The Signal Service in the European War of 1914 to 1918 – Chapter 13

«--	The Signal Service in the European War of 1914 to 1918 Chapter XIII	--»
Chapter XII		Chapter XIV

Chapter XIII.

THE SURPRISE BATTLE OF CAMBRAI AND THE WINTER OF 1917-1918.

SYNOPSIS.

First Battle of Cambrai. — Signals in the Surprise Attack. — Cavalry Line System in Addition to Command and Artillery S\-stem. — Use of Area Personnel. — "Camouflage" Line System of Poled and Trench Multicore Cable. — Effect of Tanks on the Line System. — Precautions to Ensure Secrecy. — Testing with Galvanometers and Fullerphones. — Alternative Methods. — Problems Caused by the Depth of the Advance. — -Cavalry Signals. — Decreased Telephone Facilities. — Congested Transport Prevents Line Stores from Getting Forward. — Twisted Cable in Mobilization Equipment. — Cable Dumps. — Visual: Two-way Working. — Wireless. — Other Methods of Forward Signalling. — Brigade and Battalion Signals. — Use of Enemy's Lines. — Pigeons and Runners. — Stabilisation of the Situation. — Consolidation of the Signal System. — The German Counter Attack. — S.S. 191, like S.S. 148, Ignores Possibility of a Retreat. — Emergency Divisional Route Saves the Situation. — W/T Very Useful. —Signal Personnel Engaged in Fighting. — Guards Division Counter -Attack. — Lessons of the Battle. — History of Tank Signals. — Evolution of Tank Wireless. — A.D. Signals, Tank Corps. — Trench Warfare Again. — Increase of Enemy Bombing Causes Modification of Rear Routes and Signal Offices. — Second Great Re-organization of the A.S.S. — No. 5 Section of a Divisional Signal Company. — Strength of Signal Service and Economy of Personnel. — Army Conference Considers Possibility of Retrenchment.

The action in November, 1917, now known as the battle of Cambrai, although not carried out on as large a scale as the summer offensives and not an unqualified success, had several features of particular interest from the point of view of the Army Signal Service. It was in the nature of a surprise attack long thought out by the attacking side, and it was this feature of the battle which gave to the signal preparation for the action and the signal communication during the attack a character altogether unusual. The principal characteristic of the assault was an entire absence of artillery preparation, the wire cutting, usually assigned to the latter, being carried out by the organized advance of large bodies of tanks in the wake of which the infantry were to follow. In the main, everything went as originally planned, and the striking Divisions came within an ace of breaking through the German defences and permitting the cavalry to pass through and operate along the enemy's lines of communication.

In considering the signals of the battle of Cambrai, therefore, three problems which were presented to the Signal Service for solution require particular attention. These three were: — (1) The preparation of a "secret" system of intercommunication before the battle; (2) Tank signals and the effect of the tanks on line communication; (3) The unusual speed and depth of the advance, some Brigades moving forward five miles in a few hours.

Undoubtedly the characteristic feature of the signal preparations for the battle was the laying down of a complete system of overground lines to accommodate the extra formations and units which it was proposed to move into the area immediately before the date fixed for the commencement of the action. The attack was carried out by seven Divisions in line, while two others on the left flank were engaged in subsidiary attempts to breach the Hindenburg line. If the initial assault was successful, the cavalry were to pass through the infantry Divisions resting upon their objectives and raid the enemy's communications on a huge scale.

Preparations involved the erection or laying of -13,000 miles of wire in rather less than a month. It was clear that the most careful organization was necessary, if the line system was to be ready by November 20th, the date fixed for the attack. The difficulties of the Signal Service were also increased by the necessity for strict secrecy which would naturally prevent the thorough testing of the new lines according to common practice and would also hinder the use of the circuits by the newly-arrived units prior to zero hour.

A complete system with one main forward route for each Division in the line was planned on the principles laid down in the new Staff Manual for Intercommunication in the Field (S.S. 191). Owing to the necessity for secrecy and the need for hurry, however, buried cable was prohibited and some other means had to be devised. Safety had to be sacrificed to speed and secrecy, and it was decided to rely entirely upon poled cable in the rear areas. In those areas so far forward as to invite enemy observation, the main aerial routes disappeared, being split in half and continued as trench cable to two cable-heads on each Divisional front.

For the building of the system the personnel available consisted of the Corps signal companies and attached cable and airline sections with the exception of men already employed on the maintenance of the existing system. In addition, were certain area parties supplied by the Army signal company and such small parties as could be spared by the signal companies of the Divisions in the line and those which would later be called upon to take up their battle positions in the area. Each small Divisional party was under a senior officer — usually the second in command of the signal company or the Divisional artillery captain — and formed the nucleus of the party working on what was to become their own system. The exceptional importance and urgency of the work was fully realized, and both Army, Corps and Divisional personnel worked long hours during the day, while at night special parties tested each portion of the system that had been laid during the previous day.

For the rear portion of the system lead-covered cables were used, and the routes were poled as far forward as possible. One obstacle to the quick building of the rear routes consisted in the fact that the paper-core cables were too heavy to be carried in lengths of more than 200 yards, and a large amount of work which could only be carried out by skilled jointers was thus required. Forward, where trench cable had to be resorted to, armoured cables were used and a portion of these were laid in existing open traffic-proof ditches one foot deep. Other portions were laid along the bottom of communication trenches and yet others along the ground. The latter was camouflaged as much as possible by taking advantage of hedges, ditches and tracks, and does not seem to have attracted any undue attention from the enemy. It was, however, cut on many occasions by the normal enemy shelling and the necessity for re-laying damaged sections made it very difficult to complete the forward position of the system in the time allotted. Another great difficulty was the damage done by the tanks proceeding to their hiding places during the last few nights before zero, and from their hiding places to the rallying point immediately before the attack was to begin.

This difficulty had been anticipated to some extent, but for two reasons the action taken was not altogether successful. The special crossing places of poled cable were not made high enough to clear supply tanks or tanks loaded with fascines for crossing moats and streams. In addition, through lack of liaison between Tanks and Signal Service, the former often failed to use the crossing places provided for them. This was in part due to lack of information as to the existence and location of these local buries and overhead crossings, partly due to tanks getting lost and wandering about the country indiscriminately during the night, partly also to a failure to take into consideration the limitations of the 1917 tanks. Whatever the reason, however, the result was very serious. Poled cable of normal height was caught up in the upper works of the tanks and scattered broadcast over the country. Poles were snapped and cables torn apart. Ground cables were churned into the earth and the insulation ripped off them for many yards. If the lines were not completely interrupted, they were rendered useless for speaking and could only be used with difficulty for buzzing. In one case, even, a tank was propelled straight over a Divisional cable head, crushing in the inefficient shelter which had been the best that could be provided in the short time available, and destroying the connecting frame inside.

