Chapter XIII: Cambrai

I.

The Tank Corps had to accomplish an immense amount of work in a very short time before the battle of Cambrai. This preparation consisted of five main features:--

1. The arrangements for secrecy.

2. The concentrating of three brigades of tanks at Wailly, and
their subsequent dispersal to detraining centres behind the
Cambrai front.

3. Rehearsals with the infantry of the plan of attack.

4. The construction of certain devices for crossing the wider
trenches of the Hindenburg Line, for bringing up supplies, and
for clearing the wire for the cavalry.

5. The establishment of petrol and ammunition dumps.

We will take these features in order.

The whole 3rd Army was involved in the question of secrecy. The bringing up of numerous additional batteries to positions close behind the front line, the assembling of several cavalry divisions, and all the thousand and one preparations for an advance on a wide front had to be completed without a suspicion reaching the enemy. This was accomplished successfully. The normal artillery activity was maintained, while dozens of new batteries were brought up and camouflaged. None of these were able to register before the battle opened. The Tank Corps, meanwhile, had its own especial difficulties. The withdrawal of one brigade from the Salient and another from the Lens front could not be wholly concealed, but the myth of winter training deceived the enemy, if he heard of it, as thoroughly as it deceived the troops concerned; and the activities of one or two wireless tanks, left at the front to send bogus messages, easily intercepted, to imaginary brigade headquarters, fostered the belief that the latter had not moved. The forward H.Q. established with the 3rd Army at Albert was disguised as the Tank Corps Training Office. In the beginning only four members of General Elles’s staff knew of the forthcoming attack. The information was next revealed to battalion and company commanders and reconnaissance officers; and last of all, only two or three days before 20th November, zero day, section and tank commanders were admitted to the secret. Reconnaissance was carried out with every precaution, distinctive badges being hidden or removed, and senior officers visited the front disguised in steel helmets and trench-coats.

Thirty-six tank trains were needed for the almost simultaneous removal of the nine battalions. Our supply of K.T. trucks being insufficient, it was necessary to borrow a number of the smaller French type. A vast amount of work was thrown on the railway staff, all the entraining and detraining being carried out at night. In four or five nights upwards of 450 machines were entrained at Beaumetz. Twenty-seven train-loads were concentrated at Plateau Station, near Albert, by the 14th November, and dispersed to their final detraining points during the three following days, the remaining nine trains reaching Plateau on the 15th. Only the most careful management could have succeeded in disposing of these slow and heavy trains, amid all the normal traffic, in so short a time without a hitch. The 1st Brigade detrained finally at Bertincourt and Ruyaulcourt, behind Havrincourt Wood, the 2nd Brigade at Sorel and Ytres, to the south, and the 3rd Brigade at Heudicourt, near Epehy. At all these railheads new ramps had to be built or old ones repaired. With one or two minor accidents, 378 fighting tanks and 98 administrative machines were assembled at points from two to four miles behind the lines on zero minus two days--18th November.

The plan of attack and its rehearsal with the infantry at Wailly cannot be described more lucidly than in the account given in ‘Weekly Tank Notes,’ the official publication already referred to. I will quote this account entire.

“Before the infantry assembled for training a new tactics had
to be devised, not only to meet the conditions which would
be encountered, but to fit the limitations imposed upon the
tank by it only being able to carry one tank fascine. Once
this fascine was cast it could not be picked up again without
considerable difficulty.

“Briefly the tactics decided on were worked out to meet
the following requirements: ‘To effect a penetration of
four systems of trenches in a few hours without any type of
artillery preparation.’ They were as follows:--

“Each objective was divided up into tank section attack areas
according to the number of tactical points in the objective,
and a separate echelon, or line, of tanks was allotted to each
objective. Each section was to consist of three machines--one
advance-guard tank and two infantry tanks; this was agreed to
on account of there not being sufficient tanks in France to
bring sections up to four machines apiece.

“The duty of the advance-guard tank was to keep down the
enemy’s fire and to protect the infantry tanks as they led the
infantry through the enemy’s wire and over his trenches. The
allotment of the infantry to tanks depended on the strength of
the objective to be attacked and the nature of the approaches;
their formation was that of sections in single file with a
leader to each file. They were organised in three forces:
trench clearers to operate with the tanks; trench stops to
block the trenches at various points; and trench supports to
garrison the captured trench and form an advance-guard to the
next echelon of tanks and infantry passing through.

“The whole operation was divided into three phases: the
assembly, the approach, and the attack. The first was carried
out at night-time and was a parade drill, the infantry falling
in behind the tanks on tape lines, connected with their
starting-points by taped routes. The approach was slow and
orderly, the infantry holding themselves in readiness to act
on their own initiative. The attack was regulated so as to
economise tank fascines, and was carried out as follows:--

“The advance-guard tank went straight forward through the
enemy’s wire, and turning to the left, without crossing the
trench in front of it, opened right sponson broadsides; the
infantry tanks then made for the same spot, the left-hand one,
crossing the wire, approached the trench and cast its fascine,
then crossed over the fascine, turned to the left, and worked
down the fire trench; the second infantry tank crossed over
the fascine of the first and made for the enemy’s support
trench, cast its fascine, and crossing did likewise. Meanwhile
the advance-guard tank had swung round, and crossing over
the fascines of the two infantry tanks, moved forward with
its own fascine still in position ready for the third trench
line. When the two infantry tanks met they assembled behind
the advance-guard tank and awaited orders. The leading wave of
infantry planted a red flag at the gaps in the wire.

