Updated 17-Oct-2007

This document is a modern transcription of Cabinet Office record CAB 16/147. It highlights considerations and factors concerning the vulnerability of large warships to attack by aircraft. It was transcribed by David Chessum on behalf of the the Royal Navy Flag Officers 1904-1945 website. The original file is held at the The National Archives at Kew, London. This Crown Copyrighted material is reproduced here by kind permission of The National Archives.



(CAB 16/147)


Printed for the Committee of Imperial Defence. July 1936.



(Also Paper No. V.C.S. 27.)



Report by a Sub-Committee

2 Whitehall Gardens, S.W. 1,
July 30, 1936


Terms of Reference.

1. We were appointed by the Prime Minister -

“to consider the experiments that have taken place or are proposed in connection with the defence against aircraft and the vulnerability from the air of capital ships.”

2. A strict interpretation of these terms of reference would narrow the field of enquiry to the purely technical and material aspects of the question. We have considered and discussed other matters arising fro the terms of reference which were evidently relevant to the general object of the enquiry. It would, for example, be unsatisfactory to deal with the vulnerability of capital ships to air attack without taking into account the extent to which they may become liable to this form attack; at the same time we have tried to confine within necessary limits our survey of these wider questions, which indeed, involve the functions of the Navy in War, possible theatres of war, and the strategical use of Air Forces. These problems are beyond our terms of reference and, if they were to be adequately considered, would need an enquiry of a different nature, such as was carried out by Mr. Bonar Law’s Committee in 1921.

Procedure of the Inquiry.

3. We have been greatly in our Inquiry by our Expert Advisers, the Chief of the Naval Staff and the Chief of the Air Staff, who have supplied us freely with all the information at their disposal, much of it of an extremely secret nature, and have throughout co-operated with us, and with each other, in presenting us with all the necessary data and, whenever possible, with agreed views. We are glad to record that agreement exists between them on many of the matters which have been considered.

4. We felt, however, that we should be right to get also opinions of others, and we therefore invited a number of gentlemen to give us the benefit of their views, either by attendance at a meeting or by the submission of Memoranda. A list of those who accepted this invitation is attached as Appendix A. It will be observed that all of them have had personal experience of the matters under enquiry. We invited, however, some of them to give us the benefit of their views not so much because we had reason to think that their experience was very great or up to date but because they had made speeches or written articles in the Press expressing confident opinions as to the liability of the capital ship to destroyed or seriously damaged by air attack. For the most part the witnesses under this description recognised that they had not the information which is naturally only possessed by the Service Departments, but they gave us all the help at their disposal and we are grateful to them for their contribution to our enquiry. Others who responded to our request have special technical qualifications, and in one or two cases have held high positions in one or other of the Services. To them also we are under an obligation for the trouble they have taken to answer our enquiries.

5. We also considered that we should ascertain whether certain shipbuilding firms of high standing had any views which might be of assistance to us. We accordingly approached representatives of Cammell Laird, Vickers and John Brown & Co. with an enquiry as to whether they were in possession of any information not likely to be available in the Service Departments, particularly as to whether the designs of British battleships are fully up to date and in no way inferior to those of foreign countries. These firms have all replied to the effect that, owing to the fact that they have had no recent experience of building battleships, they have no special information of their own not in possession of the Admiralty. At the same time two of the firms paid tribute to the knowledge and efficiency of the Royal Corps of Naval Constructors.

6. With these preliminary observations we pass to a survey of the questions raised by our Terms of Reference.


Nature of experiments carried out.

7. During the war bombing attacks on ships were made with small bombs which were comparatively ineffective. The protection which necessarily existed in warships to withstand shell attack was more than sufficient to withstand bombing attack. Immediately after the war the Admiralty anticipated that there would be steady progress not only in the size of the bomb, but also in the methods of carrying out attacks. Bombing trials on actual ships being generally impracticable on the score of expense, it was decided to attack the problem in a similar way to that employed for the problem of protection by armour against shell attack.

8. The value of armour against shell attack is determined at the proof ranges by firing the shell from a gun placed close to the plate. The shell is made to strike the plate at the appropriate velocity and angle at which it would hit when fired fro a gun at the long range of a fleet action. The first trials were designed to test whether a bomb could similarly be projected from a gun or howitzer at the appropriate velocity and angle that it would have if dropped from a height by aircraft. These trials, which commenced in 1920, showed that the arrival of bombs on the deck could be simulated in this manner. The next step was to ascertain the penetrating capacities of various bombs and the thicknesses of deck armour necessary to resist them.

9. These trials commenced in 1921 and have been continued to the present day. The types of bombs on which experiments have been carried out include the light case and the semi-armour-piercing bomb, both of which carry a comparatively large explosive charge for their weight, and the armour-piercing bomb which, being of stronger construction, has a greater penetration but a smaller bursting charge than either of the former. We have had laid before us detailed descriptions of these numerous experiments, a list of which is given in Appendix B. We do not think it necessary, or desirable, to record them in detail, but we note that the experiments have been carried out with a large number of different bombs up to 2,000 lb. in weight. We are satisfied that the Admiralty, as a result, have been able to make detailed calculations as to the necessary thicknesses of deck to keep out these various kinds and weights of bombs, if dropped from various heights.

10. This process of laboratory experiments has been carried further by practical tests. Before starting the reconditioning of individual capital ships, two targets were erected to represent an existing battle cruiser and a battleship respectively, and further experiments were carried out on these in order to determine the lines on which improved protection could best be given. As ships become due to be modernised the lessons learned from previous experiments have been taken into account and the necessary modifications introduced.

11. But the problem does not end with the determining of the probably thickness of armour which a bomb can penetrate. There are other matters to be considered, such as the damage which will be effected by a bomb after it has effected the maximum penetration of which it is capable, which in the case of instantaneously fused light case bombs is on the weather deck. The Admiralty and Air Ministry have carried out experiments in all these matters continuously and methodically. Having determined as a result of firing or dropping unfused bombs how far, under given conditions of weight and height, they can penetrate into a ship, they have subsequently placed similar bombs in the positions so reached, in vessels to be scrapped, and have detonated the bombs I those positions in order to gauge their effect. They have thereby acquired valuable information as to the best methods of construction for reducing the damage done by the individual bomb, and localizing its effects, and there is agreement between the Admiralty and the Air Ministry as to the defence properties of the new capital ships, in their constructional aspects, against bombs of various weights.

Effect of instantaneously fused bombs.

