The struggle against the depredations of the enemy submarines during the year 1917 was two-fold; offensive in the direction of anti-submarine measures (this was partly the business of the Anti-Submarine Division of the Naval Staff and partly that of the Operations Division); defensive in the direction of protective measures for trade, whether carried in our own ships or in ships belonging to our Allies or to neutrals, this being the business of the Trade and Mercantile Movements Divisions.

Prior to the formation of the Mercantile Movements Division the whole direction of trade was in the hands of the Trade Division of the Staff.

The difficulty with which we were constantly faced in the early part of 1917, when the effective means of fighting the submarine were very largely confined to the employment of surface vessels, was that of providing a sufficient number of such vessels for offensive operations without incurring too heavy risks for our trade by the withdrawal of vessels engaged in what might be termed defensive work. There was always great doubt whether any particular offensive operation undertaken by small craft would produce any result, particularly as the numbers necessary for success were not available, whilst there was the practical certainty that withdrawal of defensive vessels would increase our losses; the situation was so serious in the spring of 1917 that we could not carry out experiments involving grave risk of considerably increased losses.

On the other hand, the sinking of one enemy submarine meant the possible saving of a considerable number of merchant ships. It was difficult to draw the line between the two classes of operations.

The desire of the Anti-Submarine Division to obtain destroyers for offensive use in hunting flotillas in the North Sea and English Channel led to continual requests being made to me to provide vessels for the purpose. I was, of course, anxious to institute offensive operations, but in the early days of 1917 we could not rely much on depth-charge attack, owing to our small stock of these charges, and my experience in the Grand Fleet had convinced me that for success in the alternative of hunting submarines for a period which would exhaust their batteries and so force them to come to the surface, a large number of destroyers was required, unless the destroyers were provided with some apparatus which would, by sound or otherwise, locate the submarine. This will be realized when the fact is recalled that a German submarine could remain submerged at slow speed for a period which would enable her to travel a distance of some 80 miles. As this distance could be covered in any direction in open waters such as the North Sea, it is obvious that only a very numerous force of destroyers steaming at high speed could cover the great area in which the submarine might come to the surface. She would, naturally, select the dark hours for emergence, as being the period of very limited range of vision for those searching for her. In confined waters such as those in the eastern portion of the English Channel the problem became simpler. Requests for destroyers constantly came from every quarter, such as the Commanders-in-Chief at Portsmouth and Devonport, the Senior Naval Officer at Gibraltar, the Vice-Admiral, Dover, the Rear-Admiral Commanding East Coast, and the Admiral at Queenstown. The vessels they wanted did not, however, exist.

Eventually, with great difficulty, a force of six destroyers was collected from various sources in the spring of 1917, and used in the Channel solely for hunting submarines; this number was really quite inadequate, and it was not long before they had to be taken for convoy work.

Evidence of the difficulty of successfully hunting submarines was often furnished by the experiences of our own vessels of this type, sometimes when hunted by the enemy, sometimes when hunted in error by our own craft. Many of our submarines went through some decidedly unpleasant experiences at the hands of our own surface vessels and occasionally at the hands of vessels belonging to our Allies. On several such occasions the submarine was frequently reported as having been sunk, whereas she had escaped.

As an example of a submarine that succeeded not only in evading destruction, but in getting at least even with the enemy, the case of one of our vessels of the "E" class, on patrol in the Heligoland Bight, may be cited. This submarine ran into a heavy anti-submarine net, and was dragged, nose first, to the bottom. After half an hour's effort, during which bombs were exploding in her vicinity, the submarine was brought to the surface by her own crew by the discharge of a great deal of water from her forward ballast tanks. It was found, however, that the net was still foul of her, and that a Zeppelin was overhead, evidently attracted by the disturbance in the water due to the discharge of air and water from the submarine. She went to the bottom again, and after half an hour succeeded in getting clear of the net. Meanwhile the Zeppelin had collected a force of trawlers and destroyers, and the submarine was hunted for fourteen hours by this force, assisted by the airship. During this period she succeeded in sinking one of the German destroyers, and was eventually left unmolested.

For a correct appreciation of submarine warfare it is necessary to have a clear idea of the characteristics and qualities of the submarine herself, of the numbers possessed by the enemy, and of the rate at which they were being produced. It is also necessary, in order to understand the difficulty of introducing the counter measures adopted by the Royal Navy, to know the length of time required to produce the vessels and the weapons which were employed or which it was intended to employ in the anti-submarine war.