Where damage occurred during the initial concentration of the tanks in the area, it was, of course, possible to remedy it, though much extra strain was thrown upon the signal personnel responsible for the perfection of the system. Such breaks were, however, recurrent throughout the operations and the damage done by the tanks during the forming-up period immediately preceding zero hour and during the battle itself, was much more serious and, indeed, often irreparable. Thus, the first great lesson of the surprise battle which involved the use of massed tanks was the necessity for closer liaison between the Signal Service and the Tank Corps. The lesson was driven well home to a distracted Signal Service during the next ten days and it needed no repetition. The result was seen in certain recommendations made for the modification of the Tank Corps signal units.

The new line system was to be superimposed upon the old one and a feature of the preparations on which great stress was laid, both by the Intelligence Staff and the Signal Service, was that all forward traffic right up to the opening of the battle must be carried on over the lines of the old system. It was, of course, impossible to keep traffic down entirely to normal, and the original Divisions in the line reported their "operations" telegrams as rising from a normal average of between 500 and 700 to 1,600 and even 2,000 messages in the 24 hours. Every effort was made to keep down the number of messages, however, and the above total is not excessive when it is realized that no less than four Divisions were accommodated on the lines of one, and that, in addition, the essential portion of the work of many extra units was also included.

In order to ensure secrecy, the offensive signal system was divided into two distinct parts, the boundary line being the recognized limit of enemy overhearing — 3,000 yards behind the front hue. At this position, all the more forward routes were disconnected and, while the rear portion was taken into use immediately upon its completion, the forward part beyond this line was not employed at all until "Z" day. New units arriving in the area were allotted their lines but not allowed to connect until just before zero. The use of telephones forward of Brigade headquarters was absolutely forbidden except for artillery observation, and observers were instructed to take their telephones up with them and to remove them immediately their observation was completed.

In the meantime, it was necessary that some special means of testing the forward system should be devised. It was out of the question that it should be left untested. Not only must a thorough initial test be made immediately the cables were laid, but the tests must be frequent and repeated. Breaks were constantly occurring both in the rear and the forward portions. Tanks, shells, and the normal traffic of a congested area, played havoc with the lines night after night. Shells continued their work in the daytime also. It was equally out of the question that the system should be tested by the normal means — lineman's telephone and buzzer. That would at once betray to the enemy the presence of a greatly reinforced line system opposite to this portion of his lines. A knowledge of the presence of the lines would not only nullify the precautions taken to restrict their use, but would convert those very precautions into a further source of suspicion. .

To realize the danger of this, it is only necessary to refer back to one or two examples already given in this narrative, of deductions made by our own Intelligence Staff — and even b}' the personnel of listening posts— from similar indiscretions on the part of the enemy.

A suitable compromise was, however, effected and the whole of the forward line system was tested, each section as it was completed, by means of galvanometers and fullerphones used with the minimum possible battery power. Both these instruments involved the use of small direct Hue currents only, and there was, therefore, no danger of giving anything away. A final test was carried out before the lines were taken into use on "Y-Z" night and proved that the work of both construction and maintenance parties had been conscientious and thorough in the extreme. Very few pairs existed which were not through. Any notable exceptions to the rule could be traced to specific damage to the routes at the last moment by shells, traffic, or tanks. Zero day broke to show a completed system with ample circuits for the essential needs of all Divisions, the cavalry, and other formations and units normally foreign to the area.

The demands of the tanks and artillery could be catered for upon an adequate scale and the system resembled that for an ordinary position warfare battle, with two significant exceptions. It was composed practically entirely of overground lines and was, therefore, particularly vulnerable. It was of somewhat slender dimensions and, therefore, it involved the elimination of many of the less important subscribers usually accommodated upon the Divisional and Corps telephone exchanges. The R.F.C., Casualty Clearing Stations, and anti-aircraft organizations, all had their intercommunication considerably reduced. Such officers as the C.R.E., D.A.D.O.S., and A.D.M.S. of Divisions, and all purely administrative units except those immediately concerned with the supply of rations and ammunition, were forced off the main telephone system and were obliged to make shift temporarily with less facilities. In other words, an approximation to mobile warfare conditions had been brought about by an entirely different set of circumstances. Later on in the battle, telephone facilities were to be still further cut down, this time by semi-mobile warfare itself.

The relative vulnerability of the battle line system made the provision of alternative communication more necessary than ever before. Efforts were made to duplicate the lines as far back as Divisional headquarters with chains of visual stations and with wireless. Visual schemes were formulated by Divisions and coordinated by Corps. Wireless sets were used as far as possible to reinforce the signal system between Divisional and Brigade headquarters and forward of this a good system of power buzzer and amplifier communication was planned. At the same time, arrangements were made to replace the rear cables as early as possible with airline routes, though these could not be erected before "Z" day for fear of arousing enemy suspicion. Forward dumps of airline stores of all descriptions were made and cable dumps from which Divisions might be fed in case of a far-reaching advance were also a marked feature of the very thorough signal preparations.

By the morning of November 20th, the day fixed for the attack, the whole pre-arranged scheme of signal communications was completed. Already tanks were playing havoc with the forward lines while proceeding to their rallying points, but this was a development which could not be entirely countered under the circumstances of the action. Circuits were everywhere tested through and the signal staff of the attacking Divisions had taken over. All units and formations who had been allotted a place on the main system were connected up and awaiting the raising of the interdict of silence. For the first time, signal secrecy and silence had been organized on a huge scale with decisive success and the result showed that little if any knowledge of the forthcoming attack had leaked through to the enemy. Certainly signal restrictions had been so efficiently supervised and so loyally carried out that indiscretions had been practically eliminated. The signal preparations for the battle of Cambrai, indeed, had proved that such indiscreet use of buzzer and telephone as had been prevalent in the British Army for so long, was so unnecessary as to border upon the criminal. They also proved conclusively that Staffs could do efficient work with less telephone facilities than they had been accustomed to have given them. This was to be seen still more decisively in the battle itself, for after a deep advance, speaking on many of the long cable lines became practically impossible. Message traffic increased till it averaged between 1,200 and 1,500 messages a day in single Divisions, but telephone conversation was far less common than in the position warfare of the previous summer. The effect of this lesson was not lost. It was to be further emphasized in the spring, and both experiences had a distinct bearing" on a development of Staff methods which made for greater efficiency with still less telephone communication in the summer and autumn offensives of the advance in 1918.

At zero hour on November 20th, the barrage opened, and, preceded by battalions of tanks, the British infantry advanced to the attack. From the first, the result was never in doubt, and, on almost the whole front affected, the enemy resistance was quickly overcome and his defences captured. By 1030 hours, the Hindenburg reserve line had been overrun, the general advance to the final objectives had begun,, and the cavalry was moving up behind the infantry. By the end of the day, the advance had reached a depth of four and a half miles over a wide front.