12. In order to judge the results of instantaneously fused bombs bursting on the deck, bombs of this nature have been laid in various positions on deck and detonated, and the damage to superstructure which was to be expected from explosive charges of the various weights could be observed. The Admiralty have sated in this connection that the effect so far as the fighting efficiency of the ship is concerned was small, but we think that further trials should be carried out with a view to determining the possible effect of this kind of attack on anti-aircraft equipments and their crews, since these will necessarily be more exposed that the main armament of the ship, the control stations, engine rooms and magazines. The Admiralty and the Air Ministry have agreed to carry out experiments of this nature.

The “near miss.”

13. It has been stated in various quarters that, even if a bomb fails to hit a ship directly, the explosive effect of its detonation in the water close to a ship’s side will have serious results. The Admiralty have carried out experiments also in this regard in order to test the effect of the “near miss.” It has been found that the “bulge” with which capital ships are provided to protect them against torpedo attack provides a large degree of security also against the “near miss.” In fact, the Admiralty are of the opinion that the results of the experiments confirm that the “near miss” is not a vital menace to a modern battleship. It has been represented to us that the portions of the ship most likely to be damaged are the propellers and the condenser inlets, but this is not the view of the Admiralty.

Experiments with a moving ship.

14. It has been suggested to us that it would be useful to carry out a bombing experiment against a moving ship to ascertain whether a hit under the conditions of a moving ship would cause greater damage than in the case of a stationary ship. We are not in a position to judge the necessity, or the practicability, of such a test, but we call the attention of the Admiralty to the suggestion.

The American Experiments.

15. A great deal of publicity has been accorded to the American experiments carried out between 1922 and 1925 against three warships which had been scrapped; some publicists, in the absence of knowledge of the series of experiments conducted here, rely very largely on these American experiments to argue their case that the day of the capital ship is over. We have received information regarding these American experiments, and their results, from the account prepared by the American Navy and subsequently published, and we are clear that these trials, which were in many respects peculiar to themselves, do not by any means constitute the last word on the subject. On the contrary, the Admiralty, taking into account the nature and results of those early experiments, have since gone into the matter much further and much more thoroughly. The American experiments were a valuable beginning and useful guide and have been used as such. We attach as Appendix C a report on these experiments which was laid before us. It may be noted in passing that, although the Americans are, of course, in possession of all the information regarding their own trials, they have not thereby been induced to cease building capital ships, as some of the critics here, on less complete information, would wish us to do.


16. A list indicating the types of experiments which have been carried out is given in Appendix B. These experiments have been carried out by the Admiralty and Air Ministry in close co-operation. We are satisfied that the Admiralty realise how important it is, if only in their own interests, to ascertain as accurately as possible the amount of damage which may be incurred by aircraft attack and we consider they have taken all proper steps to that end. We are assured that the lessons which have been learned will be applied especially in the construction of the two new capital ships and that lessons arising out of later experiments, which will continue as a matter of course, will be embodied as found necessary. It may be that the contact between the two Departments is capable of being further developed, and we trust that no effort will be spared to do so. It is obvious that there are two aspects, the defensive side which affects the Admiralty, the offensive side with affects both the Admiralty and the Air Ministry, and from every experiment there will be lessons to be learned of value to each of them in these two different aspects.

17. Certain experiments have already been agreed upon to take place in the autumn upon particular aspects of the problem which will appear late in the Report. We recommend the continuance of experiments in this and in other matters which will be referred to subsequently, and we think that the provision of funds and material should not be allowed to stand in the way.


Forms of Attack.

18. The various methods of attacking ships from the air are as follows:-

Level Bombing is attack delivered from a high altitude with the assistance of a precise bomb-sighting apparatus. It is the only method by which any considerable penetration of deck armour can be effected. It can, of course, only be used when the cloud base is high enough to allow the aeroplane to sight in the target at the requisite altitude. In order to get penetration, even with a 2,000 lb. armour-piercing bomb, of the deck armour of one of our new capital ships an aeroplane would have to fly at a height of about 10,000 feet.

Dive Bombing is a method of attack that can be used when the cloud base is too low for level bombing. The bombs are dropped from a low altitude, by judgment, from a steeply diving aeroplane. There is insufficient striking velocity to obtain penetration of armour and (so far as capital ships are concerned) the effects will be limited to the superstructure and to anti-aircraft guns and crews on the upper deck. This method of attack may be used in combination with other methods of attack in order to give cover to them.

Torpedo Attack consists of dropping torpedoes from aircraft at a very low altitude fairly close abeam the ship. Its drawback is that it strikes the most heavily protected part of the ship, the bulged and armoured sides. It also exposes the aircraft to the easiest shot from the anti-aircraft guns. Against heavily armoured ships it is not considered by the Staffs to be as good a method of attack as other methods.

Level Bombing.

19. The Staffs are agreed that, under peace conditions at the present time with average personnel using the automatic bomb sight, from 10,000 feet at a speed of 100 knots, half the bombs dropped would fall inside a circle of 100 yards radius, the centre of the circle being the point of air. This represents 11 per cent of hits against a ship the size of the Centurion, a ship of 23,000 tons. There are, however a number of qualifications and external factors which may well affect deductions drawn from the present information.

20. Apart from the effects of anti-aircraft fire which are considered later, there are other factors, such as the effect of avoiding action on the part of the target, the loss of accuracy due to increased height, the effect of the use of faster aircraft, and errors which may arise fro the fatigue of personnel after long flights. The extent to which these factors may affect the probable percentage of hits given in the preceding paragraph is at present a matter of opinion, but we do not think it need so remain. We recommend that experiments, jointly agreed upon in advance and jointly analysed, should be carried out to determine the facts. The Staffs are agreed that the methods of carrying out and analysing practices are capable of improvement and that practices will be carried out in future not only for training purposes, but to reproduce as accurately as possible the tactics that would be employed in actual practice; analyses will then show the consequent variations in accuracy to be expected.

Dive Bombing.

21. Under peace conditions it has been found that half the bombs dropped by trained personnel, dive-bombing without sights from 1,500 – 2,000 feet, will fall inside a circle of 50 yards radius, the centre of the circle being the point of aim. This corresponds to 27 per cent. of hits on Centurion. This has been agreed by the Service Departments concerned, but further experiments are needed to determine whether these figures will be affected by height of release (to allow for increase in diving speed), decrease in angle of dive, and possible errors after long flights.

22. These questions appear to be determinable after further experiments in which the Staffs have agreed to co-operate.

The Effect of Anti-Aircraft Fire on Accuracy of Bombing.

