of Gordon's last survey, particularly the azimuths and heights, were somewhat disappointing, though mainly as a result of errors in the two connected systems, and in 1930 the International Union for Geodesy and Geophysics decided to re-survey the connection.