The signal units with the Corps and the attacking Divisions were faced on this day with a problem quite different from those of the position battles of the previous summer. Enemy resistance had been short and his guns had been captured or forced to withdraw early in the day. It was no longer a case of achieving as good a signal system as was possible under heavy retaliatory fire. Cable wagons could be used freely and there were no restrictions on the use of poled cable beyond that of the difficulty of transport of the supplies required. This situation, however, satisfactory as it was, presented special problems for solution, the main difficulties being the supply of cable to the Divisions in the line and the damage done to the line system bv the tanks. To this must also be added a considerable amount of destruction of the same system by the cavalry in their advance towards the front line and still more in their subsequent withdrawal.

It was the first occasion on which an attempt had been made to utilize cavalry on a large scale since the battle of Loos in 1915. For their use, a separate line system had been reserved. This consisted of two forward routes, one laid by the cavalry Corps themselves, the other carried out by Third Army. The former consisted of four pairs of poled twisted "D5" cable and this gave satisfactory service during the short time it was required. The second was laid from a tank but was experimental in its nature and was never through satisfactorily. The only novel feature of the cavalry signal communications as compared with those of the infantry Divisions was, of course, the use of wireless sets of higher power and greater range than the trench sets usual in forward work. The pack wireless stations with the cavalry Divisions and the Crossley lorry set with the cavalry Corps headquarters were in constant communication throughout the operations and imparted a distinctly novel tone to the sound of the operations as heard by the Intelligence stations. After the failure of the mounted action and the withdrawal of the main body of the cavalry, such formations and units as took part in the later fighting played the role of dismounted troops and were accommodated with signals of the ordinary type used by the infantry Divisions.

During the initial advance, lines were extended from cable head by cable wagon, extra transport being used to carry forward the light hop-poles which were necessary if the system was to be at all safe. One poled cable route of two or three pairs was carried forward on each Brigade front and little difficulty was found in laying these initial hues.

Two troubles soon made themselves felt, however, and between them effectually prevented any large amount of telephone traffic between the Divisions and Brigades in these first two or three days of the offensive. The poled cable was strongly built and, as in the preparatory system, special crossings were made which were intended to be tank-proof. The tanks, however, paid little attention to these crossings and their lateral movements across the front from flank to flank when carrying out a "rally" were so frequent that the lines were carried away and torn to pieces almost as fast as they were laid. It was impossible to build lines which would be tank-proof throughout their length. When a tank ran into a poled route or even a ground cable line, it meant — not the destruction of the short section actually run into or over — but the dislocation of the whole line and often the complete tearing away of several hundred yards. Until the tanks ceased their movements or kept strictly to the special crossing places made for them, telephone communication was intermittent at best and often impossible.

The movements of large bodies of cavalry across the open country" which characterises the Cambrai district was also destructive of any routes but those suspended well above the height of a mounted man. The height of the routes was, of course, dependent on the means of suspension and it was only in the few places where trees afforded opportunity of keeping the lines well up that this trouble could be avoided. For the first two or three days, until the situation settled down, all the available linemen and cable detachments were engaged in continual attempts to replace lines which had thus been swept away. It was not until the tanks ceased to take part in the fighting; on a large scale and the cavalry had retired out of the area of conflict, that the Division-Brigade routes could be relied upon.

Another difficulty which confronted the Divisional signal company was a direct result of the narrow frontage held by Divisions and the depth of the advance. The authorized scale of equipment of the cable sections and No. 1 section of a Divisional Signal Company had recently been altered from single to twisted "D5" cable. This alteration had been caused by the prevalence of metallic circuits which was one of the principal precautions against enemy overhearing. It incidentally decreased the cable carried by each cable wagon from ten miles to five miles of route and this had an unfortunate effect on_ the present occasion. The cable detachment endeavouring to advance with the Brigade to which it had been attached had to lay at least, two pairs if the route was to approach in any degree to the ideals of position warfare. His available cable supply permitted the erection of two miles of such route. For any further extension fresh supplies were required, and it was almost impossible to get forward supplies along the few congested roads, each crowded with the transport of one or two Divisions, and closed for several hours in each day for necessary repairs. Thus, one of the most important lessons learnt by the Signal Service in this action was the necessity for dumps of a good type of light cable — twisted "D3" for choice — well forward in the fighting area. Another lesson, based on the same experiences, was the advisability of a certain measure of priority being given to signal transport engaged in feeding the forward detachments and sections with cable and other essential supplies.

At the same time, some doubt was thrown on the newly-adopted policy of replacing single by twisted cable on the cable wagons. These were essentially likely to be used in fairly far-reaching advances only. Under such circumstances the maximum length of cable line was required and overhearing precautions were less necessary

One other result of the long twisted cables was that they made speaking very faint and often impossible. It was this fact which caused an unusual proportion of the Staff work to be sent over the lines as message traffic. Conversation was difficult and the energy expended upon it could usually be used to better advantage in the framing of orders and instructions in the form of messages which were dealt with by sounder, fullerphone, or in the last resort by buzzer. The overhead lines were pushed ahead by Corps and Army as rapidly ns possible, but here, again, congestion along the roads prevented a good speed being maintained.

The trouble with the Divisional line routes had, of course, the effect of much increasing the importance of the subsidiary methods of communication. Visual schemes were in existence with well sited stations right back to the immediate neighbourhood of Divisional headquarters. The country in general was open and eminently suited for the employment of visual on a large scale. Only in such places as were heavily wooded, was it impossible to use this method of signalling over long distances. The disorganization and demoralization of the enemy once more made two-way working possible The principal limit to the use of visual, where lines were not working, was often set by the disinclination of regimental commanders and Staffs to make full use of the stations provided.

The same drawback operated most seriously against the achievement of a full measure of success with the excellent chain of wireless communication that was also a feature of the signal system of most Divisions. Wireless was, indeed, used to a greater extent than ever before and was perhaps the main means of bridging the gap between Divisions and Brigades during the frequent intervals when the cable was down. A comparison of the totals of messages passing over an uninterrupted line system, with those dealt with by wireless stations keeping communication between the same headquarters when the lines were down, is most instructive. Some allowance must be made for the more cumbersome procedure of the latter method, but the totals of messages sent are even then ridiculously out of proportion. It was very rarely that the wireless sets, even on such favourable occasions as the present, were worked to full capacity, while it was seldom indeed that the signal office did not contain a pile of urgent messages awaiting the attention of the Morse operators even when the line system was working at its best. Wireless stations reported doing 16 and 20 messages a day and this was a great improvement on their record during previous operations. The waiving of field cipher regulations, however, would have much increased their usefulness. Officers would not bother to encipher messages, and if the wireless staff had to do so themselves it enormously decreased the speed of working of the stations. The relaxation of cipher and code regulations was clearly pointed to if this means of signalling was to be of full use in future mobile operations. In this action, its importance had been clearly demonstrated, while, at the same time, its usefulness had been crippled.