The German submarines may be divided into four classes, viz.: Submarine cruisers, U-boats, U.B.-boats, U.C.-boats. There were several variations of each class.

The earlier submarine cruisers of the "Deutschland" class were double-hulled vessels, with a surface displacement of 1,850 tons, and were about 215 feet long; they had a surface speed of about 12 knots and a submerged speed of about 6 knots. They carried two 5.9-inch guns, two 22 pounders, two torpedo tubes, and 12 torpedoes. They could keep the sea for quite four months without being dependent on a supply ship or base.

The later submarine cruisers were double-hulled, 275-320 feet long, had a surface speed of 16-18 knots, and a submerged speed of about 7 to 8 knots. They carried either one or two 5.9-inch guns, six torpedo tubes, and about 10 torpedoes. They had a very large radius of action, viz., from 12,000 to 20,000 miles, at a speed of 6 knots. A large number (some 30 to 40) of these boats were under construction at the time of the Armistice, but very few had been completed.

There were two or three types of U-boats. The earlier vessels were 210 to 220 feet long, double-hulled, with a surface displacement of about 750 tons, a surface speed of 15 to 16 knots, and a submerged speed of about 8 knots. They carried one or two 4.1-inch guns, four to six torpedo tubes, and about 10 torpedoes.

Later vessels of the class were 230 to 240 feet long, and of 800 to 820 tons surface displacement, and carried six torpedo tubes and 16 torpedoes. Some of them, fitted as minelayers, carried 36 mines, and two torpedo tubes, but only two torpedoes. A later and much larger class of minelayers carried a 5.9-inch gun, four torpedo tubes, 42 mines, and a larger number of torpedoes. The earlier U-boats could keep the sea for about five weeks without returning to a base or a supply ship; the later U-boats had much greater sea endurance.

The smaller U.B.-boats were single-hulled, and about 100 feet long, had a surface speed of 7 to 9 knots and a submerged speed of about 5 knots, and carried one 22-pounder gun, two torpedo tubes and four torpedoes. These boats could keep the sea for about two weeks without returning to a base or supply ship. A later class were double-hulled, 180 feet long, with greater endurance (8,000 miles at 6 knots), a surface speed of 13 knots and a submerged speed of 8 knots; they carried one 4.1-inch gun, five tubes and 10 torpedoes.

The earliest U.C.-boats were 111 feet long, with a surface displacement of 175 tons, a surface speed of 6-½ knots, and a submerged speed of 5 knots. They carried 12 mines, but no torpedo tubes, and as they had a fuel endurance of only 800 miles at 5-½ knots, they could operate only in southern waters.

The later U.C.-boats were 170 to 180 feet long, double-hulled, had a surface speed of 11 to 12 knots and a submerged speed of about 7 knots, carried 18 mines, three torpedo tubes, five torpedoes, and one 22-pounder gun, and their fuel endurance was 8,000 to 10,000 miles at a speed of 7 to 8 knots.

At the end of February, 1917, it was estimated that the enemy had a total of about 130 submarines of all types available for use in home waters, and about 20 in the Mediterranean. Of this total an average of between one-half and one-third was usually at sea. During the year about eight submarines, on the average, were added monthly to this total. Of this number some 50 per cent, were vessels of the mine-laying type.

All the German submarines were capable of prolonged endurance submerged. The U-boats could travel under water at the slowest speed for some 48 hours, at about 4 knots for 20 hours, at 5 knots for about 12 hours, and at 8 knots for about 2 hours.

They were tested to depths of at least 180 feet, but many submerged to depths exceeding 250 feet without injury. They did not usually lie on the bottom at depths greatly exceeding 20 fathoms (120 feet).

All German submarines, except possibly the cruiser class, could dive from diving trim in from 30 seconds to one minute. The U.B. class had particularly rapid diving qualities, and were very popular boats with the German submarine officers. Perhaps the most noticeable features of the German submarines as a whole were their excellent engines and their great strength of construction.

Prior to the month of February, 1917, it was the usual practice of the enemy submarine in the warfare against merchant ships to give some warning before delivering her attack. This was by no means a universal rule, particularly in the case of British merchant vessels, as is evidenced by the attacks on the Lusitania, Arabic, and scores of other ships.

In the years 1915 and 1916, however, only 21 and 29 per cent. respectively of the British merchant ships sunk by enemy submarines were destroyed without warning, whilst during the first four months of the unrestricted submarine warfare in 1917 the figure rose to 64 per cent., and went higher and higher as the months progressed.