“In training the infantry the following exercises were carried
out:--

(1) Assembly of infantry behind tanks.

(2) Advance to attack behind tanks.

(3) Passing through wire crushed down by tanks.

(4) Clearing up a trench sector under protection of tanks.

“To enable them to work quickly in section single files and
to form these into section lines a simple platoon drill was
issued, and it is interesting to note that this drill was based
on a very similar one described by Xenophon in his ‘Cyropædeia’
and attributed to King Cyrus (_circa_ 500 B.C.).”

The fascines above mentioned were designed to assist the tanks over the wider trenches of the Hindenburg system. Some of these were known to be 12 feet wide and of great depth, the actual width being increased by the inward slopes of a parapet and parados 2 feet 6 inches high. A Mark IV. tank could only cross a trench 10-11 feet wide, a fact of which the Germans were well aware. The real obstacle was the depth of the trench, as when the tank was halfway over, but not yet balanced on the parados, the tail left the parapet and dropped to the trench floor, shifting the centre of gravity and leaving the machine at an angle so near the vertical that the tracks could not grip against all the deadweight below. The object of the fascine, therefore, was not to fill the trench and form a bridge, as seems commonly to be supposed, for this was both impossible and unnecessary, but to keep the tail of the tank from dropping so far that the machine was unable to recover. These tank fascines were made by binding together with chains, under great pressure, seventy-five ordinary brushwood bundles 10 feet long. Each big fascine was 4 feet 6 inches in diameter and weighed a ton and three-quarters. This monstrosity was carried on the cab of the tank, and was dropped into a trench over the nose by means of a release mechanism from within. It was an abomination to man-handle, fifteen to twenty Chinamen being required to push each fascine through the mud at Erin; and during a train journey it had to be lowered on the truck and hoisted up again before detrainment. But in action the fascines fulfilled their purpose admirably. The widest and deepest Hindenburg trenches proved no obstacle at all.

Four hundred of these fascines were made at Central Workshops during the feverish three weeks before Cambrai. This involved the trimming to the correct length of 30,000 ordinary fascines, and the welding of 2000 fathoms of steel chain. Eighteen tanks, working in pairs, pulled on the chains in opposite directions after they had been wound several times round the huge bundles. In the same period were constructed 110 heavy timber sledges, to be towed by tanks with loads of petrol and ammunition. 127 tanks were repaired and issued, and a number were fitted with large grapnels for wire-clearing. Owing to the necessity of secrecy, the personnel at Erin were not informed of the extreme urgency of these tasks; yet during these three weeks the shops were working for 22¹⁄₂ hours out of every 24 without a break. Nor was this pressure confined to the workshops. Between the 10th and 25th November, 28 lorries covered between them 19,000 miles, or 744 miles per lorry, and some small box-cars averaged 1200 miles apiece.

Three main supply dumps were established at the lying-up points of the three brigades--Havrincourt Wood, Dessart Wood, and Villers-Guislain and Gouzeaucourt. Among the stores collected here were 165,000 gallons of petrol, 55,000 lbs. of grease, 5,000,000 rounds of S.A.A., and 500,000 rounds of 6-pounder shell. The splendid system of light railways behind the 3rd Army front alone made this huge accumulation possible in so short a time. In the field, stores were to be carried from these dumps by supply-tanks and gun-carrying tanks towing sledges.

This brief account will give some idea of the work required before a full-dress tank battle, and of the sudden problems with which the technical staff had to grapple at short notice. It must be remembered that there was no precedent whatever for this operation. Four times as many tanks as went in on the first day at Ypres were to be employed at Cambrai. There had been little time for preparation and less for reconnaissance. The Hindenburg trenches, until they were met and conquered, at least were impressive impostors; and there was a great weight of responsibility, as well as a sense of gratification, attached to the fact that for the first time in the career of the corps, tanks were to be the dominating factor in a big battle. Their method of use determined the whole tactical scheme. Confident though all ranks might be of gaining that triumph to which they had looked forward through a series of discouragements, lasting over many months, the importance of the occasion brought infinite anxiety and strain.

On the other hand, everything, at long last, had been done for them that could be done. They were to have their own way on ground of their own choosing. For over a week at Wailly tanks and infantry rehearsed over the old German trenches the exact tactics they were to employ over the new ones. To give the infantry additional confidence, they were asked to construct their own defences and wire entanglements, which the tanks then proceeded to obliterate. The deep German front line at Wailly, enlarged to a veritable gulf by the passage or ditching of innumerable tanks during previous months of training, was an admirable model on which to prove the efficiency of the fascines. The masses of rusty wire provided practice for the special machines whose grapnels were to clear the way for the cavalry. Short though the time was, nothing was left undone. Tank personnel and infantry practised together and talked together, and for the first time was reached an approximation to that mutual trust and understanding which these two arms had so long and so especially required.

II.