Power buzzers were useful during the initial stages of the attack, but the phase when they were valuable was a very short one. In an attack with a depth of 4^ miles their range was hopelessly out-distanced. Later, when the advance was slowed down, they could not be obtained from the Brigades to which they had originally been allotted. In some cases they could not be traced at all. The provision of accumulators also had not been organized on a sufficient scale. Another lesson of the Cambrai advance was that this must be arranged under Divisional auspices, if these instruments were to be of use during the fleeting phases of the action when the situation was momentarily favourable to them.

Of the various message-carrying agencies, motor cyclists, mounted orderlies, and runners, were all employed between Division and Brigade during the Cambrai operations. The roads were so congested and in such bad repair that the motor cyclists in some Divisions were of little use forward of Divisional headquarters. Mounted orderlies had been allotted to the Divisions, but not in sufficient numbers, and extra men had to be borrowed from the cavalry. In at least one Division a chain of runners was organized on the forward route between Brigade headquarters and Advanced Divisional headquarters and under the peculiar circumstances of the action this unusual step was a success.

Forward of Brigade headquarters, the enemy's line system was utilized to some extent to give line communication, but this again was much hindered by the action of our own troops. The system was much damaged by our own barrage and still more by the tanks which accompanied the advancing troops. In spite of this, however, lines were improvised, partly of our own field cable and partly of German cable, though the efforts of the forward parties were neutralized to a great extent by well-meaning infantry officers and men who cut British and German cables alike, in the effort to prevent information passing forward to the enemy. The cutting of British field cables might have been prevented by an organized attempt to familiarize the infantry with the types of cable in use by the British Signal Service, these being easily distinguished from the German field cables by anyone who gave them more than a cursory glance. The armoured cables of the two Armies were more alike, but even these could have been distinguished in great measure by the insulation tape wrapped round the outside of the German types. As it was, however, all cables were alike to an infantry intent only on doing as much damage as possible to the enemy circuits. Not only were they cut, but, with praiseworthy but annoying thoroughness, the severed ends were dragged and tangled into snarls which defied the British linemen.

Yet another lesson learnt during this most instructive battle was that enemy airline should not be interfered with at all by uninformed personnel, since it is extremely unlikely that the preliminary shelling has left a single wire continuous. Enemy cable, if touched at all, should be cut and left in position, so that the least work would be required should our own Signal Service desire to reconstruct portions of the enemy cable routes.

Where enemy lines could not be used, circuits of twisted or single "D2" cable were laid by the British forward parties but these lines were not immune from the attentions of cavalry and tanks and were much more exposed to damage from enemy shells than the rearward lines. Altogether, from the 20th to the 23rd — when the situation stiffened a good deal — forward intercommunication was mainly by runner, other subsidiary means being much less in evidence than usual.

All available wireless was concentrated upon the Divisional-Brigade routes, power buzzers and amplifiers had gone astray in some cases and been left behind in others, messenger dogs were out of the question in such mobile warfare, and pigeons were of less use than was usually the case.

There were two reasons which acted against the usual efficiency of the forward pigeon service: (1) The secrecy and short period of preparation prevented the concentration of lofts in the neighbourhood in good time to train the birds to their new area, and this decreased the available supply of birds; (2) the failure of forward units to return pigeon baskets, and the congestion of the forward roads, practically confined the effective use of pigeons to the earlier days of the battle.

The initial advance of the 20th and 21st of November was followed by a period of exploitation when local attacks with special limited objectives were carried out in the attempt to widen the salient occupied by the British troops. Increasing enemy resistance had slowed down the advance to a pace normal to previous position battles. The elimination of the cavalry and the comparative quiescence of the tanks combined with the increasing intensity of the artillery fire to give an aspect to the later stages of the battle which caused a. resemblance to the battles of Arras and Ypres. Buried cable on a large scale was not possible, and all efforts of the Signal Service were directed towards the consolidation of the existing poled and ground cable system and to the perfection of the alternative schemes of intercommunication. A useful variation of the normal forward Divisional system at this stage of the battle was a lateral connection between the forward end of the two Brigade forward routes so that Divisions were working upon loops of ground or poled cable. This doubled the security of the forward line communication for, while breaks in one "leg" or other of the loop were common, it was seldom that simultaneous breaks on both sides of either Brigade effected a complete severance of line communication.

During the 27th and 28th of November, the situation remained unchanged with maintenance increasing in difficulty under a gradually swelling volume of enemy fire. On November 29th, the enemy gunfire increased to the hurricane bombardment which preceded the great counter-attacks to the north and south of the salient which had been bitten deep into his line. The attack to the north presents few features of special interest from the point of view of the Signal Service. What little progress the enemy made was strictly local in its effects and signal communication was as good as could be expected over an overground system of lines under heavy artillery fire. Incessant activity of the maintenance parties secured at least intermittent line communication. Wireless, power buzzer and visual all helped to reinforce this.

To the south, however, the situation was much more grave, and at one time it looked as if retreat might become general and far reaching. Here, the Signal Service was faced with new experiences and suffered heavy casualties. Following a whirlwind bombardment, masses of the enemy surprised the British front line posts and, passing over them in their stride and without serious interruption to the speed of their advance, penetrated deeply and rapidly toward Divisional headquarters. "S.S. 148, and its successor S.S. 191, did not legislate for a retreat and the signal practice to meet the new conditions had to be improvised. Forward lines were cut in all directions, and at one time the fate of the Brigade and Divisional headquarters of more than one Division was in doubt. The forward signal system was perforce abandoned to the Germans and the rear system was largely destroyed with one significant exception.

The exception which proved the salvation of the situation was a two-pair emergency route of armoured cable which had been laid through Gouzeaucourt by a Division as an alternative to the multicore cable route provided for it by the Corps. This route providentially remained uncut throughout the operations and became the main channel of communication for no less than three Divisions. It was over this cable — at that time partly in the area occupied by the advancing German infantry — that the messages passed which apprised the Corps of the urgency of the situation and resulted eventually in the counter-attack by the Guards Division which restored the situation. On no single occasion during the whole war, was the value of an alternative route demonstrated in such a striking manner. The line passed in and out of German hands, but was overlooked by the enemy and remained through for two critical days.

By the date of the German counter-attack the pigeon service had been largely reconstituted, but the 2gth was misty and birds were much hampered in their flight by the fog. Many were away for several hours and the situation was complicated on the most critical portion of the front by the necessity of withdrawing a loft to prevent its capture by the enemy who had already overrun the headquarters of one Division. Visual was also interfered with to a great extent by the same mist, and the weather conditions on part of the front included a snowfall which made it difficult to locate breaks in the lines.

