A draft of this paper was circulated for the open meeting of the Technical Committee on 17 May 1972, when the forthcoming revision of the Collision Regulations by Imco was discussed (see page 427).

Radar has been in general use in merchant ships since the end of the Second World War and during this time the Regulations for the Prevention of Collision at sea have undergone two amendments at International Safety Conferences, in 1948 and 1960. The amended regulations were ratified and brought into use in 1954 and 1965 respectively. The 1948 regulations made no reference to the use of radar whatsoever—radar for commercial use was still in its infancy in 1948 and indeed little was then known about its capabilities and limitations, nor of the operational difficulties it was going to present even before these 1948 regulations were finally ratified in 1954.

Listening to Captain Brandenburg's talk on his ideas about the future navaids in Rotterdam (this Journal, 25, 67) I was left with the conviction that even the most ardent opponents of change will at least be better informed having heard it. Although in his introduction he excluded direct reference to Rotterdam, it is well known that the new Auto Com/Radar Data system when introduced there will be a real milestone for historians of port entrance techniques. I would like to think that it will lead us in this country one step closer to the day when we take the option out of progress on our own coasts and waterways and get on with it as they are doing in Rotterdam. There is of course one great difference—in Holland and Germany the Governments and Municipalities are involved, so that the necessary finances are seen in a clearer light.

An automatic collision warning system for use at sea, using present-day lower power techniques and capable of handling up to 250 ships at a time is described. The equipment does not involve expensive computer facilities.

During the past twenty years I have sailed my ketch Mara (12 tons) in the English Channel and have been involved in several close-quarter situations with large merchant ships, fortunately without accident. Any competent navigator is aware that if the bearing of the other vessel remains constant, there is a risk of collision. In Mara's case, moving at a fast walk, the threat on 090° coming at 20 kt. may pass ahead or astern. If the threat signals her intention by one or two blasts there is no problem; it is the ship that holds her course in blind silence that gives the yachtsman a cause for heart failure.

Yachtsmen, like other mariners, are being encouraged to ‘think metric’—to become accustomed to metres in Tide Tables, Pilots and Light Lists, and to make use of the new metric charts as and when these become available. Along with these changes to books and publications we can expect the introduction of metric tide gauges, and of echo sounders calibrated not in feet or fathoms but in metres. In a few years time even those yachtsmen who normally never correct their charts, or who do not think about renewing them until they are covered with cocoa stains, will be forced to work in metres.

John Hadley (1670–1744), Vice-President of the Royal Society of London, communicated his ‘Description of a new Instrument for taking Angles’ to the Society on 13 May 1731. Hadley's invention for the first time provided the navigator with an instrument by which he could measure altitudes of celestial bodies with ease and accuracy on board a lively ship at sea. It was not however until about 1750, when the instrument was to be found on board vessels of the East India Company, that Hadley's quadrant (or octant as it is sometimes called) rapidly came into general use.

It may come as a surprise to many navigators that the interesting and valuable, although seemingly little-used, method in which three successive bearings of a single fixed mark are used to find course made good, has a history of not more than about half a century.

When two or more suitably placed marks are in sight the course made good may readily be found from a series of fixes by cross bearings, and this technique is commonly used by coastal navigators. When, however, a single mark alone is available the reliability of running fixes depends upon the accuracy of estimations of courses and distances made good in the intervals between successive bearings. A table for finding the distance of a mark from two bearings and the distance run in the interval between observations was inserted in the first edition of Raper's Practice of Navigation in which the author informs his readers that the table was constructed at the suggestion of Sir Francis Beaufort and that it first appeared in the Nautical Magazine.

It is doubtful if at any period in the history of seafaring a greater effort has been made to organize the safe conduct of ships about the seas than that which is being made today, and it is equally doubtful if the need was ever more urgent. At the present time a large amount of thought and work is being put into the task of trying to formulate rules and systems aimed at providing for the safer navigation of vessels at sea. Excellent ideas have been proposed for conduct in traffic, such as those put forward by Dr. Calvert, Admiral García-Frías and many others, which, together with the systems of traffic separation which have been established in areas of heavy shipping concentrations, could bring about a decrease in the present rate of marine casualties.

Prompted by a recent paper, in which a brief reference is made to Piri Reis (the Turkish term ‘Reis’ is a title corresponding approximately to ‘Naval Commander’), I am persuaded to bring to the notice of readers of the Journal—many of whom perhaps are unacquainted with the name of Piri Reis—the following brief account of the life and work of a brilliant sixteenth-century Turkish admiral.

Graduating error distributions by families of curves has a long and distinguished history. In seeking a class of frequency distributions which graduate navigational data, Anderson and Ellis are motivated by the well-known shortcomings of the normal distribution which so often fails to do justice to the data in the tails of the distribution. They generalize the one parameter (σ) zero-mean gaussian family to a two parameter (α, β) family which is in fact the Pearson Type VII class. They then observe that this class graduates published navigational distributions very wells.