Prior to February, 1917, the more general method of attack on ships was to "bring them to" by means of gun-fire; they were then sunk by gun-fire, torpedo, or bomb. This practice necessitated the submarine being on the surface, and so gave a merchant ship defensively armed a chance of replying to the gun-fire and of escaping, and it also gave armed decoy ships a good opportunity of successful action if the submarine could be induced to close to very short range.

The form of attack on commerce known as "unrestricted submarine warfare" was commenced by Germany with the object of forcing Great Britain to make peace by cutting off her supplies of food and raw material. It has been acknowledged by Germans in high positions that the German Admiralty considered that this form of warfare would achieve its object in a comparatively short time, in fact in a matter of some five or six months.

Experienced British naval officers, aware of the extent of the German submarine building programme, and above all aware of the shadowy nature of our existing means of defence against such a form of warfare, had every reason to hold the view that the danger was great and that the Allies were faced with a situation, fraught with the very gravest possibilities.

The principal doubt was as to the ability of the enemy to train submarine crews with sufficient rapidity to keep pace with his building programme.

However, it was ascertained that the Germans had evidently devoted a very great number of their submarines to training work during the period September, 1915, to March, 1916, possibly in anticipation of the unrestricted warfare, since none of their larger boats was operating in our waters between these months; this fact had a considerable bearing on the problem.

As events turned out it would appear either that the training given was insufficient or that the German submarine officer was lacking in enterprise.

There is no doubt whatever that had the German craft engaged in the unrestricted submarine warfare been manned by British officers and men, adopting German methods, there would have been but few Allied or neutral merchant ships left afloat by the end of 1917.

So long as the majority of the German submarine attacks upon shipping were made by gun-fire, the method of defence was comparatively simple, in that it merely involved the supply to merchant ships of guns of sufficient power to prevent the submarine engaging at ranges at which the fire could not be returned. Whilst the method of defence was apparent, the problem of supplying suitable guns in sufficient numbers was a very different matter. It involved arming all our merchant ships with guns of 4-inch calibre and above. In January, 1917, only some 1,400 British ships had been so armed since the outbreak of war.

It will be seen, therefore, that so long as ships sailed singly, very extensive supplies of guns were required to meet gun attack, and as there was most pressing need for the supply of guns for the Army in France, as well as for the anti-aircraft defence of London, the prospect of arming merchant ships adequately was not promising.

When the enemy commenced unrestricted submarine warfare attack by gun-fire was gradually replaced by attack by torpedo, and the problem at once became infinitely more complicated.

Gun-fire was no longer a protection, since the submarine was rarely seen. The first intimation of her presence would be given by the track of a torpedo coming towards the ship, and no defence was then possible beyond an endeavour to manoeuvre the ship clear of the torpedo. Since, however, a torpedo is always some distance ahead of the bubbles which mark its track (the speed of the torpedo exceeding 30 knots an hour), the track is not, as a rule, seen until the torpedo is fairly close to the ship unless the sea is absolutely calm. The chance of a ship of low speed avoiding a hit by a timely alteration of course after the torpedo has been fired is but slight. Further, the only difficulty experienced by a submarine in hitting a moving vessel by torpedo-fire, once she has arrived in a position suitable for attack, lies in estimating correctly the course and speed of the target. In the case of an ordinary cargo ship there is little difficulty in guessing her speed, since it is certain to be between 8 and 12 knots, and her course can be judged with fair accuracy by the angle of her masts and funnel, or by the angle presented by her bridge.

It will be seen, then, how easy was the problem before the German submarine officers, and how very difficult was that set to our Navy and our gallant Mercantile Marine.

It will not be out of place here to describe the methods which were in force at the end of 1916 and during the first part of 1917 for affording protection to merchant shipping approaching our coasts from the direction of the Atlantic Ocean.

The general idea dating from the early months of the war was to disperse trade on passage over wide tracts of ocean, in order to prevent the successful attacks which could be so easily carried out if shipping traversed one particular route. To carry out such a system it was necessary to give each vessel a definite route which she should follow from her port of departure to her port of arrival; unless this course was adopted, successive ships would certainly be found to be following identical, or practically identical, routes, thereby greatly increasing the chance of attack. In the early years of the war masters of ships were given approximate tracks, but when the unrestricted submarine campaign came into being it became necessary to give exact routes.

The necessary orders were issued by officers stationed at various ports at home and abroad who were designated Shipping Intelligence or Reporting Officers. It was, of course, essential to preserve the secrecy of the general principles governing the issue of route orders and of the route orders themselves. For this reason each master was only informed of the orders affecting his own ship, and was directed that such orders should on no account fall into the hands of the enemy.