The alternative method which was most useful of all was wireless. The aerials of Brigade stations were shot away, but were re-erected again and again and urgent messages dealt with in considerable numbers. In the emergency, code and cipher regulations were to a great extent swept away. Extra wireless stations were rushed up to the supposed locality of missing Brigades and by this means the signal system was largely reconstructed without the use of lines, which were, however, laid as soon as correct locations were obtained

One interesting aspect of the situation as it affected signals was the absorption of a large proportion of signal personnel into the fighting line where they shared the fortunes of their infantry comrades. A particular instance which is deserving of record is that of a forward wireless station serving a battalion headquarters which was surrounded by the enemy. One of the operators remained in the dug-out to destroy the apparatus, the remaining two took their rifles and helped to man an adjoining trench. One was killed and one was captured. The third man, having thoroughly destroyed his station, joined the infantry and retreated with them fighting a rearguard action for 24 hours. He remained with them until the company was relieved, when he proceeded to join the signal company to which he belonged. It was such emergencies, occurring from time to time, which gave Signal Service personnel an opportunity of showing that they possessed the same qualities in action as the men of other fighting branches of the Army. The maintaining of lines under shell fire, the reckless riding of motor cycles over crowded slippery roads, the manning of insufficiently protected visual stations or signal offices, all called for courage of a high order. None had, however, the same effect of drawing the Signal Service nearer in sympathy to the infantry as such isolated instances when the wearers of the blue and white armbands threw aside their technical instruments and took up rifle and bomb to do their share of direct damage to a triumphant enemy. It is to the credit of the Signal Service that similar instances can be quoted whenever opportunity for such action occurred. It is equally to its credit that so few instances are recorded of unnecessary fighting by forward signal parties during an advance.

The experiences of the Guards Divisional signal company in the counter-attack which restored the situation, merely emphasized the lessons already learnt in the earlier phases of the battle. Special attention had been paid to mobility. All surplus stores and exchanges, many magneto telephones, all power buzzers and amplifiers, and all sounders but two, had been left behind before the company took over the signal office from which the signal system of the advance was extended. Here is seen the result of the experience of the previous few days. For almost the first time since the beginning of position warfare a signal company went into action stripped for battle. The result was a skeleton but highly efficient system in which none but essential needs were accommodated. The only difficulty proved to be in the moving of personnel for advanced signal offices. The men were actually taken forward in the 30-cwt. lorry. Had the roads been too bad to permit the lorry to be used, G.S. wagons would have had to be employed and the establishment of the signal offices would have lagged behind the general advance. The best solution was pointed out as the addition of a box car to the signal company. A further problem which was indicated as likely to arise in the event of an extended advance was the impossibility under present circumstances of keeping touch between Divisional artillery headquarters and the artillery Brigades. This had been met in some Divisions by the attachment of a cable detachment from No. 1 section for artillery use only. The place of this detachment in the Divisional scheme of communication was sometimes supplied by the attachment of a cable detachment from the Corps. This was the method usually adopted in the advances of the following year.

A review of signal practice during the Cambrai battle shows at once that, even if the greatest interest the operations held for the Signal Service was the secrecy and thoroughness of the preparations, yet their greatest importance lay in the lessons drawn from the near approximation to mobile warfare, both in the advance and the retreat. It was, indeed, singularly fortunate, bearing in mind the type of the two decisive phases of the war in 1918, that the Signal Service should have had such an opportunity of studying both problems on a comparatively small scale as regards space, time, and degree of decision. Many adjustments were made as the result of the relatively small action described in the foregoing pages. In particular, one of the principal outcomes of the battle was a report which embodied in its main outlines practically the entire policy which was later to be adopted by the forward signal units during the advance in October, 1918.

Little has yet been said of the special signal system which was rendered necessary by the evolution of the Tank organization which played a prominent part at Cambrai and at all the subsequent British offensives. The growth of Tank signals is, perhaps, best outlined in the present connection, since the use of tanks was the outstanding feature of the battle as originally planned. In many ways, forward Tank signals were necessarily experimental. Tank wireless, indeed, was in the nature of an "experiment inside an experiment." It was, therefore, liable to constant adjustment and change, and doomed to many initial failures in the early days of its development.

The development of the Tank signal organization and practice divides naturally into two distinct parts. One system was required for intercommunication purposes between the headquarters of the major Tank formations and units. The other, and more specialized (and therefore more interesting) system was that required to keep touch between the fighting tanks, the supply tanks, and the infantry and forward artillery, with which they worked.

The question of special signal communication for tanks first arose after their initial use in the battles of the Somme in 1916. In these early operations, tanks were only partially successful and, amongst other things, the need for a greater degree of intercommunication was proved. As the result of the experience of these battles the first provisional establishment of signal stores was proposed in December, 1916, when the Director of Signals put forward the suggestion that each battalion should be given two trench telephones and each company one trench telephone, and that tank battalion headquarters should be provided with a live-line magneto exchange. To work this primitive system it was proposed to attach four signallers to each company and eight to the battalion. In January, 1917, with the formation of a headquarters and the brigading of the tanks as Brigades of the Heavy Branch, M.G. Corps, a further establishment was proposed of one sergeant and six telephonists for headquarters and for each Tank Brigade. The Army Signal Service was still to provide and maintain all lines.

With the rapid extension of the use of these machines and the formation of the Tank Corps in 1917, it soon became clear that the original establishments were hopelessly inadequate and that Tank signal units must be formed on a scale commensurate with the growing importance and independence of this arm. The whole question was reconsidered by the Signal Directorate and as a result the formation of special Tank Corps signal units was approved in principle and recommended. The final form taken by this reform was a headquarters signal company at Tank Corps headquarters and Tank Brigade signal companies with all Brigades. The Tank Brigade signal company was planned somewhat on the lines 01 a Divisional signal company and possessed a headquarters with four sections for dealing with each of the battalions of the Brigade.

As the essence of the employment of Tanks lay in their sudden concentration wherever offensives were to take place, it was impossible to allot to them any recognised system of lines in each Army. Their needs were, however, accommodated on the general command and administrative systems of Armies, Corps and Divisions according to the temporary location of the Brigades. These lines were usually in the maintenance of the formation to whose system they belonged and in consequence only a small proportion of the new Tank signal establishments consisted of lint men. By far the greater proportion of the rear units consisted of operators, and of a pool of wireless operators and telegraph operators "B" who were held in readiness to man the wireless stations which were the mainstay of the Tank signal system during operations. Another feature of Tank signal units which was also directly dependent upon the particular conditions to be met was the unusually large proportion of instrument repairers contained in the establishment. Signalling instruments are of delicate constitution and not well adapted to withstand without injury the rough usage inevitable in the interior of a wildly-gyrating machine whose chief title to fame is a disregard for obstacles and unevennesses in its path. Breakages were, therefore, unusually numerous and an efficient staff of instrument repairers was required to minimise damage as much as possible.