An examination of a layered map shows very clearly the conformation of the country between the Hindenburg Line and Cambrai. The 3rd Army, lying south-west of that town, was attacking north-east. The battlefield can be divided into two parts, north and south of a line drawn westward from Marcoing through Flesquières along the Flesquières-Havrincourt Ridge. The southern part is a triangle, bounded to the north by this ridge, to the east by the Canal de L’Escaut, which zigzags southward from Cambrai, and to the south-west by the Hindenburg Line, which almost touches the canal near Banteux. In this part of the field we should be fighting downhill to the canal valley. The northern part, beyond the Flesquières Ridge, is lowish ground dominated by one prominent landmark, Bourlon Hill and Wood, and delimited to the west by the Canal du Nord, which the Hindenburg trenches crossed immediately north of Havrincourt Wood, thence following its course northward to Mœuvres. The Flesquières Ridge itself is an important tactical feature, running from east to west against the general lie of the country and overlooking the whole northern part of the battlefield; and where it was crossed by the German line, west of Havrincourt, the latter made a pronounced salient, taking in a portion of Havrincourt Wood. This portion was held as an advanced position in front of the trench system, and farther south the hamlet of La Vacquerie and other posts were strongly occupied in a like manner. The Hindenburg Line itself, taking it correctly as the first line of trenches, was sited on reverse slopes, hidden from our observation, and covered in front by a dense belt of wire, nowhere less than 50 yards deep. The Hindenburg wire was, in fact, a remarkable sight. You saw it winding away over the ridges in three broad belts, clean cut and separated by narrower avenues of grass; miles upon miles of it, with great patches red with rust after any rain, and looking in the distance where it climbed some final crest to the horizon like zigzag stripes of plough or trimmed and level vegetation. Behind the front system of trenches were three others, the Hindenburg Reserve, wired almost as thickly as the first line, and two more. These last were not so formidable. The series varied in depth from three and a half to four and a half miles. The front line and reserve were pierced in every bay by the shafts of deep dug-outs, although some of these had not been finished. In places they were joined up to form considerable tunnels, forty feet underground. The chalk of which the higher land is formed, although hard to work, is admirable for such subterranean use. The shafts and dug-outs were timbered, walls and roof, and when, a year later, I lived in several of them, were always dry. The whole system was knitted together by a maze of communication trenches.

The third and fourth lines, which in fact formed one system called variously the Marquion, Masnières, or Beaurevoir Line, crossed the Canal de L’Escaut south of Cambrai and covered another important tactical feature, the Niergnies-Seranvillers Ridge, east of the canal. This ridge, like that of Flesquières, overlooked all the northern part of the battlefield as far as Bourlon, making any attack on that height from the south a matter of great difficulty. The Niergnies Ridge, therefore, was to be carried if possible as soon as the canal crossings had been seized, but this effort proved beyond our resources.

The infantry at the disposal of Sir Julien Byng for his main attack amounted only to six divisions, with two cavalry divisions, and upwards of 1000 guns, for a front of six miles from east of Gonnelieu to a point north of Havrincourt Wood. Subsidiary attacks to secure the flanks were to be carried as far as Bullecourt to the north, and from Gonnelieu southward to Vendhuile. The whole operation, as we have seen, was under a time limit of forty-eight hours, while the Commander-in-Chief intended to call a halt before that period had expired if sufficient progress was not made. The first twenty-four hours was expected to show, one way or another, how much might be expected. But one rather startling factor resulted from the inadequate means to be employed. Every tank and infantryman in the six divisions was to go forward at zero. There were no reserves other than local ones; nothing was left for the second day--except the cavalry. The Guards and one or two other divisions were within a few marches of the battlefield, but not near enough to come up in time; nor were they intended to be used. The plan, in fact, was stamped plainly with that spirit of compromise due to ignorance and consequent mistrust of the capabilities of tanks. It was not really believed that the latter could do all that they claimed. Six divisions, therefore, were enough to risk on the gamble. On the other hand, in case the unexpected should happen, a most ambitious programme was drawn up, for which six divisions were hopelessly inadequate. The last German positions, beyond the Canal de L’Escaut, were to be carried in the first phase of the attack, after which cavalry was to cross the canal, occupy Cambrai and villages to the north and east, and move forward with its right on Valenciennes. In the meantime, the 4th Corps, on the right, was to form a defensive flank until the cavalry had cut the Valenciennes-Douai line nine or ten miles north of Cambrai (an incredible supposition), when the corps was to continue its advance in a north-easterly direction. The second and third phases of the battle were in the hands of the 3rd Corps on the left, whose business it was to capture Bourlon and Inchy-en-Artois (on the Canal du Nord), and then, by crossing the Arras-Cambrai road toward the marshes of the Sensée, to cut off the enemy’s troops in the angle which that river formed with the Canal du Nord. All this, with six divisions, reads like a dream, as in fact it was. Admitted that it represents the utmost possible and includes all that optimistic phase of “exploitation” which never throughout the war was reached in any battle, it remains an astonishing proposition. For if it was possible, three times six divisions would barely be enough to fill the enormously extended line we should have won; while as in the actual attack the infantry were found to be exhausted after an advance of four or five miles, it is difficult to see how, in the same time, they could be expected to cover ten or fifteen. Most extraordinary of all is the almost supernatural rôle allotted to the cavalry. It takes one back to the days of Marlborough and Prince Eugene, or to the triumphs of Murat. This venerable fallacy about cavalry would be pathetic if it were not so mischievous. For as an offensive weapon a man on a horse is an anachronism in any modern battle fought in an enclosed and cultivated country where everybody else is under cover. He cannot see his enemy, but can himself be seen a mile away, and, what is more pertinent, can be shot at that distance. He is tied to his horse, which he cannot hide. Even the pursuit of a routed enemy, supposed to be his especial _métier_, is governed now by conditions which render him useless. In every rout of a disciplined army some stout fellows will always be found as a rearguard, and two or three machine-guns can hold up a whole brigade of cavalry among villages and hedgerows. No ten feet high combination of man and beast, in short, has any chance whatever against modern small arms. It cannot be expected, however, that this ever will be admitted in an army controlled largely by influential cavalry soldiers; and as at Cambrai the mounted troops, with few exceptions, proved incapable even of arriving in time to attempt anything, the less said about them the better.