The route orders were compiled on certain principles, of which a few may be mentioned:

(a) Certain definite positions of latitude and longitude were given through which the ship was required to pass, and the orders were discussed with the master of each vessel in order to ensure that they were fully understood.

(b) Directions were given that certain localities in which submarines were known to operate, such as the approaches to the coast of the United Kingdom, were, if possible, to be crossed at night. It was pointed out that when the speed of the ship did not admit of traversing the whole danger area at night, the portion involving the greatest danger (which was the inshore position) should, as a rule, be crossed during dark hours.

(c) Similarly the orders stated that ships should, as a rule, leave port so as to approach the dangerous area at dusk, and that they should make the coast at about daylight, and should avoid, as far as possible, the practice of making the land at points in general use in peace time.

(d) Orders were definite that ships were to zigzag both by day and at night in certain areas, and if kept waiting outside a port.

(e) Masters were cautioned to hug the coast, as far as navigational facilities admitted, when making coastal passages.

The orders (b), (c) and (d) were those in practice in the Grand Fleet when circumstances permitted during my term in that command.

A typical route order from New York to Liverpool might be as follows:

"After passing Sandy Hook, hug the coast until dark, then make a good offing before daylight and steer to pass through the following positions, viz:

"Thence make the coast near the Skelligs approximately at daylight, hug the Irish coast to the Tuskar, up the Irish coast (inside the banks if possible), and across the Irish Channel during dark hours. Thence hug the coast to your port; zigzag by day and night after passing, Long. 20° W."

Sometimes ships were directed to cross to the English coast from the south of Ireland, and to hug the English coast on their way north.

The traffic to the United Kingdom was so arranged in the early part of 1917 as to approach the coast in four different areas, which were known as Approach A, B, C, and D.

Approach A was used for traffic bound towards the western approach to the English Channel.

Approach B for traffic making for the south of Ireland.

Approach C for traffic making for the north of Ireland.

Approach D for traffic making for the east coast of England via the north of Scotland.

The approach areas in force during one particular period are shown on Chart A (in pocket at the end of the book). They were changed occasionally when suspicion was aroused that their limits were known to the enemy, or as submarine attack in an area became intense.

[Transcriber's note: Chart A is a navigational map of the waters southwest of England, with approach routes marked.]

The approach areas were patrolled at the time, so far as numbers admitted, by patrol craft (trawlers, torpedo-boat destroyers, and sloops), and ships with specially valuable cargoes were given directions to proceed to a certain rendezvous on the outskirts of the area, there to be met by a destroyer or sloop, if one was available for the purpose. The areas were necessarily of considerable length, by reason of the distance from the coast at which submarines operated, and of considerable width, owing to the necessity for a fairly wide dispersion of traffic throughout the area. Consequently, with the comparatively small number of patrol craft available, the protection afforded was but slight, and losses were correspondingly heavy. In the early spring of 1917, Captain H.W. Grant, of the Operations Division at the Admiralty, whose work in the Division was of great value, proposed a change in method by which the traffic should be brought along certain definite "lines" in each approach area. Typical lines are shown in Chart B.

[Transcriber's note: Chart B is a navigational map of the waters southwest of Ireland, with approach routes marked.]

The idea was that the traffic in, say, Approach Route B, should, commencing on a certain date, be ordered by the Routeing Officer to pass along the line Alpha. Traffic would continue along the line for a certain period, which was fixed at five days, when it would be automatically diverted to another line, say Gamma, but the traffic along Gamma would not commence until a period of 24 hours had elapsed since discontinuance of the use of the line Alpha. This was necessary in order to give time for the patrol craft to change from one line to the other. During this period of 24 hours the arrangement for routeing at the ports of departure ensured that no traffic would reach the outer end of any of the approach lines, and consequently that traffic would cease on line Alpha 24 hours before it commenced on line Gamma. After a further period of five days the line would again change automatically.

It was necessary that Shipping Intelligence Officers should have in their possession the orders for directing traffic on to the various lines for some considerable time ahead, and the masters of ships which were likely to be for some time at sea were informed of the dates between which the various lines were to be used, up to a date sufficient to cover the end of their voyage. There was, therefore, some danger of this information reaching the enemy if a vessel were captured by a submarine and the master failed to destroy his instructions in time. There was also some danger in giving the information to neutrals.