The reorganization which brought the establishments of tank signals to their final form was also marked by the appointment of an A.D. Signals Tank Corps, to supervise and co-ordinate the whole. The desirability of a central signal authority with high rank and considerable power had been shown by the lack of efficient liaison between the Tank Corps and the Signal Service in the battle of Cambrai. The destruction of lines by wandering Tanks during that action did incalculable harm to an otherwise efficient signal system. The future employment of Tanks on a large scale was assured by their success on this occasion, and efficient measures had to be taken at once to prevent similar disasters in the future.

This was done by mutual arrangement. Improvements in methods of laying lines, earlier dispositions and a better system of dissemination of information about available crossings, voluntary restriction by the Tank Corps of lateral movement of Tanks in the area between the F.O.O.'s and Brigade headquarters, all assisted to remove a dangerous menace to intercommunication in battle. The appointment of A.D. Signals, Tank Corps, made certain of efficient liaison in future. The multiplication of crossings, the use of slack ground cable or cable laid in open shallow trenches, and special arrangements for maintenance during actions in which Tanks were employed, again minimised the danger to a still greater extent. The success of the precautions taken is witnessed by the fact that these machines never again caused a wholesale disruption of the signal system of any of the higher formations.

In some respects the evolution of Tank signal methods lay along the lines already familiar from a study of the general evolution of stationary warfare signal practice. The exceptions of interest were all concerned with the solving of intercommunication problems during the actual fighting. In the training area, communication was normally by telephone and telegraph, the lines serving the higher formations being allotted from the ordinary signal system, those serving the lower units being laid as required by the Tank signal personnel.

It was in battle that some special signal system was needed. By their very nature the Tanks, to be of any use, must have their place in the forefront of the action. They advanced to the attack and fought under circumstances where the protection afforded by their armour alone enabled them to survive. The ideal signalling apparatus was therefore that which could be operated from within the Tank itself.

In the early days of 1916, forward Tank signals were conspicuous by their absence rather than by their efficiency. A simple visual code and the use of red, green and white discs displayed on the side of the Tank were the first primitive means employed. The infantry accompanying them replied to their signals and indicated their own wishes by gesture with rifles or with helmets held on the end of their bayonets. These methods were later reinforced by the Aldis lamps which gave very good results but which could not be answered by the forward troops without attracting enemy attention. A later innovation was the use of Tanks fitted with ploughs for burying cable. The parts of the discarded mechanical excavators and cable ploughs were borrowed from the Signal Service for experimental purposes.

The use of pigeons from Tanks was first attempted on a large scale in the offensives in the summer of 1917. They were so successful that in future operations all Tanks were supplied with these birds. Their employment from Tanks, however, does not involve any novel features other than a special organization to ensure that the higher Tank formations secured their own messages in good time. This was arranged by the employment of Tank Corps despatch riders at the lofts. The prevalence of grease and oil in the interior of the Tank and the shaking experienced by the birds were the greatest obstacles to the new departure, but they were successfully surmounted, and good averages both as regards messages and items were achieved.

In June, 1917, a distinct departure in Tank signal practice was initiated by the experimental use of special signal Tanks with wireless, visual and pigeons. A variety of experiments with wireless followed and, in October of the same year, continuous wave stations were employed in battle. Old out-of-date Tanks of the Somme period were first fitted out as "wireless" Tanks and these were sufficiently successful to secure for continuous wave wireless a permanent place in Tank signal organization.

The next step was the use of Tanks as carriers for wireless stations which were set down at suitable places in "No Man's Land." Later, when this method had not proved altogether satisfactory, wireless aerials were fitted to modern Tanks. Special short masts were devised which could be lowered on a swivel when railway journeys were undertaken. The sets themselves were carried within the Tank and a small dynamo fitted to the engine provided the electrical power. Wireless sets were used in the battle of Cambrai with good effect. They were of the ordinary type in use with the artillery and were carried forward by fighting tanks. They proved to be too immobile when they had been set down by the Tank and in addition suffered from the disadvantage common to all forward wireless, the fact that it was necessary for the officers and men with the stations to solicit messages. What messages they did receive were, however, despatched promptly. In this action and the actions of the following summer, continuous wave wireless with the Tank Corps justified itself to such an extent that, in August, 1918, 288 C.W. sets Mark III. and 96 120 watt sets were asked for for the Tank Corps alone.

The last development in Tank signals occurred in July, 1918, when experiments with wireless telephones were carried out successfully both between Tank and Tank in the field and between Tank and contact aeroplane. Had the war continued into the next year,, wireless telephony would doubtless have played a prominent part in this as in some other departments of army signals.

The battle of Cambrai was followed by a short lull which lasted through the winter months of 1917-1918, to be succeeded on March 2 1st by the German offensive which once more entirely changed the aspect of the war. This period, otherwise quiescent, was marked by a great increase of enemy bombing and long-range shelling which gave considerable food for thought to the officers responsible for the signal communications in rear of Divisional headquarters.

The general tendency in 1917 had been to continue the grid system of buries to the rear with a chessboard system of heavy semi-permanent and permanent airline routes. This policy was already giving a characteristic appearance to those portions of the front where the situation had been more fluid and reconstruction of communications had been made necessary by the advance of the Armies. Now, however, the system had to be modified considerably by this new factor which was bulking more and more largely in the minds of signal officers.

In the summer of 1917 a great extension of enemy long-range shelling in the rear portions of Divisional areas first made itself felt. This was accentuated by a disposition evinced by his artillery to search out Divisional headquarters and to pay special attention to the main nodal points of the fine system. This situation was countered, as has been shown, by the rearward extension of the buried system to the neighbourhood of Divisional headquarters. In Divisions where the situation did not permit of this, nearly every day was marked by the destruction of one or other poled cable or airline route by a direct hit from a shell or the blast from a projectile falling in the immediate vicinity. Signal camps and signal offices also began to suffer, and casualties to signal personnel in rear areas increased considerably. In the autumn of 1917, the enemy bombing squadrons commenced to single out Divisional headquarters for special attention and, after considerable damage had been caused, the order was issued that sufficient bomb-proof shelters must be constructed to house all ranks.

