¹Russell, H. N.(1919). On the Navigation of Airplanes. Publications of the Astronomical Society of the Pacific, Vol. 31, pp. 129–149. The writer is indebted to D. H. Sadler for this gem.Google Scholar

² In Alcock’s account the fix was from ‘ the Pole Star, Vega and the Moon ’ but the Moon sight does not appear on Brown’s chart and, the following day, the Daily Mail quoted Brown as describing his fix as ‘ a cut on Polaris and Vega’. The only other sights plotted were two Sun sights providing ground speed checks, one in the late afternoon four hours out of St John’s and the other the following morning about an hour before sighting the Irish coast. The overcast was almost continuous. These four sights probably constituted the first time an aeroplane was really navigated by astro.Google Scholar

³ The prototype of the twentieth-century marine sextant is generally considered to be Hadley’s Octant built in 1730.Google Scholar

⁴ Gaspard Gustav de Coriolis published the equations of motion relative to a rotating frame of reference in 1835.Google Scholar

⁵ In the ninth edition of his Wrinkles in Practical Navigation.Google Scholar

⁶ By the accounts of the instability of The Spirit of St Louis, it seems impossible for one man to fly it and take an accurate sight even if Lindbergh had been proficient in AA. On considerations of weight he did not even take a radio and crossed the Atlantic entirely on DR. A. J. Hughes (a leading sextant manufacturer) tells us how uninterested aviators were in ‘ astro ’ in those days.Google Scholar

⁷ Variously measured 0° – 360° to the West of PZ or 0° – 180° East or West of PZ depending on the system.Google Scholar

⁸ The position line must be plotted on a conformal projection if angles are to be true and scale at any point the same in all directions. Great circles are straight lines on conformal projections at any point at which the convergency of the meridians is sine latitude although sights have, with due care, been successfully plotted on mercator projections at least up to latitude 65° but not, of course, with the long intercept methods described in Sections 4.3 and 4.10.Google Scholar

⁹ In a poll of RN and MN navigators in 1948, the overwhelming majority of the respondents opted for a GHA almanac and GHA was subsequently adopted in nautical almanacs, another example of the influence of AA on NA.Google Scholar

¹⁰ A note in the 1946 Éphémèrides Aéronautiques said ‘si les instants d’observation sont notés à l’aide d’un sidérometre ce qui est le cas en navigation aérienne’ – The words in italics was news to the rest of the world but in answer to inquiry by the Superintendent of the British Nautical Almanac Office, the sad reply from the Bureau de Longitude was ‘ les Francais tiennent a reserver l’avenir de 1’usage des sidéromètres… Cette méthode n’a jamais pu faire ses preuves en dehors des centres d’essais…’ There was a little-used sidereal chronometer at the RAF School of Air Navigation in 1936.Google Scholar

¹¹ In July 1947 the Dirección del Transito Aereo della Republica Argentina produced an Almanaque Aeronautico although at that date all qualified navigators in the Flota Aerea Mercante Argentina were ex-RAF navigators.Google Scholar

¹² Produced jointly in the Nautical Almanac Office of the Royal Greenwich Observatory and the Nautical Almanac Office of the US Naval Observatory. Identical publications (GMT not GCT) were published separately in the US and UK. There were many changes of detail, generally accepted as improvements.Google Scholar

¹³ From 1953, a common GHA interpolation table was used for the Sun, γ and planets.Google Scholar

¹⁴ Haversine tables have been available to navigators at least since 1805, versines since the eighteenth century, notably in Maskelyne’s requisite tables.Google Scholar

¹⁵ Aquino listed in 1951 (The universal sea and air navigation tables. This Journal, 4, 109) some short methods to which his own tables can be applied. These are in chronological order, Thomson (1876), Souillagouet (1891), Delafon (1893), Le Blanc (1893), Fuss (1901), Borgen (1902), Aquino (1902–46), Wedesneyer (1910), Newton (1912–24), Immler (1917–26), Bertin (1918), Modena (1919), Wendt (1925). Aquino’s list is far from complete!Google Scholar

¹⁶Ogura, GingrichMyerscough, and Hamilton, , all used a method known as ABC tables (first described in 1846) for azimuth. Weems used the Rust diagram (see Section 4.6). The method of Dreisonstok is described under Hughes.Google Scholar

¹⁷ R. Adm. R. de Aquino, Brazilian Navy, Cdr A. A. Ageton, USN.Google Scholar

¹⁸ In an obituary to Comrie, L. J. the compiler of Hughes Tables. This Journal, 4, 207.Google Scholar

¹⁹Cassini, Cesar Francois, Director of the Paris Observatory, and grandson of the great Giovanni, produced such tables in 1770 for a limited band of whole degrees of h, Ф and δ.Google Scholar