A further serious handicap to General Byng lay in the fact that most of his six divisions had come from the Salient, where they had been strained to the uttermost and had suffered heavy casualties. Their ranks were filled up with drafts who had no experience, while it had been impossible in the short time available for more than a small percentage to practice at Wailly with the tanks. But having enumerated these various adverse considerations, there remains much that was encouraging to be put against them. The divisions included some of the best in the army. The 29th had won an almost fabulous reputation in Gallipoli, although few of its original rank and file were left. The 6th was another formation with the traditions of regular service behind it. The 51st Highland Territorial Division (with which G Battalion had been associated in the Salient) had the distinction of coming second on a German list of those units most feared by the enemy. Among the troops taking part in the subsidiary attacks to the north, and likely to be involved in the main battle if all went well, were the 3rd Division, which headed the German list referred to, and the 40th London and 36th Ulster Divisions. On the right flank, where the 3rd Army linked up with the French, a strong force of French infantry and cavalry was put at our disposal. Then there was the known fact that the enemy’s position, held by Von der Marwitz’s 2nd Army, had been much depleted of troops, and especially of guns, while our own people must have gained confidence from the marked efficiency of our preliminary arrangements. Everything had gone well; every man, so far as was possible, knew what he had to do. Tanks and guns were in position; and the Germans in their deep trenches half a mile away, undisturbed by any breath of doubt or rumour, still performed mechanically the routine duties of a quiet sector, and reported that we were similarly employed.

The three Tank Brigades were disposed as follows: The 1st Brigade, of D, E, and G Battalions, with 42 tanks apiece, lay hidden in the western end of Havrincourt Wood, a large mass of timber of which the Germans actually held the eastern extremity, 3000 yards away. Colonel Hankey, of G Battalion, had under his command for the operation tanks from both the other units, giving him a total of 60 machines. The 2nd Brigade lay in Dessart Wood, two miles south of Havrincourt Wood. It consisted for the time of only two Battalions, B and H, A Battalion being attached to the 3rd Brigade. The tanks of the latter (A, C, F, and I Battalions), having no convenient wood in which to hide, were concealed under brick-coloured camouflage netting in and about Gouzeaucourt and Villers-Guislain. The battalions of the 2nd and 3rd Brigades had also 42 tanks apiece. Each unit set aside a small number as a reserve to replace mechanical breakdowns. In addition to the whole total of 378 fighting machines, there were 32 wire-clearing tanks with grapnels to open passages for the guns and cavalry; 2 for carrying bridging material; 1 with a telephone cable for the 3rd Army signals; 54 supply-tanks, some of them adapted gun-carriers; and 9 with wireless apparatus.

The 1st Brigade, on the left, was operating with the 62nd and 51st Divisions, its first objectives being Havrincourt village and Flesquières, with subsequent exploitation toward Graincourt, Anneux, Fontaine-Notre-Dame, and Bourlon.

The 2nd Brigade in the centre, leading the 6th Division, was to carry the Blue and Brown lines and Ribécourt, thence making for Marcoing on the Canal de L’Escaut and the Bois des Neufs further north.

The 3rd Brigade on the right, its 4 battalions working with the 29th, 20th, and 12th Divisions, had in hand the capture of La Vacquerie and the Blue and Brown lines beyond, then the bulging line of the canal from Marcoing to Masnières and the crossings at those places, and finally Crèvecoeur on the east bank at the foot of the Niergnies ridge.

Zero was fixed for 6.20 A.M. on the 20th. After dusk on the previous evening tank routes would be taped to the assembly positions. Before 6 o’clock all the tanks were to be ready in one long line from Trescault to Gonnelieu, in front of our trenches and within 500-1000 yards of the German outposts; the “advance-guard” tanks 150 yards ahead, and the main body deployed at approximate intervals of 100 yards. At 6.10 the whole mass would begin to move forward, followed by infantry in section columns. At zero, ten minutes later, a barrage of shrapnel, H.E. and smoke shells from upwards of 1000 guns would fall on the enemy’s outpost line, lifting after one minute, and moving forward in jumps of 250 yards, but remaining on certain objectives for stated periods. In addition, smoke-screens were to fall on the Flesquières ridge and along the right flank of the 4th Corps, together with the usual counter-battery work and bombardment of roads and bridges.

The weather had been fine, but so misty that little aeroplane observation had been possible. This did not affect us, as motives of secrecy had kept our aerial reconnaissance within the normal routine, and it prevented awkward prying on the enemy’s part. On the night of the 17th-18th November the Germans made a small raid near Havrincourt Wood and captured a few prisoners. These men were carried to Army Headquarters for investigation, a delay which had fortunate results for us, for some information about the forthcoming attack eventually was extracted from them. It came too late. A telegraphic warning reached the German front line only a few minutes before zero on the 20th, and was found half transcribed in a signaller’s dug-out.

III.