23. In the next section of this Report we deal with the probable results of anti-aircraft fire on attacking aircraft. But, apart fro the number of casualties which such fire may inflict, it seems to us that there is bound to be some, and perhaps considerable, loss in bombing accuracy on the part of aircraft which have escaped material damage. There may be a physical effect caused by disturbance on the aircraft due to close shell bursts, which may deflect the aim, and also personnel who are heavy fire may well be affected in their determination to carry out an effective attack.

24. We think the two points of view on this subject, which represent opinions held not only in the Service Departments most concerned, but by others as well, should be set forth.

25. On the one hand it is held by the Air Ministry that the inaccuracy to aim caused by disturbance of the aircraft by close shell bursts is not likely to be appreciable. There are records to show that the oscillations of aircraft are substantially less under automatic than under manual control under the same conditions of weather. The psychological effect of anti-aircraft fire will no doubt vary at different stages of the campaign and with the types of personnel available, but the only safe assumption is that the enemy would use the best trained and most intrepid of his airmen.

26. The other point of view, which is taken by the Admiralty is that non-vital gunfire will have a physical effect not less serious than the “bumpiest” and most difficult weather conditions. Experience shows that this is very considerable, and it is not considered that automatic controls will modify the effect, since the reactions of a human pilot to such sudden (as opposed to gradual) variations are more immediate than the reactions of any existing automatic control. It is therefore expected that effective gunfire will considerably increase errors. In addition, the psychological effect, which it is not possible to asses quantitatively, is likely to lead to an increase in height of release which will result in an additional loss of accuracy.

27. It is impossible for us to give judgment as to which of these opinions is likely to be correct. There is no evidence to guide us, or others, as to the possible physical effects of non-vital anti-aircraft fire. The psychological question will remain a matter of opinion until it is tested out in war. However high the standard of training of attacking pilots, and it would be unwise to assume anything but a high standard, the excitement of action, the deterrent effect of close shell bursts, the fact that the airman is carrying out an attack on the gun itself and not avoiding the gun in order to attack some objective beyond, all these are factors which would seem certain to affect statistics arrived at under peace conditions.


28. In no part of our enquiry have we found more difficulty than in the assessment of the probable results of the anti-aircraft fire by which ships defend themselves against bombing attack. It is, of course, impossible to simulate in time of peace all the conditions of war, and experiments carried out under peace conditions afford only a guide, form which it is easy to draw false deductions. The first type of anti-aircraft fire which we now discuss is that directed against “level bombing” attack, as distinct fro “dive’ bombing and torpedo attack.

Against Level Bombing Attack.

29. Naval anti-aircraft training against “level bombing” is normally carried out against a “sleeve target” towed by an aeroplane, and against the “Queen Bee” target, a small aircraft without a pilot and controlled by wireless. The “sleeve” target is comparatively slow, and the towing aeroplane has to fly, without any considerable alteration of course, or height, across the arc of the gun. The Queen Bee target, a comparatively recent development, is capable of manoeuvre to simulate the kinds of avoiding action which an attacking pilot might take. It, however, is also relatively slow; its maximum speed being about 90 m.p.h., whereas modern aircraft can fly at speeds up to 300 m.p.h., though they may require to slow down below that speed in order to get accuracy of aim. The difference in speeds between the target used in peace and the target to be engaged in war is important chiefly for the following reasons:-

(i) Crews trained in peace may need to gain some experience to accustom themselves to bring fire to bear on faster-moving objects in war. Crews are, however, being trained to deal with faster targets and as the Queen Bee target develops, its speeds should approximate more closely to those of service aircraft. Training is also carried out using service aircraft at full speed, but without actually firing.
(ii) The aircraft, flying faster, imparts a greater initial forward velocity to the bombs, the result is that it can release the bomb at a greater distance from the ship and thereby force the gun to try and hit it earlier and at longer range.

30. Nor is actual speed the only difference between peace time and war targets. The Queen Bee is expensive, and has, very naturally, been manoeuvred rather with a view to give training and to avoid damage than to simulate closely the probably action of a determined attacker. We consider that, even at the risk of greater damage than heretofore it is important that Queen Bee targets, improved pattern of which will have a better performance, should on occasions be manoeuvred as they would if carrying out an attack in war, and at the probable height at which aircraft would seek to attack. More accurate information will then be obtained as to the probability of hitting, and more realistic training will be afforded to the anti-aircraft gun and director crews.

31. We have been informed that it is part of naval anti-aircraft fire discipline to withhold fire for effect until the aircraft is making the approach to bomb, since for a short time before the moment of release the aircraft is bound to fly on a straight, or slightly curvilinear course and cannot take avoiding action. In assessing probable results the length of that time is obviously a very important question. The Air Ministry are of the opinion that with the new bomb sights which will soon be available the time of straight flight can be reduced to some 20 seconds; the Admiralty question whether that will be possible, and estimate that the period, including slightly curvilinear flight, during which effective fire can be brought to bear, will be about 90 seconds, and do not consider that even with the new sights the period of straight flight can be reduced to so low as 20 seconds.

32. We no not think that there should be any great difficulty, once the new sights are available, in testing the matter out, and we recommend that this should be done. That will eliminate at least one unknown quantity for the future.

33. There are other difficulties in drawing deductions from peacetime practices such as the fact that the speed of the Queen Bee is a known factor, whereas an attacking pilot can I practice alter his speed and so upset the gunner’s previous calculations; as also the fact that practices are necessarily pre-arranged and the element of surprise is therefore absent. It the heat and excitement of action, and especially if a naval action is taking place simultaneously, there will certainly be some loss of accuracy in anti-aircraft fire as compared with the results of practice in the less trying conditions of peace, when the gun is not being made the object of attack.

34. The Admiralty have provided us with figures, which cannot be more than estimates, as to the probable number of hits which anti-aircraft guns will be able to effect against attacking aircraft, judged from the results hitherto obtained and taking into account the various uncertain factors which are referred to above. These figures indicate a considerable probability of hits.

35. The Air Ministry, whilst agreeing with the method of estimating casualties that has been adopted, do not consider that it is sound to assume that the results obtainable against the Queen Bee targets can be assimilated to those likely to be obtained against actual bombing aircraft. We are disposed to agree that the unreal conditions under which anti-aircraft practice has hitherto taken place, although largely unavoidable, makes it difficult, if not impossible, to accept any figures until much ore information is available. It will be realised that any figures arrived at from peace-time practice will always be largely theoretical; only under war conditions, impossible to simulate in peace, can there be a completely satisfying test. But the difference between the two points of view can, we think, be considerably narrowed down by making, as we recommend, peace-time practice assimilate ore closely to war condition. To this end we think the Admiralty and the Air Ministry, for their joint benefits, should work in very close touch. The Staffs have agreed that future practices designed to provide data as to the efficacy of anti-aircraft fire should be carried out against Queen Bee targets controlled to simulate as closely as possible that tactics of the type of attack the are intended to represent, and to consult together as to what these tactics should be.

