resumed in the Atlantic in 1916 when the " Moewe " was out,
and in the course of that year some 15 convoys of two to three
transports each sailed from Halifax. But these were special
escorts intended to protect their convoys from surface craft.
The adoption of a general convoy system was still outside the
pale of contemporary naval thought. A considerable propor-
tion of the troop service across Channel was escorted; but this
was a local service arranged by the admirals at Dover (for
Folkestone to Boulogne) or Portsmouth (for Southampton to
Havre). There was a tendency to regard the loss of merchant
ships with little concern during that period, and a number of
large ships were even fitted out at great expense to act as dummy
battleships to be torpedoed by the enemy instead of the vessels
they tried to counterfeit. The idea of using destroyers to escort
the ordinary trade would have received short shrift at the Ad-
miralty and in the Grand Fleet, nor was it necessary at the time.
In March and April 1917, when the British losses in merchant shipping assumed alarming proportions, the idea of a convoy system came again to the front. Previous ocean convoys had been directed against the surface raider; it was the submarine that now formed the principal menace. The system was in use in the case of what was called the French coal trade, a cross- Channel traffic from Portsmouth and Falmouth performed by small ships, where it had worked very successfully, and it was now suggested to extend it to the ocean routes. The system of protection in vogue at the time may be called the patrolled route system. There were three main approach routes to the British Isles, one N. of Ireland for ships making the Clyde or Liverpool by the North Channel (Route C); one S. of Ireland for ships making the British Channel and Liverpool by St. George's Channel (Route B) ; one towards the Scillies for ships making the English Channel (Route A). These were called the Tory I., Fastnet (S.W. point of Ireland) and Scillies approaches, from the lights sighted by the ships as they made the coast. These approaches may be regarded as three great triangles gradually narrowing to three apexes at or near the points men- tioned. They were patrolled with trawlers and a sprinkling of destroyers, and when any area was threatened by submarine activity the routes in it were changed.
In March 1917, the system was slightly modified. Half a dozen different routes were specified in ach approach triangle, and it was proposed to switch the traffic from one route to the other every five days. As the routes in each triangle could lie some 150 m. apart in long. 15 W., there was considerable scope for dispersion, and the system was in effect a system of protection by means of dispersion and routeing. The patrols were a mere pretence, for the routes were on an average some 240 m. long, and to patrol them in strength with two destroyers and 16 trawlers was impracticable. The scheme was in its essence an endeavour to circumvent the submarine by routeing and it failed. Its advocates could not possibly maintain that it was as efficient as an escort system, for all important ships were actually escorted, and it must be regarded merely as an attempt to burke the significance of the fact that was beginning to assert itself, that every ship had to be escorted.
The idea of general convoy met with strong opposition from every side at the Admiralty, in the Grand Fleet and amongst the masters and owners of ships. The Admiralty saw that it would involve the creation of a new organization; the Grand Fleet saw its destroyers being taken away; very few recognized the fact that the battlefleets were now becoming merely comple- mentary factors in a guerre de course. The policy of the fleet being ready at any moment to rush out and join battle still held sway. It was a policy resting chiefly on the basis of intelligence supplied by wireless directionals, which made it possible to know when the German fleet was at sea. It meant that the fleet had to be ready at any moment to put to sea in battle array, and in these circumstances the commander-in-chief clung tenaciously to every one of his destroyers. These may be called the strategi- cal objections to convoy, but other strong arguments could be urged against it. Ships would incur delay in assembling, instead of sailing direct; fast ships would incur further delay by having
to reduce their speed to that of the convoy. Convoy meant congestion of ports on departure and arrival, and congestion of labour due to the simultaneous arrival of a number of ships. These objections were as old as the days of Duguay Trouin and Jean Bart, but there were other objections of a more modern type. The masters would never be able to keep station, and were at first much in favour of independent sailings.
On the other hand strong arguments could be marshalled in favour of convoy. Why string out 15 armed trawlers 10 m. apart to supply feeble protection on a line 1 50 m. long, when four destroyers attached to a convoy could give it continual and effi- cient protection over double the distance? It might appear at first that a convoy gave the submarine a massed target, but the danger of approaching it was greater, and submarine com- manders preferred to attack unescorted ships. The real obstacle in the way of a convoy system was the difficulty of finding the destroyers required for the escort. More than half the modern destroyers were absorbed by the Grand Fleet and the Harwich flotillas. In Feb. 1917 10 were detached from the Grand Fleet to Devonport to assist in escorting important ships, and the use of armed trawlers was extended, but the latter were too slow and too ill-armed to be of much value."
The weekly returns issued by the trade division at this time conveyed a misleading idea of the situation. They gave the number of arrivals and departure of all nationalities, with the number of British ships of over 1,600 tons lost. The first set of figures had little to do with the real issue, for a small Dutch coaster making three voyages a week to France would figure six times in the arrivals and departures, which ran into several thousands, whereas the number of big ships arriving and leaving daily was very much less. The situation was much worse than it appeared, and the idea of general convoy gained ground.
In April 1917 the British had 3,534 ships over 1,600 tons, of which 1,125 were required for naval and military purposes, leaving only 2,409 available for civil purposes. There were not more than 15 patrols in each area of approach, and in March and April 1917 the number of ships passing through them was about 300, of which 24 were sunk, at which rate, giving each ship a round voyage of two months, practically one-half would have been sunk by the end of the year. Again those who argued in favour of the patrolled route system were arguing in direct opposition to their own policy, for escorts were always provided for all valuable munition ships and ships of national importance carrying Government cargo (some three or four a day in 1917). The patrolled route system was thereby acknowledged to be an inferior sort of makeshift for ships that were not of national importance. But in April 1917 it began to be seen that every ship was of national importance, and that a loss of 373 ships a month meant that the navy would lose the war before the army could win it. The great advantage of the patrolled route system was that it gave much less trouble and required very few ships, but the same virtues were inherent in no system at all.
One other argument was marshalled against the system, namely, that it would be better to use destroyers directly against submarines. The reply was that no likelier spot could be chosen for seeking them than in the vicinity of a convoy, and from the date convoys commenced to run in May 1917 to the end of the war some 15% of submarines sunk were actually sunk in the vicinity of or when attacking convoys. The losses of April brought the question to an acute stage. The centralization of the control of shipping in the Ministry of Shipping facilitated the inauguration of a general system of convoy. On April 26 the director of the anti-submarine division urged its introduc- tion, and on May 17 1917 a convoy committee was appointed to arrange the details of a specific scheme. The volume of trade in the Atlantic daily at that time amounted to about 400 vessels, of which 300 were British and 87 neutral. As the area of convoy only extended to about long. 20 W., only some 30 vessels had to be convoyed daily, and it was decided to start with a convoy from the United States and Canada every three days, from Gibraltar every four and from Dakar every five days. The in- itiation of the system fell largely to Comm. Reginald G.