General Elles, on the evening of 19th November, issued his “Special Order No. 6,” in which he said: “To-morrow the Tank Corps will have the chance for which it has been waiting for many months--to operate on good going in the van of battle.... I propose leading the attack of the centre division.” Enough has been said by others about the better-known newspaper version: “England expects every tank to do its damndest”--but I cannot refrain from pointing out once more that this impudent lie was only in accordance with all the traditions of modern journalism, which knows neither scruples, education, nor good taste. Herodotus was in the habit of putting imaginary speeches into the mouths of his historical characters; but we have indeed sunk low when some illiterate scavenger of Fleet Street is permitted, and even encouraged, to employ his own characteristic phraseology for the same purpose in the columns of a newspaper.

Before midnight that evening, from Havrincourt Wood and Dessart Wood, from Gouzeaucourt and Villers-Guislain, scores of tanks had uncamouflaged and moved out and were already on their way. It was such a gathering of these singular machines as never had been seen before, and was to be seen only once again--on 8th August of the following year. Section after section, company after company (twenty-seven companies were there, as against two on the Somme a year earlier), they crawled forward in long columns that split up after a while into smaller ones, and then again into twos and threes, until toward five o’clock the whole were deploying into a single line six miles long--a threatening, silent, curving line that faced and corresponded to the larger salients and re-entrants of the Hindenburg trenches. So silently had this approach been carried out that many of the infantry, assembling behind the gaps in our own wire, heard no tanks at all, and inquired with some anxiety if they had arrived. A few had broken down, and were replaced at once from the mechanical reserve. Upwards of 350 were in position. The night was very dark, with a dense ground mist, but so complete were all arrangements that this caused no inconvenience, while it helped to screen the massing of this unparalleled armada. A sudden burst of shelling and trench-mortar fire from the German lines about 4.30 startled every one and provoked suspicions of a premature discovery, but it died away in half an hour. From five o’clock until zero the whole front was quiet. At ten minutes past six tanks and infantry began to move; and at 6.20, with the sky lightening rapidly above the mist, the barrage exploded with a shattering crash of sound along the German outpost line, some 250 yards in front of the “advance-guard” machines. In the centre of the six-mile line, in a tank called the “Hilda” of H Battalion, General Elles was leading into the most revolutionary battle of the war the corps which had made it possible, and which he had controlled almost from its infancy. It was not the ordinary post for the commander of a large organisation, but this was not an ordinary occasion. It was the consummation of two years of struggle and disappointment. There can have been little doubt in his mind, or in that of any other man in the 350 tanks, as to what the result would be now that the adventure was fairly launched.

The immediate onset of the tanks inevitably was overwhelming. The German outposts, dazed or annihilated by the sudden deluge of shells, were overrun in an instant. The triple belts of wire were crossed as if they had been beds of nettles, and 350 pathways were sheared through them for the infantry. The defenders of the front trench, scrambling out of dug-outs and shelters to meet the crash and flame of the barrage, saw the leading tanks almost upon them, their appearance made the more grotesque and terrifying by the huge black bundles they carried on their cabs. As these tanks swung left-handed and fired down into the trench, others, also surmounted by these appalling objects, appeared in multitudes behind them out of the mist. It is small wonder that the front Hindenburg Line, that fabulous excavation which was to be the bulwark of Germany, gave little trouble. The great fascines were loosed and rolled over the parapet to the trench floor; and down the whole line tanks were dipping and rearing up and clawing their way across into the almost unravaged country beyond. The defenders of the line were running panic-stricken, casting away arms and equipment. The Hindenburg Reserve, with its own massive entanglements, went the way of the first trenches; and so far our following infantry had found little to do beyond firing on the fugitives and rounding up gangs of half-stupefied prisoners. It was now broad day, the mist was thinning, and everywhere from Havrincourt to Banteux on the canal was rout and consternation. The Grand Ravin, a gully which runs from Havrincourt to Marcoing, was choked with an unarmed mob streaming eastward. The garrison of Havrincourt had put up a fight, but tanks closed on it from the north, and it was cleared by the 62nd Division.[20] La Vacquerie had gone in the first onset; and now Ribécourt was carried, and tanks and infantry were approaching the vital canal bridges at Marcoing and Masnières. Over the whole southern half of the battlefield, the triangle I have described, the defence had collapsed; and this area was virtually cleared by midday. Only at one or two points had there been any serious resistance, the most notable being Lateau Wood, on a hill-top overlooking the canal, where a number of batteries were posted. Fighting continued here during most of the morning, and included an encounter between a tank and a 5.9 howitzer, in which the tank for once was the victor. The gun blew away the tank’s right-hand sponson, but failed to injure the engine, and was overrun before it could be reloaded. Eventually the wood was cleared and the guns captured. By two o’clock Marcoing was in our hands, and all but a portion of Masnières. At the former place a tank of A Battalion arrived in time to drive off the German engineers who were in the act of destroying the bridge, which was on the far side of the village. At Masnières, however, the bridge was partially wrecked, and a tank of F Battalion, endeavouring to cross, completed the destruction and fell with the wreckage into the canal. This accident helped to delay our advance at this point upon Rumilly and the Niergnies ridge. In the meantime a hitch had occurred in the northern portion of the field. The 51st Division, advancing against Flesquières, as soon as the fall of Havrincourt and Ribécourt secured its flanks, had adopted a formation of attack which does not seem to have combined successfully with the tanks. The latter entered the village, but the infantry were too far behind and were held up by machine-gun fire. Throughout the day tanks were in and out of Flesquières, driving the Germans into cellars from which they emerged again as the danger passed; but the 51st could do no more than obtain a grip on the outskirts of the place. The château, surrounded by a stout wall and impregnable to tanks, was the centre of resistance. Night fell with the village still untaken.