36. One experiment has been suggested to us as being capable of providing some provisional data until such time as more information is obtained as the result of developments in training and material. It is that a modern high performance bomber should be flown over the existing ballistic apparatus at Orfordness, where a mobile anti-aircraft gun with the necessary range-finding and predicting apparatus should be installed. The ballistic apparatus is capable of accurately recording the position in space of an aeroplane at any moment of time, and after the procedure of laying and firing is carried out, a close estimate of the position in space of a shell burst at the same instant could be made. The experiment, it is claimed, would give some indication of the capacity of anti-aircraft guns to deal with modern high performance aeroplanes, if that claim is well-founded it will obviously be very valuable in checking data at present in possession. We as a Committee have not the technical knowledge to give an opinion as to the possibilities and usefulness of an experiment on these lines, be we commend the suggestion as one which should certainly be pursued by those concerned.

Against dive bombing and torpedo attack.

37. Anti-aircraft fire to meet dive bombing and torpedo attack requires separate consideration. The Admiralty have developed special weapons to deal with this form of attack, the eight-barrelled Mark M. Pom-Pom firing a 2-lb. shell and the four-barrelled ½-inch machine gun. In addition, the barrage fire long-range guns comprising the anti-ship armament of the vessel would be employed against short-range air attack.

38. Trials have been carried out over a considerable period, but they are necessarily based on photographic analysis, as it is only quite recently that it has been found technically possible for pilotless aircraft to execute diving attacks.

39. We have been provided by the Admiralty with estimates of the number of hits likely to be obtained by anti-aircraft weapons against short-range attacks of this kind. They are derived from photographic analysis of the accuracy of gun-laying against a large number of simulated attacks carried out by actual aircraft. Whilst the Naval and Air Staffs are again agreed that the method of analysis used for the estimation of casualties is a reasonable one, they are not agreed tat the results are applicable to the conditions to be expected in the Fleet in practice. Here, again, we see difficulty in accepting as firm figures deductions made from tests which took place under conditions necessarily very remote from those of war, and we doubt if any real advantage would be gained in recording this Report figures so obtained.

40. It will probably always be extremely difficult in time of peace to simulate realistic attacks of these natures; the difficulties are even greater than in the case of the level bombing target. Whilst we invite the two Departments to co-operate as far as possible, we desire to record the impression which the evidence has left on our minds, that the steps taken by the Admiralty to engage short-range attack will result in a very powerful concentration of fire which must, though to what precise extent it is impossible to say, affect the accuracy and the morale of the attacker.

41. There are, however, various means by which attacking aircraft would attempt to gain a measure of surprise and so to reduce the likelihood of casualties, such as attacking from the direction of the sun or from different directions at the same time, night attacks assisted by flares, or attacks covered by smoke screens. These are tactical methods of which the Services are fully aware, and we do not doubt that they will continue to experiment, and to co-operate, in their practice.

Increase in volume of anti-aircraft fire.

42. Whatever may be the precise figure to be allotted to the accuracy of the anti-aircraft fire, there can be no doubt as to the great volume of anti-aircraft fire which a number of ships can now develop, and the number of guns can still further be increased if found necessary. Those who rely for their criticisms of anti-aircraft fire merely on experience in the lat war (we do not include the Air Ministry in the category) are working on a basis quite unreliable. Whilst, apart from increased speed and power of manoeuvre, aircraft remain as vulnerable as before, and the target they offer increases with their size, there have been great developments in numbers and calibres of guns, weight of shell, accuracy and rapidity of fire.

Ammunition Supply.

43. The rates of fire of anti-aircraft guns have been much increased of recent years, partly by improved training and partly by the adoption of new weapons. When air attack on ship was only probable when opposing fleets were at sea, i.e., from carrier-borne aircraft, the quantity of ammunition to be carried was relatively small. But if, as is quite possible, our fleet, whether at sea or in harbour, is in future to be ready for heavy attack from shore based aircraft, there is no doubt that the amounts of ammunition to be carried on board will have to be much greater than in the past. We have been informed that the ammunition at present carried for the long-range gun is 250 rounds per gun, which will suffice to engage fourteen attacks, allowing 90 seconds for the period of engagement of each attack. Similarly, the Mark M. Pom-pom (provided with 1,800 rounds a barrel) can engage 85 attacks, allowing 12 seconds for the period of engagement of each attack. In harbour, with facilities for rapid replenishment, no great anxiety need be felt, but we think that the question may now require re-examination from the point of view of ships operating for considerable periods in narrow waters within range of shore-based attack.


New dangers met by new methods of defence.

44. It is, of course, no new matter to the Admiralty to have to consider the vulnerability of ships in the light of some new weapon and to adopt new measures for their defence and protection. The last 40 years have seen the advent of the submarine, the torpedo and the mine. Each of these have had great effects on naval warfare, but new forms of attack have always, sooner or later, produced new forms of defence. The question is whether attack from the air, unlike the others, is incapable of being met. The work of the Committee on Air Defence Research is evidence of the efforts which are being made, in another sphere, to create new forms of defence against the menace of air attack.

45. It is one of the main characteristics of a battleship that she is built to resist attack both above and below water fro guns, torpedoes and mines. Against air attack rearrangement and improvement of her protection is needed, the Admiralty make no claim that a ship, however designed, can be invulnerable in all circumstances to every form of attack. But they do think that the capital ship of the future can be so designed as to distribution of her armour on decks and sides, and as to interior subdivision, that she will not be the subject of fatal damage from the air. The effect of hits by aerial bombs is analogous to plunging long range fire from guns. The Admiralty view is that there is no reason why the ship cannot be designed to meet air attack just as in the past she has been designed to meet other dangers.

46. These views, of course, are not those of the critics of the capital ship, who maintain that a concentrated air attack on ships at sea or in harbour will be so effective that they cannot survive. It is a point of view that has yet to be tested in practice.

47. Whatever the future may have in store the truth at present probably lies between these two extremes. Given the greatest possible measure of protection on the constructional side against air attack much, of course, will depend on the scale of attack which can be brought to bear. In that connection the size of the air force of a potential enemy and the distances of his aerodromes from the areas in which our Fleet will operate and fro the bases that Fleet will use, are two important factors.