About this time, however, it became evident that the enemy's attention was not going to be confined to Divisional areas. His long-range guns commenced work on the Corps communication system and on Corps headquarters, while his long-distance bombing squadrons began the first of the raids which were destined throughout 1918 to add a distinct element of uncertainty to fife in the back areas generally. The effect on the rear signal system was at once evident. The system had been built up entirely with regard to two considerations; economy of personnel and stores, and convenience to subscribers. Heavy main routes were the rule and these were usually carried through, or terminated in, camps and towns, where billeting accommodation was plentiful and where, therefore, the administrative units had their headquarters. These were naturally the spots raided by the enemy airmen, and the lives of the operators housed in unprotected signal offices became harassed and anxious. Communication was often interrupted by the destruction of whole routes or of heavy junction poles in towns and villages. Signal offices were not safe, and at any time a wholesale catastrophe might paralyse the administrative services of the whole army.

Breakdown parties, who had until now been accustomed to judge the amount of their work by the state of the weather, were kept in a continual state of activity. In the past, wholesale damage had been done by storms of unusual fury or even by heavy accumulations of rime, hoarfrost, or snow, upon the wires. These visitations were, however, rare occurrences; the visits of the enemy bombing machines depended solely on the state of the moon and took place many times each month. Signs were not wanting, either, that they would shortly become independent of this controlling factor and be of nightly occurrence in fine weather.

The situation needed to be taken in hand at once and a policy was outlined which aimed at four ideals. These were, (1) the adequate protection of signal offices, (2) the burying of leads in and out of signal offices for considerable lengths, (3) the diversion of routes round towns, and (4) the substitution of fighter parallel routes 150 yards or more apart for the heavier routes which had been the rule until this date.

The protection of the personnel in signal offices was achieved in several ways. In tents, hutted camps, or other temporary buildings, splinter-proof walls three feet high were raised all round the billet to give protection against the spreading splinters of the bombs. In the case of brick and stone buildings of considerable strength, additional protection was achieved by building up windows with sandbags, and sometimes, also, by the deposit of extra layers of earth and cement upon the roof of the signal office. Basements were utilized whenever possible and, where suitable cellars existed, signal offices were always established downstairs in preference to the better-lighted and more airy rooms on the ground floor.

In important offices the problem of the protection of intercommunication facilities and instruments ranked only second to that of protection of personnel, and, during those periods in 1918 when stationary warfare conditions held sway, Army signal offices were sometimes duplicated in heavily protected dug-outs. The wiring of these was ingeniously arranged in every way as a duplicate of the signal office above stairs and, in the event of a raid, the whole of the essential circuits could be switched through to the subterranean signal office by one movement. Dislocation of traffic was thus limited to a period of minutes only. Similarly, at G.H.O., a duplicate signal office was built in the ramparts of Montreuil, and on completion the office staff moved in and the old office was abandoned.

The protection of the permanent routes was assisted by an order that future routes would not exceed 28 wires, thus making necessary two or three separate routes for all the superior formations. By this means alternative circuits were assured, the principle aimed at being to make it impossible for one bomb to sever entirely the vital communications of any formation without registering a direct hit on the signal office itself. Large junction poles were a great source of danger and were done away with whenever possible, their place being taken by protected dug-outs into which the lines were led by cable buried for 200 yards on either side. In the same way, buried multicore cable of one description or another was used to terminate the routes, which were also in many cases rearranged so as to deceive any enemy desiring to use them as a clue to the location of the signal office into which they ran. Diversions round all main towns were also contemplated and in urgent cases were carried out, cases in point being those round St. Pol and Amiens. These caused the Lines of Communication signal units a great deal of heavy extra work in 1917 and 1918.

Finally, special arrangements were made for maintenance with a view to reducing the dislocation caused by catastrophic interruptions as much as possible. The maintenance personnel was increased in number and redistributed. Special lengths of two-pair interruption cable 150 yards long were prepared to put through emergency circuits, and a re-echo of the experience of 1914 is heard in the instructions that in future the most important circuits should be given positions upon the top arms.

The period between the first great reorganization and the retreat in March, 1918, was marked by two main changes in organization. The first was the reorganization which took place as a direct result of the position battles of 1917, the second was the formation of a special section of the Divisional signal company to deal with the intercommunication of the machine-gun battalion which now formed an integral portion of each Division.

At the time of the first great reorganization, the Director of Army Signals was given to understand that no further great increase to the Signal Service could be permitted. Any future reorganization which involved an increase in one unit must be compensated for by a decrease in another. When, therefore, the position battles of the summer of 1917 and, in particular, the operations which ended in the capture of Passchendaele, betrayed clearly an absence of flexibility in the position warfare signal service, the necessary adjustment had to be made by a redistribution of personnel amongst the signal units. The great need was more constructive power on active fronts, and it was decided that this would have to be made good at the expense of the formations in the more quiescent areas.

A review of signal units showed one direction, only, in which a considerable saving might be affected without loss of efficiency. The question of the loss of man-power due to the horses of the cable section has already been referred to in connection with No. 1 Section of the Divisional signal company, and the reasons which turned the scale against a radical reorganization of these sections have been discussed. The need for mobility was not, however, of such paramount importance in the case of construction and maintenance personnel attached to Corps and Armies. At any rate, it was quite evident that under present circumstances and, indeed, under any circumstances that could be foreseen, certain of the cable sections could be spared. At the same time, a review of the establishments of both airline and cable sections showed ways in which the individual units might be slightly reduced in strength.

It was therefore decided to attempt a considerable increase in the number of airline sections — the most mobile construction signal unit in country where motor transport could be used — at the expense both of the number of cable sections and the strength of the individual airline and cable sections.

A table showing the strength of the Signal Service on September 1st, 1917, at the time when the question of this reorganization first came under consideration is given below.

Table V
Unit.	Number.	Estab.	Total.	Total.
G.H.Q		3,665	3,665	3,665
"L" Signal Battn.
Pigeon Service
Signal Depot
Cavalry Corps Signals		899	899	899
Tank Corps Signals		436	436	436
Army Signals H.Q.	(5)	274	1,370
Airline Sections	(15)	61	915
Cable Sections	(15)	38	570
Area Signal Detachments	(40)	16	640
A.F.A. Bde. Sig. Sub-sections	(47)	20	940
Extra for Corps above 3			50
Construction Company	(5)	117	585
Light Rly. Sig. Sections	(5)	41	205
Intelligence W/T Groups	(6)	47	282	5,557
Corps Signals H.Q.	(20)	140	2,800
Airline Sections	(20)	61	1,220
Cable Sections	(64)	38	2,432
Corps H.A. Sections	(20)	37	740
H.A. Group Sub-sections	(90)	27	2,430
Extra for Divisions above 3			154	9,776
Divisional Signals	(62)	289	17,918	17,918
Grand total (including attached), 38,251

It will be seen that at that date 79 cable sections and 35 airline sections were distributed between five Armies and 20 Corps. On the new basis it was decided to give to each Army and each Corps the two cable sections which would suffice for their needs and to abolish the remaining 29. The men thus produced by the abolition of these units, together with those found by the slight reduction of the individual establishments, were then pooled and formed into airline sections on the new establishment, thus increasing the number of the latter from 35 to 63. The new airline sections were then re-allotted to formations, each Army taking three and each Corps two. The remaining eight were kept in reserve as a G.H.O. pool for employment as considered requisite from time to time.