Largely on account of the inability of the 51st to give adequate support, the tanks had very serious losses in and about Flesquières. A number of batteries lay just over the ridge, and tank after tank was hit and disabled as it appeared on the crest, the easiest of targets at almost point-blank range. One gun in the west end of the village itself was believed to have knocked out no fewer than sixteen machines of D and E Battalions. This was the celebrated piece served single-handed by a very gallant German officer after all his men had abandoned him. He was killed beside his gun in the end by the infantry. Had the latter been able to support the tanks more closely, most of these mischievous batteries could have been silenced before they had a chance of doing any damage. A few riflemen pushed ahead would have driven off the gunners while the tanks remained hidden by the crest. It is no part of a tank’s duty to make itself a helpless target for field-guns. As it was, while the wings of the army had swept on triumphantly to their third objectives and beyond, an unfortunate check had occurred in the centre. On the left of the 51st Division the 62nd, having cleared Havrincourt with the help of G Battalion’s tanks, pushed on northward over the narrow neck of the Flesquières ridge up the line of the Canal du Nord. Some of our tanks were caught by the same batteries that were holding up D and E. Muirhead, of 20 Company, was killed here, and five or six machines were disabled. Chaddock, who had his section commander, Fearnley, on board, had an extraordinary escape. A shell came through the cab, took off the driver’s head, and flung it on the knees of Fearnley at the brakes, killed or wounded the two right-hand gunners, and passed out at the stern without exploding. But the majority of the tanks were covered by a spur from the view of the batteries, and elsewhere there was no stopping this combination of G and the 62nd. Both the Hindenburg main and reserve lines were crossed in many places without the use of fascines, and tanks and infantry rolled swiftly down the slope to Graincourt, beyond which were the double row of trees along the Route Nationale No. 29, the Bapaume-Cambrai road, and the heavy mass of Bourlon Wood. Eight 6-pounder tanks advanced in line direct upon Graincourt, firing every gun they could bring to bear. Two German field-pieces concealed near the village were happily spotted by Charles Baker from the dust thrown up by the blast of their first discharge, and a few 6-pounder rounds drove away the gunners. By one o’clock Graincourt was captured. Still pushing on, tanks and infantry entered Anneux. This village was not wholly cleared that afternoon, for the 62nd were now exhausted. They had made the biggest advance of any division--4¹⁄₂ miles from their original front; but they would be the first to acknowledge that the tanks alone had made this possible, and had led them throughout. Some of G Battalion even had not finished yet. They crossed the Bapaume-Cambrai road and entered Bourlon Wood. There was little or no opposition: the great wood, which was to be the centre of such desperate fighting during the next week, appeared to be empty of defenders. It is certain that with a brigade of fresh infantry it could have been occupied and consolidated that evening. “The possession of Bourlon Ridge”--to quote Sir Douglas Haig’s despatch--“would enable our troops to obtain observation over the ground to the north, which sloped gently down to the Sensée river. The enemy’s defensive lines south of the Scarpe and Sensée rivers would thereby be turned, his communications exposed to the observed fire of our artillery, and his positions in this section jeopardised. In short, so great was the importance of the ridge to the enemy that its loss would probably cause the abandonment by the Germans of their carefully-prepared defence systems for a considerable distance to the north of it.” Bourlon Wood itself, at the southern end of the ridge, was also on the highest part, and dominated the remainder. We could have captured it with ease on the evening of the 20th, and continued the advance next day, but there was no infantry capable of this final effort. The tanks unsupported could not penetrate far into the gloomy wood, still less attempt to hold it; and after a while they withdrew.

The first day’s battle, in fact, was now at an end. It was after four o’clock, and the November twilight was falling. In the eastern outskirts of Anneux and Masnières, our furthest points of penetration, street-fighting was still in progress; and at Flesquières, far behind, the stubborn garrison continued to resist, although almost completely surrounded; but all around this isolated conflict the routine tasks of consolidation and clearance and preparation for the morrow were already in full swing. The meagre infantry reserves--some three or four brigades,--masses of belated cavalry and endless columns of guns, were pushing forward through the dusk; and everywhere the tanks were rolling home to their rallying-points. Except for the unfortunate check at Flesquières, the day had been one series of startling triumphs. In ten hours three defence systems, believed to be all but impregnable, had fallen like things of buckram: an area of ground as large as the whole Ypres salient, containing a dozen villages and hamlets, had been overrun; and 7000 or 8000 prisoners, with upwards of 100 guns, already were in our hands. Great masses of material had been captured; quartermasters’ stores, field post-offices, canteens, cinema theatres, hospitals--all these were among the miscellaneous acquisitions of that day. The surprise had been overwhelmingly complete, and the German infantry, with few exceptions, had been routed with an ease that had no precedent in the war, because the correct use of the determining factor had no precedent either. The panic spread far to the rear, for in Cambrai and beyond all was confusion and dismay.

These great results had been gained at a relatively trifling cost to ourselves; and it is impossible to reflect upon this beginning without deploring the inadequate resources which caused it, in the end, to remain no more than a brilliant but disappointing fragment. There is no need, however, to go into this subject again. We have seen already that there were no fresh reserves available, or none to speak of. Sir Julien Byng had in hand a few brigades of infantry, tired but hitherto not engaged, a couple of hundred tanks, whose stipulated work was more than accomplished, and whose crews and supplies alike were exhausted, and an entire cavalry corps which practically had done nothing at all. There should also perhaps be included the French troops on his right, and the divisions engaged in the subsidiary attack at Bullecourt, which had been highly successful; but none of these were in a position to reinforce him to a degree comparable to his opportunities. With another three or four divisions, instead of three or four brigades, behind his main battle front, there would have been some chance even of the optimistic schemes of exploitation being realised without any miracles by the cavalry. There were still thirty-six hours before the enemy could hope to collect his own reinforcements.