Strategical questions.

48. Here we meet the strategical questions which though they are, strictly speaking, outside our terms of reference, have a bearing on the general object of our enquiry. Putting on one side the question of air attack from carriers which are themselves far more vulnerable to air attack than battleships, it is clear that if the areas in which our ships may operate are so far form enemy aerodromes as to make air attack impossible, the vulnerability of ships does not arise. At the other end of the scale, for example is the heavy scale of air attack which could be launched against our Fleet in Malta by Italian aircraft based on Sicily. Whether or not an enemy will decide to use his air power against our naval forces is a matter which cannot be determined in advance, but the fact that we are dependent for our existence on our sea communications behoves us, more than any other power, to anticipate that kind of attack.

49. Although our Fleet may have to operate in comparatively narrow waters such s the Mediterranean and the North Sea, it may also have to do so in the larger oceans, the Atlantic and the Pacific, much of whose spaces are at great distances from the air bases of any potential enemy. But that is only one aspect of the matter, for those oceans, of which the free passage is essential to our national existence, cannot at present be covered by our own land-based air forces. It would not be impossible for the capital ships of a hypothetical enemy, in the absence of a powerful British Fleet, to place themselves, without interference, across our trade routes in such a way that we should be powerless to do anything unless we could send a superior naval force, supported by capital ships, to drive the enemy off.

50. There is a great part of the oceans of the world where this applies. Even where a space of sea is actually within range of aircraft it by no means follows that ships therein will necessarily be subject to attack. The extreme upholders of “The Air” view will centre one leg of pair of dividers on an air base, and with the presumed radius of aircraft will mark out an area into which, they claim, a ship cannot enter without detection and destruction. That is a one-sided and specious representation of the case; the matter is by no means so simple. The hours of darkness, weather, the special training needed for over-sea navigation an the great areas which aircraft will have to reconnoitre in order to find their enemy are all factors which in practice will have great influence.

The substitution of aircraft for the capital ship.

51. It has been argued that the functions of the capital ship can at least equally well be carried out by aircraft. The defence of our territory and of our trade routes are, of course, two of the most important functions of the Navy. It has been put to use, though not by the Air Staff, that this can be done by our light naval forces, that our battleships are unnecessary as cover to those light forces since if the enemy possesses battleships they will be destroyed by our own air forces, properly distributed in advance to deal with such a situation.

52. We do not consider that the conditions prevailing to-day or likely to prevail in the near future justify these opinions, nor do we consider that there is sufficient evidence to affirm that aircraft can perform the role of our own capital ships by holding in check, and if possible destroying, enemy capital ships. We need ships equal in fighting power to those of the enemy, for there are large areas of ocean which are out of range of service aircraft based on British territory. Of our merchant ships sunk by surface craft during 1914-1918 well over half were more than 500 miles from British territory. Although the radius of action of aircraft will increase in the future, the provision all over the world of sufficient aircraft to deal with all probable contingencies in the defence of trade and territory would necessitate something approaching a two-Power standard in the air, with an immense provision of facilities of all kind.

Possibility of attack on harbours and bases.

53. We have discussed the possibility of heavy air attacks carried out against our Fleet in harbours within range of an enemy’s shore-based aircraft. The Fleet in such circumstances is able to bring to bear its maximum concentration of anti-aircraft fire and the recent naval concentration in Alexandria has made the occasion for a serious consideration of the subject from which useful lessons can be learned. Nevertheless, the prospect of heavy air attacks under these conditions is disquieting and with the ever increasing potentialities of aircraft may necessitate the use of harbours and anchorages other than those which were suitable in the last war. In the same connection we would draw attention to the increasing liability to air attack of naval ports and docking facilities. The latter are few in number, not easy to defend, and if damaged cannot rapidly be replaced. Yet a capital ship must have full facilities of maintenance and repair, and although our enquiry is concerned with the vulnerability of the ships themselves we think it right to call attention to the necessity of providing a full scale of defence for the bases on which capital ships will have to rely in a future war.

Air Power and naval vessels other than Capital Ships.

54. An interesting feature of the evidence of those who hold that the day of the capital ship is doomed by reason of its vulnerability to air attack is that they do not carry the argument to its logical conclusion. The capital ship is much more strongly constructed and much better equipped with anti-aircraft guns than any other ship. Cruisers and light craft are necessarily far more vulnerable; indeed, it can be said that the capital ship is the one remaining surface craft which, if hit severely by an air bomb, is not liable to be sunk. Yet no partisan of the air has put forward the suggestion that all naval forces are doomed to destruction, that the day of navies is over and that air forces can and must take over the tasks for which the Navy re now responsible. Yet this is the conclusion which properly flows from the arguments which they adduced against the capital ship.

The argument of relative costs.

55. No doubt one reason which has led to the selection of the capital ship for criticism is its high cost. Unofficial persons have estimated that for the cost of one such ship a great number of aeroplanes could be built; these estimates, sometimes covering only capital cost, sometimes only maintenance and sometimes both, have varied from 100 to 1,000. Such comparisons, however made, are not, in our opinion, of any value or validity, but we thought it desirable to get at the truth on official calculations. The Admiralty and Air Ministry have collaborated in this investigation and have given us an agreed figure of 43 twin-engined medium bombers as the nearest approximation possible to the equivalent in cost of one capital ship, taking into account all those overhead, maintenance, and replacement and similar charges which should be included to make an effective comparison.

56. We might mention in passing that one witness who appeared before us, and used this financial comparison as part of his argument, was asked by us to work out his own estimate of the cost of providing the number of squadrons which, according to his ideas, could replace our capital ships in their role of defence of trade and territory. The figure that he arrived at, after a sufficiently complex calculation, and omitting the cost of a large number of new aerodromes over the Empire, was the same as the capital cost of 15 capital ships.

57. The fact is that the relative costs of battleships and aeroplanes have not, in themselves, any bearing on the matter. If capital ships are essential to our security, we must have them, or go under. We are dependent, as is no other nation, on the maintenance of our overseas trade, we have an Empire sprawling over the face of the globe. We have more to lose by making a false decision in so vital a matter than has any other Power. Yet no other great Naval Power, though with less risk that we ourselves should run, proposes to do away with capital ships. Should we be the first to do so? Surely not, unless the question is settled beyond all possible doubt. We do not find that the question is so settled. It may never be settled without the test of war, but the information at present at our disposal leads us to believe that the day of the capital ship is not over, now or in the near future; to assume that it is, and to cease to build them, would lead to grave risk of disaster.