Commander P. C. H. Clissold, in a contribution to Forum in the April 1971 issue of the Journal, makes a point concerning the collision-avoidance handling of very large single-screw ships. Commander Clissold writes:

‘Because of their unwieldiness an avoiding action must be initiated while still at a considerable distance from the threat, if it is to have any effect. This distance is beyond that at which the eye of the navigator can accurately assess the risk of collision or the need to manœuvre. He must, therefore, depend upon instrumental information for making his decisions; in other words, in clear weather as in thick, he must use his radar and plot continuously if he is not to hazard his ship. If the argument is put forward that this will require two men on watch together and the state of manning does not permit this to be done, the answer is that a change in organization must be made to make it possible.’

The controversy over the introduction of corrections to standard time, so as to force agreement between Atomic Time and Mean Solar Time seems to have its roots in an erroneous concept of the meaning or concept of the entity we call Time. For example, the preamble to H. M. Smith's paper on ‘Time Data and Navigation’ suggests that ‘…Greenwich Time and longitude are inseparable…’ and in D. H. Sadler's paper on U.T.C. there is an odd confusion in noting that Time is a measure of the rotation of the Earth on its axis and at the same time that the rotation is not uniform.

It is interesting to consider navigation and the problem of traffic control with the common underlying concept of randomness, though Anderson's presentation seems more oriented to thermodynamics than to information theory. But the underlying philosophy is the same: entropy never increases.

In order to decrease entropy without the help of Maxwell's demon we have to introduce work or energy, that is we measure the state of a specific object and control that state with the aid of the information gained.

The principal purposes of this introductory paper are to identify and describe the main aspects of marine traffic engineering, and to suggest ways in which traffic engineering principles and methods developed for land and air transport systems may be applied to marine problems. Traffic engineering is a relatively new branch of engineering science, but it was quickly and widely recognized as essential in ensuring that all forms of land and air transport operate safely and efficiently. However, although sea transport is a long established activity, it is only during the past few years that marine traffic engineering has begun to emerge as a separate aspect of this broader scientific discipline.

The following comments on Captain C. H. Cotter's note (this Journal, Vol. 24, page 569) on the use of addition and subtraction logarithms in navigation may be of interest. Captain C. Carić was not the first to introduce gaussian logarithms to navigators, since their use was advocated, at least to Portuguese navigators, in 1920 by Admiral Gago Coutinho (see my review of Precision Astrolabe by Francis M. Rogers, 25, 135), and actually used by him on his pioneering transatlantic flight in 1922. They are included in several collections of nautical tables, including Tábuas Náuticas by Fontoura and Coutinho, as well as those of Friocourt and Hoüel; but I have been unable to obtain the dates of their first inclusion. However, it is most unlikely that either Coutinho or Carić, who incidentally used quite different basic formulae, were the first to advocate the use of addition and subtraction logarithms, or possibly to publish methods and tables. They were in use to a small extent in astronomical calculations, and to a somewhat larger extent in surveying. Dr. L. J. Comrie, the computer and table-maker, who was Superintendent of H.M. Nautical Almanac Office (N.A.O.) from 1930 to 1936, was an enthusiastic advocate and designed many computational forms incorporating their use.

In 1970 the Institute formed a Working Party to formulate views on the revision of the collision regulations. Its Interim Report is published in the Journal for October 1970 and two further contributions, by Calvert and García-Frías, were published in a subsequent number (24, 413–20). The Interim Report referred to the value that a manœuvring diagram along the lines originally proposed by Calvert and Hollingdale would have in helping the mariner to assess the effects of collision situations on alterations of course and speed. It has not been possible to secure unanimity about the form of this diagram, but the one illustrated below is based on the contributions of several members and has met with the approval of the majority of the Working Party.

In his Presidential Address Admiral Ritchie referred to the two series of charts in 79 small-scale sheets which 16 member states of the International Hydrographic Organization have undertaken to produce; some of them will be published this year. This new development, and how it will affect mariners who now use our Admiralty charts, may be of some interest.

The two small-scale charts for ocean and offshore navigation require a comparatively limited effort and with the enthusiasm and interest shown by all participating nations it is expected that both could be completed within a reasonable period, probably before the end of 1974. When these charts are published it will be possible to withdraw many small-scale Admiralty charts, some of which were originally published more than a century ago and have subsequently received only partial correction, and much criticism as being out-of-date. But it will not usually be possible to replace them chart for chart as each individual new International Chart is published, because replacement of existing charts can best be made when the whole scheme or at least a complete regional area has been completed. This applies not only to the United Kingdom but to all the world charting authorities and will result in their still having to maintain many existing charts until the new coverage is complete. This problem of ‘block’ replacement, rather than individual chart replacement, will assume serious proportions in the subsequent larger-scale stages of the International Chart.

A general programme of work with the overall aim of reducing the incidence of collisions in the Dover Strait has been commissioned by the Marine Division, Department of Trade and Industry. As part of the programme the Operational Research Unit in DTI is developing mathematical models of traffic flow, using inputs of data provided from the surveys mounted by NPL and described in the paper by Johnson and Wheatley.

At a first glance, one of the two accompanying diagrams might be taken to represent a three-station Decca or Loran chain, while the other offers familiar instruction on how to interpolate between the lattice lines and plot the fix. In fact, these are illustrations in a German Army handbook entitled An Introduction to Sound Ranging and the Work of the Sound Ranging Section, published in July 1918.