What, in the meantime, of the cavalry, who were to be galloping for the Sensée? The great proportion of these troops, whose talk (in Amiens or Boulogne) was always of speed and dash and initiative, and for whose benefit, that these qualities might be utilised, the special tanks had cleared wide passages in the wire, had arrived upon the field four hours late. Any one who has seen cavalry on the march trying to find their way anywhere by the use of a map will understand readily how this occurred; and it is probable that these laggards, even in such a _débâcle_, would very soon have ceased to be of use as mounted troops; but they could have fought very usefully on foot, as they were forced to do when they did arrive. One small unit, a squadron of the Canadian Fort Garry Horse, less pedantic or more skilled in map-reading than its regular colleagues, pushed early to the front, managed to cross the Canal de L’Escaut by an extemporised bridge, rushed the half-finished Beaurevoir-Masnières line on the slope beyond, and dispersed some infantry and captured a battery behind it. This quite exceptional feat may be held to prove that cavalry are still the great winners of battles that they were in the time of Turenne. It remains, however, an isolated performance; and the Canadians in the end met their inevitable fate. All their horses were killed or wounded, and they were reduced to fighting, like any infantrymen, behind the bank of a sunken road. Their gallantry, skill, and good fortune only served to emphasise the case of the numerous regular regiments who were unable to achieve anything at all.

It is necessary to descend now from the sublime to the ridiculous, and to consider briefly the condition of the nine tank battalions at the end of this first day’s fighting, to the success of which they had mainly contributed. “By 4 P.M. on the 20th November,” to quote ‘Weekly Tank Notes,’ “one of the most astonishing battles in all history had been won, and as far as the Tank Corps was concerned, tactically finished, for no reserves existing, it was not possible to do more than rally the now very weary and exhausted crews, select the fittest, and patch up composite companies to continue the attack on the morrow.” I have pointed out before the actual physical exhaustion, due to extreme heat, cramped quarters, jolting, and other causes, entailed in any prolonged operation in a tank. Of the machines which rallied on the evening of November 20th nearly all had been running for at least sixteen hours. Some had covered distances undreamt of in any previous tank action. All were out of petrol, water, and ammunition, and many had mechanical troubles that needed attention. The crews to a man were dog-tired, but before they could think of rest every remediable want had to be satisfied. Petrol and water tanks must be refilled, expended ammunition replaced, and minor repairs effected. For the brigade and battalion staffs there was the task of calculating losses and effectives, and of arranging for the formation of composite companies for the continuation of the battle next day. There was not an additional man or tank to be had: all had been in action; and, as the quotation above points out, the battle of Cambrai tactically was at an end for the Tank Corps. It had done all and more than it had promised to do, but there remained the necessity of continuing to help the no less exhausted infantry to round off, if possible, the first day’s victory. The losses in personnel had been considerable, and in tanks, from various causes, very serious. The 1st Brigade alone had nearly 40 tanks hit and disabled, mostly on the Flesquières ridge. It is doubtful if half of the original 378 fighting machines were now available, and for immediate use the proportion was far smaller. Nevertheless, early on the morning of the 21st, detachments from all three brigades were in action again. On the right, tanks of F Battalion moved out from Marcoing against the Beaurevoir-Masnières line, but the infantry, a fresh brigade unaccustomed to tanks, failed to co-operate, and the attack was a failure. Twenty-five tanks of B and H Battalions of the 2nd Brigade, with infantry of the 51st Division, rushed Cantaing, Noyelles, and Fontaine-Notre-Dame. Cantaing was carried by tanks alone, and handed over to the infantry by noon. Fontaine, which is virtually an outlying suburb of Cambrai, was occupied after severe fighting later in the day. In the meantime detachments from D and G completed the capture of Anneux, and carried a trench line to the north of it. Tanks once more pushed up the grass drives in Bourlon Wood, and once more had to return through lack of support. The wood was now occupied in force by the enemy, and was full of machine-guns. With the close of the day the original time limit of forty-eight hours expired; but the Commander-in-Chief, for reasons which he explains in his despatch, decided to allow the operation to continue. Great progress had been made, the enemy was badly shaken, and the situation in Italy, where the disaster of Caporetto had occurred, required further efforts to pin down the German reserves. Some additional divisions, including the Guards and two detailed for the Italian front, were at this late hour given to General Byng, and he was ordered to proceed with the attack. The 22nd, therefore, was spent mostly in reorganisation, and in preparing for a renewal of the battle. This involved a reshuffling of tank units among others. The centre of interest having shifted to Bourlon Ridge and its neighbourhood, the 3rd Tank Brigade was attached to the 3rd Corps, and proceeded to trek northward from Villers-Plouich, where all available tanks had concentrated. Further evidence of the importance of the northern area was forthcoming in the afternoon, when the Germans, whose own reserves were coming up, drove the 51st Division out of Fontaine-Notre-Dame. Nearly 60 tanks of the 1st and 2nd Brigades were collected for the attack on the 23rd. A furious struggle took place in Fontaine, lasting throughout the day. Eight times the tanks of the 2nd Brigade pushed through the village under an intense hail of bullets and grenades from the upper stories of the houses. As soon as they withdrew, the infantry, who had followed, were driven out again, and at length were unable to give any support at all. Field-guns had been brought into the village, and 11 tanks were hit and left there when the assault finally was abandoned. The 1st Brigade, however, admirably followed by the 40th Division, succeeded at last in clearing Bourlon Wood and village, in spite of the fact that G Battalion’s tanks were delayed till noon by bad traffic control holding up their supplies. There was very severe fighting in the wood and village, the tanks expending every round of 6-pdr. shell and most of their S.A.A. The village itself could not be held, but the great wood remained in our possession. Further to the west, about Mœuvres, on the Canal du Nord, progress was also made.