However, the system, which was adopted, did result in a reduction of losses during the comparatively short time that it was in use, and the knowledge that patrol craft on the line would be much closer together than they would be in an approach area certainly gave confidence to the personnel of the merchant ships, and those who had been forced to abandon their ship by taking to the boats were afforded a better chance of being picked up.

Various arrangements were in existence for effecting rapidly a diversion of shipping from one route to another in the event of submarines being located in any particular position, and a continual change of the signals for this purpose was necessary to guard against the possibility of the code being compromised by having fallen into enemy hands, an event which, unfortunately, was not infrequent.

Elaborate orders were necessary to regulate coastal traffic, and fresh directions were continually being issued as danger, especially danger from mines, was located. Generally speaking, the traffic in home waters was directed to hug the coast as closely as safe navigation permitted. Two reasons existed for this, (a) in water of a depth of less than about eight fathoms German submarines did not care to operate, and (b) under the procedure indicated danger from submarine attack was only likely on the side remote from the coast.

Here is an example of the instructions for passing up Channel:

From Falmouth to Portland Bill.—Hug the coast, following round the bays, except when passing Torbay. (Directions followed as to the procedure here.)

From Portland Bill to St. Catherines.—Pass close south of the Shambles and steer for Anvil Point, thence hug the coast, following round the bays.

And so on.

As it was not safe navigationally to follow round the bays during darkness, the instructions directed that ships were to leave the daylight route at dusk and to join the dark period route, showing dimmed bow lights whilst doing so.

Two "dark period routes" were laid down, one for vessels bound up Channel, and another for vessels bound down Channel, and these routes were some five miles apart in order to minimize the danger of collision, ships being directed not to use their navigation lights except for certain portions of the route, during which they crossed the route of transports and store ships bound between certain southern British ports (Portsmouth, Southampton and Devonport) and French ports.

Routes were similarly laid down for ships to follow when navigating to or from the Bristol Channel, and for ships navigating the Irish Sea.

Any system of convoy was at this time out of the question, as neither the cruisers to marshal the convoy to the submarine area, nor the destroyers to screen it when there, were available.

There was one very important factor in the situation, viz., the comparative rate at which the Germans could produce submarines and at which we could build vessels suitable for anti-submarine warfare and for defence of commerce. The varying estimates gave cause for grave anxiety. Our average output of destroyers was four to five per month. Indeed, this is putting the figure high; and, of course, we suffered losses. The French and Italians were not producing any vessels of this type, whilst the Japanese were, in the early part of 1917, not able to spare any for work in European waters, although later in the year they lent twelve destroyers, which gave valuable assistance in the Mediterranean. The United States of America were not then in the war. Consequently measures for the defence of the Allied trade against the new menace depended on our own production.

Our submarines were being produced at an average rate of about two per month only, and—apart from motor launches, which were only of use in the finest weather and near the coast—the only other vessels suitable for anti-submarine work that were building at the time, besides some sloops and P-boats, were trawlers, which, whilst useful for protection patrol, were too slow for most of the escort work or for offensive duties. The Germans' estimate of their own submarine production was about twelve per month, although this figure was never realized, the average being nearer eight. But each submarine was capable of sinking many merchant ships, thus necessitating the employment of a very large number of our destroyers; and therein lay the gravity of the situation, as we realized at the Admiralty early in 1917 that no effort of ours could increase the output of destroyers for at least fifteen months, the shortest time then taken to build a destroyer in this country.

And here it is interesting to compare the time occupied in the production of small craft in Great Britain and in Germany during the war.

In pre-war days we rarely built a destroyer in less than twenty-four months, although shortly before the war efforts were made to reduce the time to something like eighteen to twenty months. Submarines occupied two years in construction.

In starting the great building programme of destroyers and submarines at the end of 1914, Lord Fisher increased very largely the number of firms engaged in constructing vessels of both types. Hopes were held out of the construction both of destroyers and of submarines in about twelve months; but labour and other difficulties intervened, and although some firms did complete craft of both classes during 1915 in less than twelve months, by 1916 and 1917 destroyers averaged about eighteen months and submarines even longer for completion.

The Germans had always built their small craft rapidly, although their heavy ships were longer in construction than our own. Their destroyers were completed in a little over twelve months from the official date of order in pre-war days. During the early years of the war it would seem that they maintained this figure, and they succeeded in building their smaller submarines of the U.B. and U.C. types in some six to eight months, as U.B. and U.C. boats began to be delivered as early as April, 1915, and it is certain that they were not ordered before August, 1914.

