The geometry of collision at sea has been dealt with in a series of papers published in the Journal, notably by Sadler (10, 306), Calvert (13, 127), Garcia-Frias (13, 316) and Morell (14, 163); and a further contribution from Calvert (to which reference is made in this paper) will be published in the next number.

The object of the present paper is to examine whether anti-collision manœuvres, here considered for the case of two craft moving in a plane, can be formulated on a rigorous logical basis. If they can, then clearly a proper appreciation of the geometry of collision is a prerequisite to the formulation of any rules and regulations. In this important paper the author gives a precise and complete answer to the basic problem, and he proves mathematically that it is the only answer. As Dr. Hollingdale freely acknowledges, however, this can only be a contribution to the study of the collision problem, which involves innumerable operational factors in addition to the geometry of the situation.

The convention adopted in the present paper is that each craft shall manœuvre so that if the other craft stands on, the sight line always rotates in the anti-clockwise direction. The analysis shows that a simple set of manœuvres can in fact be developed on this basis and that such manœuvres are the only ones that, geometrically, meet all the specified requirements. In all cases the combined manœuvre converts a collision situation into a ‘miss’ of at least a specified magnitude.

This number of the Journal goes to press just before the third international Convention sponsored by the Institute with the Ausschuss für Funkortung and the French Institute of Navigation. As members will know, the theme of the present series of meetings is the prevention of collisions, at sea and in the air, by systems or methods based ashore or on the ground. An account of the meeting and summaries of the papers presented will be published in the next number of the Journal, and a selection of the papers themselves will be published in future numbers. Wing Commander Anderson's paper, in which he introduces the concept of a collision index by means of which travel risks may be related to the risks of other human activities, forms a suitable introduction to the series.

This paper by the Director of Control and Navigation at the Ministry of Aviation, reviews the present state of planning for air traffic control in the United Kingdom. It indicates that whilst progress is evolutionary and continuous, big steps take place approximately every ten years. The author concentrates on four principal facets of air traffic control which cause the most acute problems, namely the medium-haul operations in densely populated areas; the terminal problem; long-distance operations over the Atlantic; and low-level, short routes in the south-east of England.

The paper outlines the basic air traffic control system of the future and indicates the ways that data processing and radar may be used. It makes a short reference to the specific problems of the supersonic aircraft and concludes by referring to two principal experimental projects now in hand to enhance our knowledge of data processing with respect to air traffic control.

A method is presented for quick calculation of the frequency of cross-wind components which limit the landing of aircraft on runway layouts at aerodromes ánd for deriving the best runway layouts for a site with a known or deducible wind distribution. The method depends on some initial study of the relevant wind rose but an exact solution is obtained by the application of the empirically derived formula nv = Ce–bv2 where nv denotes the percentage of time that a wind exceeds a velocity v and C and b are constants for a given site. This formula was originally verified only for the British Isles but later it has been found to be valid elsewhere where there is a wide range of wind speeds. Allowance can be made in the calculations for bad weather, e.g. dense smoke or low cloud approaching the aerodrome from certain directions, for use of one end only of a runway and for a permissible tail wind component.

The paper was originally issued in 1944. The method described has not been changed and the description is only slightly modified. The chief changes are the substitution of numerical integration using Meteor computer instead of the original planimeter measurement of curves and the introduction of the ‘bisector’ correction. Certain details of the calculation notably four-figure tables of L(i) (eqn (5)) and of the bisector correction, a complete example of the method, the way to remove ‘bias’ in wind summaries and the derivation of the bisector correction are given in a duplicated Addendum to this paper which is available on request from the Institute.

The problem of presenting barometric height in aircraft altimeters is discussed. The alternative methods of presenting a numerical value are outlined and their characteristics analysed with reference to available comparative evidence providing a quantitative assessment. The properties of various methods of presentation are examined with reference to the varied uses of the altimeter. The conclusions are discussed with relation to the recommendations of the U.K. Altimeter Committee.

It is concluded that no single means of presenting height information will meet all of the roles in which the altimeter is used and that the recommended solution is a combination of a digital counter presentation giving an unambiguous height value combined with a circular scale and pointer rotating once per 1000 ft. providing rate and angular cue information for control purposes. Freedom of choice in presentation can only be achieved by the use of a servo-operated display.

Before commencing an analysis of the subject matter at hand, it will be as well to present some definitions which we have adopted, in our own work, even if they represent in part arbitrary interpretations of our own.

(a) Pressure altitude is, of course, nothing but barometer pressure at any given elevation, converted into equivalent altitude (feet) units; hence a variable.

(b) Sea-level in connection with altimetry is only a theoretical reference value which does not necessarily conform with mean-high water; in terms of pressure altitude, needless to say, it is also a variable.