²⁰ The most elegant solution is not to correct H but to correct the fix by the amount of the precession of an imaginary star coincident with the observer’s zenith since the epoch for which the tables were calculated. In minutes of arc latitude correction = 0.33T cos LHA γ longitude correction = [0.77 + 0.33 sin LHA γ tan Ф]T where T is years elapsed since the epoch of the tables. For such small numbers a correction table is an easy matter. It can be used also with astrographic methods (see Section 4.5).Google Scholar

²¹ In a light aircraft the weight of ANTS was prohibitive. In 1948 on a flight around the world between latitudes 15°N and 65° N in a Proctor IV on which weight considerations were paramount, Michael Townsend carried Driesonstok (less weight than Hughes), a Mk IXA sextant and a Mk II Astrocompass.Google Scholar

²² The Reichsluftfahrtministerium provided Höhentafeln nach Sternzeit, giving altitude and azimuth of six stars for each even degree of latitude and every minute of sidereal time from 1944. The Kriegsmarine issued the graphical equivalent, star charts similar to Weems (see Section 4. 5), from 1940.Google Scholar

²³Anderson, E. W. and Sadler, D. H.(1953). The genesis of EANTs. This Journal, 6, 333.Google Scholar

²⁴Meneclier, V. and Chevalier, R. (1945). Calculo del Punto por Alturas de Estrellas. Secretaria de Aeronautica, Argentina. Both authors were retired Argentine Navy Capitanos de Frigata.Google Scholar

²⁵ It is interesting that Hohne described himself as an ‘air navigation instructor’. All previous tables were the work of nautical almanac offices, hydrographic offices or others with a marine rather than aeronautical background. The Japanese Celestial Air Navigation Tables which appeared in 1940 seem to be the first to arrange selected stars in order of azimuth but as the argument was h not LHAγ the advantage is lost.Google Scholar

²⁶ Hoehne had proposed his tables to the US Hydrographic office in October 1941. The basis of HO 249 was canvassed by Hutchings, Ageton and Weems from late 1942.Google Scholar

²⁷ In fact only 16 volumes of ANTS, covering 79° S to 79° N, were ever published.Google Scholar

²⁸ tan y = tan δ /cos h where y is the declination of N in Fig. 6. Y, the complement of NZ in Fig. 6 is calculated. Then tan Z = cos y. tan h/cos Y, tanH = cos Z. tan yGoogle Scholar

²⁹ H. E. Wimperis was an early Director of Research at the Air Ministry in London. Bygrave was a close associate of his. Wimperis published in 1920 a remarkably foresighted Primer of Air Navigation (Constable).Google Scholar

³⁰ P. V. H. Weems was an influential thinker on AA from the late 1920s to the late 1940s. He retired from the US Navy at a Lieutenant-Commander and formed his own business, The Weems System of Navigation. He also produced tables (see Section 4.1) and books.Google Scholar

³¹ Not a scale one would choose for long range flights since each 1000 n.m. is represented by 6 ft but the Ruhr is not far from East Anglia.Google Scholar

³² The ARG is illustrated in this Journal, 4, 14–16 and 28 486–487.Google Scholar

³³ An excellent description of them is given by Herrick, S. (1946). Instrumental solutions in celestial navigation. Navigation, 1, No. 2.CrossRefGoogle Scholar

³⁴ Both the Goerz and the Bumstead Sun compasses followed similar principles to the astrocompass which they preceded. The hour circle was driven by clockwork so that the aircraft could be steered by it but the frequent adjustment to latitude and longitude together with the absence in most aeroplanes of a site which is always in sunlight and within the pilot’s range of vision dampened acceptance. Byrd and Floyd Bennet used the Bumstead on their flight to the North Pole in 1926.Google Scholar

³⁵ Astrodomes became unpopular when one was sucked out of an early pressurized airliner complete with sextant and navigator. Periscopic sextants quickly became standard, spelling the demise of the astrocompass almost by accident.Google Scholar

³⁶ ANTS provided a table tabulating the azimuth of Polaris from 1940 to 2000 AD.Google Scholar

³⁷ See Sadler, D. H. (1949). Tables for astronomical polar navigation. This Journal, 2, 9.Google Scholar

³⁸ However it must be mentioned that Ex-Meridian Tables (Inman, Norie, etc.) were certainly used by some airmen to find latitude well into the thirties. Noon sights were certainly practised with marine sextants in slow low-flying aircraft. The noon sight was key to the navigation plan of Harold Gatty on the first flight to Hawaii.Google Scholar

³⁹ This became 57 when the unified almanac was introduced in 1953.Google Scholar