The time taken by the Germans to build submarines of the U type was estimated by us at twelve months, and that of submarine cruisers at eighteen months. German submarine officers gave the time as eight to ten months for a U-boat and eighteen months for a submarine cruiser.

(It is to be observed that Captain Persius in a recent article gives a much longer period for the construction of the German submarines. It is not stated whether he had access to official figures, and his statement is not in agreement with the figures given by German submarine officers.)

It is of interest to note here the rate of ship production attained by some firms in the United States of America during the war.

As I mention later (Vide Chapter vi, p. 157), the Bethlehem Steel Company, under Mr. Schwab's guidance, produced ten submarines for us in five months from the date of the order. Mr. Schwab himself informed me that towards the end of the war he was turning out large destroyers in six weeks. The Ford Company, as is well known, produced submarine chasers of the "Eagle" type in even a shorter period, but these vessels were of special design and construction.

I have dealt so far with the question of anti-submarine measures involving only the use of destroyers and other small surface craft. There were, of course, other methods both in use and under consideration early in 1917 when we took stock of the situation.

For some time we had been using Decoy vessels, and with some success; it was possible to increase the number of these ships at the cost of taking merchant ships off the trade routes or by building. A very considerable increase was arranged.

The use of our own submarines offensively against enemy submarines had also been tried, and had met with occasional success, but our numbers were very limited (the total in December, 1916, fit for oversea or anti-submarine work was about forty). They were much needed for reconnaissance and offensive work against surface men-of-war in enemy waters, and only a few were at the time available for anti-submarine operations, and then only at the cost of other important services.

The hydrophone had been in the experimental stage and under trial for a considerable period, but it had not so far developed into an effective instrument for locating submarines, and although trials of the different patterns which had been devised were pushed forward with energy, many months elapsed before it became a practicable proposition.

One of the best offensive measures against the enemy submarines, it was realized, was the mine, if laid in sufficiently large numbers. Unfortunately, in January, 1917, we did not possess a mine that was satisfactory against submarines.

Our deficiency in this respect was clearly shown in the course of some trials which I ordered, when one of our own submarines was run against a number of our mines, with the result that only about 33 per cent. of the mines (fitted, of course, only with small charges) exploded. The Germans were well aware that our mines were not very effective against submarines.

We possessed at the time mines of two patterns, and whilst proving unsatisfactory against submarines, they were also found to be somewhat unreliable when laid in minefields designed to catch surface vessels, owing to a defect in the mooring apparatus. This defect was remedied, but valuable time was lost whilst the necessary alterations were being carried out, and although we possessed in April, 1917, a stock of some 20,000 mines, only 1,500 of them were then fit for laying. The position, therefore, was that our mines were not a satisfactory anti-submarine weapon.

A new pattern mine, which had been designed on the model of the German mine during Sir Henry Jackson's term of office as First Sea Lord in 1916, was experimented with at the commencement of 1917, and as soon as drawings could be prepared orders for upwards of 100,000 were placed in anticipation of its success. There were some initial difficulties before all the details were satisfactory, and, in spite of the greatest pressure on manufacturers, it was not until November, 1917, that mines of this pattern were being delivered in large numbers. The earliest minefields laid in the Heligoland Bight in September and October, 1917, with mines of the new pattern met with immediate success against enemy submarines, as did the minefields composed of the same type of mine, the laying of which commenced in November, 1917, in the Straits of Dover.

When it became possible to adopt the system of bringing merchant ships in convoys through the submarine zone under the escort of a screen of destroyers, this system became in itself, to a certain extent, an offensive operation, since it necessarily forced the enemy submarines desirous of obtaining results into positions in which they themselves were open to violent attack by depth charges dropped by destroyers.

During the greater part of the year 1917, however, it was only possible to supply destroyers with a small number of depth charges, which was their principal anti-submarine weapon; as it became feasible to increase largely the supply of these charges to destroyers, so the violence of the attack on the submarines increased, and their losses became heavier.

The position then, as it existed in the early days of the year 1917, is described in the foregoing remarks.

The result measured in loss of shipping (British, Allied, and neutral) from submarine and mine attack in the first half of the year was as follows in gross tonnage:

Because of the time required for production, it was a sheer impossibility to put into effect any fresh devices that might be adopted for dealing with submarine warfare for many months, and all that could be done was to try new methods of approach to the coast and, as the number of small craft suitable for escort duty increased, to extend gradually the convoy system already in force to a certain extent for the French coal trade and the Scandinavian trade.

In the chapters which follow the further steps which were taken to deal with the problem, and the degree of success which attended them, will be described.