It is possible to state the matter in the simplest possible terms. The advocates of the extreme air view would with this Country to build no capital ships (other Powers still continuing to build them). If their theories turn out well founded, we have wasted money; if ill founded, we would, in putting them to the test, have lost the Empire.


58. We now state the general conclusions which we have reached on the matters in question.

The experiments that have taken place have been extensive and prolonged and the information thus obtained has enabled the Admiralty to adapt the design of capital ships in such a way as to offer greater resistance to attack from the air. Further experiments are, however, essential. Indeed, we are definitely of opinion that, as developments in the power and range of aircraft take place, there should be no cessation in making experiments which should reproduce as far as possible war condition. To this end the fullest collaboration between the Admiralty and Air Ministry should take place. We do not suggest that anything has been lacking in ingenuity or perseverance in the experiments that have been made, but it is possible that the closer co-operation of the two Services in an attempt to make the experiments as realistic as possible would be fruitful.

59. We have given full consideration to the information already available as a result of the various experiments and it is plain to us that capital ships cannot be constructed so as to be indestructible by bombing from the air. This would probably be true even if factors of speed, armament and weight were disregarded in the design of the ship so that the thickness of the defensive armour could be indefinitely increased. If this fact, which we regard as being beyond the possibility of doubt, is kept in mind, it throws light on some of the more extravagant criticisms of the policy of building capital ships. The critics have written and spoken as if the issue was between those who deny and those who assert the vulnerability of capital ships to air attack. That is not the real question. In circumstances favourable to an attack from the air which could be driven home by a large and powerful force the most heavily armoured capital ship could no doubt be destroyed, or at least seriously crippled. It is impossible to do more than speculate as to the chances of such a combination of circumstances or as to the degree of success likely to be attained by the attacking force. The real question that arises on the assumption that capital ships are indispensable is whether their design is such as to secure the maximum of immunity from air attack. We have, as above stated, made enquiries in quarters other than the Admiralty as to whether the capital ships of His Majesty’s Navy are so designed and we are satisfied that everything is done that experience could suggest, or skill and money could provide.

60. Our recommendations for the future, as contained in the various relevant sections of the Report, are as follows:-

(1) Further trials should be carried out with a view to determining the possible effect of instantaneously fused bombs on anti-aircraft equipments and their crews (paragraph 12).
(2) We call the attention of the Admiralty to the suggestion to carry out experiments against a moving ship in order to ascertain whether any special type of damage may be caused to a moving ship as compared with a stationary one (paragraph 14)
(3) The closest touch should be maintained between the Admiralty and the Air Ministry in all experiments connected with the effect of bombs on ship construction (paragraph 16)
(4) In the experiments to take place in the Autumn, and in other trials in this connection, we consider that the provision of funds and material should not be allowed to stand in the way (paragraph 17).
(5) We recommend that further experiments should be carried out, jointly agreed upon and jointly analysed, to determine certain factors, at present unresolved, which affect the accuracy of level bombing (paragraph 20).
(6) We note that the Staffs have agreed to co-operate in further trials to determine the accuracy of dive bombing (paragraph 22)
(7) Queen Bee targets should on occasions be manoeuvred in the manner and at the height at which air attack is likely to be carried out in war (paragraph 30), and the Staffs have agreed to co-operate in the matter (paragraph 35).
(8) When the new bomb sights are available, data should be obtained as to the length of time, necessary for sight-setting, that an aircraft must fly on a straight course before dropping a properly aimed bomb (paragraph 32).
(9) We commend to the notice of the interested departments a suggestion for a mechanical and ballistic experiment to check the accuracy of anti-aircraft fire (paragraph 36).
(10) We have no doubt that the Services will continue to experiment, and to co-operate, in the various tactical methods by which attacking aircraft will seek to gain surprise (paragraph 41).
(11) We suggest that the question of ammunition supply and replenishment should be re-examined from the point of view of ships operating in narrow waters within range of shore-based attack (paragraph 43).
(12) We call attention to the necessity for adequate air defence of our naval ports and docking facilities, on which our capital ships rely (paragraph 53).

(Signed) T.W.H. INSKIP (Chairman).

(Signed) H.R. POWNALL
Secretary to the Sub-Committee.

2 Whitehall Gardens, S.W.1,
July 30, 1936.

–Page 15–



List of Witnesses who attended before the Committee or who submitted Memoranda.

Air Commodore J.A. Chamier, C.B., C.M.G., O.B.E., D.S.O.
Air Commodore L.E.O. Charlton, C.B., C.M.G., D.S.O.
Air Commodore P.F.M. Fellowes, D.S.O.
Admiral Sir William Fisher, G.C.B., G.C.V.O., C.V.O.
Brigadier-Gneral P.R.C. Groves , C.B., C.M.G., D.S.O.
Captain Norman Macmillan, M.C., A.F.C.
Marshal of the Royal Air Force Sir John Salmond, G.C.B., C.M.G., C.V.O.. D.S.O.
Lord Strabolgi.
Rear-Admiral Sir Murray Sueter, C.B., M.P.
Sir Eustace Tennyson-D’Eyncourt, K.C.B, C.B., LL.D.
Marshal of the Royal Air Force Lord Trenchard, G.C.B., G.C.V.O., K.C.B., D.S.O.

Invitations were also sent to the following, who did not, however, desire to submit any views other than those they had already publicly expressed:-

The Right Hon. Winston S. Churchill, C.H., M.P.
Lieut.-Colonel J.T.C. Moore-Brabazon, M.C., M.P.



Numbers and Types of Bombs Used.

I. – Proof Range Experiments.



No Fired.

120 lbs.

General Purpose


250 lbs.



500 lbs.



450 lbs.



1,500 lbs.



2,000 lbs.



II. – Proof Range Firings at Targets representing Ships.



No Fired.

250 lbs.



450 lbs.



II. – Ship Trials.

(a) Placed on board and detonated at rest.



No Fired.

112 lbs.



250 lbs.

heavy-case (early design)


250 lbs.



450 lbs.



500 lbs.



520 lbs.

light-case (early design)


550 lbs.

heavy-case (early design)


1,400 lbs.



1,800 lbs.



(b) Dropped during Bombing Trials of Ships or a Target representing Ships.



No Fired.

250 lbs.

heavy-case (early design)


520 lbs.

light-case (early design)


550 lbs.

heavy-case (early design)


1,650 lbs.



250 and

500 lbs. Modern semi-armour piercing


(c) Near Misses.

