The Institute was planned to bring together not merely the professional navigators, but all who could contribute to the exchange and advance of knowledge bearing on navigation. The founders were wise when they gave effect to the thought that the small-boat sailor and the sailplane aviator could contribute to the progress of our science and art, and could greatly strengthen the hand of the professional provider and the professional user of the processes and material of navigation. The reasons for exchanging and advancing our knowledge are as diverse and yet as closely interacting and mutually reinforcing as are the individual interests of our members.

During the Antarctic summer season of 1949 the Norwegian sealer Norsel, which had been chartered by the expedition, landed the Norwegian-British-Swedish Antarctic Expedition in Queen Maud Land in the Antarctic. In January 1951 the ship took down relief personnel and stores and for this trip the Admiralty, at the invitation of the planning committee, appointed the author as an observer. The decision to appoint a member of the navigation branch was made partly so that the techniques of ship handling and navigation in ice could be studied; and since the captain of Norsel, Captain Jakobsen, was extremely experienced in every aspect of ice navigation, a great deal of valuable information was acquired. The Institute had also suggested a number of astronomical observations that might usefully be made in polar areas and the Admiralty gave its approval for the observer to make them.

The history of marine navigation in the area to the north of Hudson Straits and west of Greenland dates back to Martin Frobisher in 1567, John Davis in 1585 (who reached N. 72° 15′), and William Baffin, who got as far as Smith Sound (N. 77° 45′) in 1616. All of them were searching for a short route to the East. There was little exploration in the area for the next two-hundred years, until, in 1818, the British Government recommended explorations for the Northwest Passage: This activity continued throughout the Franklin era up to 1875, when Captain George Nares proceeded to N. 82° 25′ on the northern end of Ellesmere Island. After this, the role of exploration gradually passed to American hands, largely culminating in 1909 with Robert E. Peary's attainment of the North Pole.

No account of the observations described in this paper has previously been published, though they were made nearly eighteen years ago. The author has kindly presented the Institute with all the data he obtained from them and with a detailed description of the investigation. We print here a short account of the observations and of the principal results obtained, which it is felt should go on record.—Ed.

Largely as a result of discussions with Dr. Per Collinder of the Swedish Hydrographic Service and Dr. Knut Lundmark, Professor of Astronomy at Lund, the author undertook, between June and August 1934, a series of sextant observations to determine the degree of refraction at low altitudes, the normal accuracy of sights at sea, the accuracy of twilight and night observations and possible elevations of the horizon at noon. 1908 observations were taken during the period.

After taking a latitude at noon and obtaining a position, a course and distance to a position nearer the destination is usually required. A course and distance from an estimated noon position may be worked prior to noon and then corrected at noon by the observed latitude and morning position line, by means of a plotting chart, giving reliable results quickly and accurately.

In Fig. 1, BA and CA represent the course and distance from B to A and C to A respectively. BD and EC are arcs of concentric circles with centre A; and AC = AE and AB = AD.

If BD and CD are small in comparison with AB, arcs BD and EC may be considered straight lines and the figure BECD a rectangle. If the rectangle can be solved, the distance AC can be deduced from AB, i.e. AC = AB - DC or AC = AB + DC if C lies the other side of BD.

There is rich material for studying navigation technique in the fifteenth and early sixteenth centuries in the volume A Marinharia dos Descobrimentos by the late Fontoura da Costa, which deals in great detail with the evidence for early sailing methods; the facsimile reprints of rare documents bearing on early Portuguese navigation lately published by Joaquim Bensaude for the Portuguese Government also contain much information on this period. The object of this article is two-fold: to establish that in an Addendum to one of these documents, the Reportorio dos Tempos of Valentim Fernandez (1st edition 1518), we have extracts from an archaic type of navigating manual; and hence to argue that, judging from his own words, Columbus (whose skill in navigation is a matter of controversy) had studied rules and instructions for navigation just such as the Addendum contains, but that owing to his lack of elementary mathematical training he had only imperfectly understood them.

The fact that the de Havilland Comet cruises at nearly double the flying speed and nearly double the flight altitude of comparable piston-engined airliners is a measure of the great leap forward in air transport which the introduction of gas-turbine engines has brought about. It also illustrates an inherent feature of jet aircraft, for only at high speed and high altitude do turbo-jet engines, and to a less critical extent turbo-prop engines, operate at their maximum efficiency, expressed in distance flown per gallon of fuel consumed. With reduction of altitude and speed, however, the fuel consumption does not fall off proportionately and efficiency decreases rapidly. For example, even when taxying on the ground the Comet's fuel consumption is about 70 per cent of that at cruising altitude, and when flying at sea level the still-air distance flown on a given quantity of fuel is less than half of that at cruising altitude. In other words you get the best out of a jet aircraft when you are going somewhere fast, and when the length of the flight is sufficient for the aircraft to climb to its optimum cruising altitude and remain there for a considerable proportion of the flight time.

The purpose of an instrument let-down is to bring an aircraft past all hills and obstructions to a particular position at low height. If the pilot follows the correct procedure, the aircraft will obviously clear the hills and obstructions. But owing to various interruptions and distractions, difficulties of reception or the need to give partial attention to other duties, he is often unable to follow the procedure exactly; he will then want to know whether departure from the procedure requires immediate rectification, or whether he is at liberty to return gradually to the correct path. A badly working radio aid, or some other difficulty, may make it necessary to abandon a partly completed approach; he will then want to know the best way to extricate his aircraft from a possibly dangerous situation. Furthermore, the pilot often has to improvise an ‘intermediate procedure’, to connect between his arrival over the aerodrome and the beginning of the let-down. For these reasons he needs to know the positions of the hills and obstructions in relation to the let-down; and the only way to present this information to him quickly and unambiguously is by making hills and obstructions the principal features of let-down charts. The local knowledge possessed by pilots of scheduled services does not lessen the need for an effective presentation of this feature on the charts, as it is precisely when landing at unfamiliar aerodromes that the charts will be most used.

This paper was written in 1946 when Mr. Kiewiet de Jonge was a student at Harvard College Observatory. Subsequent developments in astronomical navigation, particularly the publication by the U.S. Hydrographic Office of H.O.249, have overshadowed the practical advantages of the three-star position tables. But the principles of the method, which have not previously been described in print, are of sufficient importance and novelty for permanent record.

In this paper we will develop a fast tabular solution of the position deduced from three ‘simultaneous’ sights. This is achieved in a novel way, through principles not generally associated with the traditional approach to the problem. The solution is based on the fact that the measured altitudes of a set of three stars uniquely define the most probable position of the observer projected on the celestial sphere.

Amongst the schemes for simplifying the processes of astronomical position fixing the family of multi-star sextants contains some interesting material. By the simultaneous observation of two or more stars through an optical system coupled to a gravity reference, it is possible to determine the observer's position directly in latitude and longitude, with little or no computation.

The number of instruments on these lines which have been designed is very large and it would not be possible to describe them in detail in the course of a short paper. A few remarks about their principles may, however, be of interest since they are, in general, sound in theory and fall short of practical requirements only because, when used on an unstable platform such as a ship or aircraft, they make too heavy a demand on the skill of the observer.