(c) True altitude (above sea-level) cannot be obtained with satisfactory accuracy by merely correcting for sea-level barometer pressure. On the other hand, we must dispel the notion—not uncommon—that one can obtain more accurate altitude values by correcting for temperature and humidity at a particular flight level. There are such things as inversions and convection currents; we do not need cold weather to have a high barometer, and sometimes it is warmer a few thousand feet up than at ground-level.

(d) Performance of pressure instruments is properly expressed in parts of the pressure range—not in per cent of altitude—and the reason is simple: at, say, 60,000 ft. you have about 16 times as may feet per unit of pressure as at sea-level.

(e) Sensitivity is—technically speaking—the smallest part of the pressure range that will produce a measurable response.

This paper develops a method for the prediction of wave height, period and direction on a 200-mile grid over the North Atlantic Ocean, using a wave spectrum technique. The method could be useful in conjunction with optimum ship routing across the ocean.

AS discussed recently by Hanssen and James, ship routing across the ocean, using the optimum route with respect to weather conditions, is now becoming a practical possibility. One essential requirement is to predict the wave heights and lengths over all parts of the area concerned. This would require first, of course, a forecast of the wind conditions. At present the Meteorological Office issues prognostic weather charts 24 hours ahead and wave predictions must be tied to them. Wave conditions are not determined instantaneously by the weather conditions at a given time, but tend to be an integration of the weather conditions over two, three, or even more days before and so wave predictions 24 hours ahead are likely to be more accurate than the wind forecasts.

A description is given of sea trials of two types of narrow-beam echo-sounder developed with a view to obtaining information as regards both the depth and the slope of the seabed. One system involves a two-channel interferometric arrangement, which provides a family of several spaced receiving beams; the other uses an eight-element hydrophone array with electronic sector-scanning of the resultant narrow beam. Photographs of typical displays show that comparable values may be derived for the slope of the seabed beneath the ship for both systems in simple situations. Difficulties in interpretation of the interferometer record with more complex slopes render the use of the electronically-scanned system advisable in such circumstances.

The Decca method of conducting speed trials and its considerable advantages over other methods have been described in this Journal and elsewhere. Decca is nowadays widely and successfully used for acceptance trials (speed, calibration of (Sal) log, turning circles and (crash) stopways) by Royal and Merchant Navy ships of Britain, the Netherlands, France, Denmark, Sweden and probably of other nations.

The results of speed trials of Netherlands Merchant Navy ships are given in the present paper. Those of the equally successful turning circles and stopways are not given here, since their numerical results would be of little interest to the reader; a pictorial representation—among other things allowing for a comparison between various interesting turning habits of different ships—would be more suitable, but would occupy too much space.

Short Description of the Method. Each speed determination consists of 4 to 6 runs, alternatively in opposite directions. In each run 20 photographs of the decometers are taken at 30-second (or, for very high speeds of 30 kt. or more, 15-second) intervals and one single run therefore takes 9½ minutes. Specially calibrated Decca receivers are used to ensure a much higher accuracy than is achieved (and needed) for normal navigational purposes.

With the aid of special, large-scale, Decca charts or special tables, all these decometer readings—80 in 4 runs or 120 in 6 runs—are converted into rectangular coordinates X, Y.

When reading articles on radar-navigation in various nautical magazines, one is constantly struck by the difficulty of comparing the different targets on the screen and of interpreting the changes that take place in the right way. So plotting becomes still more necessary in order to get: (a) the nearest approach, and (b) the present course and speed of other vessels, before we can manœuvre to avoid collision, especially in thick weather. When plotting, comparisons are hampered by the fact that the echoes must be recorded on a piece of paper or on the reflection-plotter, and that the different distances, courses and speeds must be measured very accurately. And nearly all these tasks depend upon the skill and accuracy of one man.

Following the publication of E. S. Calvert's Manœuvres to Ensure the Avoidance of Collision (Journal, 13, 127) a number of people closely concerned with the problem of collision at sea were invited to comment on Mr. Calvert's ideas. This comment was published in Vol. 13, Nos. 3 and 4 (pp. 350–352 and 455–464). Mr. Calvert here replies to some of the criticisms. The paper he refers to as his latest will be published in the October number of the Journal.

to Captain H. D. Harries

Without a more precise definition of ‘open sea’ and ‘crowded waters', I doubt if any figure for the proportion of collisions in each has much meaning. (The same kind of difficulty arises in defining a ‘near miss’.) More precise figures might possibly be obtained by giving the proportion of collisions which occur within so many miles of the mouth of a channel, a channel being defined as an area where local or ‘edge’ rules are in force, or where the traffic is controlled. However, I doubt whether the use which could be made of such statistics would justify the trouble of collecting them. In my latest paper, i.e. the one to be published in the next issue of the Journal, I have given what I believe to be the only possible solution to the crowded-water situation, in so far as rules can provide a solution.