Weight of
Explosive Charge.



47 lbs*



260 lbs.



500 lbs.



760 lbs.



2,100 lbs.



*These are charges in bombs dropped on “Marlborough.” The bombs are already included under (b) above.



Authority – O.U. 5348 (1925).

“Aviation and the U.S. Navy.”

Target Ship New Jersey. – Laid down 1901.

No special preparation. No watertight doors in place.

1st attack. – 20 x 600 lb. bombs. 15 exploded. Damage not great – no underwater injury.
2nd attack. – 10 x 2,000 lb. bombs. No direct hits, but several near misses (one very near). Result was a 5º list after two hours.
3rd attack. – 7 x 2,000 lb. bombs. “Fell at the bow.” No damage.
4th attack. – 3 x 2,000 lb. bombs. One fell 100 feet off. One scored a hit or else fell right alongside. (One failed to explode.) Ship sank in 5 minutes.

Target Ship Virginia . – Sister ship to New Jersey.

1st attack. – 14 x 1,100 lb. bombs. “Ten were effective” (presumably near misses) “at varying distances, principally on starboard side.” One direct hit. Ship listed 10º and sank after 20 minutes.

Ostfriesland. – A German battleship built in 1911. She was not watertight before bombing commenced and had a list of two degrees, and was a foot deep in draft before the bombing began. Bombing continued for two days. During the first day 69* bombs, varying form 230 to 2,000 pounds were dropped and 13 hits were scored. Little damage was experienced. Ship increased list slightly and deepened her draft by 3 feet. No action was taken to stop leaks.

On the second day 11x1,000 lb. bombs were dropped and 3 direct his obtained. The result was no vital injury, but a lead started which caused the ship to settle 3 feet forward and 1 foot aft in 4 hours.
After these 4 hours, 6 x 2,000 lb. bombs were dropped. No hits, but 3 near misses, one very close.
Ship sank in 10 minutes, having sustained 16 direct hits from 69 bombs over a period slightly exceeding 24 hours.

*The report leaves doubt whether 69 or 52 bombs were dropped on the first day; 52 appears more probable.

Apart from the fact the planes were not fired at, had the crew been aboard it is most doubtful whether the cumulative effect of all this bombing would have sunk the ship. If the ship had been in commission, bombing would not have prevented crew from taking steps to stop leaks and pump out.

At a later date U.S.A. carried out a further series of tests of the effect of underwater explosions against the Washington, an uncompleted battleship.

Three 2,000 lb. bombs and two 400 lb. warheads were exploded at the most effective depths under water in contact or close to the ship’s bottom. Afterwards ship was inspected and rode out a gale of wind for three days; she had a 5º list. Main condensers were not affected. Had ship been in commission, the crew could easily have plugged leaks without use of pumps and kept her on an even keel and fit for service in the line of battle.

The jar of these explosions did not even put out candles stuck on deck at the end of the ship. In the end the ship remained afloat for four days, after suffering these five explosions, without repairs or pumps. She was finally sunk by 14 hits from 14-inch gunfire.

The British Admiralty have also carried out extensive trials both of above-water and under-water explosions against warships. These tests have shown that from untamped explosions the ship’s superstructure suffers surprisingly little damage. The result of tests of under-water explosions coincides with the experience of the U.S.A. trials and with estimates made from previous experience with mines and torpedoes.

Extract from Conclusions of Report of U.S.A. Special Board, 1925.

266. The battleship is the element of ultimate force in the fleet, and all other elements are contributory to the fulfillment of its function as the final arbiter in sea warfare.

* * * * * *

268. From time to time apparent threats to the supremacy of the battleship have appeared. Each has resulted in some modification of its design and in the methods of its employment in war, but its supremacy remains.

269. With the invention and development of offensive weapons have always come the counter-invention and development of defensive means and methods, so that in the end a fair balance is struck between them. The history of the gun and armour, and of the torpedo and interior subdivision, merely repeats the process by which offence always begets defence.

270. Aviation has introduced a new and highly important factor in warfare both on the land and on the sea. It was utilised on an enormous scale and with great effectiveness in land operations during the World War, but did not seriously influence sea operations. Its influence on naval warfare undoubtedly will increase in the future, but he prediction that it will assume paramount importance in sea warfare will not be realised.

* * * * *

274. Airplanes cannot occupy territory, nor can they exercise control of the sea.

* * * * *

279. The battleship of to-day, while not invulnerable to airplane attack, still possesses very efficient structural protection, as shown by the experiments on the Washington. The battleship of the future can be so designed as to distribution of her armour on decks and sides, and as to interior subdivision, that she will not be subject to fatal damage from the air. The effect of plunging long-range gun projectile hits on a ship’s deck has now become closely analogous to the effect of hits by heavy aerial bombs. By armouring the battleship’s deck with 6 to 7 inches of armour, we at one and the same time effectively meet any practicable attack from the air and also the attack by gun projectiles fired at the greatest probable battle ranges. The interior subdivision will resist any mining effect from aerial bombs. It cannot be said, therefore, that air attack has rendered the battleship obsolete.

* * * * *


 Comparative Costs of Capital Ships and Aircraft.

 Joint Memorandum by the Chief of the Naval Staff
and the Chief of the Air Staff.


At the First Meeting of the Sub-Committee on the Vulnerability of Capital Ships to Air Attack, held on the 31sth March, 1936, the Chief of Naval Staff mentioned that various statements had appeared in the Press and elsewhere to the effect that a large number of aircraft, varying in the different statements between 200 and 1,000, could be provided for the cost of one battleship, and produced a paper showing that, according to calculations made in the Admiralty, the actual number of aircraft which could be continuously maintained for a sum equivalent to that required for the building and maintenance of a battleship was in fact between 40 and 50. As, however, the Air Ministry had not had an opportunity of examining the calculations on which this conclusion was based, the Chairman expressed a wish that the problem should be reviewed again by the two Departments and that agreed figures should be submitted in due course to the Sub-Committee.

2. The attached note contains the figures which have been arrived at by the Admiralty and Air Ministry in consultation and represents the best that the two Departments can do by way of an agreed Statement.

3. The paper concludes that approximately 43 twin engine medium bomber aircraft can be maintained for the same annual expenditure as a battleship of the NELSON class. It should, however, be pointed out that this conclusion which deals with the comparative costs of maintenance is on a different footing from the original statements on which the Chief of Naval Staff commented, namely, that between 200 and 1,000 aircraft could be provided for the cost of one battleship, which, however misleading, is substantially correct.