By means of this reorganization about half of the constructive power of the signal units with the Expeditionary Force could be moved about without upsetting the situation on the quiet fronts, and the value of the reconstruction was to be very clearly perceived during the campaigns of the last year of the war.

A study of the foregoing rearrangement will show at once that a saving of a considerable number of officers was effected, but this was to a certain extent offset by the creation of certain specialist officers which had now become a matter of great importance. It was the saving incidental to the reconstitution of the signal units which permitted of the appointment of an officer in each Army as D.A.D. Signals, of a second wireless officer for each Corps, and of a subaltern for each heavy artillery section. The urgency of all these modifications was incontestable and they were all approved. In addition, the new reorganization included the grouping of Army signal units under the Army signal company and the appointment of one of the officers of that company as Adjutant to assist the O.C. company in his administrative duties.

The final result of the above redistribution is shown in the following comparative table of the increases and decreases involved.

Table VI
Decrease.			Increase.
4	officers		25	box cars
794	other ranks		25	3-ton lorries
362	horses		16	bicycles
145	cable and limber wagons
26	30-cwt. Lorries
29	motor cycles

At the same time, a review was made of other signal units and it was seen that certain small economies in men and horses could be made which resulted in a net saving of a further 104 men and 124 horses. It was found necessary, however, to add certain tradesmen to G.H.Q. and Army signal companies, while the field artillery signal sub-sections to which no transport had been allotted and which had been greatly handicapped by this fact, were accommodated with a limbered G.S. wagon and the necessary horses and drivers.

Other amendments carried out at the same time were the addition of a box car to the Divisional signal company, and the substitution of the 30-cwt lorry at present on the establishment of that unit by a 3-ton lorry The latter had been made necessary mainly by the great increase in the wireless equipment of a Division which could no longer be carried in the P.E.L. lorry as had formerly been the case.

The second reorganization referred to in an earlier paragraph — that of the Divisional signal company to improve machine-gun signals — involved a small addition only to each signal company, though the total increase of establishment was not inconsiderable. The additional men were, however, offset in great measure by a reduction of the signallers already with the machine-gun battalion. The machine-gun signal system in a Division was comparable with that of an infantry Brigade and the old state of affairs when all hues were run by the M.G. Corps personnel was not satisfactory. The new section of one officer and 20 other ranks was needed, if the machinegun communications were to become incorporated in the Divisional system. Certainly much greater economy both of material and of personnel might be expected from such an amalgamation.

In the event, the new scheme was not always successful. It depended very much on the personality of the machine-gun signal officer who had been transferred to the Signal Service as part of the reform; upon the relations of the latter with his commanding officer, and upon the extent of his willingness to identify himself with the signal company to which he now belonged. There is no doubt that the absorption of the machine-gun signals was an unqualified blessing to the efficiency of the signal system generally, but the unfavourable reports of various officers should be given due prominence in an unbiassed examination of the result of the change. Certainly, the additional personnel was of great value to the O.C. Divisional signals in the forthcoming mobile warfare. Equally certainly, in individual cases, the machine-gun battalion was not so well served under the new conditions. Weighing both sides of the question, however, the balance was strongly in favour of the reform.

The period during which these reorganizations took effect was also a period of thorough revision of all existing establishments throughout the Expeditionary Force with a view to economy of man-power. The British Armies in France had reached their maximum size and the available manhood of the Empire was insufficient to replace casualties and keep up the strength of the fighting forces on the various fronts. The situation had been relieved to a great extent by the entry of America into the fray, but was still grave, and every effort was made to keep the administrative services within reasonable bounds and to free more men for the infantry. The Signal Service shared in the careful scrutiny given to these services and in August, 1917, an Army signal conference had been appointed to consider ways and means of reducing the signal facilities given to the Armies in the field. With the D.D. Signals of an Army as President, an officer of the Adjutant General's Staff and an officer appointed by the Director of Signals as the other members of the conference, the interests of the Signal Service were safeguarded, while the man-power question was yet given due weight in a dispassionate examination of the facts.

The conference passed in review the whole question of signals in relation to the Staff. Signal communication was discussed under three heads: — (1) The delivery of written messages to the addressee (D.R.L.S., orderly, runner, etc.); (2) the reproduction of the written message for the addressee (phonogram, telegram, fullerphone, visualy wireless, etc.); and (3) methods where no record is kept (telephone). The advantages of the three methods were tabulated: the slow speed but accuracy of the first, the rapidity and relative accuracy of the second, the supreme rapidity but relative inaccuracy of the third, where, however, the latter quality is offset to a great extent by the freedom of discussion made possible. It was decided that the first method was needed to carry all matter too urgent for post but not exceptionally urgent, that the second was equally needed for all very urgent traffic, and that the third should be reserved exclusively for conversations between staff officers. This left the situation very much as it was before the conference was held. An examination of lists of officers who were connected as subscribers to the telephone system showed no way in which economy could be achieved. The policy of giving as few direct lines as possible was confirmed and emphasized. In effect, the conference was unable to suggest any radical decreases of establishment as justifiable or desirable. The result was to vindicate once and for all the conservative policy as regards increases of establishment which had been pursued consistently by the Signal Directorate, who had always subordinated the interests of the Signal Service to those of the Army as a whole.

Scrutiny of signal establishments was not confined to those of the Armies, but was exercised also on the units working the Lines of Communication signal system. Here, also, no case could be made out for decreases of establishment, but the recommendation was made that still further women telegraphists and switchboard operators should be employed. The employment of women on outdoor services was deprecated by the Deputy Director of Signals, at any rate until considerable further savings should have been made in the home signal establishments which appeared to be unduly swollen in comparison with those of the Signal Service overseas.

Indeed, the substitution of the indoor tradesmen on the L. of C. by women was itself not easily completed, since the terms of service offered did not attract women with suitable qualifications and educational attainments. The employment of native labour on the L. of C. and even as drivers with the Divisional signal companies, was advocated by the Signal Service itself and might have taken place had stationary warfare been much further prolonged. The employment of Boy Scouts as orderlies in rear offices had been recommended twice by the Director of Signals but refused by the home authorities. The use of women further forward than on the Lines of Communication was rendered impossible by lack of suitable accommodation and for other reasons.

The final outcome of the struggle waged by the advocates of expansion and reduction was thus to leave the Signal Service much as it was before, if anything slightly increased.

Footnotes

«--	The Signal Service in the European War of 1914 to 1918 Chapter XIII	--»
Chapter XII		Chapter XIV