The Tank Corps was now fought virtually to a standstill. Rest and recuperation were imperative for officers, men, and machines, or for what was left. “On the 25th and 27th November further attacks were made by tanks and infantry on Bourlon and Fontaine-Notre-Dame with varying successes, but eventually both these villages remained in the hands of the enemy. During the attacks which had taken place since the 21st November, tank units had become terribly disorganised, and by the 27th had been reduced to such a state of exhaustion that it was determined to withdraw the 1st and 2nd Brigades. This withdrawal was nearing completion when the great German counter-attack was launched early on the morning of the 30th November.... It is an interesting point to remember that in this battle the attacking infantry were assisted by 690 officers and 3500 other ranks of the Tank Corps, a little over 4000 men or the strength of a strong brigade, and that these men replaced artillery wire-cutting, and rendered unnecessary the old preliminary bombardment. More than this, by keeping close to the infantry they effected a much higher co-operation than had ever before been attainable with artillery. When on the 21st November the bells of London pealed forth in celebration of the victory of Cambrai, consciously or unconsciously to their listeners they tolled out an old tactics and rang in a new--Cambrai had become the Valmy of a new epoch in war, the epoch of the mechanical engineer.”[21]

(German Photo.)]

(German Photo.)]

IV.

On 30th November, as everybody knows, the Germans launched two tremendous attacks on the north and south-east flanks of the pronounced salient which we now held in front of Cambrai. On the north, from Bourlon to Mœuvres, where the main blow was delivered, it ended in a sanguinary repulse. Far away to the south at Gouzeaucourt, however, where the line was very attenuated in consequence of our preoccupation with the Bourlon area, a very serious breakdown occurred. Assisted by a great number of low-flying aeroplanes and preceded by a short but intense bombardment, the attack for a time carried everything before it. Villers-Guislain and Gonnelieu were carried at the first rush, and by 9 A.M. Gouzeaucourt itself, the Headquarters of the 29th Division, was captured. It was a repetition on a smaller scale of our surprise ten days before. Hospitals, stores, and dumps of every kind of material, together with a great number of guns, were lost in the first three or four hours, and the roads westward were choked with stragglers and escaping transport. The nearest Tank Brigade to the danger point was the 2nd, concentrated about Fins railhead. A Battalion had rejoined, and the whole brigade was preparing to entrain for Plateau station and the winter quarters at Bray. Rumours of the attack arrived early, but it was not until 9.55 that a telephone message came through from the 3rd Corps. Although all the machines were more or less dismantled and totally unfit for action, within three hours 36 from A and B Battalions were moving toward Gouzeaucourt, followed two hours later by a reserve of 20 from H Battalion. The Guards having recaptured Gouzeaucourt, the tanks patrolled beyond as a screen. On the following morning a composite detachment from A and B helped the dismounted troopers of the 5th Cavalry Division to clear Gauche Wood, but attempts on Villers-Guislain and Gonnelieu--the latter by the Guards and tanks from H Battalion--were eventually held up. The tanks, however, did tremendous execution among the enemy, who was in great force, and this spirited rejoinder stopped effectively any further attempts by the Germans to advance. The rapidity with which these tank counter-attacks were organised and carried out with stripped and unready machines, and in circumstances of peculiar difficulty, was in itself highly creditable. It provided also decisive evidence on a point hitherto doubted by sceptics--the value of tanks in such an emergency.

As a result of the German surprise, the Cambrai Salient was contracted and Bourlon Wood abandoned. There was no further fighting for the Tank Corps. The 1st Brigade remained for a dreary fortnight in a defensive position at Dessart Wood, passing the nights in towing heavy guns back from Villers-Plouich. The other two brigades had withdrawn to Bray, and the 1st followed to Meaulte in the latter half of December.

NOTE.--Two authoritative German comments on the battle of
Cambrai show how fully the enemy realised that he had been
saved from a great disaster by our lack of faith in tanks.
Hindenburg, in his ‘Memoirs,’ writes as follows: “By neglecting
to exploit a brilliant initial success they had let victory
be snatched from them, and indeed by troops which were far
inferior to their own both in numbers and quality.... Moreover,
his High Command seemed to have failed to concentrate the
resources required to secure the execution of their plans
and their exploitation in case of success. Strong bodies of
cavalry assembled behind the triumphant leading infantry
divisions failed, even on this occasion, to overcome the last
line of resistance, weak though it was, which barred the way
to the flank and rear of their opponents. The English cavalry
squadrons were not able to conquer the German defence, even
with the help of their tanks, and proved unequal to decorating
their standards with that victory for which they had striven so
honourably and so often.”

General Otto von Moser, commanding the 14th Reserve Corps
during the battle, points out that the British “failed because
they luckily had not the dozen or two dozen divisions at hand
to exploit the undeniable success of the 20th November.”