4. We doubt if any useful deduction can be made from this equation of the costs of maintenance of a battleship with those of 43 first line aircraft. It may be mentioned, for example, that the aircraft costs take account of the workshop and immediate reserve which are provided to ensure that the 43 aircraft can be kept continuously in the air, but a similar adjustment in the case of a battleship costs is not practicable. There can in fact be no valid basis of comparison between a single battleship and a force of aircraft, and we do not consider, therefore, that this note will be of practical value to the Sub-Committee. It may, however, be useful in discounting impressions arising from certain statements in the Press and Parliament and, as suggested by the Chief of the Naval Staff in his evidence at the First Meeting (1 st Minutes, Page 25), the Sub-Committee should be sure they are correctly informed on this important point.


(Signed) ERNLE CHATFIELD. Chief of the Naval Staff.
“ E.L. ELLINGTON. Chief of the Air Staff.


11th May, 1936



We understand that the question on which we are asked to express an agreed opinion is: ‘How many first-line aircraft of medium-bomber type can be maintained for the same average annual expenditure as is entailed by maintaining one battleship?

2. We are told that no attention should be paid to strategical considerations e.g. whether a battleship an exist as a purely independent unit without its accompanying cruisers, destroyers, etc. (which might tend to suggest that the cost of a battleship cannot be separated from that of the composite Fleet unit). We also understand that we are not to consider the application of any figures which we can give to any specific problem, e.g. whether it would be cheaper to add one battleship or X aircraft to existing numbers, or whether it would be cheaper to employ battleships or aircraft on any particular service.

3. Any answer that we can give must, therefore, be of very little practical utility. Indeed the figures are necessarily a highly misleading statement of the problem at issue without an elaborate commentary from other aspects, e.g. the strategic. We understand, however, that an opinion is desired merely in order to refute the suggestions made in irresponsible quarters that the number of aircraft which can be maintained for the same annual expenditure as one battleship is between 200 and 1,000. It requires no elaborate calculations to dispose of any such suggestions. To give an actual figure which may reliably be quoted is however, an almost impossible task, for the reason that the main factors on which any assessment of comparative cost can be made, must largely be derived by guesswork, admitting of a very large percentage of error.

4. The actual cost of maintaining equipment, whether a battleship or an aircraft, may be divided into two parts, viz:-

(a) Direct Cost
(b) Indirect, or overhead, cost.

The former comprises all charges which can be related directly to the particular equipment, such as the cost of personnel directly employed for purposes of operation, annual repair, allowance for ultimate replacement, etc.; but although the particular services can be readily specified, the figure of cost to be taken for each must in many instances be speculative. The indirect costs comprise all those charges inherent in the general organisation which cannot be specifically attributed to a particular piece of equipment, such as hospitals, store organisation, general administration at head office and locally, etc. Indirect costs can only be assessed in relation to a defined establishment, either or ships or of aircraft, as a Whole, and their distribution to particular units most obviously be largely in the nature of a guess.

5. The Admiralty has estimated that the average annual cost of maintaining a battleship, including the cost of construction, is of the order of £961,500, consisting of £695,200 for direct charges (as defined above) and £266,300 for indirect charges; but these figures, in arriving at which many arbitrary and possibly erroneous assessments have had to be made, have reference only to a battleship of the NELSON Class. They also postulate that the total estimates (excluding non-effective services) based on a Fleet of stated composition will amount to 63 millions, a figure which may prove widely under estimated. The most that can be said, therefore, is that a battleship of NELSON Class forming a unit of a Fleet of definitely stated dimensions would, on the assumption that the total Navy Vote would be 63 millions, be of the order of £961,500.

6. A similar calculation in the case of aircraft is beset by even greater difficulty. Aircraft are changing in type and cost almost from day to day, air organisation is in a fluid state, and it is difficult to predict with any degree of confidence the ultimate number of aircraft on any given date; any attempt to calculate what would be the stabilised Air Vote for a stabilised Air Force can, therefore, be little better than a hopeful speculation. Under present-day conditions, an estimate even of the direct cost of a squadron of given type is very much more dependent upon assumptions than is that of a battleship; but it is in the allocation of indirect costs, which are proportionately much heavier for the Air Force than for the Navy, that the great margin of error must exist. Nevertheless, and subject to these reservations, the Air Ministry is prepared to assess the all-in cost of a squadron of 12 first-line aircraft of twin engine medium bomber type, as a unit of a complete Air Force (excluding the F.A.A. and the R.A.F. in India) of 2,072 first-line aircraft, as being of the order of £265,000 a year. This figure allows for the immediate and workshop reserves of aircraft ( as explained in the following paragraph) for the replacement of the aircraft every five years, for pay of all personnel, upkeep of the station, and for all other expenses direct and indirect so far as they can be envisaged. On this basis approximately 43 twin engine medium bomber aircraft can be maintained for the same annual expenditure as a battleship of NELSON Class.

The immediate and workshop reserve of aircraft referred to above are 75% of first-line numbers. They are provided in order that immediate replacement may be made of any first-line aircraft unserviceable for any reason whatever, so that at any moment the squadron may have its complete quota of first-line aircraft ready for service. In the estimated cost of battleship, however, nothing is included of any reserve vessel during the time the ship may be undergoing repair or refit, for it is not practicable to make any assessment of cost on this account.

The calculations take no account of the cost of upkeep and turnover of war reserves of aircraft, inasmuch as no such provision is made in respect of battleships; but war reserves of aircraft are being provide, on the scale of 150% of first-line numbers, to enable losses in war to be made good until such time as the industry can supply such deficiencies. If account is taken of the cost of keeping and turning over these war reserves, the equivalent number of first-line aircraft becomes 39 instead of 43, but it is to be remembered that there are behind these 39, reserves of 225 (or about 88 aircraft) giving a total of 127 aircraft at the outbreak of war.

3. While we should see not great objection to these figures being used for internal purposes, subject to the reservations in paragraphs 2 and 3, we would deprecate any attempt to give official utterance to any exact estimate, for, as stated, the calculations on which it is based are necessarily extremely speculative. It is, however, doubtful whether any more reasonable figure could be arrived at without very considerable research and consideration of innumerable connected problems which could only be undertaken by an expenditure of time and labour out of all proportion to the practical utility of the result.

We do not think it necessary to take into consideration such possible facts as interest on capital. We have also ignored the question of non-effective liability for personnel employed, as the calculation of a true non-effective allowance for each type of ship or aircraft would involve considerable labour, quite unjustified by the comparatively small adjustment of relativity between the figures which might be expected to result.